https://bugs.kde.org/show_bug.cgi?id=372023
Nikita Melnichenko <[email protected]> changed: What |Removed |Added ---------------------------------------------------------------------------- Assignee|[email protected] |[email protected] Keywords| |reproducible Status|CONFIRMED |ASSIGNED --- Comment #10 from Nikita Melnichenko <[email protected]> --- Theo, thanks for sharing all your investigation results and a potential fix. (In reply to Theo from comment #8) > What would this setup roughly look like? Using my distributor's > debuginfo/debugsource packages and investigating with GDB Some people hesitate to do this and to compile the modified code. It's great that you can do it by yourself! > I start to believe that KCompressionDevice is broken, and unless someone > explains to me why I'm wrong I doubt we have KArchive experts here. If you feel there is a bug in KCompressionDevice implementation, could you please file a bug against the product "frameworks-karchive" to discuss with the devs? This may help other apps too. Some thoughts why KCompressionDevice might have been used instead of a plain file. If you look into KIso::KIso and KIso::prepareDevice, there is a code to guess mime type and take a different action. AFAIK, ISO format doesn't support any internal compression, so I guess it's for supporting .iso.gz, .iso.bz2 on the fly. > Insert dev->seek(0); The workaround seems to be harmless as there is another seek that follows. I'll check more and prepare a code review for the team when I have the right amount of free time. -- You are receiving this mail because: You are watching all bug changes.
