Hello, I was wondering: why this ugly --disable-nls configure argument is hardcoded in so much portfiles? There should be a "nls" variant deleting this of configure.args. Localization is nice and already available in a lot of packages. I think getting ride of it is a big loss.
Le 07-04-22 à 12:41, N_Ox a écrit :
Hello, I was wondering: why this ugly --disable-nls configure argument is hardcoded in so much portfiles? There should be a "nls" variant deleting this of configure.args. Localization is nice and already available in a lot of packages. I think getting ride of it is a big loss.
I can't talk for everybody but I know that : 1- It once helped gnome-panel and inkscape to not crash at startup (this is now fixed) 2- Removing it from the aspell port will make EVERY GNUstep ports crash at startup 3- Removing it from the denemo port triggers some odd behaviours and does not work anyway. 4- cross/avr-gdb/Portfile:# without --disable-nls the build process also breaks So much portfiles ... Um, that leaves about 25 ports out of 3887 with 10 of them being gcc ports yves
Le 22 avr. 07 à 19:11, Yves de Champlain a écrit :
I can't talk for everybody but I know that :
1- It once helped gnome-panel and inkscape to not crash at startup (this is now fixed) 2- Removing it from the aspell port will make EVERY GNUstep ports crash at startup 3- Removing it from the denemo port triggers some odd behaviours and does not work anyway. 4- cross/avr-gdb/Portfile:# without --disable-nls the build process also breaks
So much portfiles ... Um, that leaves about 25 ports out of 3887 with 10 of them being gcc ports
yves
Well... these seems to be a very good reason. "So much portfiles" were in fact nano and coreutils', I've removed -- disable-nls and try to install them: that works. This disabling should be done on a case by case basis, no? N_Ox. PS : Sorry Yves for the first mail, I've forgotten to put the ml address in the cc header.
participants (2)
-
N_Ox
-
Yves de Champlain