Interest in Mythbackend Builds

[mr_tea]

I wanted to gauge if there is interest in MacOS builds of mythbackend and related programs. While I have the backend compiling based on the Ansible frontend script, before I take it to a completely working state, I'd like to know if others are interested in a standalone application or if there is interest at all. That will define how much effort I put into it (if there isn't interest, I'll just keep it simple and use the MacPorts install locations).

[pvr4me]

Which version(s) are you targetting?

Craig

[jhoyt]

I think if it's a great idea if you've got the time to put in the effort. I've received multiple requests for mythbackend (v31 and master) off list. I hadn't gone down the path myself as bundling mysql/mariadb into an application looked like more trouble than I had time to support. I suspect bundling the mythtv-backend application should be pretty straightforward and very similar to what I needed to do to mythfrontend (...although you may need to bundle perl - definitely will need to bundle python).

I'd be very happy to walk you through what I learned about application bundling (and why I do things in build script, why the particular order) if it'd be helpful.

[mr_tea]

I'm testing with 31 right now. I need to get the library copy and load lookups configured right, but I haven't taken the time as I was curious first if others would want to use this too before I picked a direction (if it is just for me, I'll probably hack it together).

My initial plan is to try to get it up as a bring-your-own MariaDB and web server (for mythweb) design so that I don't have to package those in as well. I think that will be considerably easier to get up and running initially from my perspective, but certainly more complicated for any user who doesn't know how to do it themselves. If we continue the package route, I'm interested in trying a single app package with everything you need, but only after the core server works.

Is there a preference towards trying to package the backend as an app, or would a backend only version in MacPorts be okay to go along with the packaged version of the frontend? Just curious what the preferences are. I think I'd prefer a MacPorts backend as it better matches the servers design, but I've been doing that for many years now, so it feels very natural to me.

[davef]

I would very much like a backend build. Since frontend and backend versions need to match I've been stuck at .28. What makes it a bit more difficult is my server is an M1 machine. Either a MacPorts or app would be fine by me. My PITA is that I'm running MySQL 5.6.27 so I will need to update that also. I'm not looking forward to that!

[mr_tea]

I’m working with Craig’s test port for 31. I haven’t had a lot of time recently, but plan to get back to it next month. The biggest pain point seems to be XMLTV. It is now required, and setup has to be done from the command line for now as I’m not aware of any GUI config tools. I’m considering posting a test version to get feedback if anyone is interested and doesn’t mind the command line requirements.

[footlooseboss]

I'm interested and don't mind the command line requirements

perhaps put it on git/bit/hub/bucket/lab

[nxnw]

I'm very interested in a more up to date backend that runs in the background, rather than as an application. I have been running craig's excellent .28 build for years, but .28 was never really ready for prime time (webfrontend, in particular) and it would be great to have a newer build. Command line is fine.

[jhoyt]

The compile script in the packaging repo does successfully build and install mythbackend and mythfrontend as command line applications if you specify the prefix. For example, to install 33 in /usr/local/bin simply run:

Code: Select all

compileMythtvAnsible.sh --version=fixes/33 --repo-prefix=/usr/local/bin

Mysql / mariabdb are installed via ansible/macports to compile, so after the install you'd need to get your db up and running then do any manual config necessary to get mythbackend up and running.