bug-automake
[Top][All Lists]
Advanced

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

bug#7655: conditional _TEXINFOS should be supported


From: Jack Kelly
Subject: bug#7655: conditional _TEXINFOS should be supported
Date: Fri, 17 Dec 2010 09:12:40 +1100

On Fri, Dec 17, 2010 at 7:02 AM, Ralf Wildenhues <address@hidden> wrote:
> [ no need to keep bug-automake@ in Cc:; debbugs takes care of that ]

Noted.

> * Jack Kelly wrote on Thu, Dec 16, 2010 at 08:42:52PM CET:
>> > Disagree. foo.info should always be built, because it goes into `make
>> > dist'.
>
> Ah, a piece of logic I didn't think of.  Hmpf.
>
> I need to reread all the long comments in automake.in, to also take the
> info-in-srcdir-or-not complication into account.
>
>> > If you want to limit building for non-specific `make', then
>> > make sure that the .info files and such all depend on `make dist', so
>> > the tarball is correctly generated.
>
> I want to be able to say "no, I do not want this info file to be built
> nor installed" for the user.  This is what (at least some) users need.

Sure, but .info files are built as part of make dist, so just allow
maintainers to say "no, I don't want this info file to be installed"
(and no pdfs, dvis, &c. built, either).

> Also, some users need to be able to generate .texi files.  I'm not sure
> if it is always sufficient to let the maintainer generate them (and ship
> them plus their .info files).

That's currently how it's done. I just checked with automake 1.11.1 on
Ubuntu 10.10. I've maintained packages where generated .texi files are
used, and it's all fine as long as they're also distributed.

>> Further thoughts on the above: The maintainer should always have
>> access to `texinfo', and the user should never have to rebuild .info
>> files, so it seems OK to me if they're not blocked when COND is false.
>
> I don't understand what you mean with "blocked" nor with "access to
> `texinfo'".

.info files are packaged as part of `make dist', which makes `texinfo'
a maintainer tool (it's also supported by `missing', so that's more
evidence in favour, IMHO). Therefore, it seems to me that it is
reasonable that the maintainer has `texinfo' installed and the user
should never have to actually rebuild a .info file. Then it's OK to
package it and just not install it if COND is false.

-- Jack





reply via email to

[Prev in Thread] Current Thread [Next in Thread]