| 1 |
Mike Frysinger posted on Sat, 30 Jun 2012 11:16:51 -0400 as excerpted: |
| 2 |
|
| 3 |
> On Saturday 30 June 2012 07:22:39 Zac Medico wrote: |
| 4 |
>> On 06/30/2012 04:07 AM, Pacho Ramos wrote: |
| 5 |
>> > I would like to discuss a bit more issues like: |
| 6 |
>> > https://bugs.gentoo.org/show_bug.cgi?id=423087 |
| 7 |
>> > |
| 8 |
>> > Even if there are "a lot" of packages that can cause this breakage |
| 9 |
>> > when downgraded, I think it should be prevented and package managers |
| 10 |
>> > shouldn't try to downgrade this kind of packages as they will later |
| 11 |
>> > cause a total breakage. People is not supposed to know that |
| 12 |
>> > downgrading some package system will, for example, have an unusable |
| 13 |
>> > gcc. |
| 14 |
>> |
| 15 |
>> It seems like a die in pkg_pretend would serve pretty well. |
| 16 |
> |
| 17 |
> doing it on a per-ebuild basis doesn't make much sense. a simple |
| 18 |
> version compare (like we do in glibc as an exception to this rule |
| 19 |
> because of its much wider implication) is incorrect: the new version |
| 20 |
> might not introduce any new symbols compared to the old one, and even if |
| 21 |
> it has, other packages might not have been linked against the new |
| 22 |
> symbols. |
| 23 |
|
| 24 |
glibc was the example that popped up in my mind as well. Two comments: |
| 25 |
|
| 26 |
* Please make sure there's an override in case the user needs it, like |
| 27 |
glibc has. |
| 28 |
|
| 29 |
* Glibc's override is unfortunately broken in one specific case: binpkgs. |
| 30 |
The binpkg must have been built with the override (I_KNOW_WHAT_IM_DOING=1 |
| 31 |
or whatever) set, or it won't work. |
| 32 |
|
| 33 |
I found this out the hard way, when I upgraded a glibc and immediately |
| 34 |
found it broken for something or other. I hadn't rebuilt anything else |
| 35 |
yet (I try to do glibc upgrades by themselves, /because/ of their |
| 36 |
critical nature), so everything was still built against the old glibc and |
| 37 |
there would have been no breakage. Since I run FEATURES=buildpkg by |
| 38 |
default, I thought "no problem, I'll just emerge the old binpkg." |
| 39 |
WRONG! Because it wasn't built with the override variable set, it |
| 40 |
wouldn't let me downgrade using the binpkg! |
| 41 |
|
| 42 |
But whatever breakage I had wasn't letting me rebuild, either! |
| 43 |
|
| 44 |
Which is exactly why I have multiple bootable rootfs (with everything |
| 45 |
portage installs on it) setup, so I can boot into a bakup root if my |
| 46 |
normal working one gets hosed, somehow. So I just booted into it, |
| 47 |
mounted my normal working rootfs, set ROOT= to point to it, and went from |
| 48 |
there. |
| 49 |
|
| 50 |
It would be MUCH easier if a glibc binpkg downgrade could look only at |
| 51 |
the existing portage environment for that variable, not check the one in |
| 52 |
the binpkg! |
| 53 |
|
| 54 |
-- |
| 55 |
Duncan - List replies preferred. No HTML msgs. |
| 56 |
"Every nonfree program has a lord, a master -- |
| 57 |
and if you use the program, he is your master." Richard Stallman |
Archives