Fran Burstall <[email protected]> writes: > 4. Maybe an alternative to emms-print-metadata would be more > attractive. It seems that there are a zillion wrappers to taglib out > there. One example is https://pypi.org/project/pytaglib/ which comes > with a tag reader pyprinttags which (a) works and (b) was easy to > install: pip install pytaglib (assuming that libtag is already > present on the system). All that would need to be done to use this > in emms would be to get emms-info-libtag to parse the output.
I've also written to emacs-devel asking them if they have any input on how we can streamline access to taglib for Emms. Unless the people from emacs-devel come up with some brilliant idea we should try instead, we should try pytaglib on people after releasing 5.4 (first of May). > ---Fran > > > On Fri, 10 Apr 2020 at 19:47, Yoni Rabkin <[email protected]> wrote: > > Bruno Félix Rezende Ribeiro <[email protected]> writes: > > > Yoni Rabkin <[email protected]> writes: > > > >> I've personally never used ELPA and don't understand the point > of it. > > > > Why do you say that you “don't understand the point of it”? > > ...because I don't (I'm usure of what you are asking here > though.) > > However, I'll try to get Emms into ELPA if that is what people > want. I'm > maintaining Emms on behalf of GNU; not just for my own personal > benefit. > > > >> One issue would be to figure out how to ship the src/ > directory with > >> the non-elisp parts and how to get those compiled. > > > > How these work? I presume they are compiled and used as > sub-processes. > > Would not the new modules[1] support be more appropriate here? > Indeed, > > we’d have to figure out how those architecture-dependent parts > are > > supposed to get distributed. I think packages that use them > have a > > compilation script written in Lisp that is triggered on-demand > in the > > first use. > > > > > > Footnotes: > > [1] (info "(elisp) Writing Dynamic Modules") > > There are a few ways of doing this. > > Ideally, I would want to get rid of emms-print-metadata in favor > of > calling: > > * a library linked to emacs when it was compiled, or > dynamically > called by emacs > > * a command line tool that can be easily installed on an > endorsed > gnu/linux distribution > > The current situation of needing to compile a bespoke piece of > software > for taglib is suboptimal and confuses many people (based on > questions > I've received over the years). Moreover, "automating" the > compilation > won't work unless you also have a way of automating the > installation of > compilation tools onto the machine as well. > > It's still better than bongo's method of relying on filenames, > but that > isn't saying a whole lot. > > Perhaps the "modules" method is good, but depending on how new it > is, it > may leave behind everyone who wants to enjoy Emms but is stuck on > an > older emacs. I'll have to look into it. > > ...it seems like we have a worthy target for the 6.x release for > Emms. > > -- > "Cut your own wood and it will warm you twice" > > > > -- "Cut your own wood and it will warm you twice"
