The api contains lots of hard coded http urls (images etc) which leads to the infamous "mixed insecure content" whne accessed over https.
All of these http:// can likely be replaced by just //
The api contains lots of hard coded http urls (images etc) which leads to the infamous "mixed insecure content" whne accessed over https.
All of these http:// can likely be replaced by just //
Subject | Repo | Branch | Lines +/- | |
---|---|---|---|---|
Make urls protocol neutral | labs/tools/heritage | master | +5 -5 |
Status | Subtype | Assigned | Task | ||
---|---|---|---|---|---|
Resolved | JeanFred | T131714 Hold a Wiki Loves Monuments coding sprint at Wikimedia Hackathon 2016 | |||
Resolved | JeanFred | T131739 Remove hard coded http in api |
Change 281380 had a related patch set uploaded (by Lokal Profil):
Make urls protocol neutral
This seems to have fixed the mixed content issue but also have broken the html output in e.g. this example.
Most likely this is due to links being recognised through a regexp looking for http(s) rather than (http(s):)//.
Hmmm, as far as I could see, the fix was not actually deployed... I checked because I was still getting mixed content warnings.
./bin/deploy-to-toollabs.sh
ff5833f Make urls protocol neutral 4883bbb Fix import of pywikibot bits in remaining erfgoed scripts 6df1d01 Enable cover-inclusive for erfgoedbot unitests fc1ccef Add unit tests for ucfirst method 37de76d Extract method extract_elements_from_template_param from update_database 2ca107f Change all linebreaks from Windows to Unix Deploy done.
So the broken source links problem was probably there before... Probably the regex does not deal well with some of the encoded chars.
Ah, got it. I filed T132647 for this and am closing this one. Thanks @Lokal_Profil for the work :)