[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: package variant name



On Tue, Oct 22, 2002 at 09:23:30AM +1000, Glenn McGrath wrote:
> I think they raise a valid point, we do mangle our package names to make
> them unique.

Sure, that's a "valid" point. Now explain why it's a bad thing.

> It would be more idealistic if we seperate the upstream name and the
> debian extension into seperate fields to keep it clear.

*Why*? What things would this let you do that you cannot already do?

> Im not sure about the best way to do it, it could be done by introducing
> two new fields, and preserving the Package: field for use as the
> combination of the two fields.
> 
> e.g.
> Upstream-name: nano
> Variant: tiny

Note that this has broken the ability to identify uniqueness with a
single string comparison. That makes it quite a lot more complicated
to work with.

Package: nano-tiny
Source: nano

This matches a search for ^nano, and it has a clear description field
- what more do you want?

If you want to depend on it for something, that's why we have
Provides and virtual packages.

> I appreciate its probably something that is unlikely to change becasue it
> would only have a superficial benefit, thought it was worth mentioning
> anyway.

If you can't come up with a real-world case where it is an advantage,
then it is not an advantage - it's random dribbling from the marketing
department.

-- 
  .''`.  ** Debian GNU/Linux ** | Andrew Suffield
 : :' :  http://www.debian.org/ | Dept. of Computing,
 `. `'                          | Imperial College,
   `-             -><-          | London, UK



Reply to: