[gentoo-dev] Re: Last rites: net-nntp/inn

[Duncan <1i5t5.duncan@cox.net>]

Arnaud Launay posted on Tue, 12 Jan 2010 00:30:24 +0100 as excerpted:

> Hello,
> 
> Le Mon, Jan 11, 2010 at 11:05:16PM +0200, Markos Chandras a écrit:
>> # Markos Chandras <hwoarang@gentoo.org> (11 Jan 2010) # Fails with
>> -Wl,--as-needed
>> # bug #182782. Removal in 30 days
>> net-nntp/inn
> 
> As a newsmaster, I'm a bit concerned by this. By viewing bug #182782 ,
> it seems to me that only inn <=2.4.* is concerned by this bug, and that
> >=2.5 does not have it.
> 
> But, if I understand this announce correctly, the complete inn port will
> be dropped to oblivion.
> 
> Wouldn't it be better to stabilize inn 2.5 (there's even a 2.5.1 release
> out there, with a quite reactive support/dev team, to which this "bug"
> could be pushed and corrected), than to just drop inn entirely ?
> 
> Or maybe I'm just plain wrong, in which case I will happily stand
> corrected.

While I'm an --as-needed user myself, add my post here too.

A work-around is a work-around and shouldn't close the bug, agree with 
flame-eyes there, but it seems a bit much to yank something as standard 
as INN from the tree over this, especially when there's no indication it 
applies to the ~arch version 2.5, only 2.4.x, in either bug (182782, 
248145, refer to the log for the latter).

Not only is there no indication that 2.5 is affected, but if it is, 
where's the mention of the upstream bug number?

Replacements?  There's leafnode, and checking, I see openntpd4 and 
twisted news.  Still, inn is rather like sendmail for news.  Do we 
/really/ want to kill it?

IMO, the backing on this one seems thin and the timing premature.  
Perhaps it does need to go, but if so, there's certainly a stronger case 
to be made for it than those couple bugs and the last-rites masking 
comment.

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman