Re: [Gajim-devel] gsoc: file transfer over jinge

[Yann Leboulanger]

Zhenchao Li wrote:
> Hello everyone!

Hi man,

>   I am a graduate student in China and I am very interested in
> implementing file transfer over jingle for gajim. For the last month
> I've been exploring gajim's source code, reading documentations about
> gajim's architecture, reading xmpp's specifications on jingle and file
> transfer(XEP-0096, XEP-0234, XEP-0166) and getting some hands-on
> experience by trying to write a patch for a bug(#5614
> ). So I want to share with you my
> thoughts here.

Very good idea, and very good project.

[...]

>   I also want to add some additional features that might help to enhance
> user experience:
> 
>   First: file transfer status in chat window. [...]
> Suggestions welcomed about this!

Nice, it's a long time ticket, and sure we'll discuss that when time
will come.

>   Second: drag and drop currently works perfectly, but sending of
> multiple files are not supported. [...]

You can DnD multiple files, and send them in one shot. But receiver wil
have to accept them one by one. Too bad. I haven't read Jingle FT
specifications, but is it possible to do the negociation ofr several
files? If yes that's the solution, else we'll have to discuss that in
standards mailing list, so that it's added to the XEP or a new XEP
should be written to be complient with other clients. This will take
some time to define that, so not sure you'll have time to implement that.

>   If time allows I also want to add file encryption feature, which might
> be as simple as key exchange and encryption using DES or AES. But I
> don't know how to standardize this feature so it understood by all xmpp
> clients. I appreciate any advice on this.

I already saw discussions about encrypted jingle sessions ... But I
don't know the status of that at all.

>   Any advices, thoughts, suggestions, questions are welcomed!

I have to tell you that you're not the only one contacting me for this
project. I don't know how many slots XSF will have this year, and I
don't know how many valuable project they will receive, so it's
impossible to know how many slots Gajim will have (if any).
Jingle file transfer is in XSF priorities, so this kind of project will
be in the top of the list. I think.

>   PS: After discussing my idea here, should I also present it to the XSF
> mailing list?

I could be a good idea indeed to present yourself to XSF comunity so
that people who will vote for projects know a bit about you.

-- 
Yann
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

[Gajim-devel] gsoc: file transfer over jinge

[Zhenchao Li]

Hello everyone!

  I am a graduate student in China and I am very interested in implementing
file transfer over jingle for gajim. For the last month I've been exploring
gajim's source code, reading documentations about gajim's architecture,
reading xmpp's specifications on jingle and file
transfer(XEP-0096, XEP-0234, XEP-0166) and getting some hands-on experience
by trying to write a patch for a bug(#5614).
So I want to share with you my thoughts here.

  Gajim already has the ability to handle jingle singaling thanks to gsoc
2007 student Tomasz Melcer. But over the last 2 years jingle itself has also
evolved. I will first check to see if our implementation of jingle has any
major incompatibility with the current jingle protocol.(I don't foresee a
lot of it because according to jingle's revision history many of them were
just clarifications). This shoudl take about 1 week.

  Next I will implement the actual file transfer feature using the jingle
infrastructure. The actual transfer will consist of several methods: socks5
bytestream, in-band bytestream and possibly ICE-TCP if time allows). I don't
think these are extremely difficult to implement once one of those transfer
methods is implemented. However I do foresee some issues concerning pausing,
canceling file transfer etc. But the basic functionality should be completed
in reasonable amount of time.(1  week for reading and understanding what I
must know, 2 - 3 weeks for coding) Once they are completed, through tests
should be conducted.

  I also want to add some additional features that might help to enhance
user experience:

  First: file transfer status in chat window. This has already been
discussed in #1205 . Some for it some
against it. I think it's good to give users an option anyway, and it's
really a simple and useful way to track which file comes from which guy,
when you are receiving files from several friends. I do not intend to
implement a progress bar in chat window. A notice about the start, end of
the file transfer should be enough. Also the end of FT notice should contain
a link to the file. This feature will be optional, users against this will
be able to disable it. FT from users you are not chatting with does not open
a chat window. What about the current pop-up notice? Well, I think it just
feels better when you can focus your eyes around the chat window and know
everything that's happening between you and the guy you're talking to.
Suggestions welcomed about this!

  Second: drag and drop currently works perfectly, but sending of multiple
files are not supported. Users would typically archive the files or manually
send them one by one. This is not quite user friendly. I want to support
dragging a directory into gajim and letting gajim do the archiving stuff.
Also the description field will be populated with a list of files(possibly
incomplete due to the length of it). This idea is due to
#4287.
I want to clarify that this is a user experience enhancement, so the user
should be noted that what the receiver gets is an archive of the directory
that he/she sends.

