Johan Tibell
writes: The discussions about an overhauled record system also involve lots of talk about record sub-typing, extensible records, and other more advanced features. I'd like to point out that there doesn't seem to be a great demand for these features. ...
Sorry, Johan, I really have to disagree with that. There's lot's of Haskell to SQL interfaces that build on HList and its extensible record ideas (HDBC for example). But the usability is not good (as Petr points out, and as Oleg/Ralf admitted back in the paper). The type error messages are long and obscure.
... They might be nice-to-haves or might fall out naturally from a solution to the namespacing problem above, but they are in fact not needed to solve the common problem people have with the Haskell record system.
"the common problem people have" is that the record system is unusable [IMO] so doesn't get 'stretched' to see what other difficulties it has. There are all sorts of alternative systems (including Lenses) built with Template Haskell (and chewing gum and gaffer tape: that's how desperately bad is the current situation ;-). I'm saying that many people find the Haskell record system 'as is' so dysfunctional that they give up on it! I feel strongly that as soon as we get past the name collissions, there'll be other blockages to using it. I'd be interested to hear if there are any who can remember the Trex system, and how (un)usable it was? AntC