GNOME Bugzilla – Bug 580930
GTK+ 2.16.x ABI break for spinbuttons
Last modified: 2010-08-26 00:09:08 UTC
Please describe the problem: When pasting CSV data into Gnumeric, sometimes it truncates the rows. Steps to reproduce: 1. Load the attached file into Gedit, do Ctrl-A Ctrl-C to copy. 2. Open Gnumeric, press Ctrl-V to paste. 3. Do Alt-F Alt-F Alt-I for "Forward -> Forward -> Finish" 4. Only the first 490 out of 500 rows are pasted. You will notice that in the first import screen of step 3 above, in the "From line / To line" settings, "To line" is set at 490, and cannot be increased to 500. Additionally, if you go back to Gedit and copy only the last 10 lines of the file, and try to paste them at row 491 of the spreadsheet, the "To line" setting cannot be increased beyond 1, so you can only paste one more line at a time to the end of the first 490 lines. Curiously, for some other datasets (even ones with more lines), none of the above is a problem... Actual results: Expected results: Does this happen every time? Other information:
Created attachment 133688 [details] The dataset that gets truncated when pasted into Gnumeric.
I cannot reproduce this with either 1.9.6 or 1.8.2. The bahaviour with 490 strongly suggests that Gnumeric only gets the first 490 lines in your case. Could you please try another editor? Or the terminal window.
There is some discussion of this at: https://bugs.launchpad.net/ubuntu/+source/gnumeric/+bug/264167
See also https://bugs.launchpad.net/ubuntu/+source/gnumeric/+bug/369650 in brief this is the result of changes to gtk 2.14.
Yes, that looks like the problem.
I have pushed workarounds to this problem http://git.gnome.org/cgit/gnumeric/log/?h=origin/gnumeric-1-8
Created attachment 133712 [details] Script to fixup glade files The script can, I hope, fixup glade files. It zeros the last adjustment number for every GtkSpinbutton.
This is now fixed in both branches (HEAD and 1.8).
*** Bug 581014 has been marked as a duplicate of this bug. ***
*** Bug 581493 has been marked as a duplicate of this bug. ***
*** Bug 593703 has been marked as a duplicate of this bug. ***
*** Bug 594186 has been marked as a duplicate of this bug. ***
*** Bug 596868 has been marked as a duplicate of this bug. ***
This is still an issue with Fedora 12 beta. It has survived three Fedora releases (along with bug 566652). Does anyone have sway with RedHat that can help get the upstream fixes pushed into F12 in time for release? Thanks! See: https://bugzilla.redhat.com/show_bug.cgi?id=505001
Well, I assume that GTK wanted to break Gnome productivity apps for years to come, so RedHat just follows their intent.
They need to pick patches from http://git.gnome.org/cgit/gnumeric/log/?h=gnumeric-1-8
Yes, but who inside RedHat is likely to be responsive to a request to pick those patches (and/or move to Gnumeric-1.9.x) before Fedora 12 is released? I reported the RH bug months ago and there's only one guy CC'd. I don't want to be the annoying guy with the pet bug -- but this one makes Gnumeric pretty annoying to use on Fedora systems. Thanks.
You could take the stand that it's a GTK+ problem and try to bring it to the attention of mclasen@redhat.com -- if nothing else, he might know who to point at.
Emailed him, thanks for the pointer.
Got the following from Robert Scheck, FYI. -- thank you for pointing out that. I've applied the fixes from gnumeric 1.8 stable branch to F-11, F-12 and devel. As F-10 is shipping gnumeric 1.8.2 right now, I'm not sure whether the issue really applies there as well. And I am a bit unsure whether I should update gnumeric 1.5 months before EOL to 1.8.4 or whether this maybe breaks more than it helps. Let's see. Note, I'm not the gnumeric package maintainer, just a provenpackager, who updated gnumeric some time ago to 1.8.4, because the package maintainer did not do it. So this is just another provenpackager activity. -- Question: is Gnumeric 1.9.14 / HEAD stable enough for suggesting it be shipped with Fedora 12? (Rephrased: would it be better including 1.9.x in Fedora than the broken 1.8.4 that they have been shipping for almost three releases?)
*** Bug 608638 has been marked as a duplicate of this bug. ***
*** Bug 627989 has been marked as a duplicate of this bug. ***