  If time allows I also want to add file encryption feature, which might be
as simple as key exchange and encryption using DES or AES. But I don't know
how to standardize this feature so it understood by all xmpp clients. I
appreciate any advice on this.

  A bit more about myself:
  I'm pursuing a master of science degree in Fudan University, Shanghai. I'm
comfortable with C, C++, perl. And I love Python and lisp! I have some basic
experience with the linux kernel, netsurf(a small web browser, I once
implemented "read page" feature for fun). I am a heavy user of all kinds of
open source softwares, I used to write webapps using django and wrote some
fun stuff in it, like sending SMS via my homepage(click and send sms to me)
etc. I primarily use python to solve all problems that I meet. It's my
language of choice.

  My name is Li Zhenchao .My jabber account is fantastic...@jabber.com. I
have a blog: www.fantasticsid.com/blog (it's in Chinese but I will write in
English if I'm lucky enough to write reports on this year's gsoc project
later). You can contact me at cockneyke...@gmail.com or
fantastic...@fantasticsid.com.

  Any advices, thoughts, suggestions, questions are welcomed!

  PS: After discussing my idea here, should I also present it to the XSF
mailing list?

--
Homepage:   www.fantasticsid.com
EMAIL: fantastic...@fantasticsid.com
IRC: fantasticsid
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

[Gajim-devel] GSOC mentors

[Yann Leboulanger]

Hi,

This mail is for Gajim developers: is anyone interested in mentoring a 
student during this Google summer of code? about which project? we have 
some in the wiki, you can thing about others, or about xmpppy...


--
Yann
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

Re: [Gajim-devel] GSOC, redux

[Yann Leboulanger]

Brendan Taylor wrote:
> In the GC today Arc Riley asked if we wanted to get into GSoC under the
> umbrella of the Python Software Foundation. He's the PSF's coordinator
> this year, and says that we may be able to get a slot there (assuming
> there's a decent student proposal).
> 
> Do we want this? Do we have any good project ideas?
> 
> PS. he had some great encouragement too:
> 
> Arc Riley: btw, personal note, you guys have done a fantastic job with
> gajim.  the XMPP console has been invaluable for debugging concordance
> extensions

Wow that's very a generous proposition!

We have some interesting propositions, see [1], and have also some
interested students. At least 2 students contacted me directly for that.
I'll see if they are still interested.

[1] http://trac.gajim.org/wiki/GajimGSOC

Thanks Arc
-- 
Yann
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

Re: [Gajim-devel] GSOC, redux

[Brendan Taylor]

In the GC today Arc Riley asked if we wanted to get into GSoC under the
umbrella of the Python Software Foundation. He's the PSF's coordinator
this year, and says that we may be able to get a slot there (assuming
there's a decent student proposal).

Do we want this? Do we have any good project ideas?

PS. he had some great encouragement too:

Arc Riley: btw, personal note, you guys have done a fantastic job with
gajim.  the XMPP console has been invaluable for debugging concordance
extensions


pgp8R8SukncXg.pgp
Description: PGP signature
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

Re: [Gajim-devel] GSOC

[Yann Leboulanger]

Mateusz Biliński wrote:
> On Sun, Mar 1, 2009 at 10:28, Yann Leboulanger  wrote:
>> Bad news :(
>>
>> But indeed I don't plan to apply as an organisation. So no GSOC for this
>> year. I'll pay the server myself.
> 
> I don't know what's the server cost but I guess we could do some kind
> of fund-raising. Maybe via paypal?
> 

I pay 35€ per month, but I also use it for my personal usage, so it's
normal I pay it.

Of course I'm not against some help, but I want to be honest, it's not
vital.
-- 
Yann
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

Re: [Gajim-devel] GSOC

[Mateusz Biliński]

On Sun, Mar 1, 2009 at 10:28, Yann Leboulanger  wrote:
> Bad news :(
>
> But indeed I don't plan to apply as an organisation. So no GSOC for this
> year. I'll pay the server myself.

I don't know what's the server cost but I guess we could do some kind
of fund-raising. Maybe via paypal?

