Ping? This is making it very hard for me to review patches of potential SoC students and helping them getting used to the usual review process.

Nevermind, it turned out to be a user problem. :( IRC discussion if anyone is interested in details:

<wicked> hmm.. wonder if Splinter chokes on the different header syntax
<zeenix> it's a bit strange though that it's only hitting this particular person
<wicked> the one that includes the start of section
<zeenix> start of section?
<zeenix> afaik they were uploaded using git-bz
<wicked> I mean this line: @@ -49,7 +79,6 @@ public async LibvirtSystemImporter () throws GLib.Error {
<wicked> the section is the one after second @@
<wicked> but I'm not sure if I understand Splinter output at all :D
<wicked> ohhh... that patch IS wrong.. 
<wicked> @@ -9,6 +9,7 @@ 
<wicked>  private class Boxes.LibvirtSystemImporter: GLib.Object {
<wicked> that should be one line, not two ^^
<zeenix> hmm...
<zeenix> interesting
<wicked> just like my earlier example
<zeenix> i re-uploaded the patch on one of the bugs and it works 
<wicked> and does this really include the @@ -, +, @@  on top here https://bug745691.bugzilla-attachments.gnome.org/attachment.cgi?id=324838&action=diff&collapsed=&context=patch&format=raw&headers=1
<wicked> that can also be PatchReader messing up but I don't think it's used for raw output (it is for diff output, though)
<wicked> ohh, that strange bit is also on the first example
<zeenix> yeah, i think it's just one person who is doing something strange
<zeenix> now i recall that he told me he is using `git format-patch` and attaching them manually
<zeenix> not sure how got that wrong 
<wicked> maybe opening the diff in an editor, that rewraps it and then he saves it? that would explain the bleeding of line..
<zeenix> weird thing is that except for splinter, everything else works fine
<wicked> the top thing could be the file change stats messing PatchReader
<wicked> don't think Splinter uses PatchReader to read the patch but rather something own.. but I am so surprised PR doesn't choke :)
<wicked> the latter patch has actually the lines removed/added wrong.. like there's a missing empty line at the bottom
<wicked> PR doesn't use those but simply formats the diff parts it encounters so it doesn't choke on that
<wicked> zeenix: but yeah, those patches are malformed :(