Re: [racket-dev] [racket] "lab notebook on learning process" (was: Re: Macros baffle me)
2014-05-06 23:48 GMT+02:00 Stephen Chang : >> How about an extra button, a "Run Benchmark" button? > > You can already get this with the "benchmark" package, from pkg.racket-lang. A builtin button will hopefully make it obvious even for non-DrRacket users, that Run is not intended for benchmarks. For those shopping around for a language or an implementation it needs to be painfully clear, what to do in order to get a fair measurement. /Jens Axel >> 2014-05-06 19:43 GMT+02:00 Sam Tobin-Hochstadt : >>> On Tue, May 6, 2014 at 12:14 PM, Matthias Felleisen >>> wrote: > > Why does he think "Performance sucks"? > > Because here's the list of things that are slow DrRacket is an operating system running on top of your other OS to make life for Racket developers simple. It was originally developed for beginners, but I eat my own dog food, and I find it good (tm) for every day Racket work. To evaluate performance, run the programs at the command line. Measure there. Compare with other dynamically typed languages and report back what you find. If you still report performance problems, try to be precise. We are proud of Matthew and how far he has pushed Racket's performance on real software, the kind you use on a daily basis, not just minibenchmarks. >>> >>> I think ultimately that this answer isn't enough. If everyone who >>> tries out Racket in the way we suggest comes away with the impression >>> that it's really slow, suggestions on the mailing list to measure >>> differently won't eliminate the negative first impression, let alone >>> for all the people who _don't_ ask about it. >>> >>> Could we >>> - warn people when they use `time` in DrRacket? >>> - provide a "performance" mode that runs programs out-of-process, or >>> just in another place? >>> - something else? >>> >>> Sam >>> _ >>> Racket Developers list: >>> http://lists.racket-lang.org/dev >> >> >> >> -- >> -- >> Jens Axel Søgaard >> >> _ >> Racket Developers list: >> http://lists.racket-lang.org/dev -- -- Jens Axel Søgaard _ Racket Developers list: http://lists.racket-lang.org/dev
Re: [racket-dev] [racket] "lab notebook on learning process"
Just remember, any self-respecting "benchmark" feature of a platform will detect standard benchmarks and substitute hand-tuned assembler. :) Neil V. _ Racket Developers list: http://lists.racket-lang.org/dev
Re: [racket-dev] [racket] "lab notebook on learning process" (was: Re: Macros baffle me)
> How about an extra button, a "Run Benchmark" button? You can already get this with the "benchmark" package, from pkg.racket-lang. > > /Jens Axel > > > 2014-05-06 19:43 GMT+02:00 Sam Tobin-Hochstadt : >> On Tue, May 6, 2014 at 12:14 PM, Matthias Felleisen >> wrote: >>> > Why does he think "Performance sucks"? Because here's the list of things that are slow >>> >>> DrRacket is an operating system running on top of your other OS >>> to make life for Racket developers simple. It was originally developed >>> for beginners, but I eat my own dog food, and I find it good (tm) for >>> every day Racket work. >>> >>> To evaluate performance, run the programs at the command line. Measure >>> there. Compare with other dynamically typed languages and report back >>> what you find. If you still report performance problems, try to be precise. >>> We are proud of Matthew and how far he has pushed Racket's performance on >>> real software, the kind you use on a daily basis, not just minibenchmarks. >> >> I think ultimately that this answer isn't enough. If everyone who >> tries out Racket in the way we suggest comes away with the impression >> that it's really slow, suggestions on the mailing list to measure >> differently won't eliminate the negative first impression, let alone >> for all the people who _don't_ ask about it. >> >> Could we >> - warn people when they use `time` in DrRacket? >> - provide a "performance" mode that runs programs out-of-process, or >> just in another place? >> - something else? >> >> Sam >> _ >> Racket Developers list: >> http://lists.racket-lang.org/dev > > > > -- > -- > Jens Axel Søgaard > > _ > Racket Developers list: > http://lists.racket-lang.org/dev _ Racket Developers list: http://lists.racket-lang.org/dev
Re: [racket-dev] [racket] "lab notebook on learning process"
On 05/06/2014 12:18 PM, Matthias Felleisen wrote: On May 6, 2014, at 2:02 PM, Jens Axel Søgaard wrote: How about an extra button, a "Run Benchmark" button? +ω I LOLed. Were you thinking of having it launch a front-end to Josh McGrath's benchmarking library? Neil ⊥ _ Racket Developers list: http://lists.racket-lang.org/dev
Re: [racket-dev] [racket] "lab notebook on learning process"
Jens Axel Søgaard wrote at 05/06/2014 02:02 PM: Sam: Could we - warn people when they use `time` in DrRacket? - provide a "performance" mode that runs programs out-of-process, or just in another place? - something else? I like this. DrRacket has a performance mode already (disable profiling and debugging) in the (hidded part) of the language dialog. The casual beginner is not likely to recognize it as such though. How about an extra button, a "Run Benchmark" button? In addition to these, would it make sense for the REPL banner in the Interactions pane to indicate when DrRacket debugging and profiling are on? This would be a hint to people about some of the common parameters affecting performance, along with Racket version, memory limit, GC, JIT,general platform, CPU arch, and maybe number of cores Racket knows about. When showing the run of some timed code on the email list, we can even copy&paste this banner with it, as a quick way to convey some of the pertinent parameters of our setup. Maybe don't show this info in student languages, but show it everywhere else. Neil V. _ Racket Developers list: http://lists.racket-lang.org/dev
Re: [racket-dev] [racket] "lab notebook on learning process" (was: Re: Macros baffle me)
On May 6, 2014, at 2:02 PM, Jens Axel Søgaard wrote: > How about an extra button, a "Run Benchmark" button? +ω _ Racket Developers list: http://lists.racket-lang.org/dev
Re: [racket-dev] [racket] "lab notebook on learning process" (was: Re: Macros baffle me)
Sam: > Could we > - warn people when they use `time` in DrRacket? > - provide a "performance" mode that runs programs out-of-process, or > just in another place? > - something else? I like this. DrRacket has a performance mode already (disable profiling and debugging) in the (hidded part) of the language dialog. The casual beginner is not likely to recognize it as such though. How about an extra button, a "Run Benchmark" button? /Jens Axel 2014-05-06 19:43 GMT+02:00 Sam Tobin-Hochstadt : > On Tue, May 6, 2014 at 12:14 PM, Matthias Felleisen > wrote: >> >>> > Why does he think "Performance sucks"? >>> >>> Because here's the list of things that are slow >> >> DrRacket is an operating system running on top of your other OS >> to make life for Racket developers simple. It was originally developed >> for beginners, but I eat my own dog food, and I find it good (tm) for >> every day Racket work. >> >> To evaluate performance, run the programs at the command line. Measure >> there. Compare with other dynamically typed languages and report back >> what you find. If you still report performance problems, try to be precise. >> We are proud of Matthew and how far he has pushed Racket's performance on >> real software, the kind you use on a daily basis, not just minibenchmarks. > > I think ultimately that this answer isn't enough. If everyone who > tries out Racket in the way we suggest comes away with the impression > that it's really slow, suggestions on the mailing list to measure > differently won't eliminate the negative first impression, let alone > for all the people who _don't_ ask about it. > > Could we > - warn people when they use `time` in DrRacket? > - provide a "performance" mode that runs programs out-of-process, or > just in another place? > - something else? > > Sam > _ > Racket Developers list: > http://lists.racket-lang.org/dev -- -- Jens Axel Søgaard _ Racket Developers list: http://lists.racket-lang.org/dev
Re: [racket-dev] [racket] "lab notebook on learning process" (was: Re: Macros baffle me)
Right, I don't think that any of Artyom's comments are about the issue that Matthias was bringing up, but I do think that's it a real issue. Sam On Tue, May 6, 2014 at 1:55 PM, Robby Findler wrote: > I don't think those are the things being complained about. I read a > complaint about non-incremental GC, a complaint about DrRacket IO > (which is really quite slow because it uses an editor which is > overkill for a stream of text), possibly a complaint about the FFI > (but maybe there's more there?). And, possibly even hiding in there is > the usual complaint about scrolling. There's lots of work to do. :) > > Robby > > On Tue, May 6, 2014 at 12:43 PM, Sam Tobin-Hochstadt > wrote: >> On Tue, May 6, 2014 at 12:14 PM, Matthias Felleisen >> wrote: >>> > Why does he think "Performance sucks"? Because here's the list of things that are slow >>> >>> DrRacket is an operating system running on top of your other OS >>> to make life for