On 2014-11-19 21:33, Gianluca Anzolin wrote: > On Wed, Nov 19, 2014 at 09:24:27PM +0100, Felix Fietkau wrote: >> On 2014-11-19 12:58, Gianluca Anzolin wrote: >> > On Wed, Nov 19, 2014 at 08:24:49AM +0100, John Crispin wrote: >> >> >> >> >> >> On 18/11/2014 10:54, Gianluca Anzolin wrote: >> >> > So in my opinion the "better" fix would be to populate the TZ >> >> > environment >> >> > variable. >> >> > >> >> >> >> so if a process is spawned and thent he TZ changes the process is stuck >> >> on the old TZ ? that seems totally broken >> >> >> >> John >> >> >> > >> > hehe, I knew I was overlooking something :) >> > >> > So would a patch to musl that reads /etc/TZ everytime be acceptable? >> I'd like to leave that decision to musl upstream. Please submit it to >> the musl list first before you send a patch for OpenWrt. > > Hi, > > please disregard the patch I already sent... > > I don't know if I have a chance to get something like that merged, given their > pragmatic view of what a libc should do. This is the reason I tried to avoid > changes to musl in the first place and use the TZ environment variable > instead. Isn't a pragmatic view a good thing here? ;) Even if the patch isn't accepted, it's still worth a try. We can of course carry a non-upstream patch in OpenWrt if things don't work out, but I'd like to avoid that if possible.
- Felix _______________________________________________ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
