1 |
On 04/26/2012 01:03 AM, Corentin Chary wrote: |
2 |
> On Thu, Apr 26, 2012 at 9:57 AM, Zac Medico <zmedico@g.o> wrote: |
3 |
>> On 04/26/2012 12:30 AM, Corentin Chary wrote: |
4 |
>>> On Wed, Apr 25, 2012 at 6:41 PM, Michał Górny <mgorny@g.o> wrote: |
5 |
>>>> On Wed, 25 Apr 2012 09:16:05 +0200 |
6 |
>>>> Corentin Chary <corentin.chary@×××××.com> wrote: |
7 |
>>>> |
8 |
>>>>> On Tue, Apr 24, 2012 at 6:38 PM, Michał Górny <mgorny@g.o> |
9 |
>>>>> wrote: |
10 |
>>>>>> On Tue, 24 Apr 2012 16:19:11 +0000 |
11 |
>>>>>> "Robin H. Johnson" <robbat2@g.o> wrote: |
12 |
>>>>>> |
13 |
>>>>>>> On Tue, Apr 24, 2012 at 04:50:49PM +0200, Corentin Chary wrote: |
14 |
>>>>>>>>>> $ ./mirrors.py --all --count |
15 |
>>>>>>>>>> 297 ?? ?? http://pear.php.net |
16 |
>>>>>>>>>> 297 ?? ?? http://pear.php.net/get |
17 |
>>>>>>>>>> 88 ?? ?? ??http://pecl.php.net |
18 |
>>>>>>>>>> 88 ?? ?? ??http://pecl.php.net/get |
19 |
>>>>>>>>> These are already mirror bouncers. If you visit the above, |
20 |
>>>>>>>>> you'll get the closest mirror for downloading. |
21 |
>>>>>>>> And since there is already ~10 "mirrors" with only one actual |
22 |
>>>>>>>> backend, should they go to thirdpartymirrors or not ? If not, |
23 |
>>>>>>>> what about this pseudo-mirrors already present in |
24 |
>>>>>>>> thirdpartymirrors ? |
25 |
>>>>>>> I think we should add the pseudo-mirrors, but explicitly mark them |
26 |
>>>>>>> as such in the file, so that they don't get duplicate entries |
27 |
>>>>>>> added (eg adding us.pear, de.pear and the pear bouncer is bad. |
28 |
>>>>>>> Should have just the bouncer). |
29 |
>>>>>> |
30 |
>>>>>> It'd be great if we could add some kind of additional mirror |
31 |
>>>>>> entries, which would be used by repoman to signal missing mirror:// |
32 |
>>>>>> entries but won't be used for downloads. |
33 |
>>>>> |
34 |
>>>>> Yep, we could put that in it too: |
35 |
>>>>> github http://github.com/downloads/ |
36 |
>>>>> https://github.com/downloads/ |
37 |
>>>> |
38 |
>>>> Per spec, portage can choose a random mirror of the list. If we put |
39 |
>>>> entries like that, these two will be equally possible as the preferred |
40 |
>>>> cloud. URL -- while they redirect one to another. |
41 |
>>>> |
42 |
>>>> We might decide on some common syntax like preceding all extra entries |
43 |
>>>> with '-' but I don't want to be the one deciding here. |
44 |
>>> |
45 |
>>> I checked, and current portage code already handle entries starting |
46 |
>>> with a - gracefully thanks to stack_dictlist (removing them from the |
47 |
>>> list of mirrors). |
48 |
>> |
49 |
>> That means repoman will ignore them too. If you want existing versions |
50 |
>> of repoman to check for those paths in SRC_URI, you can add a line like |
51 |
>> this to thirdpartymirrors: |
52 |
>> |
53 |
>> github-bad-urls http://github.com/downloads/ https://github.com/downloads/ |
54 |
> |
55 |
> Hum, I checked repoman source code, and I didn't find where it checks |
56 |
> if SRC_URI matches something in thirdpartymirror. Any hint ? |
57 |
|
58 |
Search for SRC_URI.mirror in /usr/bin/repoman. |
59 |
-- |
60 |
Thanks, |
61 |
Zac |