* university group willing to contribute to GUIX
[not found] <87k3dhchxj.fsf@gnu.org>
@ 2014-02-01 17:53 ` Quiliro Ordóñez Baca
2014-02-02 18:25 ` Ludovic Courtès
2014-02-03 18:10 ` Quiliro Ordóñez Baca
0 siblings, 2 replies; 6+ messages in thread
From: Quiliro Ordóñez Baca @ 2014-02-01 17:53 UTC (permalink / raw)
To: guix-devel
[-- Attachment #1.1: Type: text/plain, Size: 13509 bytes --]
Hi,
This a conversation held on several mails with Civodud over the
collaboration of 9 volunteers for GUIX Distro:
-------- Mensaje original --------
From: ludo@gnu.org (Ludovic Courtès)
To: Quiliro Ordóñez Baca <quiliro@congresolibre.org>
Hi Quiliro,
Quiliro Ordóñez Baca <quiliro@congresolibre.org> skribis:
> There is a university that wants to make their own distro. I suggested
> to use GUIX. They offered to pay me to help with the project. But I told
> them I am starting my time bank. I they want my collboration, I would
> ask for them to pay me in hours. I propose the same thing to you. These
> people will collaborate in your project in exchage for your help. If you
> have another proposal, please tell me.
>
> I suggested them that they do not fork. I suggested they make their own
> ISO and desktop theme and to have a local mirror of GUIX packages. That
> way GUIX helps them and they help GUIX. They will incorporate the
> following team:
>
> * 6 students that know how to program in C, C++ and Java (40 working
> weeks, 5 days per week, 2 hours per day = 2400 person/hours)
> * 1 student from graphic design (time allotment to be negotiated)
> * 1 student from management (time allotment to be negotiated)
> * 1 student from communications (not telecommunications) (time
> allotment to be negotiated)
> * the directors of those 3 departments will help coordinate
>
> Are you willing to help us?
Yes, but only in the same way that I help anyone who tries to get
involved, as this is purely volunteer work. So this is really best
effort, so to speak.
I think this is a great project. However, getting young students
started to work on such a project probably requires a fair amount of
mentoring: probably they’ll have to learn about how free software
hackers communicate, how patch submission and review work, what the
software stack looks like, etc. I don’t think I can spend this much
effort, as this would effectively prevent me from doing anything else.
If you do the heavy lifting, then that may be easier, of course.
What do you think?
> I can do the heavy-lifting. I have some gaps about the things you
> mention. If you send links to the docs I can deal with that.
>
> The big question is: do you think, based on the knowledge level I have
> told you they have (programming in C, C++ and Java), the students will
> be able to have the distro ready for use of end users in that timeframe
> (1 year)?
Short answer: probably not. As discussed before, two things to
consider:
1. Students will most likely discover how free software development
works, which means they will probably not be “productive” at least
in the first few months. The goal should be to “get them started
with free software development”, rather than to “build a distro”
(BTW, note that Guix is 1.5 year old, to give an idea of the
effort.)
2. Work on the distro happens anyway. By “end users”, I think you
mean with a fancy desktop à la GNOME. That work may happen in the
coming year anyway, but at the same time, most early contributors
are hackers with little interest in such things.
Hope this helps.
Please think through it, and follow-up on guix-devel@gnu.org when you
have clearer ideas.
Depending on what you really want to do, it might be that students would
be better off contributing to a well-established distro like Trisquel.
Thanks,
Ludo’.
Here is a conversation held on IRC #guix channel:
(15:18:09)*Quiliro:*a university is playing with my suggestion to
collaborate with guix
(15:18:21)*Quiliro:*not with money but with 6 it students
(15:18:32)*Quiliro:*oh fosdem!
(15:18:39)*Quiliro:*live streaming?
(15:19:10)*Quiliro:*and 1 management, 1 art and 1 communication student
(15:19:27)*Quiliro:*10 hours per week....40 weeks
(15:19:52)*viric:*luckily students don't need money
(15:21:41)*Quiliro:*viric: haha...yes they do...but they need the
diploma more
(15:44:05)*Steap:*Students only need coffee.
(15:44:14)*Steap:*Pretty much like programmers.
(15:44:37)*Steap:*quiliro: what would the non-IT students do ?
(15:48:08)*Quiliro:*Steap: the management student will find the needs of
the different areas and what to include in specific groupings of
packages....for example: what does a the hotel and tourism professional
will need in his distro.....s/he will also see that the team works as a
team (the other students)
(15:49:30)*Quiliro:*the communications student will work with the art
student to see that the presentation and accessibility is good....and he
will also make a good name for the distro with press coverage and the likes
(15:50:21)*Quiliro:*the art/design student will make the artistic
par..icons, art of the distro, wallpapers, fonts, etc
(15:50:29)*Quiliro:*that is my idea
(15:50:41)*Quiliro:*what do you think? any other areas?
