[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Bug#133578: gdm bug #133578. Intend to NMU.

On Mon, Apr 26, 2004 at 11:34:13AM +0200, Mathieu Roy wrote:
> Wouter Verhelst <wouter@grep.be> wrote:
> > On Mon, Apr 26, 2004 at 09:49:57AM +0200, Mathieu Roy wrote:
> >> We talked about whether a maintainer is entitled to
> >> refuse after more than a year a bug
> >
> > He's not "refusing the bug", whatever that means.
> Refuse _a bug fix_ (the word fix was scrapped god knows why).
> > He's refusing a fix which he thinks isn't.
> Don't be hyperrelativist.

I'm not. Ryan has indicated that he does not believe the provided fix is
a clean one.

> When a bug exists and no longer occurs after applying a patch, the fix
> is indeed a fix.

Are you being deliberatively obtrusive, or are you just plain stupid?

I have something else for you to consider: throw your computer out of
the window. It'll break down, but you won't see the bug in gdm anymore.
There, fixed. And you'll do yourself and (especially) the world a favour
by doing so.

> It may not be a clean fix but it is a fix.

We're not here to create a system out of ugly hacks and bad design;
that's bound to lead to breakage. We shouldn't just do stuff for the
sake of it; to quote Colin Walters:

"[...]But if it is done, we should do it *right*.
We're Debian.  That's what we do."

> > If the bug is just "annoying", then why am I seeing an endless thread
> > about it on -devel? It seems to me it's more than just "annoying" for
> > some people.
> Something that is just annoying can easily be a real pain in the ass
> after two years. 
> But as I wrote in my previous mail, the issue is not specificly about
> this bug.

Well. Either you're interested in fixing the problem at hand, or you're
interested in fixing the outstanding bugs in the galaxy.

In the first case, this issue is specifically about this bug.

In the latter case, please take it to a different mailinglist.

> >> But a annoying bug should not be left for two years while harmless
> >> fixes exists.
> >
> > Stop this assertion. Ryan Murray, the maintainer of gdm, has stated
> > numerous times that he does not think the fix is harmless. 
> I don't know the specifics however in the message
> <20040417041342.GY13742@cyberhqz.com>, Ryan does not say this patch
> break the package, he says does not like the approach to get the
> locale setting.
> Sorry, that's harmless: it does not break anything in any way. 
> > As the maintainer of the gdm package, his opinion on that matter
> > counts more than yours
> > , unless you get the tech committee to follow you (which I think is
> > highly unlikely BTW, but perhaps that's just me).
> Unless you can get a GR removing the "Our priorities are our users",
> "We will be guided by the needs of our users", "We will place their
> interests first in our priorities", from the Social Contract, the
> maintainer is supposed to pay attention to users.

Not that argument again.

"We will be guided by the needs of our users" is *not* equal to "We will
jump to our users' every whim". Having a badly designed system is also
not good for our users.

> More than paying attention, the maintainer is supposed to place user's
> interest before his own. That could mean accepting a patch that fix a
> bug that annoys more than one user even if the approach is not the
> preferred one.

It's not just about preference; it's about having a clean upgrade path,
and a properly done configuration. Hell, we could go the Microsoft way
and store every configuration in one obsure hierarchical database; that
doesn't make it a good idea -- not even if it fixes a bug that annoys
more than one user and has done so over a long period of time.

> >> That's the point. And the fact that the Debian project is
> >> English-speaking or the fact that the bug is not critical does not
> >> really matters.
> >
> > Actually, it does. There are far more interesting and complicated bugs
> > that deserve a lot more attention than this senseless discourse.
> If you are not interested, just dont read. Why is there always someone 
> trying to explain to other they are wasting their time?

I didn't mean to say you shouldn't waste your time. What you do with
your time is yours to decide.

I meant to say you shouldn't be wasting my time and resources. By
discussing this horse to death and beyond, you will not get Ryan to
accept your non-fix; by not doing anything to help the release get any
nearer, you put the burden of doing that on other people's shoulders.

> > Also, before Sarge, Debian has never supported l10n in a default
> > install as well as it will now.
> But we dont care unless you can get Debian stop claiming to be an
> universal operating system.

You're exaggerating.


That is about all gdm has to do. Sure, the world will come to an end if
that isn't translated.

And don't tell me I'm not interested in i18n. My parents don't speak a
single (sensible) word of English; if I give them a non-l10ned system,
they are completely and utterly lost. i18n is of major concern to them,
and thus, to me; but giving them a non-l10ned *login screen*, of all
things, isn't exctly that much of a problem that Debian will cease to be
a "universal operating system" without it.

> > Indeed, the default gdm installation isn't ready yet -- so what? The
> > fix is easy, for a sysadmin:
> And you think nice to force each sysadmin to do something more ugly
> than the patch proposed as bugfix?

Whether or not that patch is more ugly is something highly subjective. I
will decline from comments in that area. Suffice to say that having a
non-localized default installation is not the end of the world; indeed,
it can be fixed locally, and the fix is easy enough for even the most
silly users to understand.

     smog  |   bricks
 AIR  --  mud  -- FIRE
soda water |   tequila
 -- with thanks to fortune

Attachment: signature.asc
Description: Digital signature

Reply to: