[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Continuations
At 3:21 AM +0000 8/11/03, Paul Graham wrote:
>--Dan Sugalski wrote:
>
>> At 2:09 PM -0700 8/9/03, Serguei Mourachov wrote:
>> >But now I have another question about continuations: how useful
>>they are for
>> >*masses* of developers?
>> >Does typical programmer need to know about and use continuations?
>> >What kind of advantages continuations bring into development process?
>>
>> No, and none, respectively. They allow for some really powerful
>> techniques, and they make some relatively bizarre things easier (like
>> implementing interpreters, compilers, and dynamic syntax extension)
>> but for most of the stuff you'll do with computers they're not
> > particularly useful.
>
>I argued this in my dissertation, but now I think I was wrong.
>They're extremely useful for overcoming the statelessness of
>CGI scripts.
You have no idea how much I'd love a good way to overcome the
statelessness of CGI (or, since you've written your share of CGI
code, perhaps you do :) but I'm not sure that it's the right way in
many circumstances, and not for the availability reasons most folks
have posited. (This is the web--either it's not a big deal if you
lose state or you restore from the data you've backed to the DB
because you know that this sort of thing is inherently unstable) Most
CGI apps would be better served with a mixed save-game/continuation
sort of deal rather than a true continuation.
I'm also not certain that going with a pure continuation in a CGI
context is the right thing to do, since it either makes the system
more complex (as you need to freeze the data to disk for proper
regeneration) or lead even more programmers down the path of bad CGI
in the face of system failure...
--
Dan
--------------------------------------"it's like this"-------------------
Dan Sugalski even samurai
dan@sidhe.org have teddy bears and even
teddy bears get drunk