Bug#60253: More clearly
From: Josip
Subject: Bug#60253:
Date: Sat,
> On Sat, Jun 17, 2000 at 08:10:56AM +0900, Atsuhito Kohda wrote:
> > > If there is Suggest: tetex-nonfree, why not suggest tetex-extra, too?
> >
> > Do you really investigate the contents of tetex-nonfree and tetex-extra?
> >
> > I think that basically tetex-extra and tetex-nonfree are supplement
> > of tetex-base so it is reasonable
>
> They are both supplements of tetex-base, yes, so they should both be
> suggested. For that matter, tetex-bin should be suggested, too.
>
> One could argue that task-tex provides this functionality, i.e. telling
> users which packages to install to get a full TeX, but the users might not
> know about task packages, and still, little more consistency can't hurt.
>
> > > tetex-bin suggests tetex-extra
> >
> > because this is a weak form of "tetex-bin depends tetex-base"!
>
> I don't understand what you mean.
Okay, let me explain a bit.
tetex-bin contains basic binaries of teTeX and they need
macros, fonts, etc. which should be contained in imaginary/virtual
tetex-base package.
So the basic relation should be
"tetex-bin depends on (imaginary) tetex-base".
But actually imaginary/virtual tetex-base package is divided
into the tetex-base, tetex-extra and tetex-nonfree.
So the relation is also divided into
"tetex-bin depends on the present (real) tetex-base"
"tetex-bin depends on tetex-extra"
"tetex-bin depends on tetex-nonfree"
but about tetex-extra and tetex-nonfree, it is too strong
to set "Depends". So the present relation
"tetex-bin suggests tetex-extra"
is, in my opinion, reasonable and this will explain the meaning
of "this is a weak form of "tetex-bin depends tetex-base"!"
In the same reason
"tetex-bin suggests tetex-nonfree"
might be reasonable as I said in the following already
> > I think, on the contrary to your request, it might be more reasonable
> > that one transfer "Suggest: tetex-nonfree" of tetex-base to tetex-bin
But tetex-nonfree might be less important than tetex-extra so
> > too but as the contents of tetex-nonfree are rather trifle so it might
> > be enought to set weak dependency as
> >
> > tetex-bin =(depends)=> tetex-base =(suggests)=> tetex-nonfree
> >
> > This is rather delicate decision of the maintainer, IMHO.
Yes, this is delicate and the decision may depend on the case or
person. But I think the present dependency is not so bad, at
least, not a bug.
I feel that you think too much of "formal" consistency.
> I was looking at it from the standpoint of a newbie user. They will think
> that tetex-base will get them a `tex', as the name may indicate, but it
> won't. If there was a suggestion (or even a recommendation) to other
> packages that do provide the full thing, it would be more clear what to
> install.
It is already clear enough what to install for TeX, IMHO.
Only selecting tetex-bin will provide fairy enough environment
of TeX for a newbie user.
But right now the splitting of teTeX packages are discussed
and when the splitting of teTeX is done then we should consider
the dependency carefully so that a newbie user can get the TeX
environment easily.
> > BTW, I think the following is a bad design of BTS.
> >
> > > No. 60253@bugs.debian.org reaches the BTS and the maintainer, whereas
> > > 60253-submitter@bugs.debian.org reaches the BTS and the submitter.
>
> I disagree. When there is a discussion about a bug report, the submitter
> doesn't need to know all about it, whereas the maintainer does. The
> Reply-To: header of the messages BTS forwards to the maintainer from
> submit@bugs.debian.org is set to user's real address plus the bug address.
> One just has to respect that Reply-To:.
I do not care this so much but in my case, I did not receive
the original mail and got the mail/file from the Web pages of
Debian so it was a bit tedious to set both bug address and
submitter address ;-)
Regards, 2000.6.18
--
Debian JP Developer - much more I18N of Debian
Atsuhito Kohda <kohda@pm.tokushima-u.ac.jp>
Department of Math., Tokushima Univ.
Reply to: