1 |
On Tue, May 01, 2007 at 03:08:56PM +0200, Piotr Jaroszyński wrote: |
2 |
> Firstly each test can be(not all categories are mutually exclusive): |
3 |
> - not existant |
4 |
> - non-functional |
5 |
> - not runnable from ebuild |
6 |
> - useful but unreasonable resource-wise |
7 |
> - useful and reasonable resource-wise |
8 |
> - necessary |
9 |
> - known to partially fail but with a way of skipping failing tests |
10 |
> - known to partially fail but with no easy way of skipping failing tests |
11 |
> Is that list comprehensive? |
12 |
|
13 |
Isn't it easier to list a set of boolean properties of _individual_ |
14 |
tests? |
15 |
- We don't need "non existent". |
16 |
- Non-functional and known to partially fail come down to "known to |
17 |
fail" for individual tests. |
18 |
- I have no idea what "not runnable from ebuild" is. |
19 |
- Unreasonable could be "resource hungry" or "needs additional deps" |
20 |
(those are two different things). |
21 |
- If a test is "necessary" I don't see why we should allow it to be |
22 |
skipped. |
23 |
- And about skipping failing tests. There is always a way to skip |
24 |
failing tests: not running any of them. It's just the granularity that |
25 |
is different, but the user doesn't care about that. |
26 |
|
27 |
> Secondly we must answer the question how precisely we want to distinguish |
28 |
> them, so users/dev can choose which categories of tests they want to run. |
29 |
> What comes to mind is: |
30 |
> - run all tests |
31 |
> - run only necessary tests |
32 |
> - run only reasonable tests |
33 |
> - don't run tests at all |
34 |
> Again, is that list comprehensive? |
35 |
|
36 |
I'd say, let the user decide based on the properties, fex: |
37 |
|
38 |
run known to fail : no |
39 |
run resource hungry: yes |
40 |
run additional deps: no |
41 |
|
42 |
if ( (known to fail == false || run known to fail == true) && |
43 |
(resource hungry == false || run resource hungry == true) && |
44 |
(additional deps == false || run additional deps == true) ) |
45 |
{ |
46 |
run the test |
47 |
} |
48 |
|
49 |
So for each test/set of tests and for each possible reason not to run |
50 |
it, either the reason does not apply to this test or the user explicitly |
51 |
said to run it anyway. |
52 |
|
53 |
You don't see the "way of skipping failing tests" in this last part. |
54 |
This is because if the decision is made not to run a test, the smallest |
55 |
set that can be skipped (possibly all tests for an ebuild) are skipped. |
56 |
|
57 |
> Please don't post solutions unless we figure out which options we really want |
58 |
> to deliver. |
59 |
|
60 |
I'm sorry if this counts as a solution, but I wasn't sure the list you |
61 |
gave was the kind of list we need and I didn't know another way of |
62 |
explaining the use of the properties I listed. |
63 |
|
64 |
Regards, |
65 |
Maurice. |
66 |
|
67 |
-- |
68 |
Maurice van der Pot |
69 |
|
70 |
Gentoo Linux Developer griffon26@g.o http://www.gentoo.org |
71 |
Creator of BiteMe! griffon26@××××××××.com http://www.kfk4ever.com |