1 |
Ciaran McCreesh wrote: |
2 |
> * There is a clean upgrade solution available that will result in |
3 |
> non-ported packages merely pulling in a load of extra unnecessary |
4 |
> packages (that non-modular users have anyway). |
5 |
> |
6 |
> * The clean solution visibly illustrates that a package is unported. |
7 |
> Users who are running ~arch can clearly see this, and can file bugs and |
8 |
> (if they care) attempt a --nodeps installation. The bugs can be picked |
9 |
> up by the package maintainers or the volunteers on this list. |
10 |
|
11 |
Yes, for all 3 people who have a clue what it means when virtual/x11 |
12 |
gets pulled in. How many users do you seriously think will have a clue |
13 |
and think "Oh, virtual/x11 is getting pulled in here. I must have a |
14 |
package that isn't ported to modular X somewhere in this list. Let me |
15 |
track it down and file a bug."? |
16 |
|
17 |
> * The clean solution is the solution originally proposed to this list, |
18 |
> and the reason we are using new style virtuals. |
19 |
|
20 |
No, this is wrong. The reason we are using new style virtuals is so we |
21 |
could have a versioning in what provides virtual/x11. Namely, 6.8 or older. |
22 |
|
23 |
> * There is an alternate upgrade solution that means that any users who |
24 |
> have an unported package will get their screen filled with several |
25 |
> pages of incomprehensible bright red crap. |
26 |
|
27 |
Several pages? How is that coming from a single unported package? Also, |
28 |
it's not incomprehensible and shouldn't be to anyone on ~arch (or |
29 |
frankly, anywhere), because packages involving blockers regularly have |
30 |
to be dealt with. |
31 |
|
32 |
> * There are currently enough unported packages that many ~arch users |
33 |
> will probably have one or two installed (assumption: many users have |
34 |
> several utterly random non-mainstream packages installed). |
35 |
|
36 |
Possible, but we can't prove this one way or the other. Certainly very |
37 |
few modular X users have encountered apps that are still unported, as |
38 |
evidenced by very few remaining blockers on #112675. And there are a |
39 |
fairly large number of |
40 |
|
41 |
> * Despite your original assurances to this list, you intend to go ahead |
42 |
> and take the alternate solution, which will lead to one of the most |
43 |
> user visible and hardest to fix breakages we've ever had. |
44 |
|
45 |
Yes, I suggested handling this differently back in early August, when |
46 |
these things were in the planning stage. But even later that month, I |
47 |
posted saying that wasn't the best way to handle it. It's not difficult |
48 |
to imagine that things can change over the course of 6 months. |
49 |
|
50 |
> * You are doing this because you believe that it is better to get every |
51 |
> package ported over extremely quickly rather than having the odd |
52 |
> package with extra unnecessary listed dependencies, and you do not |
53 |
> consider the impact upon our users to be relevant. |
54 |
|
55 |
I consider ~arch users to have agreed to help test and fix new things. |
56 |
This is included. I would not do the same thing to our stable tree, or |
57 |
if we only had a stable tree. |
58 |
|
59 |
Yes, I do think it is better to have a quick solution and let some of |
60 |
our ~arch users see a couple of blocks, for which they will file bugs. |
61 |
Then these bugs will be fixed within a day, and those users will again |
62 |
have working systems. |
63 |
|
64 |
I don't see what the big deal is. By being ~arch users, they're already |
65 |
asking to use ebuilds that have a chance of breaking their systems |
66 |
permanently, let alone a single 'emerge sync'. |
67 |
|
68 |
Thanks, |
69 |
Donnie |