I tried to test it on Windows but got mixed results :(.

First, we need more dependencies on Windows:

diff --git a/build/gtk-bundle-from-msys2.sh b/build/gtk-bundle-from-msys2.sh
index 408d727f..529fa24a 100644
--- a/build/gtk-bundle-from-msys2.sh
+++ b/build/gtk-bundle-from-msys2.sh
@@ -21,7 +21,7 @@ EXE_WRAPPER_64="mingw-w64-x86_64-wine"
 # enchant, hunspell - for SpellCheck plugin
 # lua51 - for GeanyLua plugin
 # gnupg, gpgme - for GeanyPG plugin
-# libsoup - for UpdateChecker plugin
+# libsoup3 - for UpdateChecker & GeniusPaste plugins
 # libgit2 - for GitChangeBar plugin
 # gtkspell3 - for GeanyVC plugin
 # the rest is dependency-dependency
@@ -30,6 +30,7 @@ ca-certificates
 ctags
 ctpl-git
 enchant
+glib-networking
 gnupg
 gpgme
 http-parser
@@ -39,8 +40,9 @@ libgcrypt
 libgit2
 libgpg-error
 libidn2
+libproxy
 libpsl
-libsoup
+libsoup3
 libssh2
 libsystre
 libunistring

(The diff probably won't apply cleanly against this PR but should be easy enough to adopt, sorry.)
It is important to remove "libsoup" as it cannot be installed together with "libsoup3". There is a runtime warning when both libraries are loaded.
This obviously has a direct impact on the UpdateChecker plugin, so we probably need to merge them in order and rebase the latter one after the first one is merged.

Then, I got a compiler warning which might or might not be new but still worth to look at:

In file included from C:/msys64/mingw64/include/glib-2.0/glib/giochannel.h:36,
                 from C:/msys64/mingw64/include/glib-2.0/glib.h:56,
                 from C:/msys64/mingw64/include/glib-2.0/gobject/gbinding.h:30,
                 from C:/msys64/mingw64/include/glib-2.0/glib-object.h:24,
                 from C:/msys64/mingw64/include/glib-2.0/gio/gioenums.h:30,
                 from C:/msys64/mingw64/include/glib-2.0/gio/giotypes.h:30,
                 from C:/msys64/mingw64/include/glib-2.0/gio/gio.h:28,
                 from C:/msys64/mingw64/include/libsoup-3.0/libsoup/soup-types.h:9,
                 from C:/msys64/mingw64/include/libsoup-3.0/libsoup/soup-auth.h:8,
                 from C:/msys64/mingw64/include/libsoup-3.0/libsoup/soup.h:11,
                 from geniuspaste.c:22:
geniuspaste.c: In function 'json_request_new':
C:/msys64/mingw64/include/glib-2.0/glib/gstring.h:74:5: warning: ignoring return value of 'g_string_free_and_steal' declared with attribute 'warn_unused_result' [-Wunused-result]
   70 |   (__builtin_constant_p (free_segment) ?        \
      |   ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   71 |     ((free_segment) ?                           \
      |     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   72 |       (g_string_free) ((str), (free_segment)) : \
      |       ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   73 |       g_string_free_and_steal (str))            \
      |       ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   74 |     :                                           \
      |     ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   75 |     (g_string_free) ((str), (free_segment)))
      |     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
geniuspaste.c:612:5: note: in expansion of macro 'g_string_free'
  612 |     g_string_free(str, FALSE);
      |     ^~~~~~~~~~~~~

When trying to paste some file, Geany crashes for me when it is executed natively. I don't know yet why this happens and I didn't manage it yet to debug it with gdb. This might be related to my system where I use some firewall rules to block unwanted internet access from Windows and also have proxy connections configured and after all, it is Windows 7.

When I start Geany from within the MSYS2 environment, pasting some file works to some extend, at least no crash :).
But the response look weird:
Screenshot_2024-05-04_12-48-33.png (view on web)
The link below "\1" is actually also "\1".

The response_str in pastebin_parse_response is "https://www.geany.org/p/6vzVo" which is correct. So there seems to be a problem with parsing the URL. I don't know if this related to these changes or not.

Anyway, with the bundle changes above, we should get usable installers from the CI and so others could test it as well with a less weird Windows setup :).

—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you are subscribed to this thread.