Racket developers simple. It was originally developed >>> for beginners, but I eat my own dog food, and I find it good (tm) for >>> every day Racket work. >>> >>> To evaluate performance, run the programs at the command line. Measure >>> there. Compare with other dynamically typed languages and report back >>> what you find. If you still report performance problems, try to be precise. >>> We are proud of Matthew and how far he has pushed Racket's performance on >>> real software, the kind you use on a daily basis, not just minibenchmarks. >> >> I think ultimately that this answer isn't enough. If everyone who >> tries out Racket in the way we suggest comes away with the impression >> that it's really slow, suggestions on the mailing list to measure >> differently won't eliminate the negative first impression, let alone >> for all the people who _don't_ ask about it. >> >> Could we >> - warn people when they use `time` in DrRacket? >> - provide a "performance" mode that runs programs out-of-process, or >> just in another place? >> - something else? >> >> Sam >> _ >> Racket Developers list: >> http://lists.racket-lang.org/dev _ Racket Developers list: http://lists.racket-lang.org/dev
Re: [racket-dev] [racket] "lab notebook on learning process" (was: Re: Macros baffle me)
I don't think those are the things being complained about. I read a complaint about non-incremental GC, a complaint about DrRacket IO (which is really quite slow because it uses an editor which is overkill for a stream of text), possibly a complaint about the FFI (but maybe there's more there?). And, possibly even hiding in there is the usual complaint about scrolling. There's lots of work to do. :) Robby On Tue, May 6, 2014 at 12:43 PM, Sam Tobin-Hochstadt wrote: > On Tue, May 6, 2014 at 12:14 PM, Matthias Felleisen > wrote: >> >>> > Why does he think "Performance sucks"? >>> >>> Because here's the list of things that are slow >> >> DrRacket is an operating system running on top of your other OS >> to make life for Racket developers simple. It was originally developed >> for beginners, but I eat my own dog food, and I find it good (tm) for >> every day Racket work. >> >> To evaluate performance, run the programs at the command line. Measure >> there. Compare with other dynamically typed languages and report back >> what you find. If you still report performance problems, try to be precise. >> We are proud of Matthew and how far he has pushed Racket's performance on >> real software, the kind you use on a daily basis, not just minibenchmarks. > > I think ultimately that this answer isn't enough. If everyone who > tries out Racket in the way we suggest comes away with the impression > that it's really slow, suggestions on the mailing list to measure > differently won't eliminate the negative first impression, let alone > for all the people who _don't_ ask about it. > > Could we > - warn people when they use `time` in DrRacket? > - provide a "performance" mode that runs programs out-of-process, or > just in another place? > - something else? > > Sam > _ > Racket Developers list: > http://lists.racket-lang.org/dev _ Racket Developers list: http://lists.racket-lang.org/dev
Re: [racket-dev] [racket] "lab notebook on learning process" (was: Re: Macros baffle me)
On Tue, May 6, 2014 at 12:14 PM, Matthias Felleisen wrote: > >> > Why does he think "Performance sucks"? >> >> Because here's the list of things that are slow > > DrRacket is an operating system running on top of your other OS > to make life for Racket developers simple. It was originally developed > for beginners, but I eat my own dog food, and I find it good (tm) for > every day Racket work. > > To evaluate performance, run the programs at the command line. Measure > there. Compare with other dynamically typed languages and report back > what you find. If you still report performance problems, try to be precise. > We are proud of Matthew and how far he has pushed Racket's performance on > real software, the kind you use on a daily basis, not just minibenchmarks. I think ultimately that this answer isn't enough. If everyone who tries out Racket in the way we suggest comes away with the impression that it's really slow, suggestions on the mailing list to measure differently won't eliminate the negative first impression, let alone for all the people who _don't_ ask about it. Could we - warn people when they use `time` in DrRacket? - provide a "performance" mode that runs programs out-of-process, or just in another place? - something else? Sam _ Racket Developers list: http://lists.racket-lang.org/dev
