On Nov 17, 2011, at 10:15 PM, Daniel Macks wrote:

> 
> 
> On Thu, 17 Nov 2011 20:35:05 -0500, Alexander Hansen 
> <alexanderk.han...@gmail.com> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
>> Hash: SHA1
>> 
>> On 11/17/11 8:06 PM, Hans-Christoph Steiner wrote:
>> > > On Nov 17, 2011, at 11:13 AM, Alexander Hansen wrote:
>> > >> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
>> >> >> On 11/17/11 10:57 AM, Alexander Hansen wrote:
>> >>> On 11/17/11 9:07 AM, Alexander Hansen wrote:
>> >>>> On 11/17/11 12:54 AM, Hans-Christoph Steiner wrote:
>> >>> >>>>> I'd like to package svn-bisect for fink, I can't seem to
>> >>>>> find any reference to it being in Fink anywhere. It relies
>> >>>>> on the IO:All perl module, which I also don't find any
>> >>>>> reference too. I'm not a perl person, so I could be blind
>> >>>>> to some obvious things there. >>> >>>>> 
>> >>>>> http://search.cpan.org/dist/App-SVN-Bisect >>>>> 
>> >>>>> http://search.cpan.org/dist/IO-All/
>> >>> >>>>> .hc
>> >>> >>> >>>> Indeed, those aren't currently in Fink.  They'd be
>> >>> >>>> app-svn-bisect-pm io-all-pm
>> >>> >>> >>> >>> Io:All just got added to Fink. >> >> (oops, should have been 
>> >>> >>> >>> >>> Io::All)
>> >> >> And App::SVN::Bisect has been added now, too. > > > Wow, awesome, that 
>> >> >> was fast.  I think that since the utility is
>> > called svn-bisect, the package should probably called that too. > 
>> > Something like the epydoc-py package.  It generates an epydoc-py26
>> > and then uses update-alternatives to make an epydoc command in
>> > /sw/bin.  I think this package should be svn-bisect-pm and then do
>> > the same thing. > > How does that sound?
>> > > .hc
>> > > Ah, I didn't know it was a utility. 
>> If it doesn't generate a module that other packages use, then it's OK
>> to give it a suffix other than pm, e.g. svn-bisect-perl. 
>> You could also have it use Provides: svn-bisect, in case other
>> packages down the road might need to Depend on it without caring about
>> the Perl version.  As long as the Provides: is used responsibly, e.g. not 
>> having a real "svn-bisect" package, too, this should be easier
>> than having a list like: svn-bisect-perl588 | svn=bisect-perl5100
> 
> The svn-bisect utility program is indeed bundled with the App::SVN::Bisect 
> module upstream and the module is also usable directly. I didn't realize the 
> end goal was the utility, so when I packaged the module itself, the program 
> got embedded in that package by default--*not* generating it, or generating 
> it stand-alone separately and dependent on the module is harder than just 
> letting it happen:) Install app-svn-bisect-pm5100 (or I any other variant, 
> can't remember which others I set) and you automatically already have 
> /sw/bin/svn-bisect. The Provides:svn-bisect trick or a bundle package are two 
> ways to get an invariant name to allow easy dependencies on it. Depends (heh) 
> how likely the module/program is to be changed and for others to need "a 
> specific minimum version" for a certain feature rather than just accepting 
> "any version of the command". 
> dan

I think a Provides: svn-bisect covers my concerns.  I think most people are 
going to look for it under that name rather than the module name.

.hc

----------------------------------------------------------------------------

"We have nothing to fear from love and commitment." - New York Senator Diane 
Savino, trying to convince the NY Senate to pass a gay marriage bill


------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure 
contains a definitive record of customers, application performance, 
security threats, fraudulent activity, and more. Splunk takes this 
data and makes sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-novd2d
_______________________________________________
Fink-devel mailing list
Fink-devel@lists.sourceforge.net
List archive:
http://news.gmane.org/gmane.os.apple.fink.devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel

Reply via email to