Very cool package, first time I'm seeing use of the unicode symbols package
;).

I agree that having a single datatype is preferable. In particular, the
wai-handler-snap package would get a performance boost by not needing to
translate between the two datatypes.

The one issue I have is that I've so far limited dependencies for WAI to a
bare minimum: base and bytestring. I'm going to have to consider hard
whether I should allow other packages in or not.

I'd be interested to hear what Snap has to say about this ;).

Michael

On Fri, Jul 9, 2010 at 2:39 PM, Bas van Dijk <[email protected]> wrote:

> Hello,
>
> I noticed both WAI and the Snap Framework are using the CIByteString
> type for doing case insensitive comparison on ByteStrings:
>
> http://github.com/snoyberg/wai/blob/ver0.2/Network/Wai.hs#L140
>
> http://hackage.haskell.org/packages/archive/snap-core/0.2.7.1/doc/html/Data-CIByteString.html
>
> What about sharing the same type for this?
>
> I've written a small package that generalizes the CIByteString type to
> any string-like type (String, ByteString, ShowS, etc.):
>
> http://code.haskell.org/~basvandijk/code/case-insensitive
>
> What do you think of this and would you guys like to use it? If so I
> will release it on Hackage.
>
> Regards,
>
> Bas
>
_______________________________________________
Snap mailing list
[email protected]
http://mailman-mail5.webfaction.com/listinfo/snap

Reply via email to