@L--U--P--U--S thanks for doing that, so at least your problem doesn't seem to be GTK major version dependent either.
So summarising the data:
| Reported | OS | Version | GTK | slower | |-------------|------|------------|--------|----------| | https://github.com/geany/geany-osx/issues/20 | OSX | 1.36 | ??? | no | | https://github.com/geany/geany-osx/issues/20 | OSX | 1.37-2 | ??? | yes untimed | | @L--U--P--U--S | Windows | 1.36 | GTK 2.24.32, GLib 2.60.6 | no | | @L--U--P--U--S | Windows | 1.37.1 | GTK 2.24.32, GLib 2.66.2 | yes 5-10 times | | #2639 | Windows | 1.36 | ??? | no | | #2639 | Windows | 1.37.1 | ??? | yes untimed | | @andy5995 | Linux | 1.36 | GTK 3.24.5, GLib 2.58.3 | no | | @andy5995 | Linux | 1.37 | GTK 3.24.5, GLib 2.58.3 | yes, 2 times | | @elextr | Linux | 1.36 | GTK 3.24.20, GLib 2.64.3 | No | | @elextr | Linux | 1.37 | GTK 3.24.20, GLib 2.64.3 | No |
And to summarise my conclusions from that:
1. For some people 1.37(.1) is measurably slower, but not everyone, seems from the reports at https://github.com/geany/geany-osx/issues/20 (OSX) @L--U--P--U--S (Windows) and @andy5995 (Linux) it happens on all platforms. 2. @andy5995 bisected it to https://github.com/geany/geany/commit/e42a963d6b4f1b2ef2063e59fc4139a250dfe7... on Linux 3. but neither @codebrainz nor I see any slowdown on Linux
So maybe there is something in the environment that triggers https://github.com/geany/geany/commit/e42a963d6b4f1b2ef2063e59fc4139a250dfe7... to cause slowdown.
Because that patch sets the margin width, maybe it triggers Scintilla to re-layout each file after opening which is a plausible slowdown, but why does it only happen on some systems, or is that patch a furphy, those are the questions?