[Agda] Agda environment: Hackage, Hoogle, Haddock, Cabal?

Mateusz Kowalczyk fuuzetsu at fuuzetsu.co.uk
Mon Mar 24 21:14:29 CET 2014


On 24/03/14 20:07, Andreas Abel wrote:
> Mateusz, great initiative!
> 
> We would love to a have a version of cabal/hackage for Agda!
> 
> If you are motivated to work on this, it would make sense that you
> join the Agda meeting in Paris in May (just announced!), or come for a
> visit to Chalmers.
> 
> On 24.03.2014 19:07, Mateusz Kowalczyk wrote:
>> Greetings,
> 
>> While Agda itself is great, the environment around it not so much.
>> It's difficult to find libraries online that do what we need, it's
>> hard to find the functions we need, we don't have much in the way
>> of documentation except looking up the papers and source diving and
>> we have no way to manage library versions etc.
> 
>> Are there plans to assimilate tools with functionality of Hackage, 
>> Hoogle, Haddock and Cabal into the Agda environment? Hoogle would
>> be pretty difficult to do for type signatures but it should be
>> fairly easy to look for function names. I don't imagine a
>> Haddock-like thing would be extremely hard to attach. A
>> cabal-install thing would be awesome although I'm unsure how hard
>> that would be. We could probably re-use all the Hackage code for an
>> Agda version. Maybe we could even convince the Hackage/Cabal guys
>> to spare some space for Agda packages which would be much easier
>> than having to implement and host everything ourselves.
> 
>> What do you guys think? Are there any plans already? Currently
>> trying to find a library which does what you want involves a lot of
>> looking for papers, then looking for (mostly dead) links to the
>> code, e-mailing the author when the code is dead/doesn't work… It's
>> quite a pain.
> 
> 
> 

While I'd love to attend, I doubt I'll be able to do so mostly due to
financial constraints. I'm pretty sure my exams are at that time too.

I would be interested in working on such tools but I doubt I could do it
all alone so if anyone is interested and/or has ideas then I'm
interested to hear about it! Personally I think that the best approach
would be to take existing Hackage/cabal code and adapt it to Agda's
needs but maybe there are some exciting new ideas that I don't know about.

-- 
Mateusz K.


More information about the Agda mailing list