After opening the "Preferences" dialog, the mouse cursor is not visible in the main area anymore, even if you move it around. It is visible when hovering other areas such as the thumbnailer or the "Preferences" dialog, though. After closing the "Preferences" dialog, everything works as usual. Strangely enough, this does not happen with other dialogs such as the "Library" dialog or the "Enhance image" dialog.
Sorry, but I just don't understand what the patch is supposed to do. As a user, I don't see any difference to the unpatched version. Could you please explain it to me?
Not ignore empty directories (see use case previously discussed).
Okay, thanks. The patch indeed makes it possible to enter and leave empty dirs via
PgDown. However, with the patch applied, you can also enter directories you are not allowed to read. Let's skip them. (With the patch applied, a dir without read permission also breaks navigation, so you cannot leave it withPgDownbutCtrl+Petc. Skipping them "solves" this as well.)I think we should not skip on error, no matter what. We should show an error message (invalid/unsupported archive, permission denied on directory, ...), and allow the user to switch to the next/previous archive/directory with the corresponding command.
Also, I just tested with a directory without permission, and I can use Ctrl+N/P to get out.
Note: I pushed the fixes for the 2 other issues (Ctrl+N/P when started without an input + close provider on file close).
See my branch for an updated patch that support using PgUp/PgDown for switching out of an invalid "file" too.
Last edit: Benoit Pierre 2015-03-14
I think that's fine, too. Okay, so no skip on error.
Yes,
Ctrl+Pworks. However,PgDown, for example, does not but it should.Thanks for the update on your branch. I am not sure about all commits there, but at least the way you can navigate through dirs definitely feels nice. Please go ahead and push the corresponding commits to SVN.
Done.
Note: the toolbar/menu controls should not be disabled.
Thank you for [r1216].
Besides this, it just occurred to me that
All in all, the current HEAD looks very promising. I think I will contact Oddegamra soon so he can test the Windows-specific changes, integrate
czipfileand manage the licence issues you mentioned earlier.Related
Commit: [r1216]
Right, I think a new ticket should be created, listing the various things that need to be fixed for the UI to be consistent when handling those cases.
I received a reply from Oddegamra the other day. Regarding this "force PIL on Windows" issue, he says the following:
So, what should we do? Right now, I can think of two different ways:
I think we really need someone with access to a real Windows machine to test whether the problem is present or not with PyGTK 2.24.2. I mean those messages are pretty old.