Stuart Henderson <s...@spacehopper.org> writes:
> On 2021/07/14 18:34, Omar Polo wrote: >> Hello, >> >> Straightforward update for net/tdlib from 1.7.3 to 1.7.5. telega.el >> (not in ports) works, and net/tg is happy too. > > Thanks, committed - > >> I got the following error while trying to update the port, thought. >> Probably was something strange on my end, since I had the package tdlib >> installed locally for a long time before it went into ports, but I'm >> mentioning it anyway. It seemed like the port infrastructure treated >> tdlib-1.7.3 and tdlib-1.7.5 as different packages: >> >> ----8<---- >> % make update >> ===> Updating for tdlib-1.7.5 >> Not installed, no update >> % make install >> ===> Verifying specs: c++ c++abi pthread crypto m ssl z >> ===> found c++.8.0 c++abi.5.0 pthread.26.1 crypto.46.3 m.10.1 ssl.49.0 z.6.0 >> ===> Installing tdlib-1.7.5 from /usr/ports/packages/amd64/all/ >> Can't install tdlib-1.7.5 because of conflicts (tdlib-1.7.3) > > That's likely a differing pkgpath from previous test builds, > if you still have the old package around check the @pkgpath line in > +CONTENTS. Thanks for the pointer! :) >> --- tdlib-1.7.5 ------------------- >> Can't install tdlib-1.7.5: conflicts >> Couldn't install tdlib-1.7.5 >> *** Error 1 in . (/usr/ports/infrastructure/mk/bsd.port.mk:2159 >> '/var/db/pkg/tdlib-1.7.5/+CONTENTS': @/usr/bin/env -i PKG_TMPDIR=/var/tmp >> TE...) >> *** Error 2 in /usr/ports/net/tdlib >> (/usr/ports/infrastructure/mk/bsd.port.mk:2602 'install': @lock=tdlib-1.7.5; >> export _LOCKS_HELD=" tdlib...) >> ---->8---- >> >> Oh, and another thing. I haven't bumped the solib version, but I'd like >> if another pair of eyes could verify I'm not mistaking anything. Below >> the diff of `nm -D' before and after the update. > > There's a new exported function td_set_log_message_callback so a minor > bump is warranted. I'm ashamed, I haven't spotted td_set_log_message_callback... Thanks for checking! > A major bump would be needed if there's either a removal or a change in > number/order/types of function arguments, or layout of some structs. > I started to look for these (diffing headers is usually the way) > but a lot of comments were added and it got rather hard to read so > I may have missed something, though it's relatively unlikely in a > minor update. Thanks, Omar Polo
