Stuart Henderson <[email protected]> writes:
> On 2021/05/05 16:16, Omar Polo wrote: >> >> Stuart Henderson <[email protected]> writes: >> >> > On 2021/05/05 15:34, Omar Polo wrote: >> >> Hello, >> >> >> >> Please find attached net/tdlib. It's required by another port I'm going >> >> to submit (net/tg), and it will also come in handy for the fellow emacs >> >> users (telega.el depends on net/tdlib; unfortunately I can't make a port >> >> for it because it depends on various other stuff on MELPA.) >> >> >> >> Since I've been using this for a year or so, I'd also like to take >> >> MAINTAINER. >> >> >> >> $ pkg_info tdlib >> >> Information for inst:tdlib-1.7.3 >> >> >> >> Comment: >> >> cross-platform library for building Telegram clients >> >> >> >> Required by: >> >> py3-telegram-0.14.0 >> >> >> >> Description: >> >> TDLib (Telegram Database Library) is a cross-platform, fully >> >> functional Telegram client that can be easily used from almost any >> >> programming language. >> > >> > upstream has "TDLib (Telegram Database library) is a cross-platform >> > library for building Telegram clients" which sounds more accurate? >> >> Sure >> >> (in my defense I believe they changed the description at some point) >> >> >> WWW: https://github.com/tdlib/td >> > >> > how about https://core.telegram.org/tdlib ? >> >> yep, definitely better! >> >> > : # XXX: because of GH_COMMIT >> > : PKGNAME = tdlib-1.7.3 >> > : DISTNAME = tdlib-${GH_COMMIT} >> > : >> > : GH_ACCOUNT = tdlib >> > : GH_PROJECT = td >> > : #GH_TAGNAME = v1.7.0 # XXX: keep in sync with COMMIT below >> > : GH_COMMIT = 9964d24ec8e03bd0bc4d5fe33438f056d133fc1a >> > >> > bit messy; where does 1.7.3 come from? is there a need to use these >> > development versions rather than an actual release? >> > >> > https://github.com/tdlib/td/issues/1243 >> >> While upstream has this scheme where 1.7 is stable and 1.7.X is >> considered development, there are various clients that tracks these >> "unofficial minor" releases. telega.el[0] is one of these (and >> unfortunately I cannot make a port for it since it depends a various >> other emacs libraries on MELPA...) >> >> I'm trying to follow the version needed for telega.el, while also >> ensuring everything works in net/tg. The guix package does this too. >> >> If it's something unwanted, I can submit a port for the v1.7, but I'd >> personally prefer to keep these 1.7.X releases. > > No that's fair enough. I added a comment explaining this. > >> > : COMMIT = 9964d24ec8e03bd0bc4d5fe33438f056d133fc1a >> > : SUBST_VARS += COMMIT >> > >> > if the development versions are needed, this should reuse GH_COMMIT >> > rather than a separate var with the same contents. >> >> Also true! When updating from 1.7 from 1.7.3 I forgot to de-duplicate >> the vars. >> >> Attaching a tarball with all these fixes included but still with the >> 1.7.3 version. >> >> Thanks, >> >> >> [0]: https://github.com/zevlg/telega.el/blob/master/telega.el#L14 >> > > You missed s/COMMIT/GH_COMMIT/ in patch-gen_git_commit_h_sh; but let's > go with COMMIT=${GH_COMMIT} in the Makefile and keep the patch as it was, > so it's easier to change if it moves to GH_TAGNAME for some future release > (since you can't have both GH_COMMIT and GH_TAGNAME). I dropped the > comment about running the php-based splitter, I don't think we'll want > to use that in ports builds, and set DISTNAME (for a port with GH_COMMIT > it automatically adds a shortened hash so it's clear). > > Updated tgz attached, does that look alright? If so, this is OK sthen@ > if someone would like to import this / net/py-telegram / net/tg. Sorry for the delay, Yes, net/{tg,py-telegram,tdlib} are perfect now, thanks! Apologies for forgetting to update the patch gen_git_commit_h_sh, and also for the things I missed in py-telegram. Thanks,
