Re: IMPORTANT: Do live Debian images have a future?
Hi!
Steve McIntyre:
> While the bugs are annoying, what worries me more is that they were
> only spotted in release builds. There had been testing versions of
> live images available for multiple weeks beforehand, presumably with
> the same bugs included. (Almost) none of them reported. This shows
> that we don't have enough people using these live images and/or caring
> about filing bugs.
I have a slightly different PoV on the situation here, that leads me
to different conclusions.
I had a good look at the list of bugs you pointed to, and frankly I'm
not surprised at all that these bugs were not reported earlier: very
few of them affect in a user-visible manner the use case I *guess* is
the most common one, i.e. booting a desktop Debian Live system to give
the next version a try without installing to disk. For example, we
can't realistically expect Debian + GNOME users to even notice things
like "Incorrect Volume ID used for all live images" or "not
configuring UTF-8 console correctly".
I acknowledge that those are bugs, but if nobody reported them, it's
perhaps because in practice, nobody was affected negatively enough to
bother reporting, rather than because not enough people even tested it
⇒ the perceivable quality of our Live images, from a user's PoV, is
perhaps not *that* bad :)
FTR we did test the GNOME live ISO a lot during the "more than
a Debian BSP" event in Paris back in May, and the only one we noticed
from your list was #865015… that we erroneously attributed to faulty
hardware… sorry! We also reported a number of other bugs against the
debian-live package, and they have not been acted upon so far.
So from where I stand, the lack of developer time available to resolve
issues *that users do notice* looks like a more severe problem than
the lack of testing in itself.
The good news is that despite these bugs, the live images we tested
worked pretty well :)
> We have a similar lack of involvement in terms of the content of the
> live images. […] Validation by the various desktop teams would be
> useful here.
Fully agreed.
IMO "building & distributing a Debian Live image for desktop X" should
be put explicitly as something that the team for desktop X can opt-in
for if, and only if, they feel comfortable committing to validate the
content somewhat. And then, whenever you feel they don't meet the
expectations you've set, then you can drop the problematic ones so the
problem is visible.
Is there technical difficulty involved in adding to / removing from
the list of desktops for which we build Live images? I hope not, but
who knows :)
> The current situation is *not* good enough. [...]
It looks like you feel trapped in a role you did not choose. I feel
sorry for this and I want to express support for your attempt
to escape.
> If our live images are going to be good enough to meet the standards
> that Debian users deserve and expect, we need *consistent*,
> *sustained* involvement from a lot more people.
ACK
> Please tell me if you're going to help.
(What follows was decided only a dozen days ago and I was waiting for
this decision before I replied here, hence the delay.)
By the end of the year I'll be able to tell whether I have a new,
strong incentive to put efforts into automatic testing of Debian Live
GNOME images: at Tails we will decide later this year whether we try
basing our product on quarterly snapshots of Debian testing (to start
with: at least until the end of the Buster cycle). If we do so, then
I bet that we (Tails) will want to invest into identifying bugs ASAP
i.e. in Debian testing/sid and more specifically in Debian Live GNOME
images built from Debian testing/sid.
I recognize this can replace neither involvement from desktop teams,
nor exploratory testing by humans.
> If we don't see a radical improvement soon, I'll
> simply disable building live images altogether to remove the false
> promises they're making.
This sounds entirely reasonable. I'm all for our Debian heroes &
martyrs going on strike, and for not hiding problems! Take care, and
thanks for making this move :)
Cheers,
--
intrigeri
Reply to: