On Fri, Feb 1, 2008 at 1:07 PM, Jon Harrop <
jon@ffconsultancy.com> wrote:
[...snip...]
For example, perhaps many (most?) of OCaml's issues can be addressed without
altering the language by retrofitting an enhanced standard library onto
OCaml.
I'll be honest and say I usually ignore your pontificating, but for once I agree totally.
Perhaps this can be done transparently (i.e. without changing compile
lines or code) in such a way that we can all share it compatibly (i.e.
without breaking the brittle bindings) and we are all free to contribute to
it. This would be fantastic!
That would be fantastic, but...
I believe their effort
will not implement this transparently, i.e. it will be invasive because they
do not wish to supercede the stdlib (either by patching it or by automating
the inclusion of augmentations).
I think this this is more likely the case.
I would rather have something transparent
because it would let users solve their problems more easily.
I think that backwards-incompatibility, despite what INRIA may say, is not terribly important. Let us not forget that they set a precedent when they released a incompatible and scantily documented, yet significantly improved, CamlP4 unannounced in 3.10.
I think that if the community wrote a new, significantly improved but incompatible standard library and handed it to INRIA, INRIA would be hard pressed to find a reason not to release an backwards-incompatible Caml 4.0, given there was a configure switch when building the compiler to build it using a frozen 3.x library, which was not possible 3.9 -> 3.10. After all, incompatibilities are what major version numbers are for.
From then on, the OCaml team could ship a language only tarball (plus compatability library) and concentrate on exactly what they are good at: writing a great implementation of OCaml, without us bugging them about improving the library. All of the package managers being discussed could pull the language from INRIA and the "new" stdlib from wherever it is housed and automagically put the two together.
--Jonathan
[...snip...]
[...(again)...]
[...(and again)...]
[...(yet again)...]