Git Remote Helper: Concurrency #29
Conversation
|
Notes:
|
| (renderProcessError (errError e)) | ||
| (renderSourceLoc (errCallStack e)) | ||
|
|
||
| instance Exception ConcurrentException |
There was a problem hiding this comment.
If you don't like the fact of implementing Show in a non-principled way, I guess you could try to implement displayException for ConcurrentException and see if that works out?
There was a problem hiding this comment.
I only recently realised the existence of displayException, and am confuse. Show is a superclass of Exception, so I still have to derive it for all error types - but that is something I try to avoid: one should always use the appropriate render* function.
Am I asking too much?
There was a problem hiding this comment.
Yeah, I am with you on this one. I guess the "best" thing you could do is to derive Show using the standing deriving mechanism but write your "pretty printed" version as part of displayException. I do agree that the superclass constraint is a nuisance.
There was a problem hiding this comment.
4b113d1 destroyed my hopes
This is rather naive for now, as we're essentially running every IPFS API call in its own Haskell thread. Observed speedups are significant nevertheless. Upper bounds are placed on the number of concurrent connections to the API, as well as the number of OS threads (via the `-maxN` RTS option).
Closes #23