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