pan-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Pan-users] [judgefudge tree] My eyes, they bleed, my stomach, it d


From: Duncan
Subject: Re: [Pan-users] [judgefudge tree] My eyes, they bleed, my stomach, it doesn't feel so well! (Colors screwed up.)
Date: Wed, 26 Oct 2011 08:43:31 +0000 (UTC)
User-agent: Pan/0.135 (Tomorrow I'll Wake Up and Scald Myself with Tea; GIT 045ef68 /st/portage/src/egit-src/pan2)

Duncan posted on Wed, 26 Oct 2011 06:41:45 +0000 as excerpted:

> FWIW, I do see the new "Read collapsed thread" color preference on the
> color tab of pan prefs, and its background was rather white before I
> adjusted it, but I DID adjust it and I'm still seeing the problem.
> Actually, that setting doesn't appear to do much for either collapsed or
> expanded threads, unread or not.
> 
> So I believe it's a screwup in the new read-collapsed-thread coloring
> code.  Apparently it's not applying the configured gtk+ color scheme any
> longer, thus defaulting to the rather sickening (for me) dark/black on
> near-white that I think must be the gtk default without color-scheming.

OK, now that I finally got the screenshot to post...

I think I traced the commit!  WTF?  The commit description says nothing
about changing colors in the GUI, it's all about SSL, but this has nothing
to do with SSL!  Suffice it to say this commit would never fly on the
kernel list, as a requirement there is that all commits do one thing,
only, and do it atomically.  Why is this commit changing underlines to
colors in the GUI, when it's SUPPOSED to be dealing with SSL, NOT screwing
around with colors in the GUI?

Be that what it may:

commit d2a0345a62449aa8d656e1bed531c5b119d20f69
Author: Heinrich Müller <address@hidden>
Date:   Sun Oct 23 03:30:47 2011 +0200

    fixed stupid ssl multithread pthread bug

Part of which is:

diff --git a/pan/gui/header-pane.cc b/pan/gui/header-pane.cc
index c8bd6ee..47df21c 100644
--- a/pan/gui/header-pane.cc
+++ b/pan/gui/header-pane.cc

@@ -298,7 +299,9 @@ HeaderPane :: render_subject (GtkTreeViewColumn * ,
   g_object_set (renderer,
     "text", text,
     "weight", (bold ? PANGO_WEIGHT_BOLD : PANGO_WEIGHT_NORMAL),
-    "underline", (underlined ? PANGO_UNDERLINE_SINGLE : PANGO_UNDERLINE_NONE),
+    "foreground",  unread ? p.get_color_str 
("score-color-read-fg",TANGO_ORANGE).c_str() :
+                   "black",
+    "background", unread ? p.get_color_str 
("score-color-read-bg","white").c_str() : "white",
     NULL);
 }


If I'm reading that correctly (with the hint in the new GUI pref as
mentioned), it changes the underline-if-unread-in-collapsed to ...
the preference-configured colors, falling back to TANGO_ORANGE for
the foreground and white for the background, if not configured.

That part's fine, since it at least should be configurable via the
preference dialog.

What's *NOT* so fine, however, is the fallbacks if that
unread-childposts condition does NOT hold.

Unfortunately, it hard-codes those, to something not only WAY
different than my normal color-scheme, but something that can
make me feel, literally, mildly nauseous, sick!

Those cannot be hard-coded like that, at least not without making
/someone/ unhappy (and potentially sick), since you can't hard-code both
my light-on-dark color-scheme AND the more common dark-on-light scheme.
Basically, at this point, the easiest fix is probably to setup yet
another preference, for "normal" subject colors, foreground/text
and background, and substitute those in there in place of the 
hardcoded "black" in the foreground line, and corresponding hardcoded
"white" in the background line.

-- 
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




reply via email to

[Prev in Thread] Current Thread [Next in Thread]