1 |
On 16 Sep 2008, at 17:50, Matthias Bethke wrote: |
2 |
> on Mon, Sep 15, 2008 at 10:26:56PM +0200, you wrote: |
3 |
>> Seriously: can someone more skilled than me explain why using |
4 |
>> --resume-skipfirst and then trying to solve the unmerged packages |
5 |
>> is/can be |
6 |
>> a bad idea? How can this break the system? |
7 |
> |
8 |
> Frankly I have no idea. I've heard that argument many times in the |
9 |
> Paludis discussions but never even an attempt at an explanation that |
10 |
> went beyond "it breaks your system". My understanding is that you can |
11 |
> have two kinds of situation if an upgrade fails: |
12 |
> ... |
13 |
> b) the failed package is a dependency of at least one other package |
14 |
> In case a) you get to keep the old version, no problem. |
15 |
|
16 |
The risk is that you want to install X that depends upon Y. |
17 |
|
18 |
The ebuild for X states that version >1.2.3 of Y must be used because |
19 |
there's a bug in 1.2.2. |
20 |
|
21 |
The new version of Y fails to compile, so when X is compiled it only |
22 |
has the old version of Y to work with. It may compile OK but not work |
23 |
or feature a security bug. |
24 |
|
25 |
In the ideal world X's ebuild will fail, recognizing that Y is too |
26 |
old, and I think it will - my explanation may not literally be |
27 |
correct, but I hope you get the idea. |
28 |
|
29 |
What should NOT be in dispute is that Portage is a big old mess and |
30 |
well in need of replacement. It was never designed to do what it does |
31 |
now. Hopefully the replacement chosen to be integrated into Gentoo |
32 |
itself will be chosen on its merits, whether that be Paludis or |
33 |
something else. |
34 |
|
35 |
Stroller. |