1 |
On 20/02/2013 17:45, Alec Warner wrote: |
2 |
> We could add something like PROPERTIES="network" to packages that |
3 |
> require the network. I'm vaguely sure for instance, that some |
4 |
> src_test() phases require a functioning network to work properly. |
5 |
|
6 |
This has been proposed a bunch of times before, and I still wish to see |
7 |
it implemented. |
8 |
|
9 |
On the other hand, even for those that currently require network for |
10 |
src_test, if there was a quick opt-out, I foresee trouble: quite a few |
11 |
times it's possible to test 90% (figurative number) of a library without |
12 |
network, you just need to take the time to disable the network-bound |
13 |
tests (eventually giving upstream a patch to enable/disable the network |
14 |
side, or to ignore network failures instead of consider them a FAIL — |
15 |
versus getting the wrong response which _is_ a FAIL). |
16 |
|
17 |
Having non-masked ebuilds rely on network access also makes it extremely |
18 |
hard to do things such as isolated builds (for embedded/firmware-style |
19 |
OSes), and by experience would make the whole tinderboxing effort nearly |
20 |
unfeasible — for sure I would shut down both the tinderboxes as I'd be |
21 |
debugging network issues more often than actual problems with |
22 |
tinderbox/packages. |
23 |
|
24 |
> linux-firmware[non-free] <- the use flag to toggle between free and |
25 |
> non-free licenses. |
26 |
> linux-firmware-noredist <- This one is RESTRICT="fetch mirror" |
27 |
|
28 |
+1 — It requires some work from someone to actually split the stuff |
29 |
manually though, and there is at least one problem: somebody _got_ to |
30 |
redist the firmware for it to be fetched.. unless. |
31 |
|
32 |
But yes, the first obvious solution is, when snapshotting the |
33 |
repository, to just drop the noredist files, and add the above-suggested |
34 |
USE flag for the non-free ones. |
35 |
|
36 |
-- |
37 |
Diego Elio Pettenò — Flameeyes |
38 |
flameeyes@×××××××××.eu — http://blog.flameeyes.eu/ |