[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Gnewsense-dev] Shared gNewSense/Trisquel repo (WAS: Re: GNU C-Graph: Ov
From: |
Karl Goetz |
Subject: |
[Gnewsense-dev] Shared gNewSense/Trisquel repo (WAS: Re: GNU C-Graph: Over 100, 000 Discover That Visualizing Convolution is Easy) |
Date: |
Tue, 01 Oct 2013 21:29:52 +1000 |
User-agent: |
K-9 Mail for Android |
Hugh Mayfield <address@hidden> wrote:
> On Mon, 30 Sep 2013, Karl Goetz wrote:
>
> > Date: Mon, 30 Sep 2013 09:23:03 +1000
> > From: Karl Goetz <address@hidden>
> > To: Hugh Mayfield <address@hidden>
> > Cc: Tobias Platen <address@hidden>,
> address@hidden
> > Subject: Re: [Gnewsense-dev] GNU C-Graph: Over 100,
> > 000 Discover That Visualizing Convolution is Easy
> >
> > Hugh Mayfield <address@hidden> wrote:
> >> On Mon, 30 Sep 2013, Karl Goetz wrote:
> >>
> >>> Date: Mon, 30 Sep 2013 08:23:35 +1000
> >>> From: Karl Goetz <address@hidden>
> > > > [Think i accidentally cut out whoever said this next bit, sorry]
> >>>> I think it is worth to set up a common repository for both Trisquel an
> >>>> Gnewsense that contains the latest stable releases of GNU software.
> >>>
> >>> Maintaining a repository like that is a lot of work. It would
> >> definitely be benefitial to all but time to maintain it will be the
> >> proble.
> >>
> >> Is this something I could do?
> >
> > Quite possibly, if you can provide the following:
> > * enough system resources to build packages
>
> Possibly. I've got a brace of multi-GHz machines. I don't know if
> that's enough?
A single system at 2GHz+ would be a good minimum, if you have multiple GB of
ram (4+) so much the better. As with all things you can get away with less and
builds will be slower - more will speed things up.
> > * reasonable network location for users to download from
>
> Definitely can't provide this. Cannot afford it.
It would be a requirement for such a repository to be properly useful if the
packages are to be binaries ready to install. If you plan to only provide
source packages you might be able to get away with just packaging metadata and
in that instance you would have a much smaller data requirement.
At the moment for this plan it seems data transfer is your stumbling block but
if you were to start a project like this you might be able to team up with
someone else who can provide it.
I can't speak for gNewSense/sam but I suspect if you needed somewhere to test
build some packages we could find you some room (on an adhoc basis).
> > * time to build and test the packages on the supported releases of
> the supported distros
> > * time to patch packages and update to newer versions.
> > * time to do any other supporting bits of work (eg keepingthe build
> environments up to date)
>
> Time is not a problem.
In that case you can take your time and learn how to start this project at your
own pace. To start with you could even try and package c-graph (if you know
fortran or are interested in learning it so much the better). Since upstream is
asking for it to be included in gNewSense I'm sure he will be happy to help
with questions regarding how the application works and any patches/changes you
might have questions about. Packaging questions can be asked on IRC or various
mailing lists (though which lists will depend on who and what you are working
with :))
thanks,
kk
Figured this deserves a new subject line.
- [Gnewsense-dev] Shared gNewSense/Trisquel repo (WAS: Re: GNU C-Graph: Over 100, 000 Discover That Visualizing Convolution is Easy),
Karl Goetz <=