| 1 |
On Wed, Jun 11, 2008 at 08:23:59AM +0100, Richard Brown wrote: |
| 2 |
> Also, I think you seem to be suggesting that gentoo is so well tested |
| 3 |
> that once something's marked stable, there's no point in testing it. |
| 4 |
|
| 5 |
A very good point. Just last week the *stable* perl cairo bindings were |
| 6 |
broken by a x11-libs/cairo bump. We caught this however and noticed that |
| 7 |
the new perl cairo bindings worked. Those were then stabilized at the |
| 8 |
same time and users now have a working cairo. |
| 9 |
|
| 10 |
What would have happened if that hadn't happened? Any package that |
| 11 |
depended on dev-perl/Cairo would've been broken. |
| 12 |
|
| 13 |
The lesson to learn is that once something is stable doesn't mean it's |
| 14 |
always stable. If a user finds out that way and files a bug, chances are |
| 15 |
greater that he'll get a dev-perl/Cairo that works with the new cairo |
| 16 |
version soon, rather than a dev-perl/Cairo version that breaks |
| 17 |
immediately. |
| 18 |
|
| 19 |
What would you rather have? |
Archives