From mjervis at gmail.com  Sun Aug  3 12:30:56 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Sun, 3 Aug 2008 17:30:56 +0100
Subject: [geeklog-devel] COM_makeClickableLinks
In-Reply-To: <609505460807300617n2c1321a1p15fb383d1523b3a8@mail.gmail.com>
References: <7b42e7470807281231h158dd9c9m26c57d5d97dcd428@mail.gmail.com>
	<609505460807291349j5950cdb8h25792c5e92c695d8@mail.gmail.com>
	<609505460807300617n2c1321a1p15fb383d1523b3a8@mail.gmail.com>
Message-ID: <7b42e7470808030930l3056eb4kd78619f427d46897@mail.gmail.com>

Cheers Sami,

It's looking a lot like we need to use regexp to locate/identify the
urls, but then php to parse them and turn them into links.

That gives an opportunity to shorten very long urls (within the anchor
not the href attribute (perhaps via tinyurl's API rather than
truncation))

On Wed, Jul 30, 2008 at 14:17, Sami Barakat <furiousdog at gmail.com> wrote:
> Hi,
>
> I think I've got it now, although its not a complete solution
>
> function COM_makeClickableLinks( $text )
> {
>    $text = preg_replace(
> '/([^"]?)(((ht|f)tps?):(\/\/)|(www\.))+((?=([^\s]+)&nbsp;))?(\8|[a-z0-9%&_\-\+,;=:@~#\/.\?\[\]]+)/is',
> '\\1<a href="http://\\6\\9">\\6\\9</a>', $text );
>    return $text;
> }
>
> It seems to work well with the following strings:
>
> normal link http://www.url.com
> normal link with early quote http://www.url.com/folder"stuff
> link with &nbsp; and quotes "http://www.url.com&nbsp;"
> www.url.com/ps&nbsp;
> complicated link www.sub.url.com/folder/index.php?id=foo&amp;user=bar&nbsp;
>
> it still fails however on these strings
>
> link with two &nbsp; www.url.com/ps&nbsp;&nbsp;
> link with early quote and &nbsp; "http://www.url.com/folder"stuff&nbsp;
>
> The results of the two failed strings is
>
> link with two &nbsp; <a
> href="http://www.url.com/ps&nbsp;">www.url.com/ps&nbsp;</a>&nbsp;
> link with early quote and &nbsp; "<a
> href="http://www.url.com/folder"stuff">www.url.com/folder"stuff</a>&nbsp;
>
> The second string could probably be fixed by replacing this part of
> the regular expression '[^\s]+' with this
> '[a-z0-9%&_\-\+,;=:@~#\/.\?\[\]]+'
> But really regular expressions are more helpful when validating
> strings or trying to find substrings in complicated strings, they are
> not really made to exclude parts of a string. So it might be more
> effective and less complicated to run through the expression twice.
> The first time matching urls with &nbsp; on the end and the second
> time without.
>
> Hope this helps
> Sami
>
> 2008/7/29 Sami Barakat <furiousdog at gmail.com>:
>> Hey,
>>
>> I have tried looking into this and I have come up with a partial
>> solution. From my understanding the problem is when a url has a &nbsp;
>> at the end which is getting parsed along with the url. I ask because I
>> think Gmail has filtered out some of them. Anyway the following regex
>>
>> ([^"]?)(((ht|f)tps?):(\/\/)|www\.)([a-z0-9%&_\-\+,;=:@~#\/.\?\[\]]+)(?<![&nbsp;])
>>
>> Seems to work fairly well. Here is the test code that I am using.
>>
>> echo '<pre>';
>> $string = "normal link http://www.url.com PASS\n";
>> echo htmlentities(COM_makeClickableLinks($string));
>> $string = "link with &nbsp; and quotes \"http://www.url.com&nbsp;\" PASS\n";
>> echo htmlentities(COM_makeClickableLinks($string));
>> $string = "complicated link
>> \"www.sub.url.com/folder/index.php?id=foo&amp;user=bar&nbsp;\"
>> PASS\n";
>> echo htmlentities(COM_makeClickableLinks($string));
>> $string = "problem link \"www.url.com/words&nbsp;\" FAIL\n";
>> echo htmlentities(COM_makeClickableLinks($string));
>> echo '</pre>';
>>
>> This produces
>>
>> normal link <a href="http://www.url.com">www.url.com</a> PASS
>> link with &nbsp; and quotes "<a
>> href="http://www.url.com">www.url.com</a>&nbsp;" PASS
>> complicated link "<a
>> href="http://sub.url.com/folder/index.php?id=foo&amp;user=bar">sub.url.com/folder/index.php?id=foo&amp;user=bar</a>&nbsp;"
>> PASS
>> problem link "<a href="http://url.com/word">url.com/word</a>s&nbsp;" FAIL
>>
>> As you can see the first 3 work, the problem occurs when a url ends
>> with any of the characters: '&' or 'n' or 'b' or 's' or 'p' or ';'
>>
>> So www.url.com/ps would return <a href="http://url.com/">url.com/</a>ps
>>
>> This is due to the last bit of the regex "(?<![&nbsp;])" if I tried
>> just doing (?<!&nbsp;) but it does not work at all because the
>> previous statement is being too greedy. There is also an issue with
>> the www. being removed, but thats not too much of a problem at the
>> moment.
>>
>> Also the COM_makeClickableLinks function can be simplified by removing
>> the str_replace statment resulting in simply this
>>
>> function COM_makeClickableLinks( $text )
>> {
>>    $text = preg_replace(
>> '/([^"]?)(((ht|f)tps?):(\/\/)|www\.)([a-z0-9%&_\-\+,;=:@~#\/.\?\[\]]+)(?<![&nbsp;])/is',
>> '\\1<a href="http://\\6">\\6</a>', $text );
>>    return $text;
>> }
>>
>>
>> in the original regex I was unsure why the "(\/|[+0-9a-z])" part was
>> included. I dont think its necessary so I took it out, maybe there was
>> a particular case that required it which Im overlooking.
>>
>> Anyhow I will have another crack at it later on, it really is a tough
>> one, but this is as far as ive got so far.
>>
>> Sami
>>
>> 2008/7/28 Michael Jervis <mjervis at gmail.com>:
>>> All (especially Sami!),
>>>
>>> There is a bug in the subject function. If it finds
>>> "http://www.url.com" we end up with &nbsp<a
>>> href=";http://www.url.com&nbsp">;http://www.url.com&nbsp</a>;
>>>
>>> Which isn't good.
>>>
>>> The original regexp in COM_MakeClickableLinks is:
>>>
>>> /([^"]?)((((ht|f)tps?):(\/\/)|www\.)[a-z0-9%&_\-\+,;=:@~#\/.\?\[\]]+(\/|[+0-9a-z]))/is
>>>
>>> I think the first match ([^"]?) is spurious, it matches anything other
>>> than  " before a link. So bhttp://www.foo.com" matches, but
>>> "http://www.foo.com doesn't.
>>>
>>> So that gives:
>>> /((((ht|f)tps?):(\/\/)|www\.)[a-z0-9%&_\-\+,;=:@~#\/.\?\[\]]+(\/|[+0-9a-z]))/is
>>>
>>> Resulting in:
>>> &nbsp;<a href="http:///www.url.com&nbsp">http://www.url.com&nbsp</a>
>>>
>>> So, need to add an "ignore trailing &nbsp;" bit to the clause. Closest
>>> I can get is:
>>> ((((ht|f)tps?):(\/\/)|www\.)[a-z0-9%&_\-\+,;=:@~#\/.\?\[\]]+(\/|[+0-9a-z]))(?=&nbsp;)
>>>
>>> Which results in:
>>> &nbsp;<a href="http:///www.url.com">http://www.url.com</a>&nbsp;
>>>
>>> However, unless there were quotes round the link, it won't match! So
>>> "http://www.foo.com" matches and is correctly processed, but
>>> http://www.foo.com is not matched.
>>>
>>> My head is now hurt. Any suggestions?
>>>
>>> --
>>> Michael Jervis
>>> mjervis at gmail.com
>>> 504B03041400000008008F846431E3543A820800000006000000060000007765
>>> 62676F642B4F4D4ACF4F0100504B010214001400000008008F846431E3543A82
>>> 0800000006000000060000000000000000002000000000000000776562676F64
>>> 504B05060000000001000100340000002C0000000000
>>> _______________________________________________
>>> geeklog-devel mailing list
>>> geeklog-devel at lists.geeklog.net
>>> http://eight.pairlist.net/mailman/listinfo/geeklog-devel
>>>
>>
> _______________________________________________
> geeklog-devel mailing list
> geeklog-devel at lists.geeklog.net
> http://eight.pairlist.net/mailman/listinfo/geeklog-devel
>



-- 
Michael Jervis
mjervis at gmail.com
504B03041400000008008F846431E3543A820800000006000000060000007765
62676F642B4F4D4ACF4F0100504B010214001400000008008F846431E3543A82
0800000006000000060000000000000000002000000000000000776562676F64
504B05060000000001000100340000002C0000000000


From dirk at haun-online.de  Tue Aug  5 14:05:22 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Tue, 5 Aug 2008 20:05:22 +0200
Subject: [geeklog-devel] 1.5.1 - what's left to do?
Message-ID: <20080805180522.491495310@smtp.haun-online.de>

So right now, only one bug is marked as "Target: Next Release" (#690,
and I'm not quite done with it yet).

I think we should also consider these two:

<http://project.geeklog.net/tracking/view.php?id=684>
<http://project.geeklog.net/tracking/view.php?id=698>

And FCKeditor 2.6.3 has just been released today - good timing to
include it as well.

Anything else? And any volunteers?

Can we get this done by the weekend, for a beta1 / rc1 next Monday maybe?

bye, Dirk


-- 
http://www.geeklog.net/
http://geeklog.info/



From devel at portalparts.com  Tue Aug  5 17:35:14 2008
From: devel at portalparts.com (Blaine Lang)
Date: Tue, 05 Aug 2008 17:35:14 -0400
Subject: [geeklog-devel] 1.5.1 - what's left to do?
In-Reply-To: <20080805180522.491495310@smtp.haun-online.de>
References: <20080805180522.491495310@smtp.haun-online.de>
Message-ID: <BLU0-SMTP300EBAE6F857110ADDA0E2BC7B0@phx.gbl>

I can take the FCKeditor upgrade

- blaine

Dirk Haun wrote:
> So right now, only one bug is marked as "Target: Next Release" (#690,
> and I'm not quite done with it yet).
>
> I think we should also consider these two:
>
> <http://project.geeklog.net/tracking/view.php?id=684>
> <http://project.geeklog.net/tracking/view.php?id=698>
>
> And FCKeditor 2.6.3 has just been released today - good timing to
> include it as well.
>
> Anything else? And any volunteers?
>
> Can we get this done by the weekend, for a beta1 / rc1 next Monday maybe?
>
> bye, Dirk
>
>
>   


From eakwarren at gmail.com  Tue Aug  5 22:54:30 2008
From: eakwarren at gmail.com (Eric Warren)
Date: Tue, 5 Aug 2008 20:54:30 -0600
Subject: [geeklog-devel] 1.5.1 - what's left to do?
In-Reply-To: <BLU0-SMTP300EBAE6F857110ADDA0E2BC7B0@phx.gbl>
References: <20080805180522.491495310@smtp.haun-online.de>
	<BLU0-SMTP300EBAE6F857110ADDA0E2BC7B0@phx.gbl>
Message-ID: <AB67BC3798E341CD8CEDE05F03C1261D@GeissPC>

In the spirit of collaboration... :-)

I just reported an issue to the geeklog bug tracker: 
http://project.geeklog.net/tracking/view.php?id=699 that might want to be 
assessed prior to wrapping up v1.5.1.

Thx!

Eric
----- Original Message ----- 
From: "Blaine Lang" <devel at portalparts.com>
To: "Geeklog Development" <geeklog-devel at lists.geeklog.net>
Sent: Tuesday, August 05, 2008 3:35 PM
Subject: Re: [geeklog-devel] 1.5.1 - what's left to do?


>I can take the FCKeditor upgrade
>
> - blaine
>
> Dirk Haun wrote:
>> So right now, only one bug is marked as "Target: Next Release" (#690,
>> and I'm not quite done with it yet).
>>
>> I think we should also consider these two:
>>
>> <http://project.geeklog.net/tracking/view.php?id=684>
>> <http://project.geeklog.net/tracking/view.php?id=698>
>>
>> And FCKeditor 2.6.3 has just been released today - good timing to
>> include it as well.
>>
>> Anything else? And any volunteers?
>>
>> Can we get this done by the weekend, for a beta1 / rc1 next Monday maybe?
>>
>> bye, Dirk
>>
>>
>>
> _______________________________________________
> geeklog-devel mailing list
> geeklog-devel at lists.geeklog.net
> http://eight.pairlist.net/mailman/listinfo/geeklog-devel 



From eakwarren at gmail.com  Wed Aug  6 10:36:08 2008
From: eakwarren at gmail.com (Eric Warren)
Date: Wed, 6 Aug 2008 08:36:08 -0600
Subject: [geeklog-devel] Fwd: [Geeklog 1 0000699]: Cannot un-exclude an
	author that you've previously excluded in your user preferences
In-Reply-To: <68d80951f803bbfffc580e46eec1bd27@project.geeklog.net>
References: <68d80951f803bbfffc580e46eec1bd27@project.geeklog.net>
Message-ID: <c51ed8330808060736g27d298dau67f88bd0d680b9f3@mail.gmail.com>

LOL! Yep, I suppose you're right, we do need to learn how to use our
browsers better! :-)

Thx!

Eric

---------- Forwarded message ----------
From: Geeklog Bugtracker <tracking at project.geeklog.net>
Date: Wed, Aug 6, 2008 at 1:33 AM
Subject: [Geeklog 1 0000699]: Cannot un-exclude an author that you've
previously excluded in your user preferences
To: eakwarren at gmail.com



The following issue has been RESOLVED.
======================================================================
http://project.geeklog.net/tracking/view.php?id=699
======================================================================
Reported By:                geiss
Assigned To:                dhaun
======================================================================
Project:                    Geeklog 1
Issue ID:                   699
Category:                   Bugs
Reproducibility:            always
Severity:                   minor
Priority:                   normal
Status:                     resolved
Target:                     not specified
Resolution:                 no change required
Fixed in Version:
======================================================================
Date Submitted:             2008-08-05 22:50 EDT
Last Modified:              2008-08-06 03:33 EDT
======================================================================
Summary:                    Cannot un-exclude an author that you've
previously
excluded in your user preferences
Description:
If you have excluded an author in your personal preferences, and want to
re-include them, there is no way to revert changes back. You have to
manually go into the database and change the 'aids' column in the
gl_userindex table. Refer to forum discussion at:

http://www.glfusion.org/forum/viewtopic.php?showtopic=11097 for detailed
info.
======================================================================

----------------------------------------------------------------------
 (0000252) dhaun (administrator) - 2008-08-06 03:33
 http://project.geeklog.net/tracking/view.php?id=699#c252
----------------------------------------------------------------------
Works for me.

It helps if you know how to use your browser ;-) That list is a
multi-select list, so you can select multiple entries. To un-select, you
usually need to hold down another key while clicking on the entry. It's the
Command key on my Mac (using Firefox 2). Other OS / browsers may use other
keys, e.g. Control.

Issue History
Date Modified    Username       Field                    Change
======================================================================
2008-08-05 22:50 geiss          New Issue
2008-08-06 03:33 dhaun          Note Added: 0000252
2008-08-06 03:33 dhaun          Status                   new => resolved
2008-08-06 03:33 dhaun          Resolution               open => no change
required
2008-08-06 03:33 dhaun          Assigned To               => dhaun
======================================================================
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://pairlist8.pair.net/pipermail/geeklog-devel/attachments/20080806/46fba9d8/attachment.html>

From Randy.Kolenko at nextide.ca  Wed Aug  6 10:39:47 2008
From: Randy.Kolenko at nextide.ca (Randy Kolenko)
Date: Wed, 6 Aug 2008 10:39:47 -0400
Subject: [geeklog-devel] SQL Server 2005 - MSSQL Driver
Message-ID: <063B8B70CB9DA141B2FC1DB483561B9F269FD7@nex-pluto.nextide.ca>

So where has the work on this gone?  The last communication as far as I
can see was back in May.

I have interest in this as Sami Barakat (GSOC Student) is nearing
completion of his project items -- one of which is MSSQL support.

-randy



> -----Original Message-----
> From: Kevin J. Peno [mailto:kevin at metalaxe.com] 
> Sent: May-17-08 12:40 PM
> To: Geeklog Development
> Subject: Re: [geeklog-devel] SQL Server 2005 - MSSQL Driver
> 
> 
> I wasn't the one that proposed it. There had been mention of 
> having 2 MSSQL drivers released, with separate options upon 
> install. I was wondering if this is something you guys wanted 
> to pursue or if you just want one driver option for MSSQL 
> during installation.
> 
> Kevin Peno
> 425-408-1094
> kevin at metalaxe.com
> 
> 
> -----Original Message-----
> From: geeklog-devel-bounces at lists.geeklog.net
> [mailto:geeklog-devel-bounces at lists.geeklog.net] On Behalf Of 
> Dirk Haun
> Sent: Saturday, May 17, 2008 5:56 AM
> To: geeklog-devel
> Subject: Re: [geeklog-devel] SQL Server 2005 - MSSQL Driver
> 
> Kevin J. Peno wrote:
> 
> >P.S. I never heard back about if you wanted to split release both 
> >Randy's and the MS driver I am working on. This is something I can 
> >implement in my copy while I'm waiting on feedback for the memory 
> >issues. So I figured I would ask again. It wouldn't be a difficult 
> >process. All the SQL data would be identical as far as I can see. Let
> me
> >know.
> 
> I may have missed something ... What exactly are you proposing?
> 
> bye, Dirk
> 
> 
> -- 
> http://spam.tinyweb.net/
> 
> _______________________________________________
> geeklog-devel mailing list
> geeklog-devel at lists.geeklog.net 
> http://eight.pairlist.net/mailman/listinfo/geeklog-devel
> _______________________________________________
> geeklog-devel mailing list
> geeklog-devel at lists.geeklog.net 
> http://eight.pairlist.net/mailman/listinfo/geeklog-devel
> 


From dirk at haun-online.de  Wed Aug  6 12:55:39 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Wed, 6 Aug 2008 18:55:39 +0200
Subject: [geeklog-devel] SQL Server 2005 - MSSQL Driver
In-Reply-To: <063B8B70CB9DA141B2FC1DB483561B9F269FD7@nex-pluto.nextide.ca>
References: <063B8B70CB9DA141B2FC1DB483561B9F269FD7@nex-pluto.nextide.ca>
Message-ID: <20080806165539.1253641503@smtp.haun-online.de>

Randy Kolenko wrote:

>So where has the work on this gone?  The last communication as far as I
>can see was back in May.

FWIW, I've sent an email to Kevin and Garrett (at Microsoft) 10 days ago
and only received an out-of-office reply from Garrett.

I was going to ask you next if you had seen any code ...

bye, Dirk


-- 
http://www.geeklog.net/
http://geeklog.info/



From dirk at haun-online.de  Wed Aug  6 14:05:43 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Wed, 6 Aug 2008 20:05:43 +0200
Subject: [geeklog-devel] appfs
Message-ID: <20080806180543.1197483644@smtp.haun-online.de>

I've raved about appfs (in connection with the Atompub support in
Geeklog) before, but got the impression that people didn't quite "get" it ;-)

Here's an attempt to visualize it:

  http://www.flickr.com/photos/dhaun/2739285076/

bye, Dirk


-- 
http://www.geeklog.net/
http://geeklog.info/



From kevin at metalaxe.com  Wed Aug  6 14:03:51 2008
From: kevin at metalaxe.com (Kevin J. Peno)
Date: Wed, 6 Aug 2008 11:03:51 -0700
Subject: [geeklog-devel] SQL Server 2005 - MSSQL Driver
In-Reply-To: <20080806165539.1253641503@smtp.haun-online.de>
References: <063B8B70CB9DA141B2FC1DB483561B9F269FD7@nex-pluto.nextide.ca>
	<20080806165539.1253641503@smtp.haun-online.de>
Message-ID: <9CE911D90980BB4980C2CFD886ED6CB1045E41@hive.metalaxe.com>

Hey Dirk,

I actually didn't receive your email for some reason, could you forward
it to me again? I was working on it again for about 2 weeks because they
finally found a fix for those driver issues we were having so long ago.
However I was told to stop work do to the fiscal year shuffle.

At this point I have installation working fine, just running through the
misc compatibility bugs between the mssql driver and sql-native. As soon
as I can I'll get you guys something to start looking at.

Regards,

Kevin Peno
Metalaxe, LLC
425-408-1094
kevin at metalaxe.com
sales at metalaxe.com


-----Original Message-----
From: geeklog-devel-bounces at lists.geeklog.net
[mailto:geeklog-devel-bounces at lists.geeklog.net] On Behalf Of Dirk Haun
Sent: Wednesday, August 06, 2008 9:56 AM
To: geeklog-devel
Subject: Re: [geeklog-devel] SQL Server 2005 - MSSQL Driver

Randy Kolenko wrote:

>So where has the work on this gone?  The last communication as far as I
>can see was back in May.

FWIW, I've sent an email to Kevin and Garrett (at Microsoft) 10 days ago
and only received an out-of-office reply from Garrett.

I was going to ask you next if you had seen any code ...

bye, Dirk


-- 
http://www.geeklog.net/
http://geeklog.info/

_______________________________________________
geeklog-devel mailing list
geeklog-devel at lists.geeklog.net
http://eight.pairlist.net/mailman/listinfo/geeklog-devel


From dirk at haun-online.de  Wed Aug  6 14:41:53 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Wed, 6 Aug 2008 20:41:53 +0200
Subject: [geeklog-devel] SQL Server 2005 - MSSQL Driver
In-Reply-To: <9CE911D90980BB4980C2CFD886ED6CB1045E41@hive.metalaxe.com>
References: <063B8B70CB9DA141B2FC1DB483561B9F269FD7@nex-pluto.nextide.ca>
	<20080806165539.1253641503@smtp.haun-online.de>
	<9CE911D90980BB4980C2CFD886ED6CB1045E41@hive.metalaxe.com>
Message-ID: <20080806184153.1238826028@smtp.haun-online.de>

Kevin J. Peno wrote:

>I actually didn't receive your email for some reason, could you forward
>it to me again?

On its way ...

bye, Dirk


-- 
http://www.haun-online.de/
http://geeklog.info/



From mjervis at gmail.com  Wed Aug  6 16:13:45 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Wed, 6 Aug 2008 21:13:45 +0100
Subject: [geeklog-devel] appfs
In-Reply-To: <20080806180543.1197483644@smtp.haun-online.de>
References: <20080806180543.1197483644@smtp.haun-online.de>
Message-ID: <7b42e7470808061313g4a9329c4of9d9dfca55d6be4f@mail.gmail.com>

Sounds neat. Shame it doesn't work on windows ;-)

On Wed, Aug 6, 2008 at 19:05, Dirk Haun <dirk at haun-online.de> wrote:
> I've raved about appfs (in connection with the Atompub support in
> Geeklog) before, but got the impression that people didn't quite "get" it ;-)
>
> Here's an attempt to visualize it:
>
>  http://www.flickr.com/photos/dhaun/2739285076/
>
> bye, Dirk
>
>
> --
> http://www.geeklog.net/
> http://geeklog.info/
>
> _______________________________________________
> geeklog-devel mailing list
> geeklog-devel at lists.geeklog.net
> http://eight.pairlist.net/mailman/listinfo/geeklog-devel
>



-- 
Michael Jervis
mjervis at gmail.com
504B03041400000008008F846431E3543A820800000006000000060000007765
62676F642B4F4D4ACF4F0100504B010214001400000008008F846431E3543A82
0800000006000000060000000000000000002000000000000000776562676F64
504B05060000000001000100340000002C0000000000


From dirk at haun-online.de  Wed Aug  6 16:23:10 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Wed, 6 Aug 2008 22:23:10 +0200
Subject: [geeklog-devel] appfs
In-Reply-To: <7b42e7470808061313g4a9329c4of9d9dfca55d6be4f@mail.gmail.com>
References: <20080806180543.1197483644@smtp.haun-online.de>
	<7b42e7470808061313g4a9329c4of9d9dfca55d6be4f@mail.gmail.com>
Message-ID: <20080806202310.924593848@smtp.haun-online.de>

Michael Jervis wrote:

>Sounds neat.

It is. Just think how simple it would be for less technically adept
people to update their website: Just save your article in a folder and
it magically appears on the site.


>Shame it doesn't work on windows ;-)

Would need a working port of FUSE first. The rest is only a bunch of
Ruby extensions. The actual appfs.rb script is surprisingly short.

I didn't get too far with MacFUSE either, though, last time I tried ...

bye, Dirk


-- 
http://www.haun-online.de/accu/



From furiousdog at gmail.com  Sun Aug 10 09:30:26 2008
From: furiousdog at gmail.com (Sami Barakat)
Date: Sun, 10 Aug 2008 14:30:26 +0100
Subject: [geeklog-devel] COM_makeClickableLinks
In-Reply-To: <7b42e7470808030930l3056eb4kd78619f427d46897@mail.gmail.com>
References: <7b42e7470807281231h158dd9c9m26c57d5d97dcd428@mail.gmail.com>
	<609505460807291349j5950cdb8h25792c5e92c695d8@mail.gmail.com>
	<609505460807300617n2c1321a1p15fb383d1523b3a8@mail.gmail.com>
	<7b42e7470808030930l3056eb4kd78619f427d46897@mail.gmail.com>
Message-ID: <609505460808100630u956f9f0la3f488b478e182b8@mail.gmail.com>

Hi Michael,

I dont know if you are still working on this but I came across a web
site with a good regular expression just for this situation...but its
a bit longer then what we have been working with so far....wait for
it....

^((ht|f)tp(s?)\:\/\/|~/|/)?([\w]+:\w+@)?([a-zA-Z]{1}([\w\-]+\.)+([\w]{2,5}))(:[\d]{1,5})?((/?\w+/)+|/?)(\w+\.[\w]{3,4})?((\?\w+=\w+)?(&\w+=\w+)*)?

yup thats it. Although it looks complicated its actually very well
made and can handle a wide variety of urls. More info can be found at
this site where I found it
http://geekswithblogs.net/casualjim/archive/2005/12/01/61722.aspx

That regex wont work straight away, it had to be modified a little to
take account for the problem we had originally. So here is the
modified version

((ht|f)tp(s?)\:\/\/|~\/|\/)?([\w]+:\w+@)?(([a-zA-Z]{1}([\w\-]+\.)+([\w]{2,5}))(:[\d]{1,5})?((\/?\w+\/)+|\/?)([\w\-%]+(\.[\w]{3,4})?)?((\?|&|&amp;)[\w\-%]+=[\w\-%]+)*)

It works with all the test urls we have been using and then some. I
did a little benchmarking and this one works out to be around half a
second longer when executed 10,000 times, so although it looks nasty
half a sec is nothing to complain about.

Here is the complete function

function COM_makeClickableLinks( $text )
{
    $regex = '/((ht|f)tp(s?)\:\/\/|~\/|\/)?([\w]+:\w+@)?(([a-zA-Z]{1}([\w\-]+\.)+([\w]{2,5}))(:[\d]{1,5})?((\/?\w+\/)+|\/?)([\w\-%]+(\.[\w]{3,4})?)?((\?|&|&amp;)[\w\-%]+=[\w\-%]+)*)/is';

    $text = preg_replace( $regex, '<a href="http://\\5">\\5</a>', $text );
    return $text;
}

A simple way to shorten the urls would be to just print the domain in
the anchor,
so this link
    http://www.sub.url.com/folder?user=5&this=that
would be converted to
    <a href="http://www.sub.url.com/folder?user=5&this=that">www.sub.url.com</a>
Maybe even include that little arrow icon that Wikipedia uses when
linking off site.

This can be done by changing the $replacement part of the preg function
    '<a href="\\1\\5">\\5</a>'
to
    '<a href="\\1\\5">\\6</a>'

\\1 = the http:// part of the url (or ftp:// or https:// etc.)
\\5 = the full url (without the http:// part)
\\6 = the domain name (www.url.com or url.com or www.sub.url.com etc.)

enjoy...

Sami

2008/8/3 Michael Jervis <mjervis at gmail.com>:
> Cheers Sami,
>
> It's looking a lot like we need to use regexp to locate/identify the
> urls, but then php to parse them and turn them into links.
>
> That gives an opportunity to shorten very long urls (within the anchor
> not the href attribute (perhaps via tinyurl's API rather than
> truncation))
>
> On Wed, Jul 30, 2008 at 14:17, Sami Barakat <furiousdog at gmail.com> wrote:
>> Hi,
>>
>> I think I've got it now, although its not a complete solution
>>
>> function COM_makeClickableLinks( $text )
>> {
>>    $text = preg_replace(
>> '/([^"]?)(((ht|f)tps?):(\/\/)|(www\.))+((?=([^\s]+)&nbsp;))?(\8|[a-z0-9%&_\-\+,;=:@~#\/.\?\[\]]+)/is',
>> '\\1<a href="http://\\6\\9">\\6\\9</a>', $text );
>>    return $text;
>> }
>>
>> It seems to work well with the following strings:
>>
>> normal link http://www.url.com
>> normal link with early quote http://www.url.com/folder"stuff
>> link with &nbsp; and quotes "http://www.url.com&nbsp;"
>> www.url.com/ps&nbsp;
>> complicated link www.sub.url.com/folder/index.php?id=foo&amp;user=bar&nbsp;
>>
>> it still fails however on these strings
>>
>> link with two &nbsp; www.url.com/ps&nbsp;&nbsp;
>> link with early quote and &nbsp; "http://www.url.com/folder"stuff&nbsp;
>>
>> The results of the two failed strings is
>>
>> link with two &nbsp; <a
>> href="http://www.url.com/ps&nbsp;">www.url.com/ps&nbsp;</a>&nbsp;
>> link with early quote and &nbsp; "<a
>> href="http://www.url.com/folder"stuff">www.url.com/folder"stuff</a>&nbsp;
>>
>> The second string could probably be fixed by replacing this part of
>> the regular expression '[^\s]+' with this
>> '[a-z0-9%&_\-\+,;=:@~#\/.\?\[\]]+'
>> But really regular expressions are more helpful when validating
>> strings or trying to find substrings in complicated strings, they are
>> not really made to exclude parts of a string. So it might be more
>> effective and less complicated to run through the expression twice.
>> The first time matching urls with &nbsp; on the end and the second
>> time without.
>>
>> Hope this helps
>> Sami
>>
>> 2008/7/29 Sami Barakat <furiousdog at gmail.com>:
>>> Hey,
>>>
>>> I have tried looking into this and I have come up with a partial
>>> solution. From my understanding the problem is when a url has a &nbsp;
>>> at the end which is getting parsed along with the url. I ask because I
>>> think Gmail has filtered out some of them. Anyway the following regex
>>>
>>> ([^"]?)(((ht|f)tps?):(\/\/)|www\.)([a-z0-9%&_\-\+,;=:@~#\/.\?\[\]]+)(?<![&nbsp;])
>>>
>>> Seems to work fairly well. Here is the test code that I am using.
>>>
>>> echo '<pre>';
>>> $string = "normal link http://www.url.com PASS\n";
>>> echo htmlentities(COM_makeClickableLinks($string));
>>> $string = "link with &nbsp; and quotes \"http://www.url.com&nbsp;\" PASS\n";
>>> echo htmlentities(COM_makeClickableLinks($string));
>>> $string = "complicated link
>>> \"www.sub.url.com/folder/index.php?id=foo&amp;user=bar&nbsp;\"
>>> PASS\n";
>>> echo htmlentities(COM_makeClickableLinks($string));
>>> $string = "problem link \"www.url.com/words&nbsp;\" FAIL\n";
>>> echo htmlentities(COM_makeClickableLinks($string));
>>> echo '</pre>';
>>>
>>> This produces
>>>
>>> normal link <a href="http://www.url.com">www.url.com</a> PASS
>>> link with &nbsp; and quotes "<a
>>> href="http://www.url.com">www.url.com</a>&nbsp;" PASS
>>> complicated link "<a
>>> href="http://sub.url.com/folder/index.php?id=foo&amp;user=bar">sub.url.com/folder/index.php?id=foo&amp;user=bar</a>&nbsp;"
>>> PASS
>>> problem link "<a href="http://url.com/word">url.com/word</a>s&nbsp;" FAIL
>>>
>>> As you can see the first 3 work, the problem occurs when a url ends
>>> with any of the characters: '&' or 'n' or 'b' or 's' or 'p' or ';'
>>>
>>> So www.url.com/ps would return <a href="http://url.com/">url.com/</a>ps
>>>
>>> This is due to the last bit of the regex "(?<![&nbsp;])" if I tried
>>> just doing (?<!&nbsp;) but it does not work at all because the
>>> previous statement is being too greedy. There is also an issue with
>>> the www. being removed, but thats not too much of a problem at the
>>> moment.
>>>
>>> Also the COM_makeClickableLinks function can be simplified by removing
>>> the str_replace statment resulting in simply this
>>>
>>> function COM_makeClickableLinks( $text )
>>> {
>>>    $text = preg_replace(
>>> '/([^"]?)(((ht|f)tps?):(\/\/)|www\.)([a-z0-9%&_\-\+,;=:@~#\/.\?\[\]]+)(?<![&nbsp;])/is',
>>> '\\1<a href="http://\\6">\\6</a>', $text );
>>>    return $text;
>>> }
>>>
>>>
>>> in the original regex I was unsure why the "(\/|[+0-9a-z])" part was
>>> included. I dont think its necessary so I took it out, maybe there was
>>> a particular case that required it which Im overlooking.
>>>
>>> Anyhow I will have another crack at it later on, it really is a tough
>>> one, but this is as far as ive got so far.
>>>
>>> Sami
>>>
>>> 2008/7/28 Michael Jervis <mjervis at gmail.com>:
>>>> All (especially Sami!),
>>>>
>>>> There is a bug in the subject function. If it finds
>>>> "http://www.url.com" we end up with &nbsp<a
>>>> href=";http://www.url.com&nbsp">;http://www.url.com&nbsp</a>;
>>>>
>>>> Which isn't good.
>>>>
>>>> The original regexp in COM_MakeClickableLinks is:
>>>>
>>>> /([^"]?)((((ht|f)tps?):(\/\/)|www\.)[a-z0-9%&_\-\+,;=:@~#\/.\?\[\]]+(\/|[+0-9a-z]))/is
>>>>
>>>> I think the first match ([^"]?) is spurious, it matches anything other
>>>> than  " before a link. So bhttp://www.foo.com" matches, but
>>>> "http://www.foo.com doesn't.
>>>>
>>>> So that gives:
>>>> /((((ht|f)tps?):(\/\/)|www\.)[a-z0-9%&_\-\+,;=:@~#\/.\?\[\]]+(\/|[+0-9a-z]))/is
>>>>
>>>> Resulting in:
>>>> &nbsp;<a href="http:///www.url.com&nbsp">http://www.url.com&nbsp</a>
>>>>
>>>> So, need to add an "ignore trailing &nbsp;" bit to the clause. Closest
>>>> I can get is:
>>>> ((((ht|f)tps?):(\/\/)|www\.)[a-z0-9%&_\-\+,;=:@~#\/.\?\[\]]+(\/|[+0-9a-z]))(?=&nbsp;)
>>>>
>>>> Which results in:
>>>> &nbsp;<a href="http:///www.url.com">http://www.url.com</a>&nbsp;
>>>>
>>>> However, unless there were quotes round the link, it won't match! So
>>>> "http://www.foo.com" matches and is correctly processed, but
>>>> http://www.foo.com is not matched.
>>>>
>>>> My head is now hurt. Any suggestions?
>>>>
>>>> --
>>>> Michael Jervis
>>>> mjervis at gmail.com
>>>> 504B03041400000008008F846431E3543A820800000006000000060000007765
>>>> 62676F642B4F4D4ACF4F0100504B010214001400000008008F846431E3543A82
>>>> 0800000006000000060000000000000000002000000000000000776562676F64
>>>> 504B05060000000001000100340000002C0000000000
>>>> _______________________________________________
>>>> geeklog-devel mailing list
>>>> geeklog-devel at lists.geeklog.net
>>>> http://eight.pairlist.net/mailman/listinfo/geeklog-devel
>>>>
>>>
>> _______________________________________________
>> geeklog-devel mailing list
>> geeklog-devel at lists.geeklog.net
>> http://eight.pairlist.net/mailman/listinfo/geeklog-devel
>>
>
>
>
> --
> Michael Jervis
> mjervis at gmail.com
> 504B03041400000008008F846431E3543A820800000006000000060000007765
> 62676F642B4F4D4ACF4F0100504B010214001400000008008F846431E3543A82
> 0800000006000000060000000000000000002000000000000000776562676F64
> 504B05060000000001000100340000002C0000000000
> _______________________________________________
> geeklog-devel mailing list
> geeklog-devel at lists.geeklog.net
> http://eight.pairlist.net/mailman/listinfo/geeklog-devel
>


From dirk at haun-online.de  Sun Aug 10 12:59:17 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Sun, 10 Aug 2008 18:59:17 +0200
Subject: [geeklog-devel] 1.5.1 - what's left to do?
In-Reply-To: <20080805180522.491495310@smtp.haun-online.de>
References: <20080805180522.491495310@smtp.haun-online.de>
Message-ID: <20080810165917.1984929954@smtp.haun-online.de>

Dirk Haun wrote:

> <http://project.geeklog.net/tracking/view.php?id=698>

The entire block handling is giving me headaches ...

The problem is that COM_showBlocks('right'); is called twice, at least
under normal circumstances.

From what I can see, the code in lines 1355-1371 is mostly duplicated in
lines 1410-1430, but to make things more interesting, the latter version
also involves an - otherwise undefined - variable $what.

One of those two copies should probably be removed (and that $what as
well). Not sure which one, though.

Help! Please?

bye, Dirk


-- 
http://www.geeklog.net/
http://geeklog.info/



From devel at portalparts.com  Sun Aug 10 13:35:30 2008
From: devel at portalparts.com (Blaine Lang)
Date: Sun, 10 Aug 2008 13:35:30 -0400
Subject: [geeklog-devel] 1.5.1 - what's left to do?
In-Reply-To: <20080810165917.1984929954@smtp.haun-online.de>
References: <20080805180522.491495310@smtp.haun-online.de>
	<20080810165917.1984929954@smtp.haun-online.de>
Message-ID: <BLU0-SMTP12E75C8837048A7FCB1C56BC760@phx.gbl>

Looks like the same issue is there for the leftblocks.

When/Who added this code as I don't recall using this feature before to 
have blocks appear in the footer.
$_CONF['left_blocks_in_footer']
$_CONF['right_blocks_in_footer']

I understand why one may want this feature but possibly we should be 
setting a new template variable that only would be used in the footer 
template.

- blaine

Dirk Haun wrote:
> Dirk Haun wrote:
>
>   
>> <http://project.geeklog.net/tracking/view.php?id=698>
>>     
>
> The entire block handling is giving me headaches ...
>
> The problem is that COM_showBlocks('right'); is called twice, at least
> under normal circumstances.
>
> >From what I can see, the code in lines 1355-1371 is mostly duplicated in
> lines 1410-1430, but to make things more interesting, the latter version
> also involves an - otherwise undefined - variable $what.
>
> One of those two copies should probably be removed (and that $what as
> well). Not sure which one, though.
>
> Help! Please?
>
> bye, Dirk
>
>
>   


From dirk at haun-online.de  Sun Aug 10 14:04:45 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Sun, 10 Aug 2008 20:04:45 +0200
Subject: [geeklog-devel] 1.5.1 - what's left to do?
In-Reply-To: <BLU0-SMTP12E75C8837048A7FCB1C56BC760@phx.gbl>
References: <20080805180522.491495310@smtp.haun-online.de>
	<20080810165917.1984929954@smtp.haun-online.de>
	<BLU0-SMTP12E75C8837048A7FCB1C56BC760@phx.gbl>
Message-ID: <20080810180445.1650397754@smtp.haun-online.de>

Blaine Lang wrote:

>When/Who added this code as I don't recall using this feature before to 
>have blocks appear in the footer.
>$_CONF['left_blocks_in_footer']
>$_CONF['right_blocks_in_footer']

I introduced 'left_blocks_in_footer' back in 1.3.10 as an easy way to
create two-column layouts:

>- Introduced $_CONF['left_blocks_in_footer'] config option that makes the
>  {left_blocks} variable available in footer.thtml (disabling it in
>header.thtml
>  at the same time). This is really only useful for two-column layouts where
>  you want the side blocks in the right column [Dirk]

So it effectively makes all left-side blocks appear on the right side
automatically.


Looks like Oliver added the right_blocks variant some time during 1.5.0
development, probably for symmetry:

<http://project.geeklog.net/cgi-bin/viewvc.cgi/geeklog/Geeklog-1.x/
public_html/lib-common.php?r1=1.656&r2=1.657>


>I understand why one may want this feature but possibly we should be 
>setting a new template variable that only would be used in the footer 
>template.

Not sure if anyone's using $_CONF['right_blocks_in_footer']. If not, I'd
say we remove it again.

bye, Dirk


-- 
http://www.haun-online.de/
http://geeklog.info/



From devel at portalparts.com  Sun Aug 10 14:11:23 2008
From: devel at portalparts.com (Blaine Lang)
Date: Sun, 10 Aug 2008 14:11:23 -0400
Subject: [geeklog-devel] 1.5.1 - what's left to do?
In-Reply-To: <20080810180445.1650397754@smtp.haun-online.de>
References: <20080805180522.491495310@smtp.haun-online.de>	<20080810165917.1984929954@smtp.haun-online.de>	<BLU0-SMTP12E75C8837048A7FCB1C56BC760@phx.gbl>
	<20080810180445.1650397754@smtp.haun-online.de>
Message-ID: <BLU0-SMTP195AA579FAE722A26D7287BC760@phx.gbl>

Dirk Haun wrote:

> Not sure if anyone's using $_CONF['right_blocks_in_footer']. If not, I'd say we remove it again.

Why not just have it set a new template variable like {footer_rightblocks} ?

- blaine


From dirk at haun-online.de  Sun Aug 10 14:57:50 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Sun, 10 Aug 2008 20:57:50 +0200
Subject: [geeklog-devel] 1.5.1 - what's left to do?
In-Reply-To: <BLU0-SMTP195AA579FAE722A26D7287BC760@phx.gbl>
References: <20080805180522.491495310@smtp.haun-online.de>	<20080810165917.1984929954@smt
	p.haun-online.de>	<BLU0-SMTP12E75C8837048A7FCB1C56BC760@phx.gbl>
	<20080810180445.1650397754@smtp.haun-online.de>
	<BLU0-SMTP195AA579FAE722A26D7287BC760@phx.gbl>
Message-ID: <20080810185750.567718756@smtp.haun-online.de>

Blaine Lang wrote:

>Why not just have it set a new template variable like {footer_rightblocks} ?

Actually - what is right_blocks_in_footer supposed to do? They _are_
already in the footer. The left blocks are part of the header and the
right blocks are part of the footer. 'left_blocks_in_footer' has its use
(as explained). And if you need even more control, there's the $custom
parameter.

So unless I'm missing something, I don't see a need for the right_blocks
option.

bye, Dirk


-- 
http://www.haun-online.de/
http://geeklog.info/



From mjervis at gmail.com  Mon Aug 11 14:59:39 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Mon, 11 Aug 2008 19:59:39 +0100
Subject: [geeklog-devel] CVS Broken?
Message-ID: <7b42e7470808111159t3709d43cs364de97801c9198e@mail.gmail.com>

Trying to fix further mssql install/upgrade issues:

cvs [commit aborted]: received abort signal
Assertion failed: (rev == NULL || isdigit ((unsigned char) *rev)),
function RCS_checkout, file
/usr/src/gnu/usr.bin/cvs/cvs/../../../../contrib/cvs/src/rcs.c, line
4172.



-- 
Michael Jervis
mjervis at gmail.com
504B03041400000008008F846431E3543A820800000006000000060000007765
62676F642B4F4D4ACF4F0100504B010214001400000008008F846431E3543A82
0800000006000000060000000000000000002000000000000000776562676F64
504B05060000000001000100340000002C0000000000


From dirk at haun-online.de  Mon Aug 11 15:15:46 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Mon, 11 Aug 2008 21:15:46 +0200
Subject: [geeklog-devel] CVS Broken?
In-Reply-To: <7b42e7470808111159t3709d43cs364de97801c9198e@mail.gmail.com>
References: <7b42e7470808111159t3709d43cs364de97801c9198e@mail.gmail.com>
Message-ID: <20080811191546.1351080846@smtp.haun-online.de>

Michael Jervis wrote:

>cvs [commit aborted]: received abort signal
>Assertion failed: (rev == NULL || isdigit ((unsigned char) *rev)),
>function RCS_checkout, file
>/usr/src/gnu/usr.bin/cvs/cvs/../../../../contrib/cvs/src/rcs.c, line
>4172.

Works for me (see notifications). A temporary glitch?

bye, Dirk


-- 
http://www.haun-online.de/accu/



From mjervis at gmail.com  Mon Aug 11 15:24:17 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Mon, 11 Aug 2008 20:24:17 +0100
Subject: [geeklog-devel] CVS Broken?
In-Reply-To: <20080811191546.1351080846@smtp.haun-online.de>
References: <7b42e7470808111159t3709d43cs364de97801c9198e@mail.gmail.com>
	<20080811191546.1351080846@smtp.haun-online.de>
Message-ID: <7b42e7470808111224g1f309acj8370ca5f2eb42ae2@mail.gmail.com>

Assertion failed: (rev == NULL || isdigit ((unsigned char) *rev)),
function RCS_checkout, file
/usr/src/gnu/usr.bin/cvs/cvs/../../../../contrib/cvs/src/rcs.c, line
4172.
cvs [commit aborted]: received abort signal


Still getting it...


From dirk at haun-online.de  Mon Aug 11 15:35:17 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Mon, 11 Aug 2008 21:35:17 +0200
Subject: [geeklog-devel] CVS Broken?
In-Reply-To: <7b42e7470808111224g1f309acj8370ca5f2eb42ae2@mail.gmail.com>
References: <7b42e7470808111159t3709d43cs364de97801c9198e@mail.gmail.com>
	<20080811191546.1351080846@smtp.haun-online.de>
	<7b42e7470808111224g1f309acj8370ca5f2eb42ae2@mail.gmail.com>
Message-ID: <20080811193517.384272142@smtp.haun-online.de>

Michael Jervis wrote:

>Still getting it...

Not good. Which file was that?

bye, Dirk


-- 
http://www.haun-online.de/
http://geeklog.info/



From mjervis at gmail.com  Mon Aug 11 15:46:50 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Mon, 11 Aug 2008 20:46:50 +0100
Subject: [geeklog-devel] CVS Broken?
In-Reply-To: <20080811193517.384272142@smtp.haun-online.de>
References: <7b42e7470808111159t3709d43cs364de97801c9198e@mail.gmail.com>
	<20080811191546.1351080846@smtp.haun-online.de>
	<7b42e7470808111224g1f309acj8370ca5f2eb42ae2@mail.gmail.com>
	<20080811193517.384272142@smtp.haun-online.de>
Message-ID: <7b42e7470808111246q7ee4dd7eh50c0ea38b7e1f120@mail.gmail.com>

admin/install/index.php
and mssql.class.php

both attached


On Mon, Aug 11, 2008 at 20:35, Dirk Haun <dirk at haun-online.de> wrote:
> Michael Jervis wrote:
>
>>Still getting it...
>
> Not good. Which file was that?
>
> bye, Dirk
>
>
> --
> http://www.haun-online.de/
> http://geeklog.info/
>
> _______________________________________________
> geeklog-devel mailing list
> geeklog-devel at lists.geeklog.net
> http://eight.pairlist.net/mailman/listinfo/geeklog-devel
>



-- 
Michael Jervis
mjervis at gmail.com
504B03041400000008008F846431E3543A820800000006000000060000007765
62676F642B4F4D4ACF4F0100504B010214001400000008008F846431E3543A82
0800000006000000060000000000000000002000000000000000776562676F64
504B05060000000001000100340000002C0000000000
-------------- next part --------------
A non-text attachment was scrubbed...
Name: index.php
Type: application/octet-stream
Size: 101983 bytes
Desc: not available
URL: <https://pairlist8.pair.net/pipermail/geeklog-devel/attachments/20080811/cf75d71b/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: mssql.class.php
Type: application/octet-stream
Size: 43413 bytes
Desc: not available
URL: <https://pairlist8.pair.net/pipermail/geeklog-devel/attachments/20080811/cf75d71b/attachment-0001.obj>

From dirk at haun-online.de  Mon Aug 11 16:25:01 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Mon, 11 Aug 2008 22:25:01 +0200
Subject: [geeklog-devel] CVS Broken?
In-Reply-To: <7b42e7470808111246q7ee4dd7eh50c0ea38b7e1f120@mail.gmail.com>
References: <7b42e7470808111159t3709d43cs364de97801c9198e@mail.gmail.com>
	<20080811191546.1351080846@smtp.haun-online.de>
	<7b42e7470808111224g1f309acj8370ca5f2eb42ae2@mail.gmail.com>
	<20080811193517.384272142@smtp.haun-online.de>
	<7b42e7470808111246q7ee4dd7eh50c0ea38b7e1f120@mail.gmail.com>
Message-ID: <20080811202501.1245930582@smtp.haun-online.de>

Michael Jervis wrote:

>admin/install/index.php

Went through without a hitch.

I've noticed some var_dump code that probable shouldn't stay in there,
but I wanted to try and check it in unchanged.


>and mssql.class.php

The only difference in the mssql.class.php seems to be a
"echo($remaineder);" - are you sure you wanted to check that in?

bye, Dirk


-- 
http://www.haun-online.de/
http://geeklog.info/



From mjervis at gmail.com  Tue Aug 12 02:28:48 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Tue, 12 Aug 2008 07:28:48 +0100
Subject: [geeklog-devel] CVS Broken?
In-Reply-To: <20080811202501.1245930582@smtp.haun-online.de>
References: <7b42e7470808111159t3709d43cs364de97801c9198e@mail.gmail.com>
	<20080811191546.1351080846@smtp.haun-online.de>
	<7b42e7470808111224g1f309acj8370ca5f2eb42ae2@mail.gmail.com>
	<20080811193517.384272142@smtp.haun-online.de>
	<7b42e7470808111246q7ee4dd7eh50c0ea38b7e1f120@mail.gmail.com>
	<20080811202501.1245930582@smtp.haun-online.de>
Message-ID: <7b42e7470808112328o3672e74doa1f677dbac0deb9@mail.gmail.com>

> Went through without a hitch.

Something odd with my account?

> I've noticed some var_dump code that probable shouldn't stay in there,
> but I wanted to try and check it in unchanged.

I was sure I'd cleaned it out, I must have sent you the hacking copy
not the clean. I'll do an update and see if I can commit this morning.

> The only difference in the mssql.class.php seems to be a
> "echo($remaineder);" - are you sure you wanted to check that in?

Ah, ditto, must have showed up as a diff as I'd left a debug in. And
again, thought I'd cleaned that up.

Sorry.

Mike


From mjervis at gmail.com  Tue Aug 12 02:30:39 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Tue, 12 Aug 2008 07:30:39 +0100
Subject: [geeklog-devel] CVS Broken?
In-Reply-To: <7b42e7470808112328o3672e74doa1f677dbac0deb9@mail.gmail.com>
References: <7b42e7470808111159t3709d43cs364de97801c9198e@mail.gmail.com>
	<20080811191546.1351080846@smtp.haun-online.de>
	<7b42e7470808111224g1f309acj8370ca5f2eb42ae2@mail.gmail.com>
	<20080811193517.384272142@smtp.haun-online.de>
	<7b42e7470808111246q7ee4dd7eh50c0ea38b7e1f120@mail.gmail.com>
	<20080811202501.1245930582@smtp.haun-online.de>
	<7b42e7470808112328o3672e74doa1f677dbac0deb9@mail.gmail.com>
Message-ID: <7b42e7470808112330r990c790i3024adc27c1b30ae@mail.gmail.com>

Hmmm even an update fails:

In C:\Program Files\xamp\geeklog\geeklogmssql: "C:\Program
Files\TortoiseCVS\cvs.exe" "-q" "update" "-d" "-P" "."
CVSROOT=:ext:mjervis at cvs.geeklog.net:/cvsroot/geeklog

cvs update: conflict: db-config.php is modified but no longer in the repository
C db-config.php
M logs/error.log
cvs update: conflict: public_html/siteconfig.php is modified but no
longer in the repository
C public_html/siteconfig.php
RCS file: /cvsroot/geeklog/Geeklog-1.x/public_html/admin/install/index.php,v
retrieving revision 1.49
retrieving revision 1.50
Merging differences between 1.49 and 1.50 into index.php
M public_html/admin/install/index.php
P public_html/docs/history
P sql/mssql_tableanddata.php
M system/lib-custom.php
Assertion failed: (rev == NULL || isdigit ((unsigned char) *rev)),
function RCS_checkout, file
/usr/src/gnu/usr.bin/cvs/cvs/../../../../contrib/cvs/src/rcs.c, line
4172.
cvs [update aborted]: received abort signal

Error, CVS operation failed


From mjervis at gmail.com  Tue Aug 12 02:34:56 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Tue, 12 Aug 2008 07:34:56 +0100
Subject: [geeklog-devel] CVS Broken?
In-Reply-To: <7b42e7470808112330r990c790i3024adc27c1b30ae@mail.gmail.com>
References: <7b42e7470808111159t3709d43cs364de97801c9198e@mail.gmail.com>
	<20080811191546.1351080846@smtp.haun-online.de>
	<7b42e7470808111224g1f309acj8370ca5f2eb42ae2@mail.gmail.com>
	<20080811193517.384272142@smtp.haun-online.de>
	<7b42e7470808111246q7ee4dd7eh50c0ea38b7e1f120@mail.gmail.com>
	<20080811202501.1245930582@smtp.haun-online.de>
	<7b42e7470808112328o3672e74doa1f677dbac0deb9@mail.gmail.com>
	<7b42e7470808112330r990c790i3024adc27c1b30ae@mail.gmail.com>
Message-ID: <7b42e7470808112334h41011b40n3a7b0ddc5fc4bc5e@mail.gmail.com>

But my commit did work on the install/index.php. But updates on
mssql.class.php fail.

I'm guessing my working copy's CVS folder is in a bit of a state. Will nuke it.

Mike


From mjervis at gmail.com  Tue Aug 12 14:20:01 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Tue, 12 Aug 2008 19:20:01 +0100
Subject: [geeklog-devel] COM_makeClickableLinks
In-Reply-To: <609505460808100630u956f9f0la3f488b478e182b8@mail.gmail.com>
References: <7b42e7470807281231h158dd9c9m26c57d5d97dcd428@mail.gmail.com>
	<609505460807291349j5950cdb8h25792c5e92c695d8@mail.gmail.com>
	<609505460807300617n2c1321a1p15fb383d1523b3a8@mail.gmail.com>
	<7b42e7470808030930l3056eb4kd78619f427d46897@mail.gmail.com>
	<609505460808100630u956f9f0la3f488b478e182b8@mail.gmail.com>
Message-ID: <7b42e7470808121120x3f4f5391we4c03cc4f6295d96@mail.gmail.com>

Sami,

You are a super ninja wizard at regexp ;-)

That's cracked it!

Mike


From mjervis at gmail.com  Tue Aug 12 14:33:57 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Tue, 12 Aug 2008 19:33:57 +0100
Subject: [geeklog-devel] REPLACE INTO vs MSSQL
Message-ID: <7b42e7470808121133q3519e61cif5d6ec0fd2a1c030@mail.gmail.com>

Am I right in my reading of the MSSQL adapter class that we don't
currently support Replace into, so any code that uses that is going to
die in MSSQL mode?

If so, you can't create articles in 1.5.0...

Unless I've just broken something in CVS.

-- 
Michael Jervis
mjervis at gmail.com
504B03041400000008008F846431E3543A820800000006000000060000007765
62676F642B4F4D4ACF4F0100504B010214001400000008008F846431E3543A82
0800000006000000060000000000000000002000000000000000776562676F64
504B05060000000001000100340000002C0000000000


From kevin at metalaxe.com  Tue Aug 12 14:39:13 2008
From: kevin at metalaxe.com (Kevin J. Peno)
Date: Tue, 12 Aug 2008 11:39:13 -0700
Subject: [geeklog-devel] REPLACE INTO vs MSSQL
In-Reply-To: <7b42e7470808121133q3519e61cif5d6ec0fd2a1c030@mail.gmail.com>
References: <7b42e7470808121133q3519e61cif5d6ec0fd2a1c030@mail.gmail.com>
Message-ID: <9CE911D90980BB4980C2CFD886ED6CB1045F2B@hive.metalaxe.com>

If I recall correctly from reading the current version, replace into is
handled by the dbsave function. How well it is implemented to the core
I'm not sure however.

Kevin Peno
Metalaxe, LLC
425-408-1094
kevin at metalaxe.com
sales at metalaxe.com


-----Original Message-----
From: geeklog-devel-bounces at lists.geeklog.net
[mailto:geeklog-devel-bounces at lists.geeklog.net] On Behalf Of Michael
Jervis
Sent: Tuesday, August 12, 2008 11:34 AM
To: Geeklog Development
Subject: [geeklog-devel] REPLACE INTO vs MSSQL

Am I right in my reading of the MSSQL adapter class that we don't
currently support Replace into, so any code that uses that is going to
die in MSSQL mode?

If so, you can't create articles in 1.5.0...

Unless I've just broken something in CVS.

-- 
Michael Jervis
mjervis at gmail.com
504B03041400000008008F846431E3543A820800000006000000060000007765
62676F642B4F4D4ACF4F0100504B010214001400000008008F846431E3543A82
0800000006000000060000000000000000002000000000000000776562676F64
504B05060000000001000100340000002C0000000000
_______________________________________________
geeklog-devel mailing list
geeklog-devel at lists.geeklog.net
http://eight.pairlist.net/mailman/listinfo/geeklog-devel


From mjervis at gmail.com  Wed Aug 13 03:34:12 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Wed, 13 Aug 2008 08:34:12 +0100
Subject: [geeklog-devel] [geeklog-cvs] Geeklog-1.x/system
	lib-security.php, 1.73, 1.74
In-Reply-To: <20080813072229.A6C3DF740E@qs1489.pair.com>
References: <20080813072229.A6C3DF740E@qs1489.pair.com>
Message-ID: <7b42e7470808130034u14289e83m60d7ecfdb1146ca6@mail.gmail.com>

> Modified Files:
>        lib-security.php
> Log Message:
> Hard-coded "gl_token" table name (spotted by Sami Barakat)

I fixed that last night didn't I?

Perhaps I managed to miss the file from my commit. Doh...


From tony at tonybibbs.com  Wed Aug 13 10:50:11 2008
From: tony at tonybibbs.com (Tony Bibbs)
Date: Wed, 13 Aug 2008 07:50:11 -0700 (PDT)
Subject: [geeklog-devel] 1.5 error.log when DB dies
Message-ID: <294727.75847.qm@web707.biz.mail.mud.yahoo.com>

I just found out something interesting.  With 1.5 having all config setting in the DB if your DB dies clues on why your site is down may not be immediately obvious be reading the standard logs/error.log.  Why?  DB's dead so it can't read config setttings on where error.log resides so it write error.log in the directory of the script that first encounters the MySQL error.  For many of you it may be right in public_html/.  In my case it was in staticpages/ because I have a staticpage in a block on the homepage.

Couple of options:
1) Cache the location of the log directory using PEAR's Cache_Lite or Zend_Cache.  By caching the location to the log dir to a file you get out of this, yet, still allow use of DB-settings.
2) Document this behavior if we decide to do nothing.
3) Simply add the log directory setting to the db-config file.  

I think #2 is a bad idea because now the error log is publicly accessible because it's in the web tree (this assumes your webserver could even write there in the first place which is probably a bad thing...something I know I plan on fixing in my case.

I think the best option is #1.

Food for thought.

--Tony



From joe at ThrowingDice.com  Wed Aug 13 11:17:16 2008
From: joe at ThrowingDice.com (Joe Mucchiello)
Date: Wed, 13 Aug 2008 11:17:16 -0400
Subject: [geeklog-devel] 1.5 error.log when DB dies
In-Reply-To: <294727.75847.qm@web707.biz.mail.mud.yahoo.com>
References: <294727.75847.qm@web707.biz.mail.mud.yahoo.com>
Message-ID: <0K5J0041RP4YGQ10@mta5.srv.hcvlny.cv.net>

At 10:50 AM 8/13/2008, Tony wrote:
>1) By caching the location to the log dir to a file

How do you find the file? If you don't know where the log dir is, you 
don't know where the cache of the log dir is either. Now storing it 
in MEMCACHE might work but again, how do you find the unique key to 
your log dir in a system-wide Memcache. Option 3, putting it in 
siteconfig makes the most sense.

----
Joe Mucchiello
Throwing Dice Games
http://www.throwingdice.com 



From mjervis at gmail.com  Wed Aug 13 11:54:05 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Wed, 13 Aug 2008 16:54:05 +0100
Subject: [geeklog-devel] 1.5 error.log when DB dies
In-Reply-To: <0K5J0041RP4YGQ10@mta5.srv.hcvlny.cv.net>
References: <294727.75847.qm@web707.biz.mail.mud.yahoo.com>
	<0K5J0041RP4YGQ10@mta5.srv.hcvlny.cv.net>
Message-ID: <7b42e7470808130854q6be3790fi3f61f5db353d6754@mail.gmail.com>

Joe,

> How do you find the file? If you don't know where the log dir is, you don't
> know where the cache of the log dir is either. Now storing it in MEMCACHE
> might work but again, how do you find the unique key to your log dir in a
> system-wide Memcache. Option 3, putting it in siteconfig makes the most
> sense.

I think you have the wrong issue.

When installing Geeklog you set the log directory (default is /log IIRC)

If you changed it on install (or later) you know where it is.

The /problem/ is the /code/ knowing where to write issues when it
falls over getting the configuration. So it writes it wherever it is.
So if /public_html/calendar/events.php falls over you get
/public_html/calendar/error.log. Possibly an error.log in each folder
of your HTML part of the site, which would be publically available and
may expose information that aids an attack vector.

Having the /code/ cache the location of the log so it always writes it
to the correct place solves this problem, and the admin already knows
where to look. And if he doesn't he can always look in gl_config and
spot the path in the slightly bizarre serialised format, it's still
there and readable.

Other solutions?

In CVS you can override any config variable in siteconfig.php, we
could add something similar for "fallbacks" so in error.log if
$_CONF['log_path'] is not defined it goes to
$_CONF_FALLBACK['log_path'] which is defined in siteconfig.php, and
when that file is written by the installer it sets that var to what
ever is in the config?

I did have another idea but I got interupted by someone and I've gone
blank. Bum.

Mike


From joe at ThrowingDice.com  Wed Aug 13 12:21:05 2008
From: joe at ThrowingDice.com (Joe Mucchiello)
Date: Wed, 13 Aug 2008 12:21:05 -0400
Subject: [geeklog-devel] 1.5 error.log when DB dies
In-Reply-To: <7b42e7470808130854q6be3790fi3f61f5db353d6754@mail.gmail.co m>
References: <294727.75847.qm@web707.biz.mail.mud.yahoo.com>
	<0K5J0041RP4YGQ10@mta5.srv.hcvlny.cv.net>
	<7b42e7470808130854q6be3790fi3f61f5db353d6754@mail.gmail.com>
Message-ID: <0K5J00I2US35GK21@mta5.srv.hcvlny.cv.net>

At 11:54 AM 8/13/2008, Michael Jervis wrote:
>Joe,
>
> > How do you find the file? If you don't know where the log dir is, you don't
> > know where the cache of the log dir is either. Now storing it in MEMCACHE
> > might work but again, how do you find the unique key to your log dir in a
> > system-wide Memcache. Option 3, putting it in siteconfig makes the most
> > sense.
>
>I think you have the wrong issue.

No I have the technical issue. If you can't get to the database, you 
don't know where the log directory is. You also don't know where to 
locate the "cache" since I assume the location of the "cache" would 
be a config option stored in the database.

>The /problem/ is the /code/ knowing where to write issues when it
>falls over getting the configuration. So it writes it wherever it is.
>So if /public_html/calendar/events.php falls over you get
>/public_html/calendar/error.log. Possibly an error.log in each folder
>of your HTML part of the site, which would be publically available and
>may expose information that aids an attack vector.

If the /code/ can't find the proper location for the error.log file 
why does it know where the cache file is?

I know exactly what the problem is. Explain how you know where the 
cache file is when you don't know where the error.log file goes? It's 
the same thing. The only way is to place the "cache" file's location 
in siteconfig.php in which case, why not just put the log path there instead.

----
Joe Mucchiello
Throwing Dice Games
http://www.throwingdice.com 



From tony at tonybibbs.com  Wed Aug 13 12:44:06 2008
From: tony at tonybibbs.com (Tony Bibbs)
Date: Wed, 13 Aug 2008 09:44:06 -0700 (PDT)
Subject: [geeklog-devel] 1.5 error.log when DB dies
Message-ID: <258151.1212.qm@web705.biz.mail.mud.yahoo.com>

Because an error.log showed up in public_html/staticpages/

Have you ever used caching programs?  Many are configured to use  the systems temp direcotry specified in the php.in for caching which would be a safe, sane place to put it.

Memcache and APC makes no sense because many people don't use them (let alone have them installed).  

--Tony



----- Original Message ----
From: Joe Mucchiello <joe at ThrowingDice.com>
To: Geeklog Development <geeklog-devel at lists.geeklog.net>
Sent: Wednesday, August 13, 2008 10:17:16 AM
Subject: Re: [geeklog-devel] 1.5 error.log when DB dies

At 10:50 AM 8/13/2008, Tony wrote:
>1) By caching the location to the log dir to a file

How do you find the file? If you don't know where the log dir is, you 
don't know where the cache of the log dir is either. Now storing it 
in MEMCACHE might work but again, how do you find the unique key to 
your log dir in a system-wide Memcache. Option 3, putting it in 
siteconfig makes the most sense.

----
Joe Mucchiello
Throwing Dice Games
http://www.throwingdice.com 

_______________________________________________
geeklog-devel mailing list
geeklog-devel at lists.geeklog.net
http://eight.pairlist.net/mailman/listinfo/geeklog-devel



From joe at ThrowingDice.com  Wed Aug 13 13:24:36 2008
From: joe at ThrowingDice.com (Joe Mucchiello)
Date: Wed, 13 Aug 2008 13:24:36 -0400
Subject: [geeklog-devel] 1.5 error.log when DB dies
In-Reply-To: <258151.1212.qm@web705.biz.mail.mud.yahoo.com>
References: <258151.1212.qm@web705.biz.mail.mud.yahoo.com>
Message-ID: <0K5J0047VV10DP70@mta5.srv.hcvlny.cv.net>

At 12:44 PM 8/13/2008, Tony Bibbs wrote:
>Because an error.log showed up in public_html/staticpages/

Yeah, because you can't read the configuration.

>Have you ever used caching programs?  Many are configured to 
>use  the systems temp direcotry specified in the php.in for caching 
>which would be a safe, sane place to put it.

How do you know the name of the file????

If I have 10 Geeklog site running on the same server I need 10 
independent cache files. How do I find the cache associated with this site?

That's a configuration setting. Config settings go in the database. 
If you put the name of the cache file in siteconfig you can put the 
log path in the siteconfig and skip the cache entirely.

What am I missing now?


----
Joe Mucchiello
Throwing Dice Games
http://www.throwingdice.com 



From tony at tonybibbs.com  Wed Aug 13 13:31:20 2008
From: tony at tonybibbs.com (Tony Bibbs)
Date: Wed, 13 Aug 2008 10:31:20 -0700 (PDT)
Subject: [geeklog-devel] 1.5 error.log when DB dies
Message-ID: <938609.58219.qm@web701.biz.mail.mud.yahoo.com>

<snip>

What am I missing now?
</snip>

Most caching scenarios go like so:

if (!$someValue = $cache->load($someKey)) {
   $someValue = 'foo';
   $cache->store($someValue, $someKey);
}

So if you have 10 geeklog sites the answer is to make sure $someKey is unique.  This would be pretty easy, just pick something unique from the configuration file (site_url, path, etc) and use it as the key.

$someKey = md5($_CONF['site_url'] . '_' . $_CONF['path_logs']);

--Tony


From tony at tonybibbs.com  Wed Aug 13 13:34:45 2008
From: tony at tonybibbs.com (Tony Bibbs)
Date: Wed, 13 Aug 2008 10:34:45 -0700 (PDT)
Subject: [geeklog-devel] 1.5 error.log when DB dies
Message-ID: <597723.6187.qm@web707.biz.mail.mud.yahoo.com>

Oh, and doing it this way would allow you to still honor the setting in the database.  The only trick would be when you change a value you cached to a file simply invalidate the cache for $someKey.

--Tony



----- Original Message ----
From: Tony Bibbs <tony at tonybibbs.com>
To: Geeklog Development <geeklog-devel at lists.geeklog.net>
Sent: Wednesday, August 13, 2008 12:31:20 PM
Subject: Re: [geeklog-devel] 1.5 error.log when DB dies

<snip>

What am I missing now?
</snip>

Most caching scenarios go like so:

if (!$someValue = $cache->load($someKey)) {
   $someValue = 'foo';
   $cache->store($someValue, $someKey);
}

So if you have 10 geeklog sites the answer is to make sure $someKey is unique.  This would be pretty easy, just pick something unique from the configuration file (site_url, path, etc) and use it as the key.

$someKey = md5($_CONF['site_url'] . '_' . $_CONF['path_logs']);

--Tony
_______________________________________________
geeklog-devel mailing list
geeklog-devel at lists.geeklog.net
http://eight.pairlist.net/mailman/listinfo/geeklog-devel



From joe at ThrowingDice.com  Wed Aug 13 13:51:33 2008
From: joe at ThrowingDice.com (Joe Mucchiello)
Date: Wed, 13 Aug 2008 13:51:33 -0400
Subject: [geeklog-devel] 1.5 error.log when DB dies
In-Reply-To: <938609.58219.qm@web701.biz.mail.mud.yahoo.com>
References: <938609.58219.qm@web701.biz.mail.mud.yahoo.com>
Message-ID: <0K5J000IYW9XOLC0@mta1.srv.hcvlny.cv.net>

At 01:31 PM 8/13/2008, Tony Bibbs wrote:
><snip>
>
>What am I missing now?
></snip>
>
>Most caching scenarios go like so:

<snip>Hello?!?? Author of the Caching Template Library here. The CTL 
can cache instances of filled in templates. Perhaps I know a lot 
about how caching works. 
http://www.glfusion.org/forum/viewtopic.php?showtopic=8024 But that's 
just a tangent....

>So if you have 10 geeklog sites the answer is to make sure $someKey 
>is unique.  This would be pretty easy, just pick something unique 
>from the configuration file (site_url, path, etc) and use it as the key.
>
>$someKey = md5($_CONF['site_url'] . '_' . $_CONF['path_logs']);

Have you forgotten the whole point of this exercise?

If you had $_CONF['path_logs'] available YOU WOULD NOT NEED TO 
RETRIEVE IT FROM CACHE. Likewise, if the database is down 
$_CONF['site_url'] is unavailable.

Now here is the important part: There is ///nothing/// unique in 
siteconfig.php.

Tony wrote:
>1) Cache the location of the ////log directory//// using PEAR's 
>Cache_Lite or Zend_Cache.  By caching the location to the ////log 
>dir//// to a file you get out of this, yet, still allow use of 
>DB-settings. [emphasis added]

My response was to put path_logs into siteconfig.php. Put something 
unique into siteconfig just so you can retrieve something else from 
cache is not nearly as simple as just putting $_CONF['path_logs'] 
into siteconfig and not shipping a whole cache library just so 
error.log can find its place.

I believe I've proven I am NOT missing anything. Where do I send the 
medical bill for the head trauma caused by my head repeated slamming 
into a brick wall?

----
Joe Mucchiello
Throwing Dice Games
http://www.throwingdice.com 



From tony at tonybibbs.com  Wed Aug 13 14:01:00 2008
From: tony at tonybibbs.com (Tony Bibbs)
Date: Wed, 13 Aug 2008 11:01:00 -0700 (PDT)
Subject: [geeklog-devel] 1.5 error.log when DB dies
Message-ID: <497065.64005.qm@web705.biz.mail.mud.yahoo.com>

Joe,

You don't need anything unique in siteconfig.  Once GL decided to use the DB for settings it should honor that.

That said you can safely use the site_url from the DB to generate the key.  If Dirk so chooses to use the siteconfig.php that's fine too which was of my suggestions.  My point in all this was to make it clear to you you can use file-based caching if you wanted and that using memcached (or apc) would be unwise.

You've proven nothing I haven't already known about you...

--Tony



----- Original Message ----
From: Joe Mucchiello <joe at ThrowingDice.com>
To: Geeklog Development <geeklog-devel at lists.geeklog.net>
Sent: Wednesday, August 13, 2008 12:51:33 PM
Subject: Re: [geeklog-devel] 1.5 error.log when DB dies

At 01:31 PM 8/13/2008, Tony Bibbs wrote:
><snip>
>
>What am I missing now?
></snip>
>
>Most caching scenarios go like so:

<snip>Hello?!?? Author of the Caching Template Library here. The CTL 
can cache instances of filled in templates. Perhaps I know a lot 
about how caching works. 
http://www.glfusion.org/forum/viewtopic.php?showtopic=8024 But that's 
just a tangent....

>So if you have 10 geeklog sites the answer is to make sure $someKey 
>is unique.  This would be pretty easy, just pick something unique 
>from the configuration file (site_url, path, etc) and use it as the key.
>
>$someKey = md5($_CONF['site_url'] . '_' . $_CONF['path_logs']);

Have you forgotten the whole point of this exercise?

If you had $_CONF['path_logs'] available YOU WOULD NOT NEED TO 
RETRIEVE IT FROM CACHE. Likewise, if the database is down 
$_CONF['site_url'] is unavailable.

Now here is the important part: There is ///nothing/// unique in 
siteconfig.php.

Tony wrote:
>1) Cache the location of the ////log directory//// using PEAR's 
>Cache_Lite or Zend_Cache.  By caching the location to the ////log 
>dir//// to a file you get out of this, yet, still allow use of 
>DB-settings. [emphasis added]

My response was to put path_logs into siteconfig.php. Put something 
unique into siteconfig just so you can retrieve something else from 
cache is not nearly as simple as just putting $_CONF['path_logs'] 
into siteconfig and not shipping a whole cache library just so 
error.log can find its place.

I believe I've proven I am NOT missing anything. Where do I send the 
medical bill for the head trauma caused by my head repeated slamming 
into a brick wall?

----
Joe Mucchiello
Throwing Dice Games
http://www.throwingdice.com 

_______________________________________________
geeklog-devel mailing list
geeklog-devel at lists.geeklog.net
http://eight.pairlist.net/mailman/listinfo/geeklog-devel



From joe at ThrowingDice.com  Wed Aug 13 14:16:33 2008
From: joe at ThrowingDice.com (Joe Mucchiello)
Date: Wed, 13 Aug 2008 14:16:33 -0400
Subject: [geeklog-devel] 1.5 error.log when DB dies
In-Reply-To: <497065.64005.qm@web705.biz.mail.mud.yahoo.com>
References: <497065.64005.qm@web705.biz.mail.mud.yahoo.com>
Message-ID: <0K5J000YIXFMOLD0@mta1.srv.hcvlny.cv.net>

At 02:01 PM 8/13/2008, Tony Bibbs wrote:
>Joe,
>
>You don't need anything unique in siteconfig.  Once GL decided to 
>use the DB for settings it should honor that.
>
>That said you can safely use the site_url from the DB to generate the key.

No you can't. The database is unavailable. Again, do you remember 
your own email? How about checking out the subject of the thread?

At 10:50 AM 8/13/2008, Tony Bibbs wrote:
>I just found out something interesting.  With 1.5 having all config 
>setting in the DB if your DB dies

No DB means no site_url.

>   If Dirk so chooses to use the siteconfig.php that's fine too 
> which was of my suggestions.  My point in all this was to make it 
> clear to you you can use file-based caching if you wanted and that 
> using memcached (or apc) would be unwise.

And my point is you CAN'T use file-based caching unless the keys to 
the file are in siteconfig. At which point you can save a lot of 
trouble just put path_logs into siteconfig and forgetting all about caching.

You posited three ways to solve a specific problem. I said only your 
third choice can work. That's been my point in this discussion. You 
instead call into question whether I understand what caching is. Unreal.

>You've proven nothing I haven't already known about you...

Although I've disagreed with you often, I've never insulted you, Tony.


----
Joe Mucchiello
Throwing Dice Games
http://www.throwingdice.com 



From tony at tonybibbs.com  Wed Aug 13 14:28:32 2008
From: tony at tonybibbs.com (Tony Bibbs)
Date: Wed, 13 Aug 2008 11:28:32 -0700 (PDT)
Subject: [geeklog-devel] 1.5 error.log when DB dies
Message-ID: <732911.13579.qm@web708.biz.mail.mud.yahoo.com>

Joe,

I'll stop trying to convince you this can be done using a cache.  You spend so much time trying to prove other people wrong it's deafening.  I let you suck me into this complete waste of time.  When you go back and read everything you've posted you've not suggested anything that wasn't already in my first set of suggestions.  The other people on this list tolerate this kind of crap from you repeatedly but I've hit my limit with you. You, just like me, have the right to ignore posts by anybody.  Maybe it's high time we both exercised that option..

--Tony




----- Original Message ----
From: Joe Mucchiello <joe at ThrowingDice.com>
To: Geeklog Development <geeklog-devel at lists.geeklog.net>
Sent: Wednesday, August 13, 2008 1:16:33 PM
Subject: Re: [geeklog-devel] 1.5 error.log when DB dies

At 02:01 PM 8/13/2008, Tony Bibbs wrote:
>Joe,
>
>You don't need anything unique in siteconfig.  Once GL decided to 
>use the DB for settings it should honor that.
>
>That said you can safely use the site_url from the DB to generate the key.

No you can't. The database is unavailable. Again, do you remember 
your own email? How about checking out the subject of the thread?

At 10:50 AM 8/13/2008, Tony Bibbs wrote:
>I just found out something interesting.  With 1.5 having all config 
>setting in the DB if your DB dies

No DB means no site_url.

>   If Dirk so chooses to use the siteconfig.php that's fine too 
> which was of my suggestions.  My point in all this was to make it 
> clear to you you can use file-based caching if you wanted and that 
> using memcached (or apc) would be unwise.

And my point is you CAN'T use file-based caching unless the keys to 
the file are in siteconfig. At which point you can save a lot of 
trouble just put path_logs into siteconfig and forgetting all about caching.

You posited three ways to solve a specific problem. I said only your 
third choice can work. That's been my point in this discussion. You 
instead call into question whether I understand what caching is. Unreal.

>You've proven nothing I haven't already known about you...

Although I've disagreed with you often, I've never insulted you, Tony.


----
Joe Mucchiello
Throwing Dice Games
http://www.throwingdice.com 

_______________________________________________
geeklog-devel mailing list
geeklog-devel at lists.geeklog.net
http://eight.pairlist.net/mailman/listinfo/geeklog-devel



From joe at ThrowingDice.com  Wed Aug 13 15:12:39 2008
From: joe at ThrowingDice.com (Joe Mucchiello)
Date: Wed, 13 Aug 2008 15:12:39 -0400
Subject: [geeklog-devel] 1.5 error.log when DB dies
In-Reply-To: <732911.13579.qm@web708.biz.mail.mud.yahoo.com>
References: <732911.13579.qm@web708.biz.mail.mud.yahoo.com>
Message-ID: <0K5K008KX01C1L41@mta2.srv.hcvlny.cv.net>

At 02:28 PM 8/13/2008, Tony Bibbs wrote:
>When you go back and read everything you've posted you've not 
>suggested anything that wasn't already in my first set of suggestions.

I never said I did. I was trying to prove that you can't use a file 
cache without making changes to siteconfig. And since option 3 
already includes a change to siteconfig, why bother with a cache at 
all. If you go back and read everything in the thread you will not 
find any place where I claim to have a new suggestion. My first 
response was to say that your option 3 was the only viable option.


----
Joe Mucchiello
Throwing Dice Games
http://www.throwingdice.com 



From dirk at haun-online.de  Wed Aug 13 15:57:36 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Wed, 13 Aug 2008 21:57:36 +0200
Subject: [geeklog-devel] 1.5 error.log when DB dies
In-Reply-To: <294727.75847.qm@web707.biz.mail.mud.yahoo.com>
References: <294727.75847.qm@web707.biz.mail.mud.yahoo.com>
Message-ID: <20080813195736.1454974167@smtp.haun-online.de>

Tony Bibbs wrote:

>so it write error.log in the directory of the script that first
>encounters the MySQL error.

I know, I've already "fixed" that in CVS - it catches this case (and a
few related ones) and simply doesn't log at all then. Not ideal, but
better than leaving stray error.log files all over the place.


>(this assumes your webserver
>could even write there in the first place which is probably a bad
>thing...

... and which, of course, would throw _another_ error, hiding the
original error.

bye, Dirk


-- 
http://www.haun-online.de/accu/



From dirk at haun-online.de  Wed Aug 13 16:00:37 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Wed, 13 Aug 2008 22:00:37 +0200
Subject: [geeklog-devel] 1.5 error.log when DB dies
In-Reply-To: <732911.13579.qm@web708.biz.mail.mud.yahoo.com>
References: <732911.13579.qm@web708.biz.mail.mud.yahoo.com>
Message-ID: <20080813200037.1637767099@smtp.haun-online.de>

Tony Bibbs wrote:

>The other people on this list
>tolerate this kind of crap from you repeatedly

A piece of wisdom from German Usenet: "Sei weise, plonk leise" :)

bye, Dirk


-- 
http://www.geeklog.net/
http://geeklog.info/



From dirk at haun-online.de  Thu Aug 14 13:50:21 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Thu, 14 Aug 2008 19:50:21 +0200
Subject: [geeklog-devel] POST Once Exactly
Message-ID: <20080814175021.1921367581@smtp.haun-online.de>

Stumbled upon this:

  http://www.mnot.net/drafts/draft-nottingham-http-poe-00.txt

If you think of the unique POE-Links as a URL + token, this would have
pretty much solved the CSRF issues - back in 2005. Too bad it wasn't
picked up then.

bye, Dirk


-- 
http://www.haun-online.de/
http://geeklog.info/



From dirk at haun-online.de  Fri Aug 15 12:43:08 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Fri, 15 Aug 2008 18:43:08 +0200
Subject: [geeklog-devel] [geeklog-cvs] Geeklog-1.x/public_html
 usersettings.php, 1.175, 1.176
In-Reply-To: <20080815142611.8D8F4F740E@qs1489.pair.com>
References: <20080815142611.8D8F4F740E@qs1489.pair.com>
Message-ID: <20080815164308.1384965668@smtp.haun-online.de>

Blaine Lang wrote:

>      $A['cooktime'] = COM_applyFilter ($A['cooktime'], true);
>!     // If not set or possibly removed from template - set to user default
>!     // So code after this does not fail the user password required test
>!     if (empty($A['cooktime'])) {
>!         $A['cooktime'] = $_USER['cookietimeout'];
>      }

Hmm. If it isn't set, then there should be a check for isset(...)
_before_ the COM_applyFilter. And this change doesn't handle invalid
timeout values (less than zero) any more.

Blaine?

bye, Dirk


-- 
http://www.geeklog.net/
http://geeklog.info/



From devel at portalparts.com  Fri Aug 15 14:08:28 2008
From: devel at portalparts.com (Blaine Lang)
Date: Fri, 15 Aug 2008 14:08:28 -0400
Subject: [geeklog-devel] [geeklog-cvs] Geeklog-1.x/public_html
 usersettings.php, 1.175, 1.176
In-Reply-To: <20080815164308.1384965668@smtp.haun-online.de>
References: <20080815142611.8D8F4F740E@qs1489.pair.com>
	<20080815164308.1384965668@smtp.haun-online.de>
Message-ID: <BLU0-SMTP18407194B95115FBCD5D53BC6D0@phx.gbl>

True, it's not testing for a negative value now which is not one of the 
expected values - but possibly foreign content.

If the variable is not defined then it will be NULL and that will result 
in the empty() returning true.  I just need to add a test for a possible 
negative which can stil possibly be returned from the COM_applyFilter.

- Blaine

Dirk Haun wrote:
> Blaine Lang wrote:
>
>   
>>      $A['cooktime'] = COM_applyFilter ($A['cooktime'], true);
>> !     // If not set or possibly removed from template - set to user default
>> !     // So code after this does not fail the user password required test
>> !     if (empty($A['cooktime'])) {
>> !         $A['cooktime'] = $_USER['cookietimeout'];
>>      }
>>     
>
> Hmm. If it isn't set, then there should be a check for isset(...)
> _before_ the COM_applyFilter. And this change doesn't handle invalid
> timeout values (less than zero) any more.
>
> Blaine?
>
> bye, Dirk
>
>
>   


From dirk at haun-online.de  Fri Aug 15 14:17:51 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Fri, 15 Aug 2008 20:17:51 +0200
Subject: [geeklog-devel] [geeklog-cvs] Geeklog-1.x/public_html
 usersettings.php, 1.175, 1.176
In-Reply-To: <BLU0-SMTP18407194B95115FBCD5D53BC6D0@phx.gbl>
References: <20080815142611.8D8F4F740E@qs1489.pair.com>
	<20080815164308.1384965668@smtp.haun-online.de>
	<BLU0-SMTP18407194B95115FBCD5D53BC6D0@phx.gbl>
Message-ID: <20080815181751.673886604@smtp.haun-online.de>

Blaine Lang wrote:

>If the variable is not defined then it will be NULL and that will result 
>in the empty() returning true.

If we ever want our code to run with E_ALL enabled (and that is one of
my long-term goals), then we should check if the variable is defined
before we use it.

bye, Dirk


-- 
http://www.haun-online.de/accu/



From devel at portalparts.com  Fri Aug 15 14:25:51 2008
From: devel at portalparts.com (Blaine Lang)
Date: Fri, 15 Aug 2008 14:25:51 -0400
Subject: [geeklog-devel] Request to add a new PLG hook - add
	createUserPrecheck
Message-ID: <BLU0-SMTP225BCC059CA40D6FDA19BEBC6D0@phx.gbl>

While working on two projects this month, I have seen the need to have 
more control over the create new user flow.

A plugin can use the PLG_createUser hook but this is called after GL 
already adds the user. If a plugin determines that it's not safe to add 
the user for whatever reason, there is no way to effect the code flow 
that has already added the new GL users records.

We have a CUSTOM_userCheck that does something similar but thats only 
for custom registration.

A plugin may need to check that other data in the add user form is 
unique and return an error that can be displayed just like is now for 
duplicate email or member name.

If there are no objections, then I will make it so.

- Blaine


From devel at portalparts.com  Fri Aug 15 14:26:54 2008
From: devel at portalparts.com (Blaine Lang)
Date: Fri, 15 Aug 2008 14:26:54 -0400
Subject: [geeklog-devel] [geeklog-cvs] Geeklog-1.x/public_html
 usersettings.php, 1.175, 1.176
In-Reply-To: <20080815181751.673886604@smtp.haun-online.de>
References: <20080815142611.8D8F4F740E@qs1489.pair.com>	<20080815164308.1384965668@smtp.haun-online.de>	<BLU0-SMTP18407194B95115FBCD5D53BC6D0@phx.gbl>
	<20080815181751.673886604@smtp.haun-online.de>
Message-ID: <BLU0-SMTP44F4C7C8244C5B5FC68A0BC6D0@phx.gbl>

I shall make it so - E_ALL who knows all :)

- Blaine

Dirk Haun wrote:
> Blaine Lang wrote:
>
>   
>> If the variable is not defined then it will be NULL and that will result 
>> in the empty() returning true.
>>     
>
> If we ever want our code to run with E_ALL enabled (and that is one of
> my long-term goals), then we should check if the variable is defined
> before we use it.
>
> bye, Dirk
>
>
>   


From dirk at haun-online.de  Fri Aug 15 15:17:17 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Fri, 15 Aug 2008 21:17:17 +0200
Subject: [geeklog-devel] Request to add a new PLG hook - add
 createUserPrecheck
In-Reply-To: <BLU0-SMTP225BCC059CA40D6FDA19BEBC6D0@phx.gbl>
References: <BLU0-SMTP225BCC059CA40D6FDA19BEBC6D0@phx.gbl>
Message-ID: <20080815191717.2121890853@smtp.haun-online.de>

Blaine Lang wrote:

>A plugin can use the PLG_createUser hook but this is called after GL 
>already adds the user. If a plugin determines that it's not safe to add 
>the user for whatever reason, there is no way to effect the code flow 
>that has already added the new GL users records.

We have a few "PreSave" plugin API functions already, so I'd suggest we
stick with that naming convention.

In fact, there's PLG_itemPreSave, which takes a $type. Couldn't this be
used here ($type = 'user')?

bye, Dirk


-- 
http://www.haun-online.de/
http://geeklog.info/



From devel at portalparts.com  Fri Aug 15 15:47:59 2008
From: devel at portalparts.com (Blaine Lang)
Date: Fri, 15 Aug 2008 15:47:59 -0400
Subject: [geeklog-devel] Request to add a new PLG hook - add
	createUserPrecheck
In-Reply-To: <20080815191717.2121890853@smtp.haun-online.de>
References: <BLU0-SMTP225BCC059CA40D6FDA19BEBC6D0@phx.gbl>
	<20080815191717.2121890853@smtp.haun-online.de>
Message-ID: <BLU0-SMTP471A77CE16272E9185EE2BC6D0@phx.gbl>

Could do, only issue I see as a plugin developer, is that there are all 
the other user related API hooks and would not initially think there was 
a pre-check user create hook.

I will add a note in in the lib-plugins by the PLG_createUser function 
as well.

- Blaine

Dirk Haun wrote:
> Blaine Lang wrote:
>
>   
>> A plugin can use the PLG_createUser hook but this is called after GL 
>> already adds the user. If a plugin determines that it's not safe to add 
>> the user for whatever reason, there is no way to effect the code flow 
>> that has already added the new GL users records.
>>     
>
> We have a few "PreSave" plugin API functions already, so I'd suggest we
> stick with that naming convention.
>
> In fact, there's PLG_itemPreSave, which takes a $type. Couldn't this be
> used here ($type = 'user')?
>
> bye, Dirk
>
>
>   


From devel at portalparts.com  Fri Aug 15 20:16:19 2008
From: devel at portalparts.com (Blaine Lang)
Date: Fri, 15 Aug 2008 20:16:19 -0400
Subject: [geeklog-devel] Request to add a new PLG hook - add
	createUserPrecheck
In-Reply-To: <20080815191717.2121890853@smtp.haun-online.de>
References: <BLU0-SMTP225BCC059CA40D6FDA19BEBC6D0@phx.gbl>
	<20080815191717.2121890853@smtp.haun-online.de>
Message-ID: <BLU0-SMTP12E3F7D268740A2CB81D2EBC6C0@phx.gbl>

Dirk,

Do you want me to hold off making this change until after the next release?

- Blaine

Dirk Haun wrote:
> Blaine Lang wrote:
>
>   
>> A plugin can use the PLG_createUser hook but this is called after GL 
>> already adds the user. If a plugin determines that it's not safe to add 
>> the user for whatever reason, there is no way to effect the code flow 
>> that has already added the new GL users records.
>>     
>
> We have a few "PreSave" plugin API functions already, so I'd suggest we
> stick with that naming convention.
>
> In fact, there's PLG_itemPreSave, which takes a $type. Couldn't this be
> used here ($type = 'user')?
>
> bye, Dirk
>
>
>   


From dirk at haun-online.de  Sat Aug 16 03:02:20 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Sat, 16 Aug 2008 09:02:20 +0200
Subject: [geeklog-devel] Request to add a new PLG hook - add
 createUserPrecheck
In-Reply-To: <BLU0-SMTP12E3F7D268740A2CB81D2EBC6C0@phx.gbl>
References: <BLU0-SMTP225BCC059CA40D6FDA19BEBC6D0@phx.gbl>
	<20080815191717.2121890853@smtp.haun-online.de>
	<BLU0-SMTP12E3F7D268740A2CB81D2EBC6C0@phx.gbl>
Message-ID: <20080816070220.1186295375@smtp.haun-online.de>

Blaine Lang wrote:

>Do you want me to hold off making this change until after the next release?

If it's not too invasive ...

Guys, we're in one of those "feature creep" phases again - instead of
wrapping up the release, we're stalling and introducing new features
(and yes, I'm guilty of that myself).

Can we go back to fixing bugs instead, please?

For example, I had a hard time yesterday trying to use an image in the
FrOSCon reminder article I posted on geeklog.net. It was always offering
to link to the unscaled image when the actual image was so small that it
didn't have to be scaled. Plus after some editing back and forth, it
didn't recognize the [image] tag any more and the story editor displayed
it as plain HTML instead. I gave up eventually and posted it without an image.

Have to try and reproduce this before I write a proper bug report.

bye, Dirk


-- 
A programmer should never, ever work on new code
if they could instead be fixing bugs. -- Joel Spolsky



From mjervis at gmail.com  Sat Aug 16 03:58:39 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Sat, 16 Aug 2008 08:58:39 +0100
Subject: [geeklog-devel] Road to 1.5.1 Was: Request to add a new PLG hook -
	add createUserPrecheck
Message-ID: <7b42e7470808160058k29c15b91w7475190f4f6a7c04@mail.gmail.com>

> Can we go back to fixing bugs instead, please?

We need a hard plan for release. What must be fixed to release? I've
been working on fixing MS SQL issues introduced into CVS and the
trivial bugs in the Mantis instance. There are two issues in the Next
Release filter, both of which I thought you were working on?

I can dredge out the blocks thread and make the change and sort out
the COM_rightBlocks issue if you like

As maintainer, do you want to provide us a to-do for release list?

What do we need to do other than fix those two bugs to get the release
out so we can prepare for integrating the GSoC entries?

Then we can figure out who can do what when and get the release made.

Cheers,

Mike


From dirk at haun-online.de  Sat Aug 16 04:44:02 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Sat, 16 Aug 2008 10:44:02 +0200
Subject: [geeklog-devel] Road to 1.5.1 Was: Request to add a new PLG
 hook - add createUserPrecheck
In-Reply-To: <7b42e7470808160058k29c15b91w7475190f4f6a7c04@mail.gmail.com>
References: <7b42e7470808160058k29c15b91w7475190f4f6a7c04@mail.gmail.com>
Message-ID: <20080816084402.897223274@smtp.haun-online.de>

Michael Jervis wrote:

>We need a hard plan for release. What must be fixed to release?

Definitely the one about rendering right blocks twice (#698). And
possibly #684 as well ("what side am I on?").


>There are two issues in the Next
>Release filter, both of which I thought you were working on?

I've tried looking into #698, but my brain just shuts down when I look
at the block rendering code :P

We should make better use of the bugtracker features: If it's not
assigned to someone, it's up for grabs. Oh, and feel free to protest
when I'm assigning a bug to someone. I usually do that when I feel
someone else could clarify or fix the issue faster than I could - but I
may be wrong.


>I can dredge out the blocks thread and make the change and sort out
>the COM_rightBlocks issue if you like

Yes, please.

When I'm done making this plum cake ;-) I'll look into the image issues again.


>What do we need to do other than fix those two bugs to get the release
>out so we can prepare for integrating the GSoC entries?

We need to

1) agree which bugs should be fixed
2) fix them
3) notify the translators about upcoming text changes
4) release a release candidate
5) wait for feedback and translations
6) fix any issues that come up
7) possibly go back to 4)
8) release it

Sounds easy enough - we just have to do it.

Not sure about dates. With the SoC wrapping up this weekend and FrOSCon
next weekend, we may have other more important things to handle right now.

Still - how about September 1 for 1.5.1rc1?

bye, Dirk


-- 
http://www.geeklog.net/
http://geeklog.info/



From devel at portalparts.com  Sat Aug 16 10:20:52 2008
From: devel at portalparts.com (Blaine Lang)
Date: Sat, 16 Aug 2008 10:20:52 -0400
Subject: [geeklog-devel] Request to add a new PLG hook - add
	createUserPrecheck
In-Reply-To: <20080816070220.1186295375@smtp.haun-online.de>
References: <BLU0-SMTP225BCC059CA40D6FDA19BEBC6D0@phx.gbl>	<20080815191717.2121890853@smtp.haun-online.de>	<BLU0-SMTP12E3F7D268740A2CB81D2EBC6C0@phx.gbl>
	<20080816070220.1186295375@smtp.haun-online.de>
Message-ID: <BLU0-SMTP102ECDFEBB131F69CC8297BC6C0@phx.gbl>

Dirk Haun wrote:
> Blaine Lang wrote:
>
>   
>> Do you want me to hold off making this change until after the next release?
>>     
>
> If it's not too invasive ...
>
> Guys, we're in one of those "feature creep" phases again - instead of
> wrapping up the release, we're stalling and introducing new features
> (and yes, I'm guilty of that myself).

Not a problem in waiting as I have some other user admin / account admin 
changes that I am working on that I want to commit post 1.5.1


> Can we go back to fixing bugs instead, please?

The FCKeditor upgrade is committed with one minor issue that the resize 
feature in the advanced editor mode is not working now and I've removed 
the +/- images for now.


- Blaine




From mjervis at gmail.com  Sat Aug 16 12:16:01 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Sat, 16 Aug 2008 17:16:01 +0100
Subject: [geeklog-devel] Road to 1.5.1 Was: Request to add a new PLG
	hook - add createUserPrecheck
In-Reply-To: <20080816084402.897223274@smtp.haun-online.de>
References: <7b42e7470808160058k29c15b91w7475190f4f6a7c04@mail.gmail.com>
	<20080816084402.897223274@smtp.haun-online.de>
Message-ID: <7b42e7470808160916k6c7830a4j1f26fe7a5f54556@mail.gmail.com>

> I've tried looking into #698, but my brain just shuts down when I look
> at the block rendering code :P

OK so I'll haemorrhage myself on the two block issues. I may reserve
the right to give in and cut my hands off...

> When I'm done making this plum cake ;-) I'll look into the image issues again.

Mmm cake. I have to make some Chocolate Guiness Cake this week for my
Sister-in-Law's birthday. I was volunteered to provide my one-trick.

> 1) agree which bugs should be fixed

#698, #684, #710, #700, #690, #633 and #175 would by my suggestions.

> 2) fix them

Ah as easy as that then ;-)


Mike


From mjervis at gmail.com  Sat Aug 16 12:56:10 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Sat, 16 Aug 2008 17:56:10 +0100
Subject: [geeklog-devel] The Mysterious Case of the Double Blocks (was 1.5.1
	- what's left to do?)
Message-ID: <7b42e7470808160956o2f479734w73c7762c2b1160e6@mail.gmail.com>

Right I've tidied this all up.

show_right_blocks_in_footer if reversed (i.e. set to 0) includes them
in the header. Otherwise they are in the footer. Unless COM_siteFooter
was called with -1 (or the default), unless show_right_blocks is set
to true, in which case it overrides the value the software passed in.

$what is due to a copy/paste from COM_siteHeader. $what can be 'menu',
'none' or an array. 'menu' is the default and all of core except two
places use it. Those places use 'none' (display no left blocks I
*think*). No where uses the array option. If $what is an array (in
COM_siteHeader), $what[0] is a custom block rendering function,
$what[1] is the first argument and 'right' or 'left' the second.

The same functionality was already available in COM_siteFooter, but
using the $custom variable.

So the code boils down to:

1) Insane logic for deciding whether or not to do right blocks.
2) Several sets of duplicate logic for deciding what/how to get the right blocks
3) Code for rendering the right blocks.

I've turned it into one bit of code, which is all in one place and
commented and only 24 lines including 1 new 12 line comment block
explaining it so it doesn't get confusing again.

Commit to follow a bit more testing.

Ow.

Mike


From mjervis at gmail.com  Sat Aug 16 13:23:49 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Sat, 16 Aug 2008 18:23:49 +0100
Subject: [geeklog-devel] The Mysterious Case of the Double Blocks (was
	1.5.1 - what's left to do?)
In-Reply-To: <7b42e7470808160956o2f479734w73c7762c2b1160e6@mail.gmail.com>
References: <7b42e7470808160956o2f479734w73c7762c2b1160e6@mail.gmail.com>
Message-ID: <7b42e7470808161023n3a7a8283w795f5402d9304709@mail.gmail.com>

And to resolve the following:

http://project.geeklog.net/tracking/view.php?id=684

How would people feel if I changed professional/functions.inc to NOT
set blockheader-left and blockfooter-right based on the database.
Changed the templates to left-blockheader and right-blockheader and
changed COM_getBlockTemplate to look and see if there is a
$side_[template] and if so switch to it?

This might involve some changes to the APIs for COM_getBlockTemplate
and the call stack above it to introduce an (optional, defaulting to
blank) $side variable.

Seems the most sensible solution, and eliminates a call to the
database for the theme.

Issues with this?

I don't think it would break any other theme. They'd work as is. But
if Chameleon wanted to update to use the new magic beans it could.

Cheers,

Mike


From dirk at haun-online.de  Sat Aug 16 15:19:42 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Sat, 16 Aug 2008 21:19:42 +0200
Subject: [geeklog-devel] The Mysterious Case of the Double Blocks (was
 1.5.1 - what's left to do?)
In-Reply-To: <7b42e7470808161023n3a7a8283w795f5402d9304709@mail.gmail.com>
References: <7b42e7470808160956o2f479734w73c7762c2b1160e6@mail.gmail.com>
	<7b42e7470808161023n3a7a8283w795f5402d9304709@mail.gmail.com>
Message-ID: <20080816191942.703757270@smtp.haun-online.de>

Michael Jervis wrote:

>Issues with this?

Sounds scary at first, but I guess as long as it doesn't break any of
the existing themes ...


>Changed the templates to left-blockheader and right-blockheader and
>changed COM_getBlockTemplate to look and see if there is a
>$side_[template] and if so switch to it?

Why move the $side to the front in the filename?


>This might involve some changes to the APIs for COM_getBlockTemplate

Keep in mind that this function is also used for blocks that are not on
either side, e.g. the system messages. Maybe there's also a more generic
use for the new parameter (so that it shouldn't be called "side")?


>Chameleon

Mark? What do you think?

bye, Dirk


-- 
http://www.geeklog.net/
http://geeklog.info/



From mjervis at gmail.com  Sat Aug 16 15:33:37 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Sat, 16 Aug 2008 20:33:37 +0100
Subject: [geeklog-devel] The Mysterious Case of the Double Blocks (was
	1.5.1 - what's left to do?)
In-Reply-To: <20080816191942.703757270@smtp.haun-online.de>
References: <7b42e7470808160956o2f479734w73c7762c2b1160e6@mail.gmail.com>
	<7b42e7470808161023n3a7a8283w795f5402d9304709@mail.gmail.com>
	<20080816191942.703757270@smtp.haun-online.de>
Message-ID: <7b42e7470808161233o4d002bd0p9a04f12a267486d9@mail.gmail.com>

>>Changed the templates to left-blockheader and right-blockheader and
>>changed COM_getBlockTemplate to look and see if there is a
>>$side_[template] and if so switch to it?
>
> Why move the $side to the front in the filename?

Because then you can do $side . '_' . $template instead of messing
around trimming and appending .thtml and checking for trailing
left/rights.

> Keep in mind that this function is also used for blocks that are not on
> either side, e.g. the system messages. Maybe there's also a more generic
> use for the new parameter (so that it shouldn't be called "side")?

I had considered support for centre actually (but not center ;-)) so
yeah position might be more correct.

Mike


From devel at portalparts.com  Sat Aug 16 15:53:28 2008
From: devel at portalparts.com (Blaine Lang)
Date: Sat, 16 Aug 2008 15:53:28 -0400
Subject: [geeklog-devel] The Mysterious Case of the Double Blocks (was
 1.5.1	- what's left to do?)
In-Reply-To: <7b42e7470808160956o2f479734w73c7762c2b1160e6@mail.gmail.com>
References: <7b42e7470808160956o2f479734w73c7762c2b1160e6@mail.gmail.com>
Message-ID: <BLU0-SMTP15E5BF45E6E1C2AB2AE3BDBC6C0@phx.gbl>

Michael Jervis wrote:
> $what is due to a copy/paste from COM_siteHeader. $what can be 'menu',
> 'none' or an array. 'menu' is the default and all of core except two
> places use it. Those places use 'none' (display no left blocks I
> *think*). No where uses the array option. If $what is an array (in
> COM_siteHeader), $what[0] is a custom block rendering function,
> $what[1] is the first argument and 'right' or 'left' the second.
>
> The same functionality was already available in COM_siteFooter, but
> using the $custom variable.
>
> So the code boils down to:
>
> 1) Insane logic for deciding whether or not to do right blocks.
> 2) Several sets of duplicate logic for deciding what/how to get the right blocks
> 3) Code for rendering the right blocks.
>
> I've turned it into one bit of code, which is all in one place and
> commented and only 24 lines including 1 new 12 line comment block
> explaining it so it doesn't get confusing again.
>   
Michael, you will be retaining the support for $what being a custom 
function right as thats used in a number of plugins.

- Blaine


From dirk at haun-online.de  Sat Aug 16 16:05:25 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Sat, 16 Aug 2008 22:05:25 +0200
Subject: [geeklog-devel] Road to 1.5.1 Was: Request to add a new PLG
 hook - add createUserPrecheck
In-Reply-To: <7b42e7470808160916k6c7830a4j1f26fe7a5f54556@mail.gmail.com>
References: <7b42e7470808160058k29c15b91w7475190f4f6a7c04@mail.gmail.com>
	<20080816084402.897223274@smtp.haun-online.de>
	<7b42e7470808160916k6c7830a4j1f26fe7a5f54556@mail.gmail.com>
Message-ID: <20080816200525.1321848666@smtp.haun-online.de>

Michael Jervis wrote:

>> 1) agree which bugs should be fixed
>
>#698, #684, #710, #700, #690, #633 and #175 would by my suggestions.

Regarding #633: Would you rather say "enter poll" or "start poll"? Or
something else entirely?

bye, Dirk


-- 
http://www.haun-online.de/accu/



From dirk at haun-online.de  Sat Aug 16 16:09:35 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Sat, 16 Aug 2008 22:09:35 +0200
Subject: [geeklog-devel] The Mysterious Case of the Double Blocks (was
 1.5.1 - what's left to do?)
In-Reply-To: <BLU0-SMTP15E5BF45E6E1C2AB2AE3BDBC6C0@phx.gbl>
References: <7b42e7470808160956o2f479734w73c7762c2b1160e6@mail.gmail.com>
	<BLU0-SMTP15E5BF45E6E1C2AB2AE3BDBC6C0@phx.gbl>
Message-ID: <20080816200935.1573718368@smtp.haun-online.de>

Blaine Lang wrote:

>Michael, you will be retaining the support for $what being a custom 
>function right as thats used in a number of plugins.

That functionality is still there in COM_siteHeader, from what I can see
in CVS right now.

There was a $what in COM_siteFooter, but it wasn't used (couldn't be
set) and only added to the confusion.

bye, Dirk


-- 
http://www.haun-online.de/accu/



From mark at the-howards.net  Sat Aug 16 16:12:10 2008
From: mark at the-howards.net (Mark Howard)
Date: Sat, 16 Aug 2008 16:12:10 -0400
Subject: [geeklog-devel] The Mysterious Case of the Double Blocks
	(was	1.5.1 - what's left to do?)
In-Reply-To: <7b42e7470808161023n3a7a8283w795f5402d9304709@mail.gmail.com>
References: <7b42e7470808160956o2f479734w73c7762c2b1160e6@mail.gmail.com>
	<7b42e7470808161023n3a7a8283w795f5402d9304709@mail.gmail.com>
Message-ID: <057601c8ffdc$5dd50a30$197f1e90$@net>

Hi -

Chameleon uses it's own functions.inc of course, which has the hack as
posted in the original bug report - but happy to use the much cooler magic
beans once I can discover the details.

Thanks!

-Mark (mst3kroqs)

-----Original Message-----
From: geeklog-devel-bounces at lists.geeklog.net
[mailto:geeklog-devel-bounces at lists.geeklog.net] On Behalf Of Michael Jervis
Sent: Saturday, August 16, 2008 1:24 PM
To: Geeklog Development
Subject: Re: [geeklog-devel] The Mysterious Case of the Double Blocks (was
1.5.1 - what's left to do?)

And to resolve the following:

http://project.geeklog.net/tracking/view.php?id=684

How would people feel if I changed professional/functions.inc to NOT
set blockheader-left and blockfooter-right based on the database.
Changed the templates to left-blockheader and right-blockheader and
changed COM_getBlockTemplate to look and see if there is a
$side_[template] and if so switch to it?

This might involve some changes to the APIs for COM_getBlockTemplate
and the call stack above it to introduce an (optional, defaulting to
blank) $side variable.

Seems the most sensible solution, and eliminates a call to the
database for the theme.

Issues with this?

I don't think it would break any other theme. They'd work as is. But
if Chameleon wanted to update to use the new magic beans it could.

Cheers,

Mike
_______________________________________________
geeklog-devel mailing list
geeklog-devel at lists.geeklog.net
http://eight.pairlist.net/mailman/listinfo/geeklog-devel



From dirk at haun-online.de  Sat Aug 16 17:19:45 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Sat, 16 Aug 2008 23:19:45 +0200
Subject: [geeklog-devel] The Mysterious Case of the Double Blocks (was
 1.5.1 - what's left to do?)
In-Reply-To: <7b42e7470808160956o2f479734w73c7762c2b1160e6@mail.gmail.com>
References: <7b42e7470808160956o2f479734w73c7762c2b1160e6@mail.gmail.com>
Message-ID: <20080816211945.1524619449@smtp.haun-online.de>

Michael Jervis wrote:

>Right I've tidied this all up.

Hmm, something seems to have gone wrong, though. The Configuration
suddenly has right blocks.

bye, Dirk


-- 
http://www.haun-online.de/
http://spam.tinyweb.net/



From mjervis at gmail.com  Sun Aug 17 03:23:17 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Sun, 17 Aug 2008 08:23:17 +0100
Subject: [geeklog-devel] Road to 1.5.1 Was: Request to add a new PLG
	hook - add createUserPrecheck
In-Reply-To: <20080816200525.1321848666@smtp.haun-online.de>
References: <7b42e7470808160058k29c15b91w7475190f4f6a7c04@mail.gmail.com>
	<20080816084402.897223274@smtp.haun-online.de>
	<7b42e7470808160916k6c7830a4j1f26fe7a5f54556@mail.gmail.com>
	<20080816200525.1321848666@smtp.haun-online.de>
Message-ID: <7b42e7470808170023x5dabb6di9e40f32bcdae77c5@mail.gmail.com>

> Regarding #633: Would you rather say "enter poll" or "start poll"? Or
> something else entirely?

I'd argue a multi-question poll was a survey. But, nevermind. I'd go
with start survey if it was a survey so start poll if it's a poll ;-)

Mike


From mjervis at gmail.com  Sun Aug 17 03:31:42 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Sun, 17 Aug 2008 08:31:42 +0100
Subject: [geeklog-devel] The Mysterious Case of the Double Blocks (was
	1.5.1 - what's left to do?)
In-Reply-To: <BLU0-SMTP15E5BF45E6E1C2AB2AE3BDBC6C0@phx.gbl>
References: <7b42e7470808160956o2f479734w73c7762c2b1160e6@mail.gmail.com>
	<BLU0-SMTP15E5BF45E6E1C2AB2AE3BDBC6C0@phx.gbl>
Message-ID: <7b42e7470808170031m654c9a27xede012334b0195cf@mail.gmail.com>

> Michael, you will be retaining the support for $what being a custom function
> right as thats used in a number of plugins.


Yes, of course. COM_siteHeader has $what, and I'm not planning on
changing it. COM_siteFooter tries to use $what, but that's not passed
in, it has however $custom which I have preserved.

> Hmm, something seems to have gone wrong, though. The Configuration
> suddenly has right blocks.

Ah but only ONE set of them. Doh.

> Chameleon uses it's own functions.inc of course, which has the hack as
> posted in the original bug report - but happy to use the much cooler magic
> beans once I can discover the details.

Leading side or trailing side easier for people?

Mike


From dirk at haun-online.de  Sun Aug 17 05:48:53 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Sun, 17 Aug 2008 11:48:53 +0200
Subject: [geeklog-devel] The Mysterious Case of the Double Blocks (was
 1.5.1 - what's left to do?)
In-Reply-To: <7b42e7470808161233o4d002bd0p9a04f12a267486d9@mail.gmail.com>
References: <7b42e7470808160956o2f479734w73c7762c2b1160e6@mail.gmail.com>
	<7b42e7470808161023n3a7a8283w795f5402d9304709@mail.gmail.com>
	<20080816191942.703757270@smtp.haun-online.de>
	<7b42e7470808161233o4d002bd0p9a04f12a267486d9@mail.gmail.com>
Message-ID: <20080817094853.1114066143@smtp.haun-online.de>

Michael Jervis wrote:

>> Why move the $side to the front in the filename?
>
>Because then you can do $side . '_' . $template instead of messing
>around trimming and appending .thtml and checking for trailing
>left/rights.

The current naming convention, block[header|footer]{-something}.thtml
looks more logical to me (and makes it easier to find template files
that belong together).

Aren't you building the filenames from the given parameters anyway? So
the function is in control of the filenames and shouldn't have to do any
stripping or trimming. Or am I missing something?

bye, Dirk


-- 
http://www.haun-online.de/
http://geeklog.info/



From mjervis at gmail.com  Sun Aug 17 06:49:03 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Sun, 17 Aug 2008 11:49:03 +0100
Subject: [geeklog-devel] The Mysterious Case of the Double Blocks (was
	1.5.1 - what's left to do?)
In-Reply-To: <20080817094853.1114066143@smtp.haun-online.de>
References: <7b42e7470808160956o2f479734w73c7762c2b1160e6@mail.gmail.com>
	<7b42e7470808161023n3a7a8283w795f5402d9304709@mail.gmail.com>
	<20080816191942.703757270@smtp.haun-online.de>
	<7b42e7470808161233o4d002bd0p9a04f12a267486d9@mail.gmail.com>
	<20080817094853.1114066143@smtp.haun-online.de>
Message-ID: <7b42e7470808170349t72bf1debndca90251de778cba@mail.gmail.com>

> The current naming convention, block[header|footer]{-something}.thtml
> looks more logical to me (and makes it easier to find template files
> that belong together).

I liked having left and right together in the file list, but I can see
how having all the block templates together is actually more useful.

> Aren't you building the filenames from the given parameters anyway? So
> the function is in control of the filenames and shouldn't have to do any
> stripping or trimming. Or am I missing something?

functions.php in professional:

$nrows = DB_numRows($result);
for ($i = 0; $i < $nrows; $i++) {
    $A = DB_fetchArray($result);
        if ($A['onleft'] == 1) {
            $_BLOCK_TEMPLATE[$A['name']] =
'blockheader-left.thtml,blockfooter-left.thtml';
        } else {
            $_BLOCK_TEMPLATE[$A['name']] =
'blockheader-right.thtml,blockfooter-right.thtml';
    }
}

$_BLOCK_TEMPLATE['_msg_block'] =
'blockheader-message.thtml,blockfooter-message.thtml';

Followed by many more hard-coded entries in $_BLOCK_TEMPLATE to give
particular blocks a specific format.

Then COM_getBlockTemplate just looks them up, if there isn't an entry
for a named block, fails over to blockheader.thtml.

SO...

The change will have to strip the .thtml off the end of the template
from teh array, look for -left or -right and if not test for it. Not
exactly rocket science, but at least a few lines of code more than my
solution ;-)

Still, it keeps them together. I'll go with that. STill saves that
database load and loop in professional's functions.php as it will just
default to looking for blockheader[-left|-right].thtml.

Mike


From dirk at haun-online.de  Sun Aug 17 07:20:33 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Sun, 17 Aug 2008 13:20:33 +0200
Subject: [geeklog-devel] Road to 1.5.1 Was: Request to add a new PLG
 hook - add createUserPrecheck
In-Reply-To: <7b42e7470808160916k6c7830a4j1f26fe7a5f54556@mail.gmail.com>
References: <7b42e7470808160058k29c15b91w7475190f4f6a7c04@mail.gmail.com>
	<20080816084402.897223274@smtp.haun-online.de>
	<7b42e7470808160916k6c7830a4j1f26fe7a5f54556@mail.gmail.com>
Message-ID: <20080817112033.217958816@smtp.haun-online.de>

Michael Jervis wrote:

>> 1) agree which bugs should be fixed
>
>#698, #684, #710, #700, #690, #633 and #175 would by my suggestions.

Looks like we're already down to 3 of those - that was quick :-)

#175, #684, #700

bye, Dirk


-- 
http://www.haun-online.de/
http://spam.tinyweb.net/



From mjervis at gmail.com  Sun Aug 17 08:02:10 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Sun, 17 Aug 2008 13:02:10 +0100
Subject: [geeklog-devel] Road to 1.5.1 Was: Request to add a new PLG
	hook - add createUserPrecheck
In-Reply-To: <20080817112033.217958816@smtp.haun-online.de>
References: <7b42e7470808160058k29c15b91w7475190f4f6a7c04@mail.gmail.com>
	<20080816084402.897223274@smtp.haun-online.de>
	<7b42e7470808160916k6c7830a4j1f26fe7a5f54556@mail.gmail.com>
	<20080817112033.217958816@smtp.haun-online.de>
Message-ID: <7b42e7470808170502x7fef67ddm9ade89e8394a57d5@mail.gmail.com>

> #175, #684, #700

Technically 2 then, once I change the name of $side and update it to
put the thing on the end.


From dirk at haun-online.de  Sun Aug 17 14:23:00 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Sun, 17 Aug 2008 20:23:00 +0200
Subject: [geeklog-devel] Road to 1.5.1 Was: Request to add a new PLG
 hook - add createUserPrecheck
In-Reply-To: <7b42e7470808170502x7fef67ddm9ade89e8394a57d5@mail.gmail.com>
References: <7b42e7470808160058k29c15b91w7475190f4f6a7c04@mail.gmail.com>
	<20080816084402.897223274@smtp.haun-online.de>
	<7b42e7470808160916k6c7830a4j1f26fe7a5f54556@mail.gmail.com>
	<20080817112033.217958816@smtp.haun-online.de>
	<7b42e7470808170502x7fef67ddm9ade89e8394a57d5@mail.gmail.com>
Message-ID: <20080817182300.8459143@smtp.haun-online.de>

Michael Jervis wrote:

>Technically 2 then, once I change the name of $side and update it to
>put the thing on the end.

