[Python-modules-team] Bug#567378: Using alternatives for /usr/bin/markdown?

[Damyan Ivanov]

-=| Bernd Zeimetz, Mon, Feb 08, 2010 at 10:50:28PM +0100 |=-
> Is there a reason why markdown should be kept in the archive then?

I think no. But the package is still listed as maintained as Matt 
Kraai. Matt, do you mind the proposed removal of the markdown package  
in favour of libtest-markdown-perl and python-markdown?

> Migrating to to libtext-markdown-perl sounds like the best idea. I'm 
> CCing the RFA bug for markdown to let people know about that. We 
> could make markdown a package which depends on libtext-markdown-perl 
> | python-markdown to migrate to an uptodate version and handle 
> /usr/bin/markdown by alternatives. The other option would be to drop 
> /usr/bin/markdown from the python package, I don't have a proper 
> opinion on that the best thing is yet.

Do I understand your suggestion correctly? You propose that 
libtext-markdown-perl adds a transitional 'markdown' binary package 
which depends on libtext-markdown-perl | python-markdown. Additionaly, 
'libtext-markdown-perl' and 'python-markdown' should conflict/replace 
'markdown (<< 1.0.26-2~)', right? (assuming 1.0.26-2 is the version 
introducing the 'markdown' binary package').

About alternatives, if we define "the markdown interface" to be

    markdown input > output

then both implementations seem to support it and alternatives are 
appropriate.

(Note that starting libt-m-p's markdown without arguments waits for 
input from STDIN (an can be used as 'markdown < input'), while p-m's 
markdown shows usage information. Not sure if this is an 
incompatibility that needs addressing)

Now we 'only' have to decide about priorities :)
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 835 bytes
Desc: Digital signature
URL: <http://lists.alioth.debian.org/pipermail/python-modules-team/attachments/20100216/a3c22113/attachment.pgp>