-- 
Mateusz Biliński
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

Re: [Gajim-devel] GSOC

[Yann Leboulanger]

Mateusz Biliński wrote:
> On Thu, Feb 12, 2009 at 20:35, Yann Leboulanger  wrote:
>> Mateusz Biliński wrote:
>>> Might come in handy: Alek's blog posts about implementing server-side
>>> history in Psi during GSoC 2008:
>>> http://alek.silverstone.name/en/tags/soc
>> I started a wiki page for GSOC. See http://trac.gajim.org/wiki/GajimGSOC
> 
> http://blog.xmpp.org/?p=297
> 
> I guess this pretty much ends discussion for this year's GSoC.
> 
> Unless Gajim plans to apply as separate org, which IMHO is not a good
> idea currently, keeping in mind all the program administration tasks.
> 

Bad news :(

But indeed I don't plan to apply as an organisation. So no GSOC for this
year. I'll pay the server myself.

-- 
Yann
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

Re: [Gajim-devel] GSOC

[Mateusz Biliński]

On Thu, Feb 12, 2009 at 20:35, Yann Leboulanger  wrote:
> Mateusz Biliński wrote:
>> Might come in handy: Alek's blog posts about implementing server-side
>> history in Psi during GSoC 2008:
>> http://alek.silverstone.name/en/tags/soc
>
> I started a wiki page for GSOC. See http://trac.gajim.org/wiki/GajimGSOC

http://blog.xmpp.org/?p=297

I guess this pretty much ends discussion for this year's GSoC.

Unless Gajim plans to apply as separate org, which IMHO is not a good
idea currently, keeping in mind all the program administration tasks.

-- 
Mateusz Biliński
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

Re: [Gajim-devel] GSOC

[Yann Leboulanger]

Mateusz Biliński wrote:
> Might come in handy: Alek's blog posts about implementing server-side
> history in Psi during GSoC 2008:
> http://alek.silverstone.name/en/tags/soc
> 

I started a wiki page for GSOC. See http://trac.gajim.org/wiki/GajimGSOC

-- 
Yann
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

Re: [Gajim-devel] GSOC

[Mateusz Biliński]

Might come in handy: Alek's blog posts about implementing server-side
history in Psi during GSoC 2008:
http://alek.silverstone.name/en/tags/soc

-- 
Mateusz Biliński
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

Re: [Gajim-devel] GSOC

[Yann Leboulanger]

Jonathan Schleifer a écrit :

Am 09.02.2009 um 07:04 schrieb Yann Leboulanger:


codecs are handled by gstreamer, not sure we do a lot about that. I CC
to liori who may have ideas about that.


Ok, then another project would be to get all that running on Windows. 
I'm not so sure how well gstreamer runs on Windows, a while ago, it 
wouldn't run there at all IIRC.


Yep that could be a project, but maybe for next year. We don't have it 
under linux yet.



Gajim is not an organisation.


It doesn't need to be. Last year, Audacious participated in the GSoC as 
well and we're not an organization there either AFAIK. :)



And I don't have to time to manage all that process.


Well, I don't know how much work it is to manage that, I have to admit. 
But is it really more than having a list of projects, assigning mentors 
to them and looking that the mentors do their jobs?


+ handle thing with google (payment for example). Really I don't have 
time for that. It doesn't take 5 minutes. And I thank Peter Saint-Andre 
a lot for that.


--
Yann
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

Re: [Gajim-devel] GSOC

[Jonathan Schleifer]

Am 09.02.2009 um 07:04 schrieb Yann Leboulanger:


codecs are handled by gstreamer, not sure we do a lot about that. I CC
to liori who may have ideas about that.


Ok, then another project would be to get all that running on Windows.  
I'm not so sure how well gstreamer runs on Windows, a while ago, it  
wouldn't run there at all IIRC.



Gajim is not an organisation.


It doesn't need to be. Last year, Audacious participated in the GSoC  
as well and we're not an organization there either AFAIK. :)



And I don't have to time to manage all that process.


Well, I don't know how much work it is to manage that, I have to  
admit. But is it really more than having a list of projects, assigning  
mentors to them and looking that the mentors do their jobs?



PS: I replied to the list, even if you message was only to me.


Yeah, thanks. Apple Mail often screws that up. I always have to add  
the list manually.


--
Jonathan



PGP.sig
Description: Signierter Teil der Nachricht
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

Re: [Gajim-devel] GSOC

[Yann Leboulanger]

Jonathan Schleifer wrote:

> I think adding FT should be pretty easy, whereas I think video might be
> a lot of work due to the codecs etc.

codecs are handled by gstreamer, not sure we do a lot about that. I CC
to liori who may have ideas about that.

> So, maybe, we could as well participate in the GSoC and not only XMPP?
> There's no requirement for us to be only part of the XMPP projects, we
> could be a project as well in the GSoC.