(15:52:01)*Quiliro:*so that will be 2400 IT hours and 1200 non-IT for GUIX
(15:52:40)*Quiliro:*3600 hours...of course they will cost...the
knowledge we will give them is not gratis
(15:53:03)*Quiliro:*it is free but not gratis
(15:53:11)*Quiliro:*it is libre
(15:57:53)*Steap:*call me a grumpy hacker if you like, but I'm a bit
afraid of waht the "communication" student might do
(15:58:17)*Steap:*Because we do not try to force-feed people with Guix :)
(15:58:32)*Steap:*And well, management is pretty much useless.
(15:59:04)*Steap:*The art student would probably the most useful, since
developers aren't usually extremely gifted when it comes to designing
logos and such
(16:01:48)*mark_weaver:*quiliro: it would probably be best to talk about
this over email, on the guix-devel@gnu.org mailing list. I know that
civodul prefers to communicate that way for most things.
(16:04:48)*Steap:*Yes, such important matters should probably be
discussed over email, since not everybody idles on IRC 24/7
(16:12:23)*civodul:*yes, good idea
(16:14:05)*Quiliro:*ok...the idea is to make a free distro....is devel
the right list?
(16:14:40)*Quiliro:*or gnu-linux-libre
(16:14:49)*civodul:*hey quiliro
(16:15:00)*civodul:*quiliro: if the project is specifically about using
Guix, the guix-devel is the right list, yes
(16:15:03)*Quiliro:*bonsoir civodul
(16:15:09)*civodul:*bonsoir ;-)
(16:15:30)*Quiliro:*yes...the guix 100% free distro
(16:16:50)*civodul:*right, so guix-devel is the place
(16:17:09)*civodul:*it's the only mailing list actually, so don't be
afraid just because it has "devel" in it ;-)
(16:17:11)*Quiliro:*cool
(16:17:17)*Quiliro:*haha
(16:17:59)*Quiliro:*i think guix is the friendliest developer distro for
the way it packages the programs
(16:18:11)*Quiliro:*for the little i know about programming
(16:18:28)*Quiliro:*but it loks friendly
(16:18:39)*Quiliro:*i saw the august presentation
(16:18:58)*civodul:*heh
(16:19:18)*Quiliro:*The big question is: do you think, based on the
knowledge level I have told you they have (programming in C, C++ and
Java), the students will be able to have the distro ready for use of end
users in that timeframe (1 year)?
(16:19:42)*Quiliro:*it should have a desktop
(16:19:50)*Quiliro:*:-D
(16:20:36)*Quiliro:*in "normal-people" speak (as my ex-girlfriend used
to tell me to talk to her)
(16:24:49)*civodul:*quiliro: probably not, but i do hope the rest of us
will have it "ready" by then ;-)
(16:25:12)*civodul:*for students it should be thought of as a way to get
familiar with free software development in general
(16:25:18)*civodul:*not as a way to build a product
(16:25:23)****civodul*goes back preparing his talk
(16:27:09)****Steap*hates the word "product"
(16:27:40)*mark_weaver:*+1
(16:27:51)*Steap:*quiliro: it really depends upon their background, I think
(16:28:04)*Steap:*When I was a student, most of the IT students hated IT :)
(16:28:28)*Steap:*Needless to say they were not that good.
(16:28:55)*civodul:*mark_weaver: me too, but i must have been too much
exposed to IT marketing newspeak :-)
(16:30:47)*Quiliro:*it is very important that it is usable by the end
user in 1 year....what is necessary for that to occur?
(16:31:06)*Quiliro:*will the students be of help or will they be a load?
(16:31:56)*Quiliro:*i wanted them to come to guix because i want guix to
grow
(16:32:20)*Steap:*quiliro: packaging a DE would be the biggest part
(16:32:33)*Quiliro:*but if they will not help, it is better for me that
they contribute somewhere else
(16:32:36)*Quiliro:*ok
(16:33:02)*Steap:*also, even with a DE, it's not gonna be as friendly as
Ubuntu within a year
(16:33:23)*Quiliro:*why?
(16:33:45)*Steap:*well, we'd need a nice GUI such as Synpatic on top of
the package manager
(16:33:51)*Steap:*Synaptic*
(16:34:06)*Quiliro:*don't worry about that part
(16:34:15)*Quiliro:*the important part is the usability
(16:34:16)*Steap:*and it's not easy to imagine what the distro is gonna
be in a year
(16:34:24)*Quiliro:*not installing new packages
(16:34:26)*Steap:*quiliro: what do you mean by that ?
(16:34:36)*Steap:*quiliro: I can already use the package manager
(16:34:42)*Steap:*I know other people do
(16:35:54)*Quiliro:*that the user can write emails, write text
documents, spreadsheets, save and list files and use several other
programs from the gui (with pretty little icosn,haha)
(16:36:22)*Quiliro:*yes , it is important but that is more important for
the project that university is proposing
(16:36:34)*Steap:*well, we can do that in text :)
(16:36:49)*Steap:*the GUI part, on the other hand...
(16:37:07)*Steap:*does the distro have to wokr well, or would a package
manager be enough ?
(16:37:16)*Quiliro:*Steap: yes it can be done in text, but that will
defeat that project objective
(16:37:51)*Quiliro:*Steap: what do you mean? it is the distro that is needed
(16:38:22)*Steap:*quiliro: they could work on the GUI part using the
package manager on top of a Debian system, for instance
(16:39:18)*Quiliro:*no...the idea is that they want to make a distro...i
told them not to make a fork but contribute to guix
(16:39:33)*Quiliro:*but they have to show end user results
(16:39:40)*Quiliro:*not console
(16:39:59)*Steap:*I see.
(16:40:17)*Quiliro:*unless that serves other purposes....what i mean is
that they have to make something something that someone needs
(16:40:50)*Quiliro:*but i want to use that to help freedom -> free
distro -> guix
(16:41:06)*Quiliro:*-> guix package manager
(16:41:23)*Quiliro:*how about gpm?
(16:41:31)*Steap:*gpm ?
(16:41:44)*Quiliro:*guix package manager -> GPM
(16:41:52)*Steap:*yeah
(16:42:00)*Steap:*well, the GPM is already usable
(16:42:03)*Quiliro:*but i do not want to get into that
(16:42:10)*Quiliro:*i know
(16:42:11)*Steap:*it just lacks end-user packages
(16:42:17)*Steap:*so they could work on that
(16:42:23)*Quiliro:*ok
(16:42:47)*Quiliro:*you work on making it bootable and they work on end
user packages?
(16:43:03)*Quiliro:*but someone has to learn first
(16:43:12)*Steap:*well, packaging stuff is "easy"
(16:43:15)*Quiliro:*i will be the one to teach them
(16:43:25)*Quiliro:*so....
(16:43:33)*Quiliro:*i saw the video
(16:43:39)*Quiliro:*it looks very easy
(16:43:48)*Quiliro:*civodul: ^
(16:43:52)*Steap:*well, you "just" have to fix errors one by one
(16:44:02)*Quiliro:*what errors?
(16:44:08)*Quiliro:*compilation?
(16:44:14)*Steap:*well, let's say you want to package "foo"
(16:44:26)*Steap:*there migth be compilation errors
(16:44:29)*Quiliro:*is there some doc about that?
(16:44:31)*Steap:*errors when running configure
(16:44:32)*Steap:*etc.
(16:44:46)*Steap:*and there's no real doc, because it depends on the
software you're packaging
(16:45:08)*Steap:*so, when things go wrong, this can take quite a lot of
time to fix
(16:45:39)*Quiliro:*who to ask when errors present?
(16:46:07)*Steap:*well, you have to understand what causes them
(16:46:11)*Quiliro:*where should someone start if they want to install
and then contribute to guixx?
(16:46:14)*Steap:*go through the source code of your package
(16:46:15)*Steap:*etc.
(16:46:20)*Steap:*well
(16:46:25)*Steap:*1) Install Guix from git
(16:46:29)*Steap:*2) Build "hello"
(16:46:29)*Quiliro:*i never had a problem compiling
(16:46:42)*Steap:*3) try to package something that is not yet in Guix
(16:46:46)*Quiliro:*unless i changed something in configure
(16:46:55)*Quiliro:*or missed a dependency
(16:47:20)*Quiliro:*i think i saw those steps somewhere
(16:48:20)*Steap:*they may want to try that to see if they're interested
in the project
(16:50:01)****Quiliro*is reading http://www.gnu.org/software/guix/manual/
(16:51:22)*Quiliro:*Steap: how?
(16:51:34)*Quiliro:*show them civodul's video?
(16:52:31)*Steap:*sure
(16:52:34)*Steap:*or read a recipe
(16:52:37)*Steap:*copy/paste it
(16:52:38)*Steap:*modify it
(16:52:40)*Quiliro:*teach them to install guix distro?
(16:52:41)*Steap:*see what happens
(16:52:48)*Quiliro:*ok
(16:52:48)*Steap:*come here for help
(16:52:50)*Steap:*etc.
(16:52:54)*Quiliro:*thank Steap
(17:30:19)*rgc:*Steap: I see you're the support guy :)
(18:50:57)*adadfd:*quiliro: re packaging, check this out:
https://gnu.org/software/guix/guix-ghm-andreas-20130823.pdf
(20:16:35)*Quiliro:*adadft left the chatroom
(20:17:03)*Quiliro:*is there a video with the presentation he did with
that pdf?
(20:27:49)*zerwas:*I don't think so
[-- Attachment #1.2: Type: text/html, Size: 38287 bytes --]
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 555 bytes --]
^ permalink raw reply [flat|nested] 6+ messages in thread