| 1 |
On Mon, 2009-11-09 at 14:33 +0100, Ben de Groot wrote: |
| 2 |
> I am of the opinion it is irresponsible to leave vulnerable versions of Qt with |
| 3 |
> known security bugs any longer in the tree. The Qt team therefore requests |
| 4 |
> that arches that have not done so already move quickly on stabilizing Qt |
| 5 |
> 4.5.3, see bug 290922 and 283810. |
| 6 |
|
| 7 |
It is more irresponsible and outright wrong to remove the latest stable |
| 8 |
revision of a package for some arches, despite security implications. |
| 9 |
Hard masking constitutes the same - the last stable version is not in |
| 10 |
stable visibility anymore. |
| 11 |
|
| 12 |
You can however remove the keywords of the arches from older versions |
| 13 |
that do have a newer version/revision stable as seen in all profiles. |
| 14 |
|
| 15 |
|
| 16 |
> We plan on REMOVING or at the very least HARDMASKING pending removal |
| 17 |
> all <=4.5.2 ebuilds by the end of this week. This means that arches that have |
| 18 |
> not stabilized 4.5.3 would loose their stable Qt4 version. |
| 19 |
|
| 20 |
How do you see this being acceptable for the users of these |
| 21 |
architectures? Many of these architectures that are "lagging behind" not |
| 22 |
being even security supported architectures. |
| 23 |
|
| 24 |
> Please let us know if there is any way in which we can assist arches. We |
| 25 |
> are aware that some arches are down to one active person. But if there is |
| 26 |
> no other way, maybe the status of such arches should be reconsidered. |
| 27 |
|
| 28 |
It seems most these arches that are at ~1 person are not security |
| 29 |
supported either |
| 30 |
|
| 31 |
> We especially request ppc64 to be marked as an experimental arch, as it |
| 32 |
> is the worst one lagging in stabilization. See bug 281821 for a poignant |
| 33 |
> example, a 3 months open security bug. |
| 34 |
|
| 35 |
First its security supported status should be considered, not making it |
| 36 |
an experimental arch, as that could very well throw it in a backwards |
| 37 |
spiral of getting more and more problematic due to repoman iirc not |
| 38 |
checking issues with it by default. |
| 39 |
|
| 40 |
-- |
| 41 |
Mart Raudsepp |
| 42 |
Gentoo Developer |
| 43 |
Mail: leio@g.o |
| 44 |
Weblog: http://planet.gentoo.org/developers/leio |
Archives