Sorry for the noise, accidental Ctrl-Enter.
On Tue, 16 Mar 2010 00:06:58 +1100 Lex Trotman elextr@gmail.com wrote:
Probably we can fix this to return "none" instead of "common" and adjust all calling code to still work.
Or just store it in filetypes.common, doesn't matter so long as we write to the file we read from :-) Its a separate section so it won't upset any other content.
I suppose this is OK. (I think there may be good reasons to have separate filetypes.common and filetypes.none so that filetypes.common only has common settings. But really this is yet another separate issue ;-))
But for now (if it's easier) I think we can just disable the None filetype's Execute command (and its filetype build commands if there is a situation where they're not disabled [but I can't find one myself]).
Wierd, for me they are always enabled, looking at the code the only time they are set insensitive is if something is overloading them (but you don't have a project), or if doc==NULL or doc->filetype==NULL
No, with doc open and GDB proves doc->ft is not NULL.
BTW which function(s) are you talking about (I haven't looked yet)?
Maybe we've got another uninitialised variable that is NULL on your machine but not on mine, but I can't find it. GCC gives no warnings with -Wall, but thats not authoritative.
I think you only get uninitialised warnings if you have -O2 on. I might try this later.
Are the regex fields sensitive or not?
Yes, both.
It sounds like there is something different in your setup if the dialog
sets
filetype commands insensitive somewhere.
I don't think there can be because I tried this with a new config dir (-c option). Same for the filetype None issue above.
I'm using a separate user so I don't conflict with the new stuff, so the whole .config/geany is clean, is that the difference maybe? Don't see how though?
With -c some_dir geany shouldn't read .config/geany.
...
I'm giving up for the night, if you have any sudden insight during your day let me know, otherwise tomorrow.
OK, cool ;-)
Regards, Nick