Gajim is not an organisation. And I don't have to time to manage all
that process.

PS: I replied to the list, even if you message was only to me.
-- 
Yann
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

Re: [Gajim-devel] GSOC

[Yann Leboulanger]

Jonathan Schleifer wrote:
> Am 08.02.2009 um 16:30 schrieb Yann Leboulanger:
> 
>> Esessions: already in release, and I don't see any bug.
> 
> It's a very basic implementation, not yet complete. But it's going to go
> away anyway due to XTLS, I guess.

what is not complete?

>> Jingle: Liori will work on that after his exams, in some weeks.
> 
> That's only Jingle Audio, I guess? Then we could have Jingle Video and
> Jingle File Transfers as a project.

Yep that are projects I thought about. But I don't know if it's hard and
long enough for a GSOC project (3 months full time job).As we'll already
have jingle for voice, is adding FT or video hard? I cannot answer.

>> Plugin system: I'll talk about that with Vardo. Maybe on wednesday.
> 
> Ok.
> 
>> But that doesn't mean someone else cannot work on a new feature. I
>> cannot do everything in this project: discuss and fix tickets, implement
>> new features. I need help. So anything that can bring help (GSOC for
>> example) is very welcome.
> 
> I don't think only introducing new features is a good idea. We could
> have a SoC project which is to fix bugs, do API changes, rewrite some of
> the code that needs a rewrite. We could have a project that is to
> separate the GUI code from the XMPP code (it's sometimes mixed in
> common/) etc. There can be project that help Gajim, but don't introduce
> new features before the others are stable.

I don't think such a project would be accepted by XSF members. Plugin
system last year was hardly discussed, and not easily accepted.

Not that what you said is useless, of course! But it doesn't promote
XMPP directly, showing new feature that can be done thanks to XMPP.

-- 
Yann
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

Re: [Gajim-devel] GSOC

[Yann Leboulanger]

Jonathan Schleifer wrote:
> Am 08.02.2009 um 14:55 schrieb Yann Leboulanger:
> 
>> Hi all,
>>
>> GSOC will arrive in the next monthes, so we can start thinking about
>> project we could propose.
>>
>> So what new features would you like to see in Gajim ?
>>
>> We can discuss that on Wednesday at the end of the meeting.
> 
> I think we should not introduce any new features for now, but finish and
> stabilize the features introduced in the last SoCs. I'm mainly talking
> about Jingle, the Plugin System and ESessions here.

Esessions: already in release, and I don't see any bug.
Jingle: Liori will work on that after his exams, in some weeks.
Plugin system: I'll talk about that with Vardo. Maybe on wednesday.

But that doesn't mean someone else cannot work on a new feature. I
cannot do everything in this project: discuss and fix tickets, implement
new features. I need help. So anything that can bring help (GSOC for
example) is very welcome.

-- 
Yann
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

Re: [Gajim-devel] GSOC

[Jonathan Schleifer]

Am 08.02.2009 um 14:55 schrieb Yann Leboulanger:


Hi all,

GSOC will arrive in the next monthes, so we can start thinking about
project we could propose.

So what new features would you like to see in Gajim ?

We can discuss that on Wednesday at the end of the meeting.


I think we should not introduce any new features for now, but finish  
and stabilize the features introduced in the last SoCs. I'm mainly  
talking about Jingle, the Plugin System and ESessions here.


--
Jonathan



PGP.sig
Description: Signierter Teil der Nachricht
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

Re: [Gajim-devel] GSOC

[Jonathan Schleifer]

Am 08.02.2009 um 14:55 schrieb Yann Leboulanger:


Hi all,

GSOC will arrive in the next monthes, so we can start thinking about
project we could propose.

So what new features would you like to see in Gajim ?

We can discuss that on Wednesday at the end of the meeting.


I think we should not introduce any new features for now, but finish  
and stabilize the features introduced in the last SoCs. I'm mainly  
talking about Jingle, the Plugin System and ESessions here.


--
Jonathan



PGP.sig
Description: Signierter Teil der Nachricht
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel

[Gajim-devel] GSOC

[Yann Leboulanger]

Hi all,

GSOC will arrive in the next monthes, so we can start thinking about
project we could propose.

So what new features would you like to see in Gajim ?

We can discuss that on Wednesday at the end of the meeting.
-- 
Yann
___
Gajim-devel mailing list
Gajim-devel@gajim.org
http://lists.gajim.org/cgi-bin/listinfo/gajim-devel