This is the mail archive of the
cygwin-apps
mailing list for the Cygwin project.
Re: [ITP] libargp 20110921-1 (ping)
On Tue, Sep 27, 2011 at 10:02:38AM -0500, Reini Urban wrote:
>On Tue, Sep 27, 2011 at 9:24 AM, Christopher Faylor wrote:
>> On Tue, Sep 27, 2011 at 10:06:36AM -0400, Andrew Schulman wrote:
>>>> I'd like to package and maintain libargp for Cygwin. ?argp is a C interface for
>>>> processing command line arguments.
>>>>
>>>> The libargp source files are taken from gnulib. ?gnulib is a well-known program
>>>> available in major Linux distros (e.g. http://packages.debian.org/sid/gnulib),
>>>> but I'm not aware of any distros that separately package just argp from it. ?So
>>>> I'm not sure if libargp needs to be voted on or not for inclusion in Cygwin.
>>>>
>>>> Documentation: ? ? http://www.gnu.org/s/hello/manual/libc/Argp.html
>>>> Source home page: ?http://www.gnu.org/s/gnulib/MODULES.html
>>>> License: ? ? ? ? ? LGPL
>>>
>>>Package maintainers, please consider voting in favor of libargp for Cygwin.
>>>
>>>I don't know of any rule that says I can't vote for my own package, so I'm
>>>voting +1 for libargp. ?I want it specifically to use with nosleep [1]. ?I
>>>could have just baked the argp source files statically into nosleep, but they
>>>would have made up the bulk of the code, which seemed silly. ?It seems better
>>>to make libargp available for shared use.
>>
>> I think the fact that other releases contain an implementation of argp should
>> be good enough to make its inclusion a foregone conclusion. ?I do hope that
>> someday the package can go away and this functionality can be added to the
>> Cygwin DLL though.
>>
>> So, all we need is a GTG on packaging.
>>
>> The same doesn't apply to the "nosleep" package, of course.
>
>Since argp is a gnulib extension, and gnulib cannot/should not be
>packaged as seperate library, only pulled in from source (clisp e.g.
>does so) I fail to recognize
>why nosleep does not include gnulib at all?
argp is part of glibc but we can't just pull it into Cygwin. It makes sense
to include a general implementation:
http://cygwin.com/ml/cygwin/2011-09/msg00266.html