Closed jalvesaq closed 4 months ago
I think we are ready to create the R.nvim
organization with the project cmp-r
...
@PMassicotte ? @she3o ?
For me it is in a state that is very usable. So I am up for it :)
I created the organization. Its "display name" is "R.nvim", but it seems that the URL cannot have a dot in the name. Github created https://github.com/R-nvim instead. So, maybe we should change the name of the organization to "R-nvim"...
@PMassicotte , I invited you to the organization. @she3o , I will invite you if you are still interested in being part of the organization.
The organization can't be "R.nvim", but the repository can.
@she3o , I will invite you if you are still interested in being part of the organization.
Of course, I'd be delighted to join. I must apologize for the frequent interruptions in communication. Balancing postgraduate studies and work commitments has significantly limited my availability on GitHub. However, I'll make every effort to contribute by submitting pull requests whenever possible.
Initially, I intended to copy the files from this repository to R-nvim
, starting the new repository R.nvim
as a simple conversion from VimScript to Lua. But now we have not only the conversion of old code, but new code written by @PMassicotte, @she3o, and @hongyuanjia. So, now I prefer to keep the git history of your contributions, and, in this case, we can immediately transfer the repository to the new organization. I still prefer to squash all consecutive commits by the same author in single commits, to begin with a cleaner git history.
What do you think?
Perfectly fine with me, this seems fair enough. :+1:
Sounds good to me!
yep, let's do it.
I did my first squash with gitlazy, it seems to be OK. In a few minutes, I will continue...
I managed to decrease the number of commits from about 280 to 178, but, then, some conflicts appeared again and again.
An atempt of using git rerere
broke everything. I will restore the files from my backup...
Rewriting git history is a real pain...
I restored the files. Using diff -aur
to compare my backup with the tmp-R-Nvim
repository, it seems that everything is OK.
I will transfer the repository tomorrow.
The repository is at its new address.
Here are some tasks that I foresee:
Send Lua table instead of Vim list, Vim dictionary, or white space separated list of words toDone on data sent tojob.stdout
, and uselua cmd
instead ofcall v:lua.
.cmp-r
which is the only one that requires high speed.Create theDone.R.nvim
organization with the repositorycmp-r
.Dogrep -R '\\|' lua/
and fix patterns that don't work.Usevim.regex
orvim.fn.matchlist()
to fix code that uses patterns\k
or\w
Fix
allmostFIXME
. OK, the remaining might be fixed later.Add "desc" field to maps.Temporarily forceSimpler than that: I didrnvimserver
to send strings broken in pieces to fully testjob.stdout()
.<LocalLeader>o
over a character vector with 10000 lines with R running in a remote machine. The data arrived in three parts, and the function worked properly.Test everything.
Visit open issues in. Done.Nvim-R
requesting new features and invite people to trytmp-R-Nvim
for a week.One week later:
Add the repository R.nvim to the organization.
Close issues in Nvim-R that are too old, aren't reproducible, or won't be fixed, and invite people to switch toAlready done.R.nvim
(where there is a chance of the bug being fixed).