| 1 |
Enrico Weigelt wrote: |
| 2 |
>> What sort of problems? An example backing up your claims would be very nice. |
| 3 |
> + Additional complexity (slotting) is necessary, so additional |
| 4 |
> changes of bugs. |
| 5 |
|
| 6 |
Oh please, this is so lame. That feature has been in existance for long enough |
| 7 |
to be proven useful and not faulty. The "higher probability of problems" is |
| 8 |
really not the best argument when discussing features that have been around for |
| 9 |
an incredible long time. |
| 10 |
|
| 11 |
> + Package maintainers have to both take care of slots *and* |
| 12 |
> version number *ranges* |
| 13 |
|
| 14 |
"taking care" takes you one line. I already gave you both dependency strings. |
| 15 |
Now guess what: If they were two packages, it would take you one line too! OMG! |
| 16 |
|
| 17 |
> + Different packages are treated as equal, produces confusion |
| 18 |
|
| 19 |
Aside from that guy who opened bug 143063 [1] I have yet to see anybody who got |
| 20 |
confused by this behaviour. |
| 21 |
|
| 22 |
> So, why don't you consider libxml and libxml2 equal packages ? |
| 23 |
|
| 24 |
Because that's the way upstream names them. |
| 25 |
|
| 26 |
> As said: you have to take care of version *ranges*. |
| 27 |
> Adds additional complexity. |
| 28 |
|
| 29 |
> BTW: how do you enforce an minimum gtk1 version ? |
| 30 |
|
| 31 |
You know that this wouldn't even make sense, as - you've pointed it out so many |
| 32 |
times - the API is incompatible. |
| 33 |
|
| 34 |
So, I'm asking you one last time: Do you have any actual good reasons to not |
| 35 |
package things the way upstream does it? |
| 36 |
|
| 37 |
[1] http://bugs.gentoo.org/show_bug.cgi?id=143063 |
| 38 |
|
| 39 |
-- |
| 40 |
Kind Regards, |
| 41 |
|
| 42 |
Simon Stelling |
| 43 |
Gentoo/AMD64 Developer |
| 44 |
-- |
| 45 |
gentoo-dev@g.o mailing list |
Archives