Re: Presentation + A debian-based for audio creation and production, stage technics and video blend (or "the future of TangoStudio")
Hi,
On Tue, Nov 13, 2012 at 10:25:10AM +0800, Paul Wise wrote:
> Conflicts are a bad idea since you force people to not use things even
> if they want to. In the case of pulseaudio it sounds like they just
> paper over a bug. As Adrian says, that bug (pulseaudio/jack
> interaction) should be fixed instead of papering over it with
> Conflicts.
I'd definitely call using Conflicts to hide a bug a missuse (even if
there would be some way to inject Conflicts which as I said is not.)
> I don't know how to make the phrase "Debian Pure Blend" any clearer,
> perhaps we should just use "Debian" instead. Jonas's presentations on
> Debian Pure Blends at FOSSASIA and other places would probably help
> but I can't find any videos, some links to slides here:
>
> http://wiki.jones.dk/DebianAsia2011
My talk at DebConf 11:
http://ftp.acc.umu.se/pub/debian-meetings/2011/debconf11/low/711_Blending_Debian.ogv
My archive of talks (several of them Blends related)
http://people.debian.org/~tille/talks/
Frostbyte interview (Jonas and me together explaining Blends)
http://traffic.libsyn.com/frostbitemedia/TWID_017.ogg
> DebianEdu is not yet a "Debian Pure Blend", they are still a
> derivative for now. They have a separate website, separate
> donations/funding stream, separate release schedule, separate ISO
> images, some modified packages and maybe more.
I tried to save some time and shortcutted this in my previous mails.
There are people who call Debian Edu a Blend and SkoleLinux (the product
around featuring ISO images on non-Debian servers) as something else.
I do not see a point to blur the main information that you can perfectly
cherry-pick from their work with discussion about this - please use a
different thread if you feel a need for more discussion about this.
> There is no guarantee that good (multimedia) software will remain in
> Debian and QA folks remove truckloads of software every year.
Right, there is no such guarantee. I should have better said: If you
manage to attract a solid developer base around multimedia application
inside Debian (one part of the work to run a Blend) you have pretty good
chances to hold good software in. I tried to demonstrate this fact by
pointing to graphs about the number of Debian Med developers and the
number of packages they are maintaining. The packages in principle are
similar to multimedia in so far that they are very specific for some
topic and the average user has no real development skills. In contrast
to something separate from Debian this way has to be proven as
sustainable for >10 years and we never lost any package (for sure we
intentionally droped some outdated / bad programs.)
> The only
> thing that ensures that software remains in Debian is when people put
> effort into keeping it there.
Yes. And to make sure people are doing this you need to assemble all
interested people in a team - and the Blends idea is formalising and
simplifying doing this.
> For a specialised types of software like
> media production, the set of people who want to use it does not
> intersect much with the set of people who have the skills to package
> it, leading to things like cinepaint leaving Debian, people trying to
> bring it back and then giving up.
I agree that the set of developers in those special fields is lower -
one more reason to focus them onto a common goal. Regarding skills we
try to train new people (basically developers without packaging skills).
I'm currently mentoring the fourth student in the MoM[1] effort I
started this year. Even if there was no student every month we
approached things we would never have managed without this.
> The multimedia team doesn't do sponsoring but does do team maintainence:
>
> http://wiki.debian.org/DebianMultimedia/Sponsoring
Uhmmm, this sucks, sorry. I understand that a lack of manpower might
force people to concentrate on things they regard as most important.
However, if I try to measure my Debian Med *packaging* work I think it
increased to about 60%-70% of sponsoring. As a proof of this concept
I would like you to have a look at
https://wiki.debian.org/DebianMed/Developers
which shows how many DDs joined Debian *only* *because* Debian Med
existed. This is only because we continuosely tried to involve more
people and one way to do so is sponsoring. On average we dragged one
developer per year into Debian (not counting people who currently are
only sponsees). I wished other teams would understand the power behind
this concept.
> http://wiki.debian.org/DebianMultimedia/Join
>
> d-i doesn't yet allow choice of desktop and I read somewhere that
> d-i/tasksel folks wanted to limit desktop choices but maybe they would
> be open to a "specialised tasks" sub-menu or something that would
> contain the non-default desktops plus all the blends.
I once tried to inject Blends into tasksel (#186085) but failed and did
not yet tried again (and also did not followed recent discussion on this
- shame on me.) I'd be very happy about such a sub-menu option. Do you
have any links?
> For people who want to stick with specific older versions, we have
> snapshot.debian.org.
Right, but I think we need to make picking from there a bit more handy
from a Blends perspective (I do not have a proper idea how, but it seems
to be a reasonable feature.)
Kind regards
Andreas.
[1] https://wiki.debian.org/DebianMed/MoM
--
http://fam-tille.de
Reply to: