Re: [racket-users] DrRacket Background Expansion Memory Usage
It sounds like there might be a leak somewhere. Is the program you were editing easy to share? Robby On Wed, Jan 18, 2017 at 1:56 AM Lehi Toskin wrote: > In DrRacket I had background expansion on for the longest time until I > started to notice that whenever I would work on my programs it would, well, > expand in the background and take up 100% of one of my cores and the memory > would slowly crawl upwards. A few times it got so bad that DrRacket by > itself would occupy 60% of my 8GB of memory, which is *quite* silly. > Turning off background expansion brings the memory usage down to a > cool(-ish) 230MB --- 400MB, even after editing and running my program a > number of times. > > > > What's the deal with background expansion ballooning out of control like > that? Do I just need to click on the memory usage button more often? > Background expansion is kind of a neat feature but it's not neat enough to > warrant such incredible memory consumption. > > > > -- > > You received this message because you are subscribed to the Google Groups > "Racket Users" group. > > To unsubscribe from this group and stop receiving emails from it, send an > email to racket-users+unsubscr...@googlegroups.com. > > For more options, visit https://groups.google.com/d/optout. > > -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to racket-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
[racket-users] can't open/start DrRacket v6.8.0.2 on macOS Sierra 10.12.2
I downloaded the 20170118 nightly 64bit Mac version of DrRacket v6.8.0.2, and I just can't open the application at all. It shuts down immediately. I've had this with other applications on Sierra, and rebooting the Mac fixed the problem, but this time I can't get the newest snapshot to start. My MBP is a 2012 edition(no touchbar temporarily requiring 32bit DrRacket). Just wondering if I'm the only one in this situation. I can certainly go back to the v6.7.0.1 I had been using, but I was hoping to try out the latest and greatest. Geoff -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to racket-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: [racket-users] can't open/start DrRacket v6.8.0.2 on macOS Sierra 10.12.2
> On Jan 18, 2017, at 4:45 PM, Geoffrey Knauth wrote: > > I downloaded the 20170118 nightly 64bit Mac version of DrRacket v6.8.0.2, and > I just can't open the application at all. It shuts down immediately. I've > had this with other applications on Sierra, and rebooting the Mac fixed the > problem, but this time I can't get the newest snapshot to start. My MBP is a > 2012 edition(no touchbar temporarily requiring 32bit DrRacket). > > Just wondering if I'm the only one in this situation. I can certainly go > back to the v6.7.0.1 I had been using, but I was hoping to try out the latest > and greatest. I had this problem too, but running raco setup fixed it. Also, opening DrRacket from the command line via ./bin/drracket should work within the Racket v6.8.0.2 directory. Alex Knauth > Geoff -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to racket-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: [racket-users] can't open/start DrRacket v6.8.0.2 on macOS Sierra 10.12.2
My guess would be that it's related to the nightly builds not being signed (which is a problem for Sierra), but if 6.7.0.1 (which also looks like a nightly build, unless you built it from source) is working for you, then that may not be it. 6.8 should be coming out before the end of the month (and be signed). In the meantime, 6.8 pre-release builds are available from pre-release.racket-lang.org (and should also be signed). Signing nightly builds is on the to-do list. Vincent On Wed, 18 Jan 2017 09:45:03 -0600, Geoffrey Knauth wrote: > > I downloaded the 20170118 nightly 64bit Mac version of DrRacket v6.8.0.2, and > I just can't open the application at all. It shuts down immediately. I've > had this with other applications on Sierra, and rebooting the Mac fixed the > problem, but this time I can't get the newest snapshot to start. My MBP is a > 2012 edition(no touchbar temporarily requiring 32bit DrRacket). > > Just wondering if I'm the only one in this situation. I can certainly go > back to the v6.7.0.1 I had been using, but I was hoping to try out the latest > and greatest. > > Geoff > > -- > You received this message because you are subscribed to the Google Groups > "Racket Users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to racket-users+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to racket-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: [racket-users] can't open/start DrRacket v6.8.0.2 on macOS Sierra 10.12.2
This message from Matthew (and a few others like it) are probably relevant. https://groups.google.com/forum/#!topic/racket-users/4Bl5OcU6DOI Robby On Wed, Jan 18, 2017 at 12:00 PM, Vincent St-Amour wrote: > My guess would be that it's related to the nightly builds not being > signed (which is a problem for Sierra), but if 6.7.0.1 (which also looks > like a nightly build, unless you built it from source) is working for > you, then that may not be it. > > 6.8 should be coming out before the end of the month (and be signed). In > the meantime, 6.8 pre-release builds are available from > pre-release.racket-lang.org (and should also be signed). > > Signing nightly builds is on the to-do list. > > Vincent > > > > On Wed, 18 Jan 2017 09:45:03 -0600, > Geoffrey Knauth wrote: >> >> I downloaded the 20170118 nightly 64bit Mac version of DrRacket v6.8.0.2, >> and I just can't open the application at all. It shuts down immediately. >> I've had this with other applications on Sierra, and rebooting the Mac fixed >> the problem, but this time I can't get the newest snapshot to start. My MBP >> is a 2012 edition(no touchbar temporarily requiring 32bit DrRacket). >> >> Just wondering if I'm the only one in this situation. I can certainly go >> back to the v6.7.0.1 I had been using, but I was hoping to try out the >> latest and greatest. >> >> Geoff >> >> -- >> You received this message because you are subscribed to the Google Groups >> "Racket Users" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to racket-users+unsubscr...@googlegroups.com. >> For more options, visit https://groups.google.com/d/optout. > > -- > You received this message because you are subscribed to the Google Groups > "Racket Users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to racket-users+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to racket-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: [racket-users] DrRacket Background Expansion Memory Usage
The program I've been working on is https://github.com/lehitoskin/ivy - it has a few library dependencies, but those should be easy to get. On Wednesday, January 18, 2017 at 4:04:01 AM UTC-8, Robby Findler wrote: > It sounds like there might be a leak somewhere. Is the program you were > editing easy to share? > > > Robby -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to racket-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
[racket-users] Re: can't open/start DrRacket v6.8.0.2 on macOS Sierra 10.12.2
Thanks Alex, Vincent and Robby [and Matthew]! I'll remember all your tricks. What worked this time was the Matthew-trick Robby referenced about moving the .app to the enclosing folder and then putting it back in the correct folder. But it's good to know there are other ways. Geoff -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to racket-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
[racket-users] Bugs in the Racket Bytecode Versifier (as of 6.7)
Hey all, I noticed that the following program fails to run, but only after its been compiled (on 6.7, 6.6 and earlier don't seem to have this issue): https://gist.github.com/LeifAndersen/dfad9a8f55ef671e4f3aa41b321ef7bf #lang racket/base ;; ill-formed.rkt (define ill (let ((base (string-append "a"))) (λ () base))) (ill) To get the error, run: $ raco make ill-formed.rkt $ racket ill-formed.rkt ?? 0 compiled/ill-formed_rkt.zo::369: read (compiled): ill-formed code [../../../racket/gc2/../src/validate.c:1426] context...: standard-module-name-resolver This error doesn't seem to happen if `string-append` is replaced with various other functions that create strings (or just string literals), and it seems to go away if the function is inlined. I'm not entirely sure what's causing this bug yet, but it appears to either be a bug in the versifier, or a new optimization. Has anyone else seen something like this recently? (If not I'll update the list once I get a sense as to what is going on.) ~Leif Andersen -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to racket-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: [racket-users] Bugs in the Racket Bytecode Versifier (as of 6.7)