I've updated geeklog.info and Damn Spam! now to run on the current CVS
version.

The only problem I had was with my hack to make both sites use the same
code - that's still not working quite as I would expect (and doing
"live" tests that take down 2 sites at once is not so desirable ...).

bye, Dirk


-- 
http://geeklog.info/
http://spam.tinyweb.net/



From mjervis at gmail.com  Sun Aug 17 14:50:28 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Sun, 17 Aug 2008 19:50:28 +0100
Subject: [geeklog-devel] Road to 1.5.1 Was: Request to add a new PLG
	hook - add createUserPrecheck
In-Reply-To: <20080817182300.8459143@smtp.haun-online.de>
References: <7b42e7470808160058k29c15b91w7475190f4f6a7c04@mail.gmail.com>
	<20080816084402.897223274@smtp.haun-online.de>
	<7b42e7470808160916k6c7830a4j1f26fe7a5f54556@mail.gmail.com>
	<20080817112033.217958816@smtp.haun-online.de>
	<7b42e7470808170502x7fef67ddm9ade89e8394a57d5@mail.gmail.com>
	<20080817182300.8459143@smtp.haun-online.de>
Message-ID: <7b42e7470808171150g3bde0937mf74e61ff3665efc3@mail.gmail.com>

> The only problem I had was with my hack to make both sites use the same
> code - that's still not working quite as I would expect (and doing
> "live" tests that take down 2 sites at once is not so desirable ...).

Yes, geeklog.info appears to be all written in German all of a sudden ;-)

I'm going to install a dogfood instance once there's a bundled
beta/release candidate and start to us it again properly.

Cheers,

Mike


From joe at ThrowingDice.com  Sun Aug 17 16:20:49 2008
From: joe at ThrowingDice.com (Joe Mucchiello)
Date: Sun, 17 Aug 2008 16:20:49 -0400
Subject: [geeklog-devel] Language File Copyrights
In-Reply-To: <7b42e7470808171150g3bde0937mf74e61ff3665efc3@mail.gmail.co m>
References: <7b42e7470808160058k29c15b91w7475190f4f6a7c04@mail.gmail.com>
	<20080816084402.897223274@smtp.haun-online.de>
	<7b42e7470808160916k6c7830a4j1f26fe7a5f54556@mail.gmail.com>
	<20080817112033.217958816@smtp.haun-online.de>
	<7b42e7470808170502x7fef67ddm9ade89e8394a57d5@mail.gmail.com>
	<20080817182300.8459143@smtp.haun-online.de>
	<7b42e7470808171150g3bde0937mf74e61ff3665efc3@mail.gmail.com>
Message-ID: <0K5R000IAHSTQ9C0@mta1.srv.hcvlny.cv.net>

Not that this is extremely important but is there a reason the core 
GL files list Tony, Tom W, Blaine and Dirk as the authors while the 
core language files only list Jason Whittenburg and sometimes the 
maintainer? Obviously this isn't necessarily a problem but I'm 
curious about the lack of consistency.


----
Joe Mucchiello
Throwing Dice Games
http://www.throwingdice.com 



From mjervis at gmail.com  Mon Aug 18 02:43:22 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Mon, 18 Aug 2008 07:43:22 +0100
Subject: [geeklog-devel] Language File Copyrights
In-Reply-To: <0K5R000IAHSTQ9C0@mta1.srv.hcvlny.cv.net>
References: <7b42e7470808160058k29c15b91w7475190f4f6a7c04@mail.gmail.com>
	<20080816084402.897223274@smtp.haun-online.de>
	<7b42e7470808160916k6c7830a4j1f26fe7a5f54556@mail.gmail.com>
	<20080817112033.217958816@smtp.haun-online.de>
	<7b42e7470808170502x7fef67ddm9ade89e8394a57d5@mail.gmail.com>
	<20080817182300.8459143@smtp.haun-online.de>
	<7b42e7470808171150g3bde0937mf74e61ff3665efc3@mail.gmail.com>
	<0K5R000IAHSTQ9C0@mta1.srv.hcvlny.cv.net>
Message-ID: <7b42e7470808172343g499b2f99g77ef7ffc652c37d4@mail.gmail.com>

Joe,

> Not that this is extremely important but is there a reason the core GL files
> list Tony, Tom W, Blaine and Dirk as the authors while the core language
> files only list Jason Whittenburg and sometimes the maintainer? Obviously
> this isn't necessarily a problem but I'm curious about the lack of
> consistency.

It is important to record with whom the copyright and credit lies.

People add themselves to the copyright when they have provided
significant input into some code. So for most of the core geeklog
Tony, Tom and Dirk have had significant input and get
credit/copyright/ownership. For the language files, Jason set them up,
but their content comes from the maintainer. Dirk just facilitates and
I assume doesn't feel he has ownership of them or significant input (I
don't think he even does the German translation).

Whereas other files have ownership resting with other people, for
example, story.class.php rests with me, as does the syndication
framework and remote authentication. But I'm not listed on article.php
as I've not done much to it (beyond tweaks to use story.class.php).
Ditto lib-common etc.

Files credit original authors, and those who have made significant
input. And that's the etiquette in open source.

I /was/ looking for a post/article/email on the subject that I
got/found/read when I first started to contribute to open source about
this that was really good. But neither I nor Dirk can find it. It may
have been related to my Drupal contributions, or PHPNuke not Geeklog.
But I was sure it was Geeklog. Sigh.

Mike


From mjervis at gmail.com  Mon Aug 18 13:24:39 2008
From: mjervis at gmail.com (Michael Jervis)
Date: Mon, 18 Aug 2008 18:24:39 +0100
Subject: [geeklog-devel] GSoC installer
Message-ID: <7b42e7470808181024o2c96d09an4b5feac83a2b39da@mail.gmail.com>

Fatal error: Call to undefined function curl_init() in C:\Program
Files\xamp\geeklog\geeklog-hg-dev\public_html\admin\install\lib-install.php
on line 521

Of course I don't have the curl extension installed (default in
XAMPP). So perhaps we should detect such pre-requisites before
crashing?

Also there is install and install with additional plugins, does just
install include all the ones ticked by default and the differences
with additional is just uploading extras?

Pretty neat, nice and slick looking.

Mike

-- 
Michael Jervis
mjervis at gmail.com
504B03041400000008008F846431E3543A820800000006000000060000007765
62676F642B4F4D4ACF4F0100504B010214001400000008008F846431E3543A82
0800000006000000060000000000000000002000000000000000776562676F64
504B05060000000001000100340000002C0000000000


From dirk at haun-online.de  Mon Aug 18 14:46:42 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Mon, 18 Aug 2008 20:46:42 +0200
Subject: [geeklog-devel] GSoC installer
In-Reply-To: <7b42e7470808181024o2c96d09an4b5feac83a2b39da@mail.gmail.com>
References: <7b42e7470808181024o2c96d09an4b5feac83a2b39da@mail.gmail.com>
Message-ID: <20080818184642.1350838860@smtp.haun-online.de>

Michael Jervis wrote:

>Of course I don't have the curl extension installed (default in
>XAMPP). So perhaps we should detect such pre-requisites before
>crashing?

Good point, hadn't noticed this.


>Also there is install and install with additional plugins, does just
>install include all the ones ticked by default and the differences
>with additional is just uploading extras?

"Install" is supposed to work like the current install. Or rather,
simply install all plugins contained in the bundle. The other option is
the "advanced install" option where you can configure the plugin mix at
install time.


>Pretty neat, nice and slick looking.

We want the buttons on the initial screen to look more like in 1.5.0 of
course, but that's only cosmetics :)

bye, Dirk


-- 
http://www.geeklog.net/
http://geeklog.info/



From dirk at haun-online.de  Mon Aug 18 17:03:36 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Mon, 18 Aug 2008 23:03:36 +0200
Subject: [geeklog-devel] Language File Copyrights
In-Reply-To: <7b42e7470808172343g499b2f99g77ef7ffc652c37d4@mail.gmail.com>
References: <7b42e7470808160058k29c15b91w7475190f4f6a7c04@mail.gmail.com>
	<20080816084402.897223274@smtp.haun-online.de>
	<7b42e7470808160916k6c7830a4j1f26fe7a5f54556@mail.gmail.com>
	<20080817112033.217958816@smtp.haun-online.de>
	<7b42e7470808170502x7fef67ddm9ade89e8394a57d5@mail.gmail.com>
	<20080817182300.8459143@smtp.haun-online.de>
	<7b42e7470808171150g3bde0937mf74e61ff3665efc3@mail.gmail.com>
	<0K5R000IAHSTQ9C0@mta1.srv.hcvlny.cv.net>
	<7b42e7470808172343g499b2f99g77ef7ffc652c37d4@mail.gmail.com>
Message-ID: <20080818210336.1043914889@smtp.haun-online.de>

Michael Jervis wrote:

>For the language files, Jason set them up,
>but their content comes from the maintainer. Dirk just facilitates and
>I assume doesn't feel he has ownership of them or significant input

If the translators don't add their names to the language files, then
that is entirely their decision. I do try and give proper credits in the
checkin comments and the changelog, though.

Jason's name is in there since pretty much every language file has been
derived from the original english.php and people just keep the
copyrights they find there.


>(I don't think he even does the German translation).

A German translation was actually one of my first contributions to Geeklog :)

The language file for formal German was later started by Philipp Sack.
Markus Wollschl?ger did the translation of the new texts in 1.5.0 and
also ensured (restored?) consistency.


>Files credit original authors, and those who have made significant
>input. And that's the etiquette in open source.

Yep. I've only recently decided that I'm now worthy of being listed as a
co-author for lib-webservices.php. I still consider it to be mainly
Ramnath's work.

bye, Dirk


-- 
http://www.geeklog.net/
http://geeklog.info/



From dirk at haun-online.de  Sat Aug 23 13:23:23 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Sat, 23 Aug 2008 19:23:23 +0200
Subject: [geeklog-devel] Greetings from FrOSCon
Message-ID: <20080823192323.rgyq2rjb404occkc@webmail.df.eu>

Hi there,

day 1 of FroSOCon is already over. Lots of interested people here.

And, btw, here's the ultimate comment on templating systems:

I've finally had a chance to shake hands with Kris K?hntopp (and if  
you don't know who that is, you've never looked at our template  
class). He was slightly amused that we're still using his code. His  
recommendation: Use Smarty. So there you have it.

More later.

bye, Dirk

P.S. <http://www.flickr.com/photos/dhaun/2790142606/>


-- 
http://www.haun-online.de/
http://geeklog.info/



From dirk at haun-online.de  Mon Aug 25 13:55:31 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Mon, 25 Aug 2008 19:55:31 +0200
Subject: [geeklog-devel] FrOSCon 2008
In-Reply-To: <20080823192323.rgyq2rjb404occkc@webmail.df.eu>
References: <20080823192323.rgyq2rjb404occkc@webmail.df.eu>
Message-ID: <20080825175531.350374600@smtp.haun-online.de>

Okay, let's see what I can remember ...

FrOSCon 2008 was pretty good for us, I'd say. We had a lot of interested
people at our booth on Saturday. Sunday was a lot quieter, but still
quite some interesting contacts. So it was definitely the right decision
(after last year's flop with the project room) to have a booth.

The good thing with an open source conference such as FrOSCon is that
more people know Groklaw :-) That always helps.

What were people missing in Geeklog? One word: Postgres. I think we had
more requests for Postgres on Saturday morning alone than during the
entire LinuxTag (which was a four-day event). Something for the next
Summer of Code, maybe? Or a big bounty? (Cue Tony suggesting to switch
to a different DB layer ;-)

Why the increased interest in Postgres? I forgot to ask, but I'm getting
the impression that people (in the open source world) aren't comfortable
with the MySQL buyout by Sun and are starting to look for alternatives.

Other requests I remember:

- More flexible moderation. Some people had very specific ideas about
how moderation should work, but I think we can improve some things
there, e.g. by allowing to save changes back into the moderation queue.

- Integration and migration - that always comes up: How do I integrate
system "X" into Geeklog? How do I migrate my data from system "Y" to
Geeklog? CRM and shopping carts / ecommerce systems were mentioned a few
times.

Markus, anything else I forgot?


People: Christian Weiske dropped by and introduced himself (Hi!). Sorry
that I had to cut you off due my scheduled interview with RadioTux
(which reminds me that I haven't listened to it yet ...). Btw, do you
have an URL for that VMWare installation protocol / system that you mentioned?

I also met a bunch of the usual suspects - people I know from elsewhere.
There were even a suprising number of people from the Stuttgart area,
including someone from the Chaos Computer Club Stuttgart who invited me
to their next meeting - which features a talk by Kris K?hntopp (him
again ...).

Thomas Weinert from Papaya CMS mentioned experiments he'd done with
Baysian spam filters and filters based on the distribution of letters in
a post, so that you can recognize posts in foreign languages. That
sounded very interesting. I've argued before that people don't want to
train Baysian filters on their website, but he seems to have the
training down to a minimum.


Presentations: Well, the downside of having to man a booth is that you
miss all the great talks. I've only managed to hear the one by Andrew
Tanenbaum[1].

My own presentation (on the Atom Publishing Protocol) went well, I
think, although it ran a bit short (it probably didn't help that the
last time I gave a presentation on AtomPub was a 10-minute quickie at
the local WebMonday) - should have thrown in a few more examples, I guess.


In summary: Great conference, as usual. A lot more contacts for us this
year, so the booth was a good idea. I'll be back in 2009 :)


Thanks: To the FrOSCon organizers, who managed the growth over last year
just fine. To Markus Wollschl?ger for manning the booth with me on
Saturday. To evenos.de for lending us an LCD screen on Sunday.

For photos, try

- http://www.flickr.com/search/?q=froscon2008
- http://gallery.froscon.de

bye, Dirk

[1] http://www.flickr.com/photos/stiefkind/2792284466/


-- 
http://www.geeklog.net/
http://geeklog.info/



From cweiske at cweiske.de  Tue Aug 26 03:08:05 2008
From: cweiske at cweiske.de (Christian Weiske)
Date: Tue, 26 Aug 2008 09:08:05 +0200
Subject: [geeklog-devel] FrOSCon 2008
In-Reply-To: <20080825175531.350374600@smtp.haun-online.de>
References: <20080823192323.rgyq2rjb404occkc@webmail.df.eu>
	<20080825175531.350374600@smtp.haun-online.de>
Message-ID: <20080826090805.47f7d200@bogo.home.cweiske.de>

Hello Dirk,


> People: Christian Weiske dropped by and introduced himself (Hi!).
> Sorry that I had to cut you off due my scheduled interview with
> RadioTux (which reminds me that I haven't listened to it yet ...).
> Btw, do you have an URL for that VMWare installation protocol /
> system that you mentioned?
http://www.apsstandard.com/app


-- 
Regards/Mit freundlichen Gr??en
Christian Weiske

-= Geeking around in the name of science since 1982 =-
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: not available
URL: <https://pairlist8.pair.net/pipermail/geeklog-devel/attachments/20080826/52c02590/attachment.sig>

From devel at portalparts.com  Tue Aug 26 08:20:43 2008
From: devel at portalparts.com (Blaine Lang)
Date: Tue, 26 Aug 2008 08:20:43 -0400
Subject: [geeklog-devel] FrOSCon 2008
In-Reply-To: <20080825175531.350374600@smtp.haun-online.de>
References: <20080823192323.rgyq2rjb404occkc@webmail.df.eu>
	<20080825175531.350374600@smtp.haun-online.de>
Message-ID: <BLU0-SMTP5FA90C07F7DFB28CD0581BC660@phx.gbl>

Dirk, Markus: Nice report and great job done promoting Geeklog.
<insert applause here>

The link to APS was quite interesting and would like to better 
understand whats reqiured to package up an application (or plugin) for 
APS and am wondering if it may be easier to maintain/support for even as 
an automated  plugin installer subsystem.

Dirk, was there a link to your interview?

Cheers,
Blaine



From dirk at haun-online.de  Tue Aug 26 14:02:35 2008
From: dirk at haun-online.de (Dirk Haun)
Date: Tue, 26 Aug 2008 20:02:35 +0200
Subject: [geeklog-devel] FrOSCon 2008
In-Reply-To: <BLU0-SMTP5FA90C07F7DFB28CD0581BC660@phx.gbl>
References: <20080823192323.rgyq2rjb404occkc@webmail.df.eu>
	<20080825175531.350374600@smtp.haun-online.de>
	<BLU0-SMTP5FA90C07F7DFB28CD0581BC660@phx.gbl>
Message-ID: <20080826180235.486499120@smtp.haun-online.de>

Blaine Lang wrote:

>Dirk, was there a link to your interview?

This doesn't look like a permanent URL, but it's currently at
<http://cms.radiotux.de/index6.php> together with other interviews they
did at FrOSCon.

It's in German, of course :)

The MP3 starts in mid-sentence for some reason. Still haven't actually
listened to it. We talked about what Geeklog is and also about the
Summer of Code.

And at the risk of turning this into an ego trip: There are also some
photos from my AtomPub presentation:
<http://www.flickr.com/search/?q=froscon2008+dirkhaun&m=tags&ss=2>

bye, Dirk


-- 
http://www.haun-online.de/
http://geeklog.info/