[Caml-list] [ANN] React 0.9.0
Daniel Bünzli
I'd like to announce the following module.
React is an OCaml module for functional reactive programming (frp). It
provides support to program with time varying values : applicative
events and signals. React doesn't define any primitive event or
signal, this lets the client chooses the concrete timeline.
React is made of a single, independent, module and distributed under
the new BSD license.
Project home page : http://erratique.ch/software/react
Although the code is feature complete and has been tested it has not
been in large scale programs yet. A stable 1.0.0 will be published
once I feel we have the right api.
Your reactions are welcome,
Daniel
P.S.
Various incarnations of frp (frtime, yampa, flapjax, ocamlrt,
reactive) helped to implement React, thanks to their authors for
sharing their source.
P.P.S.
Here's the answer to the forthcoming question about how React compares
to ocamlrt. Following are the points I think are different and worth
mentioning. But as I don't claim to have a thorough understanding of
how ocamlrt works feel free to correct me or make additions. ocamlrt
used to be located at http://users.wpi.edu/~squirrel/repos/ocamlrt2
but recently vanished. The version at http://code.google.com/p/ocamlrt
seems to be an older one.
* React doesn't use the object system. ocamlrt uses it internally.
* React doesn't use any global data structures. Care is still needed
in multithreaded environments (see docs), but React is more thread and
modular friendly.
* React treats constant signals separately via smart constructors,
they are not part of the dataflow graph.
* React uses weak pointers in the dataflow graph's forward
dependencies and in the update queue to prevent space leaks. ocamlrt
leaks.
* React uses client provided functions to test signal value equality.
ocamlrt doesn't deal correctly with effectful signals. Correctly means
no value change (according to equality), no update, no effect.
* React has most of pervasives' functions on ints, floats and booleans
already lifted in specific modules that may be opened in a given
context.
* React is (and will be) only a pure OCaml reactive runtime.
Interfacing with primitive events, signals or real time is left to
React's clients. In other words React corresponds to the functionality
provided by the library fr of ocamlrt.
* React has a precise description of the combinators' semantic in the
documentation.
* React doesn't depend on extlib.
* Performance ? I don't really know. Some aspects may make React
faster : doesn't leak, doesn't use a global datastore (which involves
a hashtable and universal types). Some aspects may make it slower :
use of weak references, subtleties in signal creation, correct
treatment of equality.
* Being inavailable ocamlrt seems to be a dead project.
_______________________________________________
Caml-list mailing list. Subscription management:
http://yquem.inria.fr/cgi-bin/mailman/listinfo/caml-list
Archives: http://caml.inria.fr
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
Bug reports: http://caml.inria.fr/bin/caml-bugs
khigia
server (connection drop, too long).
Anywhere else I could find the package?
On Jan 20, 11:57 am, Daniel Bünzli <daniel.buen...@erratique.ch>
wrote:
> Hello,
>
> I'd like to announce the following module.
>
> React is an OCaml module for functional reactive programming (frp). It
> provides support to program with time varying values : applicative
> events and signals. React doesn't define any primitive event or
> signal, this lets the client chooses the concrete timeline.
>
> React is made of a single, independent, module and distributed under
> the new BSD license.
>
> Project home page :http://erratique.ch/software/react
>
> Although the code is feature complete and has been tested it has not
> been in large scale programs yet. A stable 1.0.0 will be published
> once I feel we have the right api.
>
> Your reactions are welcome,
>
> Daniel
>
> P.S.
> Various incarnations of frp (frtime, yampa, flapjax, ocamlrt,
> reactive) helped to implement React, thanks to their authors for
> sharing their source.
>
> P.P.S.
> Here's the answer to the forthcoming question about how React compares
> to ocamlrt. Following are the points I think are different and worth
> mentioning. But as I don't claim to have a thorough understanding of
> how ocamlrt works feel free to correct me or make additions. ocamlrt
> used to be located athttp://users.wpi.edu/~squirrel/repos/ocamlrt2
> but recently vanished. The version athttp://code.google.com/p/ocamlrt