Pidgin

I was just trying to figure this out last night, myself.  There's a
description of the status API in the API documentation, but I wasn't =
able to
derive much from it, or the other plugins.  I know the listbox =
represents
various different account statuses you can have in various respective
accounts.  One would think, at least, that the icon set would be =
different,
so you knew which of the several 'Away' entries were which.  I haven't =
been
able to find any plugins that do any manipulation of this box, at least =
in
2.0b2.   If someone knows of one I could poke at, or some more complete
documentation, that would be awesome.  I'm trying to write a =
status-oriented
plugin, and this thing is just confusing me at the moment. :)
=20
=20

  _____ =20

From: gaim-devel-admin@...
[mailto:gaim-devel-admin@...] On Behalf Of David =
Balazic
Sent: Thursday, March 02, 2006 1:54 AM
To: Gaim-devel@...
Subject: [Gaim-devel] Gaim 2 beta: Status problems



Hi!=20

I use Gaim 2.0.0beta2 on Windows.=20

I have two related issues :=20

 - what is the purpose of that ListBox at the bottom of the=20
"Buddy List" windows ? Why does it have so many duplicate entries?=20

 - recentylI tried to send a message to a MSN buddy and I got this =
printed=20
in the chat windows : "can not send messate due to invisible mode"=20
This is not the exact text. Wha does it mean ? Who is invisible ?=20
How can I check that ? How can I fix it ?=20

Regards,=20
David Bala=BEic=20

---------------------------------------------------------------------=20
http://noepatents.eu.org/          Innovation, not litigation !=20
---=20
David Balazic                      mailto:david.balazic@...
HERMES Softlab                     http://www.hermes-softlab.com=20
Zagrebska cesta 104                Phone: +386 2 450 8846=20
SI-2000 Maribor=20
Slovenija=20
---------------------------------------------------------------------=20
"Be excellent to each other." -=20
Bill S. Preston, Esq. & "Ted" Theodore Logan=20
---------------------------------------------------------------------=20

--- Josh Edson <jedson@...> wrote:
...
> I know the listbox
> represents
> various different account statuses you can have in various
> respective
> accounts.  One would think, at least, that the icon set would
> be different,
> so you knew which of the several 'Away' entries were which.

The entries in the dropdown has different icon-emblems depending
what status it represents (ie, `Away' or `Available'). There's
also an emblem to show whether a status was saved or not.

> I haven't been
> able to find any plugins that do any manipulation of this box,
> at least in
> 2.0b2.   If someone knows of one I could poke at, or some more
> complete
> documentation, that would be awesome.  I'm trying to write a
> status-oriented
> plugin, and this thing is just confusing me at the moment. :)

You can try taking a look at mystatusbox plugin in plugin-pack of
Guifications (#guifications on IRC). It allows you to show the
per-account status-boxes (which were there in beta-1 release, but
removed for beta-2), hide the global status-box etc. (so you can
make all the status-boxes disappear). This plugin deals with the
statusbox -- which is there in Gaim. If you are looking for a
plugin to deal with the status-API, you may take a look at geam
(http://sourceforge.net/projects/geam/). I haven't worked on it
for a while, but it should still be mostly functional, and give
an idea how to deal with status-API.

> 
> From: gaim-devel-admin@...
> [mailto:gaim-devel-admin@...] On Behalf Of
> David Balazic

> Hi! 
> 
> I use Gaim 2.0.0beta2 on Windows. 
> 
> I have two related issues : 
> 
>  - what is the purpose of that ListBox at the bottom of the 
> "Buddy List" windows ? Why does it have so many duplicate
> entries? 
> 

I believe this has been fixed for next release, so you won't have
duplicate entries in the dropdown.

-- Adil

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com

On Tue, 28 Feb 2006 20:52:41 -0500, Evan Schoenberg wrote
> How is it that the plugins / prpls can be built without linking  
> against the gaim binary given that they (obviously) use symbols 
> which  are defined in gaim rather than within their own code?  I'm 
> very  ignorant of proper build processes, having depended upon an 
> IDE to  manage the magic for me in the past, but would like to learn 
> and  don't know where to start.
> 
> Thanks :)
> 
> -Evan

I'm certainly no compiler expert, but here's my take on it.

The plugins are compiled as dynamic libraries.  They #include header files
from the Gaim core.  The header files tell gcc, "this function is defined
somewhere--don't worry about where it's defined--here is the function's
signature."  Then some placeholder code is inserted into the binary.

At runtime, the operating system or glibc or something dynamically links the
plugin into the Gaim core and substitutes the placeholder code with references
to the actual functions.

It's possible that the operating system or glibc or something won't be able to
find the actual function.  This can happen if you declare a function in a
header file but never actually define the function anywhere (if you removed
aim_sncmp from src/protocols/oscar/util.c, for example).  In this case, Gaim
prints out an error about being unable to find function blahblahblah and the
PRPL is not loaded and cannot be used.

I'm sure you already knew some of that... is that the kind of info you were
looking for?
-Mark

On Tue, 2006-02-28 at 20:52 -0500, Evan Schoenberg wrote:
> How is it that the plugins / prpls can be built without linking  
> against the gaim binary given that they (obviously) use symbols which  
> are defined in gaim rather than within their own code?  I'm very  
> ignorant of proper build processes, having depended upon an IDE to  
> manage the magic for me in the past, but would like to learn and  
> don't know where to start.

I think the modules are simply linked with the -module libtool option,
which results in the linker being instructed not to error out on
undefined symbols.

On Mon, 2006-02-27 at 07:51 -0500, Luke Schierer wrote:
> Since it blurs the line between "Developers" and "Crazy Patch Writers"

> Since many of us have multiple trees anyway.

> Since several of those other-tree projects would benefit from help,
> such as gobjectification or the work towards msn 7.

> Since most of the non-cvs scm are distributed.

An alternative way to address these issues is the aggressive use of
branches and branch commit access.  The gcc project has the same issues,
and prefers a centralized scm with branches to a distributed scm because
it keeps more of gcc's development visible in one place.

Subversion's branch support is better than CVS's in some important ways:
branch creation is cheaper, the namespace of branches is versioned and
can be pared down to active branches (old branches can be moved aside or
simply removed as desired, and no information is truly lost), and
merging between branches is a simpler process because of atomic
revisions, even though there is no support for merge tracking or
advanced tree merging.  On the other side, the UI for cross-branch
comparisons is a little clunky compared to CVS's, and it's harder to
discover the history of a particular file or directory across all
branches.

Of course, you may find that distributed development suits you better,
but the space of distributed SCMs is pretty fragmented at the moment,
and it may be years before one arises which is feature-complete,
performs acceptably, has a clear development future, works well on
Windows, etc.

I migrated two of my other projects on sf.net.  Went without a hitch, thoug=
h
admittedly, they are a fraction of a fraction of the size of Gaim.


--
God Bless (Psalms 20:4),
    Travis Harris
    New Covenant Software
    322 Old Mill Rd
    Newmanstown, PA 17073

    866.213.6539

On 2/27/06, Evan Schoenberg <evan.s@...> wrote:
>
>
> On Feb 26, 2006, at 4:31 PM, Ethan Blanton wrote:
>
>  once to the marginally better svn
>
>
> I would argue that subversion is significantly better than cvs, not
> 'marginally better' (though your point remains).
>
> It looks like sf.net has set up an automatic migration script which, if i=
t
> works, would mean very little effort in the migration process for Gaim.
>
> -Evan
>
>

Evan Schoenberg spake unto us the following wisdom:
> On Feb 26, 2006, at 4:31 PM, Ethan Blanton wrote:
> > once to the marginally better svn
>=20
> I would argue that subversion is significantly better than cvs, not =20
> 'marginally better' (though your point remains).

In many important areas it's not an improvement, which is why I make
this claim.  It doesn't really fix branching and merging to speak of
(sure, branches are cheaper, but merges remain much more painful than
most distributed VCSs), and it doesn't help third-party developers
work with our sources beyond what CVS provides.

> It looks like sf.net has set up an automatic migration script which, =20
> if it works, would mean very little effort in the migration process =20
> for Gaim.

Unfortunately, this is the cheap part of migrating, because it only
has to be done once.  The expensive part is bringing all of the
developers and processes up to speed.  Since we're pretty shy on
process, at least that part wouldn't be a big deal. ;-)

Ethan

--=20
The laws that forbid the carrying of arms are laws [that have no remedy
for evils].  They disarm only those who are neither inclined nor
determined to commit crimes.
		-- Cesare Beccaria, "On Crimes and Punishments", 1764

On Mon, Feb 27, 2006 at 01:00:54PM -0500, Greg Hudson wrote:
> Subversion's branch support is better than CVS's in some important ways:
> branch creation is cheaper, the namespace of branches is versioned and
> can be pared down to active branches (old branches can be moved aside or
> simply removed as desired, and no information is truly lost), and
> merging between branches is a simpler process because of atomic
> revisions, even though there is no support for merge tracking or
> advanced tree merging.  On the other side, the UI for cross-branch
> comparisons is a little clunky compared to CVS's, and it's harder to
> discover the history of a particular file or directory across all
> branches.

Having used both cvs and svn, I'm curious to know more details regarding
the two points where you think cvs is better, since I've always thought
that svn was better than cvs in those areas?

For instance, when I create a tag (which is simply a copy) all my
history goes with it.  As an example:

http://svn.sourceforge.net/viewcvs.cgi/gaim-bs/gaim-bs/tags/release-0.9.0/s=
rc/gaim-bs.c?view=3Dlog

You'll also notice in the above history that at some point earlier, I
renamed this file (using 'svn mv').  But this history is still there,
regardless of the new name (I think cvs lacks this feature).

And if I want to compare two branches, it seems quite
easy/straightforward to do the following (all on one line):

svn diff
https://svn.sourceforge.net/svnroot/gaim-bs/gaim-bs/tags/release-0.9.0
https://svn.sourceforge.net/svnroot/gaim-bs/gaim-bs/trunk

--=20
Encrypted Mail Preferred:
    Key ID:  8527B9AF
    Key Fingerprint:  E1B6 40B6 B73F 695E 0D3B  644E 6427 DD74 8527 B9AF
    Information:  http://www.gnupg.org/
                                                                           =
    =20
ASCII ribbon campaign:
()  against HTML email
/\  against Microsoft attachments
    Information:  http://www.expita.com/nomime.html

On Mon, Feb 27, 2006 at 01:22:05PM -0500, Ethan Blanton wrote:
> Evan Schoenberg spake unto us the following wisdom:
> > On Feb 26, 2006, at 4:31 PM, Ethan Blanton wrote:
> > > once to the marginally better svn
> >=20
> > I would argue that subversion is significantly better than cvs, not =20
> > 'marginally better' (though your point remains).
>=20
> In many important areas it's not an improvement, which is why I make
> this claim.  It doesn't really fix branching and merging to speak of
> (sure, branches are cheaper, but merges remain much more painful than
> most distributed VCSs), and it doesn't help third-party developers
> work with our sources beyond what CVS provides.

My only significant experience is with CVS, Subversion, and ClearCase.
So I'm curious to know of a tool that fixes these issues to your
satisfaction, as I would be interested to read up on how it (and a
distributed VCS) should work.

Thanks.

--=20
Encrypted Mail Preferred:
    Key ID:  8527B9AF
    Key Fingerprint:  E1B6 40B6 B73F 695E 0D3B  644E 6427 DD74 8527 B9AF
    Information:  http://www.gnupg.org/
                                                                           =
    =20
ASCII ribbon campaign:
()  against HTML email
/\  against Microsoft attachments
    Information:  http://www.expita.com/nomime.html

David Everly spake unto us the following wisdom:
> On Mon, Feb 27, 2006 at 01:22:05PM -0500, Ethan Blanton wrote:
> > Evan Schoenberg spake unto us the following wisdom:
> > > On Feb 26, 2006, at 4:31 PM, Ethan Blanton wrote:
> > > > once to the marginally better svn
> > >=20
> > > I would argue that subversion is significantly better than cvs, not =
=20
> > > 'marginally better' (though your point remains).
> >=20
> > In many important areas it's not an improvement, which is why I make
> > this claim.  It doesn't really fix branching and merging to speak of
> > (sure, branches are cheaper, but merges remain much more painful than
> > most distributed VCSs), and it doesn't help third-party developers
> > work with our sources beyond what CVS provides.
>=20
> My only significant experience is with CVS, Subversion, and ClearCase.
> So I'm curious to know of a tool that fixes these issues to your
> satisfaction, as I would be interested to read up on how it (and a
> distributed VCS) should work.

OpenCM (http://www.opencm.org/), a non-distributed VCS, actually,
handles branching and merging the best of anything I've seen.  It is
unfortunately not a reasonable candidate for Gaim development as it
has performance issues with large-ish repositories, isn't quite
finished, and development seems to be dead.

Darcs (http://abridgegame.org/darcs/) basically abstracts branching
and merging to the point that they require no different developer
concerns from a typical update.  One could conversely consider that
updates are sufficiently abstract to handle branching and merging.
:-)

Monotone (http://www.venge.net/monotone/) is my current favorite in
the distributed VCS arena, but I do not like its current merging
mechanism.  It is, however, an interesting place to look for the
general lay of what I consider a good distributed VCS.  You will find
many things in common with git (http://git.cz/) and cogito
(http://www.kernel.org/pub/software/scm/cogito/), the former of which
I understand was influenced by monotone's design.

I think there's some information about all of this in the wiki.

Ethan

--=20
The laws that forbid the carrying of arms are laws [that have no remedy
for evils].  They disarm only those who are neither inclined nor
determined to commit crimes.
		-- Cesare Beccaria, "On Crimes and Punishments", 1764

On Mon, 2006-02-27 at 11:59 -0700, David Everly wrote:
> > On the other side, the UI for cross-branch
> > comparisons is a little clunky compared to CVS's, and it's harder to
> > discover the history of a particular file or directory across all
> > branches.
> 
> Having used both cvs and svn, I'm curious to know more details regarding
> the two points where you think cvs is better, since I've always thought
> that svn was better than cvs in those areas?

"cvs diff -r branch file" is an easier way to compare a file against
that same file on the head of a branch (or in a tag) than the equivalent
command in Subversion.

"cvs log file" will give you the history of a file including all changes
to that file on any branch.  Sometimes that's noise, but I hear tell
some people like that.

While I'm here, Ethan Blanton wrote:
> it doesn't help third-party developers work with our sources beyond
> what CVS provides.

I disagree.  Changing to a tree-versioned SCM is an important aid to
third-party developers, even if it has no advanced merging support.
Tools like svk and tailor have a much easier time working against a
Subversion repository than they do against a CVS repository.

Whenever I work against a third-party CVS tree, I find myself very
frustrated when I want to view the changes to a directory, or the whole
tree, over some period of time.  I have to go look for a commit mailing
list or a ChangeLog, either of which might not exist or might not be
perfectly in sync.

I promise to shut up about revision control now unless someone directly
asks me a question.  I'm a (not terribly active any more) Subversion
developer, so I can go on at much greater length on this issue than most
Gaim developers really care to hear.

On Mon, Feb 27, 2006 at 02:27:46PM -0500, Ethan Blanton wrote:
> 
> Monotone (http://www.venge.net/monotone/) is my current favorite in
> the distributed VCS arena, but I do not like its current merging
> mechanism.  It is, however, an interesting place to look for the
> general lay of what I consider a good distributed VCS.  You will find
> many things in common with git (http://git.cz/) and cogito
> (http://www.kernel.org/pub/software/scm/cogito/), the former of which
> I understand was influenced by monotone's design.
> 
> I think there's some information about all of this in the wiki.
> 
> Ethan

Note that cogito just builds on top of git, and that the two tools
use precisely the same repository.

There is some info at
http://www.faceprint.com/gaimwiki/index.cgi?Post2Point0  but it is
insufficient and incomplete.

luke

Luke Schierer spake unto us the following wisdom:
> Note that cogito just builds on top of git, and that the two tools
> use precisely the same repository.

Yes, sorry, I didn't mean to imply otherwise.  Cogito is a set of
scripts for working with git, which is a backend kind of thing.

Ethan

--=20
The laws that forbid the carrying of arms are laws [that have no remedy
for evils].  They disarm only those who are neither inclined nor
determined to commit crimes.
		-- Cesare Beccaria, "On Crimes and Punishments", 1764

It sounds like the general consensus is that using Subversion would be better
than CVS, and that using a distributed source code repository would be even
better.

I propose that we migrate our web repository to Subversion within the next
week or so, before releasing Gaim 2.0.0.  Assuming that goes well, I propse
that we migrate the gaim repository to Subversion shortly after releasing Gaim
2.0.0 (so that we don't hinder the translators).

