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

Re: FTBFS for Archetecture all package (Bug#167049)



Thank you for your comments.

On Sun, Nov 03, 2002 at 01:48:51PM +0100, Santiago Vila wrote:
> Henrique de Moraes Holschuh wrote:
> > Osamu Aoki wrote:
> > > Policy 2.4.2 Package relationship only states "should be specified" in
> > > the first paragraph.  Serious is defined as a severe violation of Debian
> >
> > Well, FTBFS is different when caused by missing build-dependencies, or wrong
> > build-dependencies.  If you don't have any build-dependencies, it is an
> > annoyance but not a "severity serious" bug (because the dependencies are
> > unknown, and policy says 'should').  If you DO have build-dependencies, it
> > is a "severe" or "grave" bug (I don't know which because I didn't go hunting
> > down in policy to verify that it states the build-dependencies MUST be
> > correct... although IMHO that'd be implied anyway).
> 
> I'm not sure. Could you please provide policy chapter and verse?

Policy 2.4.2 Package relationship only states "should be specified" in
the first paragraph.  Serious is defined as a severe violation of Debian
policy (that is, it violates a "must" or "required" directive)....
(See my original post)

> When we started to consider "fail to build from source" bugs as *RC* bugs,
> there was no testing distribution, build-dependencies or even
> "serious" severity (RC bugs were "important or higher").
> 
> We did this (i.e. considering them *RC* bugs) mainly to ensure that
> autobuilders would successfully build the package for all architectures.

You are talking only about Architecture "ANY".

> Since we now have testing, and packages do not move from unstable to
> testing until they are recompiled for all archs, there is not, IMHO,
> an absolute need to use the "serious" severity for this kind of bugs.

I am talking about Architecture "ALL". They are not going through
autobuilders.

> I would not use a serious severity for a wrong build-depends which
> only affects the Architecture: all packages, since this does not
> reflect a real problem with the debian autobuilders.

If tetex from experimental hits sid, I know it will hit me.  Perfectly
readable documentation shall not be blocked from these silly RC bug
report.  It is only affecting Junichi's autobulder.

(I admit, this gets me running to fix it.  So he did good to me and this
particular incident did no harm.  But I am thinking what should be right
thing to do, here.)

This FTBFS on these documentation package is nothing but testing
more-or-less stability of tex build environment evolution.  In local
environment, people edit tex configuration to fit their needs.  Also
upgrade script will take care default directory changes.  But
documentation script which copied those configuration into document source
tree tends to get hit by these package update.  (tetex use too small
pool size for larger documents)

I sometimes wonder locale support issue.  Does Junichi's autobuilder
offer all the locales and LC_CTYPE?

-- 
~\^o^/~~~ ~\^.^/~~~ ~\^*^/~~~ ~\^_^/~~~ ~\^+^/~~~ ~\^:^/~~~ ~\^v^/~~~ +++++
        Osamu Aoki <osamu@debian.org>   Cupertino CA USA, GPG-key: A8061F32
 .''`.  Debian Reference: post-installation user's guide for non-developers
 : :' : http://qref.sf.net and http://people.debian.org/~osamu
 `. `'  "Our Priorities are Our Users and Free Software" --- Social Contract



Reply to: