-
Notifications
You must be signed in to change notification settings - Fork 1
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Sync with deps #1
base: master
Are you sure you want to change the base?
Conversation
Incidentally trimmed some lines (automated in my emacs settings)
(looks like component ids should be coerced to map instead)
Now the project compiles well with the coq-utils version pinned as a submodule !
It works well without merging master though |
Ok it's been fixed (register was redefined as a Variant and it didn't work well with QuickChick deriving |
The issue for the derivation of variant is documented here but it doesn't look like it would be taken care of anytime soon QuickChick/QuickChick#155 |
I'm building Coq directly, never via opam. Can I still build things? |
@floriangru and I discussed the changes again earlier today. In principle, everything should be ready for a final test and merge; I'll start looking at that now. @catalin-hritcu As I understand it, you always have the choice of how to build, but you have the option to use |
This works well for me too ... without any opam beyond for installing OCaml 4.08.1 ... so we should lift the 4.06 upper bound here:
Here also the 4.02.3 lower bound seems quite hard to believe ... has anyone tried? :) |
oldest one should be 4.05.0 when compatibility with coq 8.10.0 would be possible
Have to see if the extraction behaviour for target program can be changed to avoid using Pervasives (don't exactly know if it's caused by QuickChick, coq or our extraction) cleaned more leftover files in Makefile.Test
…q 8.10.0 In proof mode under coq 8.10.0, these parameters are named according to the definition (they were strangely renamed in previous versions, with an appended 0). Auto trim
Apart from the extracted examples failing for most of them, the testing of SFI worked with 4.09.0 as well (up to 263 generated tests, didn't wait until the 600th)
We'll have to lift the lower bound to 4.05.0 if we want compatibility with coq 8.10 (which now only depends on Arthur's libraries), so we won't have to try the older versions 😌 |
This builds with extructures 0.2.0, coq-utils 0.2.0 (and the math-comp libaries 1.9.0 with coq 8.9.1). Building coq-utils can be streamlined by pinning it with the upstream repo when arthuraa/coq-utils#3 would be merged. In the meantime,
opam pin add coq-utils
and copy-pasting the content of the file added in the coq-utils PR works for me.