Bug#200461: Processed: Refinement -- Package name must not begin with 'zz'
On 08 Jul 2003 19:20:46 +0200, Thomas Hood <jdthood@yahoo.co.uk> said:
> On Tue, 2003-07-08 at 18:30, Manoj Srivastava wrote:
>> As a corollary, I suppose you would want to outlaw aalib
> No, since there are lots of strings that aren't legal package names
> that precede all legal package names in listings.
>> And what happens if two different sets of entities want a script to
>> be run last?
> Then they have to decide amongst themselves who really gets to go
> last.
Since the packages desiring the last script in run-parts would
need to coordinate anyway, why create an ineffective policy dictum in
the first place? Why not let the cooperating packages decide on the
relevantr package name(s) by just appending zzzzzzzzzz's?
>> And why does one want things to be absolute forst or absolute last?
> The aim is to make sure that one can name a script such that it is
> guaranteed to follow the default names used by all other packages.
But you can't, really, since another package may also want a
last run script.
Also, most cases where ordering is important, people prefix
numbers (20Szis-pkg, 20anothe-pkg, for example), so this restrictyion
on the name of packages is silly.
>> Isn't this a bandaid for a proper dependnecy ordering mechanism for
>> run-parts?
> Yes. Bandages can be useful for first aid.
But not hard coded into policy, no.
>> Why is policy being used to substitute for proper technical
>> solutions?
> Let the technically perfect not be the enemy of the good.
This is not good -- this is pretyy bad as a technical
solution. An ineffective, partial, bandaid does not belong in
policy.
manoj
--
Today is a good day to bribe a high-ranking public official.
Manoj Srivastava <srivasta@debian.org> <http://www.debian.org/%7Esrivasta/>
1024R/C7261095 print CB D9 F4 12 68 07 E4 05 CC 2D 27 12 1D F5 E8 6E
1024D/BF24424C print 4966 F272 D093 B493 410B 924B 21BA DABB BF24 424C
Reply to: