25 Mar 2024 |
WeeskyBDW | bugzilla is really that old ? | 14:25:43 |
Danny Colin | 25 yo 😛 | 14:25:58 |
justdave | It was actually original written in Tcl, and was ported to Perl because "that's what the cool kids use these days" at the time. | 16:06:29 |
justdave | * It was actually originally written in Tcl, and was ported to Perl because "that's what the cool kids use these days" at the time. | 16:06:38 |
WeeskyBDW | was another portage to more used langage (perl is cool but reall not that popular theses days) or not ? i know that a lot of work | 16:07:52 |
WeeskyBDW | * was another portage to more used langage (perl is cool but reall not that popular theses days) or not ? i know that a LOT of work | 16:08:03 |
ashimema | porting 25 years worth of development to a new language isn't an overnight job 😜 | 16:09:07 |
justdave | At the time of the Perl port the entire app was about 6 CGI scripts I think. | 16:09:27 |
justdave | There was talk in the early 2010s of trying to port it to Python but by then the app was so large that it would have taken years to do it and it would have stolen resources from doing anything else to it in the meantime, so it was never pursued. | 16:10:07 |
| * ashimema had no idea it was in Tcl before Perl | 16:10:14 |
ashimema | I've been there.. we ported a php app that was only ever meant to stand as a proof of concept.. it took us over 2 years to port it to a more modern codebase and that kinda killed the product | 16:11:07 |
justdave | The general long-term direction of things at the moment is to prioritize back end/API development in Perl and make the API complete enough that people could write front-ends for it in whatever language they wanted. | 16:11:38 |
ashimema | totally | 16:11:49 |
ashimema | I'm very happy with that approach | 16:11:54 |
ashimema | it's also how Koha is headed | 16:11:59 |
ashimema | more modern perl for the backend and modern JS from the front | 16:12:07 |
| * ashimema has a soft spot for Perl still though.. | 16:12:27 |
Danny Colin | In reply to @justdave:mozilla.org The general long-term direction of things at the moment is to prioritize back end/API development in Perl and make the API complete enough that people could write front-ends for it in whatever language they wanted. Yep. This is definitely the best first step. | 16:12:29 |
WeeskyBDW | Redacted or Malformed Event | 16:13:23 |
ashimema | indeed.. once there's a stable API you stand more of a chance of porting parts to different serverside stuff | 16:14:09 |
WeeskyBDW | (i deleted my message where i said that with api anyone can reimplement bugzilla api, that why ashimema said that) | 16:15:07 |
WeeskyBDW | * (i deleted my message where i said that with clear api specs anyone can reimplement bugzilla api, that why ashimema said that) | 16:15:28 |
justdave | Perl has definitely grown over the years... When I was screwing with the Docker code to get the test suite working again last week, some of the code I was having to touch (which was written by someone at Mozilla when they created the original Docker images) was async-io event code, which I hadn't seen done in Perl before. (Tying actions to events and passing futures around and so forth) | 16:16:43 |
justdave | My prior experience with async-io has all been in Python | 16:18:11 |
ashimema | Perl can do a lot these days 😜 | 16:19:02 |
ashimema | and Mojolicious is a lovely framework to work in. | 16:19:14 |
Danny Colin | I think the problem is always to find people fluent in it tho. | 16:19:46 |
ashimema | Yes.. I'd agree with that | 16:20:03 |
Danny Colin | It's the same with any programming language. | 16:20:08 |
ashimema | though we've had some success in on-boarding new devs of late.. they're learning perl and enjoying it so far | 16:20:29 |