Is this production-ready?

[3noch]

Could someone use this instead of lens? Are there downsides?

[chshersh]

I would also like to see comparison with microlens.

[mrkgnao]

Yes, you should be able to drop this in as a replacement for lens once it's done, since I'm changing as little of the API as I can, but, no, it isn't feature-complete yet. Once it is done, it should be exactly equal in power to lens since that's what I'm porting. (So, for any comparisons with microlens, just compare it to lens.)

Downsides:

• existing -lens libraries (e.g. lens-aeson) will have to be ported. This is pretty easy: any provided lenses will just have to be wrapped in the fromRaw* functions:

-import Control.Lens
+import Silica

foo :: Lens' Foo Bar
-foo = whatever
+foo = fromRawLens whatever

fromRawLens has the type R_Lens s t a b -> Lens s t a b, where R_Lens is the type that lens provides.

[3noch]

What about performance?

[chshersh]

Maybe inspection-testing library can help with ensuring good performance...

[RobertFischer]

When you're asking about performance, it's important to specify what performance you're talking about. If you're talking about the performance of developers' efficiency in producing working and adaptable code, then this library is already leaps and bounds beyond lens.

[anchpop]

I would be very interested in using this, so it would be nice if it were updated to support newer versions of GHC.

[3noch]

@RobertFischer I've not actually had any significant trouble using lens as is when it comes to developer productivity. But I would definitely prefer a more beginner friendly tool if it doesn't degrade other aspects of my code (namely, run-time performance).