avsm
Anil Madhavapeddy
Alain Frisch and Leo White: a public working group on the future of camlp4 and metaprogramming in OCaml: groups.google.com/group/fa.caml/…
jakedonham
Jake Donham
@avsm sad that chucking it seems a foregone conclusion. amazing tool, marred by INRIA neglect, lack of docs, p4/5 schism, fixable impl warts
@avsm sad that chucking it seems a foregone conclusion. amazing tool, marred by INRIA neglect, lack of docs, p4/5 schism, fixable impl warts
avsm
Anil Madhavapeddy
@jakedonham my hope is this will energise fixing it in parallel (/cc @bobzhang1988). It'll take *years* to fully decouple camlp4, if ever
@jakedonham my hope is this will energise fixing it in parallel (/cc @bobzhang1988). It'll take *years* to fully decouple camlp4, if ever
bobzhang1988
bob zhang
@avsm @jakedonham Have you take a look at Fan? it already fixes most problem Alain proposed without loosing any expressivity ;-)
@avsm @jakedonham Have you take a look at Fan? it already fixes most problem Alain proposed without loosing any expressivity ;-)
avsm
Anil Madhavapeddy
@jakedonham it was your blog post series that led me and @samoht to do so much syntax extension hacking! not sure whether to hate you or not
@jakedonham it was your blog post series that led me and @samoht to do so much syntax extension hacking! not sure whether to hate you or not
avsm
Anil Madhavapeddy
@jakedonham I wonder if the revised/original quotation choice is the tipping point to camlp4 becoming too complex
@jakedonham I wonder if the revised/original quotation choice is the tipping point to camlp4 becoming too complex
jakedonham
Jake Donham
@avsm oh yeah forgot that one :). I see the motivation—orig syntax lacks context for antiquotations—but there are better alternatives
@avsm oh yeah forgot that one :). I see the motivation—orig syntax lacks context for antiquotations—but there are better alternatives
yminsky
Yaron Minsky
@avsm @jakedonham I think the concept is wrong at the core. Too much power, too little structure. Quotations+annotations+ppx FTW!
@avsm @jakedonham I think the concept is wrong at the core. Too much power, too little structure. Quotations+annotations+ppx FTW!
jakedonham
Jake Donham
@yminsky compared to codegen with interpolated strings it is highly structured :)
@yminsky compared to codegen with interpolated strings it is highly structured :)
jakedonham
Jake Donham
@yminsky less flippant reply: why not provide a simplified API (just quot/annot/ppx) and a full one? why prevent advanced use cases?
@yminsky less flippant reply: why not provide a simplified API (just quot/annot/ppx) and a full one? why prevent advanced use cases?
bobzhang1988
bob zhang
@jakedonham I agree with you on this point. It's not too hard to provide (quot/annot/ppx) on top of Camlp4 or Fan
@jakedonham I agree with you on this point. It's not too hard to provide (quot/annot/ppx) on top of Camlp4 or Fan
jakedonham
Jake Donham
@yminsky lpw25.net/2013/01/23/cam… misses important camlp4 use cases. e.g. nested quot/antiquot, exts to pattern matching
@yminsky lpw25.net/2013/01/23/cam… misses important camlp4 use cases. e.g. nested quot/antiquot, exts to pattern matching
jakedonham
Jake Donham
@yminsky and surveying what exists then supporting only common cases is philosophically backward for an extension mechanism
@yminsky and surveying what exists then supporting only common cases is philosophically backward for an extension mechanism
yminsky
Yaron Minsky
@jakedonham If a fully programmable syntax is too much power, an alternative will give up on some things. A survey can inform those choices.
@jakedonham If a fully programmable syntax is too much power, an alternative will give up on some things. A survey can inform those choices.