Hi Nyall, On Wed, 14. Feb 2018 at 14:14:10 +1000, Nyall Dawson wrote: > If my understanding is correct, the osgeo4w architecture (and correspondingly > the windows release builds) only currently supports a single version of qt 5 > (and a single qt 4 version).
Correct. > If this is the case, it means that upgrading our qt version at some stage to > 5.9.4 (or 5.10.1) would also affect patch releases, right? E.g. updating to > 5.10.1 for a future 3.2 release would mean that the next 3.0.x build would > also jump up to 5.10.1. As 3.0 will be replaced with 3.2 once it's there (3.4 will be the next LTR), so that's not the actual case. But yes, switching to a newer version for Qt5 will require to rebuild everything that uses Qt5 (currently that's - not counting PyQt5 - only QGIS 3.x). > Is this interpretation correct? Do we consider this safe to do in a patch > mid-way through a stable release? And if not, does it mean we're effectively > stuck with the current version of Qt throughout the life of 3.x? Not necessarily, but we'd have to update all 3.x to be able to build & run with the Qt5 we have (ie. 3.4 and 3.x w/ x>4). > So I'd hate to see the Windows builds locked to 5.9 indefinitely. Windows is not the only platform. Debian unstable is also at 5.9 currently, buster/testing 5.9, stretch/stable has 5.7 (so we build without 3D there). bionic (next ubuntu lts) also has 5.9. Jürgen -- Jürgen E. Fischer norBIT GmbH Tel. +49-4931-918175-31 Dipl.-Inf. (FH) Rheinstraße 13 Fax. +49-4931-918175-50 Software Engineer D-26506 Norden http://www.norbit.de QGIS release manager (PSC) Germany IRC: jef on FreeNode
Description: PGP signature
_______________________________________________ QGIS-Developer mailing list QGIS-Developer@lists.osgeo.org List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer