[1361] in java-interest
Re: "perform:" and Java
daemon@ATHENA.MIT.EDU (Chris Warth)
Thu Aug 31 00:07:43 1995
Date: Wed, 30 Aug 1995 17:26:58 -0700
From: csw@scndprsn.Eng.Sun.COM (Chris Warth)
To: johnm@emf2-003.emf.net, clp@home.HarvardSq.com
Cc: Tako.Schotanus@bouw.tno.nl, Schotanus@home.HarvardSq.com,
Tako@home.HarvardSq.com, java-interest@java.Eng.Sun.COM
>
> As much as I applaud the above sentiment (and agree with it), in this case
> it is a little mis-directed. A dynamic language which already has some meta-
> features (reflective ones as well), which allows: new("class" + "name"),
> and separates interfaces from implementations so cleanly screams out for the
> natural complement that perform: represents. I would claim it was part of
> the same orthogonal set to which dynamically linked classes and late binding
> all belong. That's why I wanted a few Sun/Java souls to chime in about this.
>
Ok, I'll chime in. We looked at a variety of "perform:" like features
over the past few years. There is no doubt that this is a popular
feature and one that would be used a great deal. However making this
thing reasonably flexible, typesafe, and having it look anything less
than a kludge turned out to be too difficult for our meager skills.
It's a lot of work but I'll go see if I can dredge up the old email
debates that we had. I'd be willing to send it to people who have a
serious interest in this issue, but I'm not going to spew the mailing
lists with a lot of obscure detail.
Basically the issues revolved around type safety, method matching, and
the ability to queue these perform:-closures for later execution. The
debate got rather involved, but we did discuss this issue for months.
-csw
-
Note to Sun employees: this is an EXTERNAL mailing list!
Info: send 'help' to java-interest-request@java.sun.com
