On Tue, Aug 3, 2010 at 9:45 PM, Ryan Ingram wrote:

So I believe the "final" way to do this, which is not yet implemented, works something like this:

type family LeftToRight a type family RightToLeft b

class (LeftToRight a ~ b, RightToLeft b ~ a) => Bijection a b where   ...

I agree, the fact that this doesn't work is really dumb.

I think it is more simple like: class Bijection a b where ... type LeftToRight a = (Bijection a b) => b type RightToLeft b = (Bijection a b) => a -- Victor Nazarov

On 4 August 2010 03:45, Ryan Ingram wrote:

So I believe the "final" way to do this, which is not yet implemented, works something like this:

type family LeftToRight a type family RightToLeft b

class (LeftToRight a ~ b, RightToLeft b ~ a) => Bijection a b where   ...

I agree, the fact that this doesn't work is really dumb.

The reason Manuel Chakravarty gave me (if I recall correctly) that this doesn't work is as follows: Type classes are stored in dictionaries. To be able to have superclass constraints like this requires that type information also be stored in those dictionaries, which currently isn't done. They could have added this functionality in, but SPJ was going to be working on a new type-checker anyway (which I believe is now done) so this was postponed until after the new type-checker was finished. As we should hopefully be seeing 6.14 out in a few months time, there isn't time to be able to get superclass constraints in for it, so we probably won't have them until 6.16 next year. So, I wouldn't say that it's "dumb". Unfortunate, a pity, etc., but not "dumb". Don't forget, GHC is open source: if this lack really was "dumb" and annoying you, there was nothing stopping you from rectifying this situation up until now. -- Ivan Lazar Miljenovic Ivan.Miljenovic@gmail.com IvanMiljenovic.wordpress.com

Ivan Lazar Miljenovic wrote:

...

Don't forget, GHC is open source: if this lack really was "dumb" and annoying you, there was nothing stopping you from rectifying this situation up until now.

Except that, in the real world, this is actually completely infeasible. Yes, I know it's the basic tenant of OSS that you can modify the program to do whatever you want. But in reality, something like GHC is far too large and complex for this to be a realistic possibility. And this holds for most other nontrivial software too.

_______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe Fair enough, but if one can't do better one's self then one should be careful about calling the work of others "dumb", which was the original

On 8/4/10 11:40 PM, Andrew Coppin wrote: point. Cheers, Greg

Ivan Lazar Miljenovic wrote:

On 5 August 2010 16:48, Gregory Crosswhite wrote:

...

On 8/4/10 11:40 PM, Andrew Coppin wrote:

...

Ivan Lazar Miljenovic wrote:

...

Don't forget, GHC is open source: if this lack really was "dumb" and annoying you, there was nothing stopping you from rectifying this situation up until now.

Except that, in the real world, this is actually completely infeasible. Yes, I know it's the basic tenant of OSS that you can modify the program to do whatever you want. But in reality, something like GHC is far too large and complex for this to be a realistic possibility. And this holds for most other nontrivial software too.

Fair enough, but if one can't do better one's self then one should be careful about calling the work of others "dumb", which was the original point.

Exactly. Either do it yourself or be grateful that someone has done _something_, even if it isn't as good as you like. It's not like you're paying for it...

Well *I* didn't say that anything was dumb. I was merely pointing out that the much-touched "do it yourself" benefit of OSS is actually out of most people's reach. There surely can't be many people alive on Earth who actually understand type theory, and far fewer who understand it well enough to meddle with something as astonishingly complex as GHC. So really, there's little or no chance of anybody except the GHC devs fixing this. (If nothing else, the learning curve is pretty much vertical just to fix this one minor problem.) On the one hand, it's nice that we have a freely available compiler at all. (And it's one of the best pieces of Haskell software I've seen to date.) On the other hand, I've seen too many people who write open-source software answer every query and issue simply with "patches welcome". As if that's a magic bullet to solve every problem and deficiency. There needs to be a reasonable balance. (Note that in the particular case of GHC, I think us users are getting a pretty good deal. My statements are about open-source in general, not about GHC.)

On Wed, Aug 4, 2010 at 11:55 PM, Ivan Lazar Miljenovic wrote:

On 5 August 2010 16:48, Gregory Crosswhite wrote:

...

 On 8/4/10 11:40 PM, Andrew Coppin wrote:

...

Ivan Lazar Miljenovic wrote:

...

Don't forget, GHC is open source: if this lack really was "dumb" and annoying you, there was nothing stopping you from rectifying this situation up until now.

Except that, in the real world, this is actually completely infeasible. Yes, I know it's the basic tenant of OSS that you can modify the program to do whatever you want. But in reality, something like GHC is far too large and complex for this to be a realistic possibility. And this holds for most other nontrivial software too.

Fair enough, but if one can't do better one's self then one should be careful about calling the work of others "dumb", which was the original point.

Exactly.  Either do it yourself or be grateful that someone has done _something_, even if it isn't as good as you like.  It's not like you're paying for it...

My bad, no offense was meant, and perhaps I should choose my words better. I've read Manuel's papers and worked with Simon, and they're both very smart people. I wasn't commenting at all on the quality of their implementation, but rather trying to make the point that it's really a bug that it doesn't work, and, AFAIK, a bug that is going to be fixed. I think everyone can relate to the frustration that occurs when you are surprised by a feature being missing. I actually think it's a testament to the quality of GHC that things "just work" so often that I can be so surprised when they don't. -- ryan