Eli: are you saying that you think this is _not_ a bug?
(I think Racket has firmly positioned itself in the "primitive
operations in the language do the right thing or signal errors" camp,
so a performance-based concern like this seems quite out of
character.)
Robby
On Fri, Nov 9, 2012 at 10:40
Eli, if compose is 20x slower than writing out the lambda,
and respect for arity costs 5%, I think we're fine.
A warning in the docs sounds good. Something like
"Warning: Using compose, instead of plain lambda, for function composition
helps readability and facilitates re-coding the numerous
Re 3.: we should fix those, then.
:)
Robby
On Fri, Nov 9, 2012 at 10:40 AM, Eli Barzilay wrote:
> I just measured the overhead again, to see if things changed. The
> attached file has the exact implementation of the compose functions,
> and another copy with the proper reduction and renaming.
I just measured the overhead again, to see if things changed. The
attached file has the exact implementation of the compose functions,
and another copy with the proper reduction and renaming. For the
correct function I'm getting ~5x slowdown from the current code, which
in turn is 20x slower than
I think the right approach is to make the function behave correctly
first and then find the places to optimize second.
In this case, using procedure-reduce-arity (or something else with the
same effect) is not merely "nice" but, IMO, necessary. That is, we
have procedure-arity in our PL and our pr
On Tue, Aug 7, 2012 at 10:46 PM, Erik Dominikus
wrote:
> I had this conversation with DrRacket 5.2:
> [...]
On August 7th, Robby Findler wrote:
> Looks like a bug in compose1 (and compose) to me.
The source of both `compose' and `compose1' has the place where this
would be fixed -- see the "here
Glad to hear you're all set. I've submitted a PR anyways, so hopefully
this will be fixed for others.
Thanks for letting us know!
Robby
On Sat, Aug 11, 2012 at 12:59 AM, Erik Dominikus
wrote:
> After looking at collects/racket/private/list.rkt, I think this issue
> might be too hard to deal wit
After looking at collects/racket/private/list.rkt, I think this issue
might be too hard to deal with; I forgot that a procedure could have
optional and keyword arguments.
Thus, I decided to roll my own stripped-down version of 'compose'; I
only need a little of it actually.
Thank you.
On Tue,
Looks like a bug in compose1 (and compose) to me.
Robby
On Tue, Aug 7, 2012 at 10:46 PM, Erik Dominikus
wrote:
> I had this conversation with DrRacket 5.2:
>
>> (procedure-arity (compose1 (lambda (x) 0) (lambda () 0)))
> (arity-at-least 0)
>
>> (procedure-arity (compose1 (lambda (x) 0) (lambda (
I had this conversation with DrRacket 5.2:
> (procedure-arity (compose1 (lambda (x) 0) (lambda () 0)))
(arity-at-least 0)
> (procedure-arity (compose1 (lambda (x) 0) (lambda (x) 0)))
1
> (procedure-arity (compose1 (lambda (x) x) (lambda (x y) 0)))
(arity-at-least 0)
I think the arity of the fir
10 matches
Mail list logo
