Open snow-abstraction opened 4 years ago
Hi Douglas, I have this library very useful as it is. We have managed to use it to do all our scheduling in a school I was working at. I am sure there are always ways to improve but it is pretty decent as is...
Just my 2 cents.
Le mar. 31 déc. 2019 à 00:17, Douglas Potter notifications@github.com a écrit :
This is useful project but its development seems to have stalled. I hope this project can continue to improve. Does anyone know the answers to the following:
- Is there some road map?
- Some list of feature requests thought to be most useful for users?
- Is there alternative Python project that is more active?
It might be that this project is done and is complete for its scope. And then people needing faster solutions or more complex models directly use integer programming, constraint solvers, SMT solvers, meta-heuristics, etc themselves.
I write this as I am surfing github and gitlab looking for some meaningful project to contribute to.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/timnon/pyschedule/issues/90?email_source=notifications&email_token=ABXH2IB6MCYTL4R4C4QPJPLQ3JQPDA5CNFSM4KBQHVS2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4IDM74CA, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABXH2IDGTZG7O47KP6EEVNDQ3JQPDANCNFSM4KBQHVSQ .
Hi, there has indeed been little progress during the last year on this project, mostly due to timing contraints. If anybody wants to contribute or even take over, please step forward!
The initial intention of this project was to build a simple frontend to multiple backends (different contraint programming solvers, integer programming formulations, heuristics). This turned out to be quite cumbersome, since the different backends have very different strengths and weaknesses, and to support all of these is just too much work. In the end, it turned out that using a time-indexed integer-programming formulation with lots of redundancy-removals covers a lot of small to medium-sized use-cases very well, and offers maximum flexibility. So over time, everything was build around this formulation.
@timnon I sympathize that this is a huge task. I think the choice "time-indexed integer-programming formulation" is a good one.
I don't know if I will contribute but I think it would help others understand if:
2.) some list of areas known to need the most improvement. (This is something deeper than a list of the current open issues and would require knowledge of the code base and history of issues.)
Finally, do you know of any organizations that could share their use cases?
Again excellent work.
Hi guys, I noticed that the project seems to be frozen since the last release. The last commit is from early 2019, more than 18 months ago. No problem anyway, I know that maintaining an open source project can be quite a tough work !
For my own needs, I created a fork at https://github.com/tpaviot/pyschedule, which includes, in an unordered list: continuous integration and testing at Azure, online tests at mybinder, pylint, python2 drop and py3k only support, draft for plotly plotter. I have no roadmap, I just add what I need when I need it. Please @timnon let me know if ever/when you come back to this repository, I may have some PR to submit.
At last, thank you for this amazing work.
This is useful project but its development seems to have stalled. I hope this project can continue to improve. Does anyone know the answers to the following:
It might be that this project is done and is complete for its scope. And then people needing faster solutions or more complex models directly use integer programming, constraint solvers, SMT solvers, meta-heuristics, etc themselves.
I write this as I am surfing github and gitlab looking for some meaningful project to contribute to.