The bug is a mismatch between the optimizer and validator, where the optimizer can see that `string-append` is applied to a string and will definitely succeed, while the validator can't infer that behavior. The fact by itself isn't the problem, but it allows a further inference in the optimizer that's relevant to the validator. The optimizer should use a specific flag internally at a specific point to keep it consistent with the validator. I'll push that repair after testing. At Wed, 18 Jan 2017 15:58:15 -0500, Leif Andersen wrote: > Hey all, > > I noticed that the following program fails to run, but only after its been > compiled (on 6.7, 6.6 and earlier don't seem to have this issue): > > https://gist.github.com/LeifAndersen/dfad9a8f55ef671e4f3aa41b321ef7bf > > #lang racket/base ;; ill-formed.rkt > (define ill > (let ((base (string-append "a"))) > (λ () base))) > (ill) > > To get the error, run: > > $ raco make ill-formed.rkt > $ racket ill-formed.rkt > ?? 0 > compiled/ill-formed_rkt.zo::369: read (compiled): ill-formed code > [../../../racket/gc2/../src/validate.c:1426] > context...: >standard-module-name-resolver > > This error doesn't seem to happen if `string-append` is replaced with > various other functions that create strings (or just string literals), and > it seems to go away if the function is inlined. > > I'm not entirely sure what's causing this bug yet, but it appears to either > be a bug in the versifier, or a new optimization. Has anyone else seen > something like this recently? (If not I'll update the list once I get a > sense as to what is going on.) > > ~Leif Andersen > > -- > You received this message because you are subscribed to the Google Groups > "Racket Users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to racket-users+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to racket-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
[racket-users] Serializing macro transformer procedures
I'm trying to use serial-lambda in macro transformer procedures so that I can serialize them and bring them down to run-time. However, serial-lambda isn't working within a define-syntax. It says: syntax-local-lift-provide: not expanding in a module run-time body This seems to be a lie; it works fine in a begin-for-syntax (as long as you don't try to deserialize). What's the real reason it doesn't work in a define-syntax? Is there another way to bring macro transformer procedures down to run-time? Alex Knauth -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to racket-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
[racket-users] Dynamic-require a compile-time identifier?
Is there a way to dynamic-require an identifier provided at phase 1? #file x.rkt #lang racket (provide (for-syntax x)) ; x is provided at phase 1 (define-for-syntax x 3) #file use.rkt #lang racket (dynamic-require-from-phase-1 "x.rkt" 'x) ;=> 3 I could do it by defining another module that did the phase shift: #file x.rkt #lang racket (provide (for-syntax x)) (define-for-syntax x 3) (module* ct racket/base (provide (for-meta -1 (all-from-out (submod ".." (require (for-meta -1 (submod ".." #file use.rkt #lang racket (dynamic-require '(submod "x.rkt" ct) 'x) But that requires an extra module to be defined already. I could also do it with eval, but that feels wrong and shouldn't be necessary. There should be a way to do this without needing eval. Alex Knauth -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to racket-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: [racket-users] Typed Racket & Higher Kinded Types
On 01/17/2017 05:31 AM, Sam Tobin-Hochstadt wrote: > The major obstacle is that the current kind system design is not easy to > reconcile with higher kinds. This is mostly because the current system > is poorly designed, but we need to avoid breaking existing programs. Why is this an issue? #lang something/else, and the problem is solved, no? -- Anthony Carrico -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to racket-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
Re: [racket-users] Typed Racket & Higher Kinded Types
And how would components in #lang typed/racket interact with components in #lang kinded/racket interact? > On Jan 18, 2017, at 8:16 PM, Anthony Carrico wrote: > > On 01/17/2017 05:31 AM, Sam Tobin-Hochstadt wrote: >> The major obstacle is that the current kind system design is not easy to >> reconcile with higher kinds. This is mostly because the current system >> is poorly designed, but we need to avoid breaking existing programs. > > Why is this an issue? #lang something/else, and the problem is solved, no? > > -- > Anthony Carrico > > -- > You received this message because you are subscribed to the Google Groups > "Racket Users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to racket-users+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. -- You received this message because you are subscribed to the Google Groups "Racket Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to racket-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
