Bug 428169 - update of libalkimia to 8.0.4 results into craft failure
Summary: update of libalkimia to 8.0.4 results into craft failure
Status: RESOLVED FIXED
Alias: None
Product: Craft
Classification: Unmaintained
Component: Core (other bugs)
Version First Reported In: stable
Platform: Other Linux
: NOR normal
Target Milestone: ---
Assignee: Hannah von Reth
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2020-10-24 08:38 UTC by Ralf Habacker
Modified: 2020-10-24 08:56 UTC (History)
2 users (show)

See Also:
Latest Commit:
Version Fixed In:
Sentry Crash Report:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Ralf Habacker 2020-10-24 08:38:48 UTC
After releasing libalkimia 8.0.4 to https://download.kde.org/stable/alkimia/8.0.4 and updating the craft blue print https://invent.kde.org/packaging/craft-blueprints-kde/-/commit/966fff47e6b731d2944312d53ec0faadea595021 craft fails to download alkimia

STEPS TO REPRODUCE
1. inspect https://binary-factory.kde.org/job/KMyMoney_Release_win64/807/consoleText

OBSERVED RESULT
[2020-10-24T06:38:18.434Z] --2020-10-24 06:38:17--  https://download.kde.org/stable/alkimia/8.0.4/alkimia-8.0.4.tar.xz.sha256
[2020-10-24T06:38:18.434Z] HTTP request sent, awaiting response... 404 Not Found

EXPECTED RESULT
The failure should not occur

ADDITIONAL INFORMATION
Compared to the previous version at https://download.kde.org/stable/alkimia/8.0.3/ there is no difference in the memory of the source version.
When looking at other handcrafted blueprints, e.g. KDiff3 with the 'targetDigestUrls' property set, no sha256 files were added to the version at https://download.kde.org/stable/kdiff3/ either, which leads me to assume that there must be some hidden things that are responsible for this problem.
Comment 1 Ben Cooksley 2020-10-24 08:56:58 UTC
This is because Mirrorbrain (the software that runs download.kde.org and files.kde.org) only periodically regenerates it's hashes. 

Files added after the last generation run won't have them available until the next generation run is performed. This process is run once every 24 hours in our case - so the fix in this case will be to wait a few more hours.