Christian Dysthe posted on Tue, 15 Jul 2014 23:19:08 -0500 as excerpted: > Hi, > > I am having a strange problem with Pan built from git (GTK 2). One one > particular news group the threading indication/indentation is gray. On > all other newsgroups it's white. I have tried to unsubscribe to this one > newsgroup and then add it again, but I get the same result. > > Here's a screen shot of the group that get gray background: > > https://www.dropbox.com/s/s2vpfleqxs7y27t/Pan140-2.png > > Here'a a screen shot of how all other news groups look: > > https://www.dropbox.com/s/w829sjrkudffmuj/Pan140-1.png > > Any idea why that one group get the gray background? Is this a setting > somewhere that I am missing?
Heh, your color-scheme is /entirely/ different than mine. This shot's a bit old (2013) but still representative. Pan is on the middle monitor. Note that it's a triple full-HD monitor stacked desktop for a full 1920x3240 px screenshot, full-size, so you'll probably need a scaling viewer to see the full thing at once, then scale to 100% to see it at full-size and scroll around, tho even "full size" won't do it justice unless you're running 42-inch TVs as monitors as I am (for the bottom two, anyway). http://wstaw.org/m/2013/05/11/duncan-fullscreen.png Tho I've switched pan's groups and body panes so groups is on the left now, and I got new contacts and reduced the font sizes somewhat. Anyway... That is indeed a rather strange problem. Given my dark-background color prefs I've had my share of strange color issues on pan and elsewhere, but I don't recall seeing anything like that, certainly not for one group only. Based on the above you are already building pan against gtk2, which would be my first recommendation, since pan built against gtk3 mostly works, but people do report strange issues with it reasonably frequently, that go away when they build pan against gtk2, and the gtk3 port simply doesn't get the testing or wide deployment of gtk2, so... But you're already building against gtk2, so that can't be it. Tho it wouldn't hurt verifying that. Run ldd pan in a terminal window, and check that it's not loading any gtk3 stuff. As I don't even have gtk3 on my box, it can't load it here, but verifying that it's only doing gtk2 on a box with gtk3 can't hurt. In particular, my libgtk and libgdk lines look like this (gentoo ~amd64 box, hex numbers omitted): libgtk-x11-2.0.so.0 => /lib64/libgtk-x11-2.0.so.0 libgdk-x11-2.0.so.0 => /lib64/libgdk-x11-2.0.so.0 libgdk_pixbuf-2.0.so.0 => /lib64/libgdk_pixbuf-2.0.so.0 FWIW, my gtk/gdk-pixbuf/cairo versions are as follows (the -r3 on cairo is a gentoo-specific revision number indication): gdk-pixbuf-2.30.8 gtk+-2.24.24 cairo-1.12.16-r3 The other possibility is that it's a color issue. You don't happen to have any scoring set that might apply to all articles in that group, do you? You can try fiddling with pan color prefs a bit, both for scores and for "other text", and see if that affects it. But I don't see how it could for just that group, /unless/ it was related to a score you had set that happens to affect everything in that group and little or nothing else, but score colors normally only affect the score column, so that shouldn't be it either. The only really group-specific settings are visible in group preferences. There's color, which normally only affects the group pane so shouldn't have any effect on the headers pane, and charset. Charset! That might actually be it! Might there be some post in that group with a strange charset used in either the subject or author header? Since the threading indention is in the subject column, that would be the strongest candidate. If there's some exotic charset used in a subject header, it's just possible that gtk's processing of it triggers a rendering bug that results in the off-color seen in whatever pan or gtk is using for the header spacing. If pan's using an exotic space character as it's threading spacer, as it might be... Admittedly that's reaching a bit for an explanation, but it's the best one I can think of that fits the evidence we have so far. If the group was unaffected until a certain date, and now it's doing this, you might check the posts around that date... You can also try: 1) With pan not running, move pan's data dir elsewhere temporarily, and restart pan with a clean config and no groups loaded. Fill in your server info and visit just that group. Pan won't have a record for it so will prompt to download headers. Try getting just the last say two headers, then tell it to get say 10, etc, until you can see at least one threaded header. Does it show up then? If not, try getting a few days/ weeks worth of headers and see if it shows up. Then try downloading headers back as far as you had them in your normal config (assuming your server has enough retention for that, of course). Does it then show up? The idea being to see if it's in a clean config, and if the headers of some message trigger the problem. 2) Try the same thing, but start with a fresh full user config, either configuring a new login user for the test, or temporarily moving your entire user config to backup so it starts with an entirely clean user config. Of course the idea here is to see if it's something in your user config, by starting with an entirely clean one. That should give us some data on where the problem is, anyway. If it's happening with a clean user config, then it's something either at the distro level or upstream. If it's not happening there, but it's happening with a clean pan config, then it's gotta be something in your user config. If it's not happening with a clean pan config on your existing user, but starts happening when you download a bunch of headers for that group, then it's gotta be something in those headers triggering it, tho the problem's probably at the distro or upstream level. If it's not happening in a clean pan config and doesn't start happening when you download all the headers that you had in the group before, then it's gotta be something in the pan config, unless of course you couldn't get all the headers you had before due to server retention policies for that group, in which case it could still be something in the message headers. -- Duncan - List replies preferred. No HTML msgs. "Every nonfree program has a lord, a master -- and if you use the program, he is your master." Richard Stallman _______________________________________________ Pan-users mailing list Pan-users@nongnu.org https://lists.nongnu.org/mailman/listinfo/pan-users
