rocky
commented
7 months ago The archive comment in the project description is representative of activity .
It has not yet been archived in the github's sense, where it would be truly read only and irreversible.
Let me describe the reason behind:
- the archive comment in the project description
- why it is not archived in the github sense, and then finally,
- the star rating situation
The description archive comment
The reality is that no one who has been working on this before 2021 has been actively contributing to this. People who have been contributing since 2021 have decided that we would be better served by putting this under a github organization where we have several owners and maintainers. (After the Mathics3 organization was set up and underway, we did get full administrive and maintainer control.)
No rights have been removed here. Anyone who has access, can still contribute. And I believe if you or others want to contribute to this repository, I think I now have the permission to add you. And if that's what you and others want to do, let me know (after reading the rest)...
It wasn't always the case though that I could add new people. There was a period of something like 6 months where we had someone who we wanted to allow committing directly to this repository. And we could not give that person permission because we did not have the authority to do so and because of unresponsiveness from the people who had maintainer access.
Unresponsiveness in the open source community for contributors == death.
In that delay though, there was another concern among that the 2-4 current contributors/maintainers had. The code here is a bit of a monolith. It had some Django code with javascript (up to version 1.1), as well as the core interpreter. And other things; see below.
There was, and there still is, a desire to separate things into different concerns with different independent repositories which allow for finer grained access and independent development.
So a new organization, Mathics3, was created. (The 3 in Mathics3 initially represented that this is Python3 only; all we needed really was just a new name.)
Mathics3 github organization
With the addition of a new github organization that we could control, we were then able to split the code up into these separate repositories:
- mathics-scanner: tokenizer and character tables for Wolfram language. This allows for example programs that just want to syntax highlight in a text editor or printed documentation to get data here. Symja may also use these or consult these tables. (I am not totally sure here though.)
- mathics-django: the Django-based front end from this project. Possibly in the future there will be a separate github repo for Jupyter.
- mathics-three-js-backend: Javascript three.js implementation of 3D graphics. Symja definitely uses this.
- mathicsscript: A more beefed up command-line (CLI) implementation.
- mathics-development-guide: Sphinx documentation for developing Mathics3.
- mathics-core: The interpreter sans scanner, and UI except for a very minimal CLI, and reduced backend rendering code.
- pymathics-graph: network graphs via NetworkX and Matplotlib
- pymathics-natlang: NLP via
nltkandspacy - pymathics-hello: "Hello World" program as a Mathics3 module
- mathics-omnibus: packaging many of the above into one PyPI repository by requiring other PyPI packages; also it contains code for docker containers related to Mathics3.
A couple further things as a result of splitting the code.
First, by splitting code up into different and separately installable packages, we can reduce the dependency burden to install anything. In the old code base, the minimum requirements were pretty hefty; having code to optionally include things was getting out of hand. So was Continuous integration testing, since everything had to be installed and tested on all OS environments.
Second, by splitting up the code it has given the code a lot of room to grow. The combined sum of those individual repositories put here would at least double the size of this repository making it even more unwieldy for the current folks who are contributing to the code.
Why is this code not archived in the github sense?
This one is much simpler to answer. Simply because people who were given permission as maintainers are not the original owner(s).
But this repository has lots of stars and corresponding Mathics3 organization repositories do not!
This aspect was discussed among the 4 people before before moving. We decided the plusses of adding a new organization where we have full control far outweighed the downsides.
And I would maintain the differnce in ratings might not be wrong. The hard-earned deserved stars were voted up for this code repository, not Mathics3. It is up to individuals to decide whether to add stars to those other repositories as they see fit. And I suppose individuals can remove stars here if they feel this is no longer worthy of attention. But ultimately, this is the decision of the individuals, not the maintainers and current authors which have changed.
Note though that with respect to github history, if you look at the projects in the Mathics3 organization, there is one continuous history. And in sum, merging stuff back into here, at least from the current maintainer's and authors' standpoint does not make sense (even though it could be done).
ad-si
Sorry, but I don't understand the reasoning why you would archive this repo (with 2.1 K stars!) in favor of a repo with only 650 stars.
Also the whole
mathicsvsMathics3organization is super confusing for outsiders.How about you merge everything back into one organization and make the repo with the most stars the main repo? ^^