IABSD.fr/ports/misc/portroach/patches
Branch :
Log
| Author | Commit | Date | CI | Message |
|---|---|---|---|---|
 |
6b2d8ac0 | 2025-01-10 11:44:18 | change SITE_PYPI to use files.pythonhosted.org as suggested in https://docs.pypi.org/api/#integration-guide "Predictable URLs"; teach portroach to handle this. ok aja@ jasper@ | |
 |
c7ecc66c | 2023-09-11 14:45:35 | fix postgres instructions by allowing the portroach user to own its own db. Thanks jeremy@ | |
|
5ed52aa7 | 2023-09-10 09:57:12 | revert to using DISTNAME if it's defined and FULLPKGNAME only if not | |
|
8ba7d01d | 2023-09-08 17:59:57 | help portroach through recent infrastructure changes - use the new ROACH_URL and ROACH_SITES values in sqlports - create an indirection thru a temp view that makes things less painful Accordingly: use PortsQ consistently, which is WAY faster. - use FULLPKGNAME instead of DISTNAME so that no port gets bad messages because of that - shorten the progressmeter slightly so it fits on a normal terminal okay aja@ | |
|
7dd95127 | 2023-07-21 10:57:10 | teach portroach to use the cpan handler for more cpan MASTER_SITES lines (https was not handled, and it required a /CPAN/modules/ path, so also allow https and {meta,}cpan.org/modules to match). also, when stripping versions from distnames for the cpan handler to pass to a metacpan search, cope with Name-v1.23 format as well as Name-1.23. this avoids falling back to directory lists in some cases, which is particularly important for cpan because their version number ordering does not match the ordering normally used in ports/portroach, e.g. 1.1, 1.101, 1.2 vs 1.1, 1.2, 1.101, and also portroach's listing handler doesn't cope with a mixture of -1.23 and -v1.23 formats in the same port either. | |
 |
c715eab9 | 2022-03-11 19:38:05 | drop RCS Ids | |
 |
8f7fbbc1 | 2020-07-23 14:58:41 | unbreak the bitbucket site handler | |
|
3d9581a6 | 2020-07-23 14:14:25 | unbreak CPAN site handler by migrating from API v0 to v1 | |
|
c2bdc880 | 2019-03-11 08:15:31 | replace slow query with an better one, as this is used in a tight loop insane code okay aja@, jasper@ | |
|
5325a424 | 2017-04-16 09:14:52 | update to portroach-2.0.11, containing a bunch of fixes related to github. nb: in order to do a full portroach scan one needs a github token otherwise rate-limiting will block requests after the default limit of 60 p/h has been reached. |