On 11 March 2010 18:19, Jonathan S. Shapiro <[email protected]> wrote:
> On Thu, Mar 11, 2010 at 8:51 AM, Aleksi Nurmi <[email protected]> 
> wrote:
>> Now that I think of it, is there a compelling reason not to use GHC
>> for native code compilation? It is stable and in active development.
>
> Yes: the longer the dependency chain for BitC, the harder it is to adopt.
>
>> Many recent languages such as Agda actually target Haskell...
>
> Research languages can afford to do that. BitC has been intended from
> the beginning to end up a production language.
>
> I have a lot of respect for GHC, but having one strange new language
> depend on a second strange language (in the eyes of the mainstream)
> would be problematic, and it doesn't seem like we need that extra
> dependency.

My window manager also depends on GHC and until the latest Debian
packages of GHC doubled in size to about 300MB I did not really care.
At this point I am considering alternatives as 300MB just to manage my
windows seems somewhat bloated.

I guess it's the same for BitC - if it depends on exotic language
which is easily available it should not be of much problem. I happen
to have GHC on my system but no trace of Mono - not to say it cannot
be easily installed as it is also prepackaged in Debian.

Thanks

Michal
_______________________________________________
bitc-dev mailing list
[email protected]
http://www.coyotos.org/mailman/listinfo/bitc-dev

Reply via email to