From: "Denys Duchier" <duchier@ps.uni-sb.de>
> Some additional remarks about closing ports:
>
> > Oh well. It just looked soo cool to write {Map Stream Consumer}...
>
> The primary reason that close is currently absent from the port API is
> that it is not clear what should happen when you send to a closed
> port.
Yes, I do realise I am abusing the many-to-one port for a one-to-one
connection.
> > One more reason to switch to "for" loops: break!
>
> :-) of course, "break" is just using an exception to escape; so you
> could do that too if you thought it convenient.
try-raise-catch looks slighly less elegant than {Break} :-)
[...]
> Space.inject is more expensive (there is a thread outside that injects
> stuff, and then there is a thread inside that needs to deal with it)
> and it can only affect stuff that is reachable from the space's root.
But you make everything reachable anyway so you see it in explorer when you
debug :-)
Are there any plans to collect and documet such patterns? We only use what
we see in the tutorials and examles, and there is so much more to it...
The whole pattern business may look like over-explaining the obvious to the
experienced, but it is a great tool for learning and discussing design.
> > They say it's multiparadigm all right, but it's not fair they have
> > to type so much more than other paradigm people. Declaring a
> > variable "final" is just 1 extra word in Java, in contrast.
>
> Ah, give me a break! These are the same people who think nothing of
> having to write a couple of new classes and a dozen adapters everytime
> they change their underwear :-)
For every new type of underware, yes, but not for every usage of every
piece.
And they probably outnumber us by 50 dB, so penalising their paradigm is a
political statement :-)
Gabor Szokoli
Gabor Szokoli
-
Please send submissions to users@mozart-oz.org
and administriva mail to users-request@mozart-oz.org.
The Mozart Oz web site is at http://www.mozart-oz.org/.
Please send bug reports to bugs@mozart-oz.org.