-
Notifications
You must be signed in to change notification settings - Fork 97
New issue
Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? # to your account
Make it compile with MicroHs #190
Conversation
@@ -86,7 +85,7 @@ library | |||
UndecidableInstances | |||
|
|||
ghc-options: -Wall | |||
if impl(ghc >= 8.0) | |||
if impl(ghc >= 8.0) || impl(mhs) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I actually dropped the support for pre GHC-8.0, so this conditional's else-branch can be removed. Do you mind?
I'm not sure whether Hackage is using new enough Cabal
to accept impl(mhs)
.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
You can remove the else branch if you want.
Cabal accepts mhs as a valid compiler now, I believe.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hmm. Maybe the CI is using some older cabal?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Cabal accepts mhs as a valid compiler now, I believe.
newer Cabal
does. But hackage-server
still uses old Cabal
, so this change could not be released. (And it might take quite a long time).
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
At the moment Hackage uses Cabal 3.12.1.0, which knows about MHS:
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@Bodigrim good to know. But given a thought, if the thing can be written without impl(mhs)
, it should.
You help me, I help you. I'm sorry, but I don't care about MicroHS
, so if I can be unaware I will.
That said, the first moment the CI breaks, I will remove it; so adding CI for microhs is somewhat unnecessary too. I won't put any effort into it.
TL;DR my stance is: GHC is the only compiler out there. The ship for multi-compiler ecosystem has long sailed away.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
If there is no support for pre-8.0 GHC then the condition can be simplified and it will work with MicroHs.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I started this discussion with saying
I actually dropped the support for pre GHC-8.0, so this conditional's else-branch can be removed. Do you mind?
This left a bad taste. I removed the conditional, so AFAICT this patch has only MonoLocalBinds
changes which are relevant.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
For a time being, I won't accept changes using impl(mhs)
.
... but I don't really care anymore. I see there are a forks of transformers and mtl, so I guess you will have to maintain a fork of parsec too. |
The MicroHs changes to transformers have been accepted a long time ago. But since the master copy of the source is in darcs I have a copy in my GitHub repo. |
Ah, I miunderstood you comment about dropping support for 8.0. I thought you'd done it, but not yet changed the parsec.cabal file, but it was just that I hadn't synced my repo. My bad. |
Two changes