Reasons for moving away from CVS:
* Sourceforge's anonymous CVS servers are crap, and a publicly accessible
source code repository is vital to an open source project

Reasons for not moving to a distributed source code repository yet:
* There is no clear best-of-breed, and I have a feeling it'll take us some
time to find one that's acceptable to most people
* We would need to find somewhere to host it.  This isn't a major problem...
but it will definitely take a bit more time.

And we should definitely keep open the option of moving to a distributed
source code repository in the future.
-Mark

Mark Doliner wrote:

>It sounds like the general consensus is that using Subversion would be better
>than CVS, and that using a distributed source code repository would be even
>better.
>
>I propose that we migrate our web repository to Subversion within the next
>week or so, before releasing Gaim 2.0.0.  Assuming that goes well, I propse
>that we migrate the gaim repository to Subversion shortly after releasing Gaim
>2.0.0 (so that we don't hinder the translators).
>  
>
Sounds good to me.

--Tim

On 2/28/06, Mark Doliner <mark@...> wrote:
> I propose that we migrate our web repository to Subversion within the nex=
t
> week or so, before releasing Gaim 2.0.0.  Assuming that goes well, I prop=
se
> that we migrate the gaim repository to Subversion shortly after releasing=
 Gaim
> 2.0.0 (so that we don't hinder the translators).

I'm not certain Sourceforge's CVS to SVN migration process allows
separate migrations of one project's CVS repository?

Perhaps this can be hacked by getting the CVS repository image itself
and then migrating only one portion of it ...

--
ASCII ribbon campaign:
()  against HTML email
/\  against Microsoft attachments
    Information:  http://www.expita.com/nomime.html

In short, It does!

--
God Bless (Psalms 20:4),
    Travis Harris
    New Covenant Software
    322 Old Mill Rd
    Newmanstown, PA 17073

    866.213.6539

On 3/1/06, David Everly <deckrider@...> wrote:
>
> On 2/28/06, Mark Doliner <mark@...> wrote:
> > I propose that we migrate our web repository to Subversion within the
> next
> > week or so, before releasing Gaim 2.0.0.  Assuming that goes well, I
> propse
> > that we migrate the gaim repository to Subversion shortly after
> releasing Gaim
> > 2.0.0 (so that we don't hinder the translators).
>
> I'm not certain Sourceforge's CVS to SVN migration process allows
> separate migrations of one project's CVS repository?
>
> Perhaps this can be hacked by getting the CVS repository image itself
> and then migrating only one portion of it ...
>
> --
> ASCII ribbon campaign:
> ()  against HTML email
> /\  against Microsoft attachments
>     Information:  http://www.expita.com/nomime.html
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by xPML, a groundbreaking scripting
> language
> that extends applications into web and mobile media. Attend the live
> webcast
> and join the prime developer group breaking into this new coding
> territory!
> http://sel.as-us.falkag.net/sel?cmdlnk&kid=110944&bid$1720&dat=121642
> _______________________________________________
> Gaim-devel mailing list
> Gaim-devel@...
> https://lists.sourceforge.net/lists/listinfo/gaim-devel
>

Etan Reisner wrote:

> On Fri, 17 Feb 2006, David Frascone wrote:
>
>> I'm trying to write a simple Perl plugin to change a particurlarly 
>> formated string into a url (to make it easy to click defect tracker 
>> bug ids).  But, for some reason, I get immediate segv's in both the 
>> "received-im-msg" and "receiving-im-msg" callback.  I've tried doing 
>> absolutely nothing, returning 0, returning undef, but nothing seems 
>> to work.
>>
>> Any ideas?
>>
>> Should I attach the perl snippet & backtrace?  What else should I 
>> attach?
>>
>> -Dave
>
>
> What version of gaim are you writing the plugin for?
> And yes, if you want any real help the plugin and a backtrace would be 
> useful.
>
> Though if you are writing the plugin for gaim 1.5.0 then know that the 
> perl interface has changed a good bit since then and that it was known 
> to be rather broken in 1.5.0.
>
>     -Etan


Yeah, I figured it was broken.  I ended up switching to the TCL api, and 
it worked like a charm!

-Dave