remove cache from deploy CI because it always breaks #446
Conversation
|
I think because Maybe we can add this pin in owi to fix the CI: |
|
Thanks. :) I'll try this but shouldn't our dependency on smtml.0.3.1 force opam to use the development version of alt-ergo ? |
I had to remove the pinned dependencies from smtml's .opam file. I only noticed this when I made the last PR to update Owi. But I'll update smtml's .opam file in the opam repository once a new version of alt-ergo is released |
|
Ah yes sure, makes sense, they won't let you commit a pin-depends in opam-repo. I'll merge this as is for now. We can remove it later once new alter-ergo and smtml versions are released. EDIT: actually it does not seem happy about this alt-ergo version |
|
It was missing the alt-ergo solver pin as well I believe |
|
I think we just have to promote now. The backtrace is annoying, we were having issues with Colibri2's dependencies turning on the backtrace. It seems that alt-ergo might do that as well? (ref. formalsec/smtml#212 (comment)) |
|
Done. Let's see if it works now... I opened an issue (#455) so we don't forget to investigate why the backtraces are turned on. |
Co-authored-by: filipeom <[email protected]>
Co-authored-by: filipeom <[email protected]>
zapashcanon
force-pushed
the
cicicici
branch
from
44dc409
to
35efe8d
Compare
|
OK, looks like we are hitting ocaml/setup-ocaml#897 on macOS. For the linux build, I'm not sure what's going on, there's no proper error displayed... |
Looks like we're running out of diskspace in the runner. You can see the errors at the top of the CI result
We should try and run this in debug mode and see if the submodules, or something is taking up a lot of disk space. I was having similiar issues with disk space running out due to opam checking out submodules when it detects a .gitmodules in the directory. I ended up doing this hack formalsec/smtml#244 |
|
Aww, thanks, I missed it somehow... Let's see now! |
No description provided.