On 19 November 2010 06:04, Nick Treleaven <nick.trelea...@btinternet.com> wrote: > On Mon, 15 Nov 2010 08:57:51 +1100 > Lex Trotman <ele...@gmail.com> wrote: > >> >> > Also, g_file_replace_contents does have a make_backup argument we could >> >> > provide an option for. This might handle the disk exhaustion problem. >> >> >> >> To summarise the long thread for you Nick, g_file_replace_contents and >> >> g_file_replace can't be used because there is NO WAY to stop >> >> g_stream_close doing the rename of the temp file over the old file >> >> even if the write to the temp file fails. So you can get a broken >> >> output file. >> > >> > I know. But at the least the user has the backup file then. >> >> Attached is a patch to actually make the backup :-) > > Ok. We do have some backup functionality in the Save Actions plugin, > but anyway this pref may be more efficient.
Yes, save actions is fine on a local disk and much more flexible since you can choose name and location for backups, but it copies the file which is a performance problem for remote filesystems which was one of the initiating complaints on the long "safe file save" thread IIRC. > > I think either: > 1. unsafe_save_backup should be renamed to gio_unsafe_save_backup to > avoid confusion. Ok, attached. > 2. call the pref make_backup and make it work for all save > implementations, and maybe have a GUI option for it. To do this we would almost be writing the whole of GIO with the algorithm I outlined previously. So we probably should do that instead. For now lets just offer this easy option that provides improved performance on remote filesystems. Cheers Lex > > Nick > _______________________________________________ > Geany-devel mailing list > Geany-devel@uvena.de > http://lists.uvena.de/cgi-bin/mailman/listinfo/geany-devel > _______________________________________________ Geany-devel mailing list Geany-devel@uvena.de http://lists.uvena.de/cgi-bin/mailman/listinfo/geany-devel