Re: [racket-dev] flvector-copy
I'm preparing to push this patch, but also generalizing `flvector-copy' to accept start and end positions like `vector-copy'. At Thu, 16 Sep 2010 20:46:58 -0400, "Will M. Farr" wrote: > The attached patch against the current git master adds an flvector-copy > procedure (along with docs and tests); it's simple, but nice to have in the > flvector library. Let me know if there are any issues with including this in > racket. _ For list-related administrative tasks: http://lists.racket-lang.org/listinfo/dev
Re: [racket-dev] rename + changes => bad? (was: Fwd: [plt] Push #21110: master branch updated)
On Sep 17, 2010, at 11:03 AM, Sam Tobin-Hochstadt wrote: > On Fri, Sep 17, 2010 at 1:36 PM, John Clements > wrote: >> Oof... it looks like doing a rename plus a few changes pretty effectively >> hides those changes, though I can dig them out with 'git annotate'. Do we >> have a policy (or should we have) a policy stating that you shouldn't put a >> rename and modifications to that file in the same commit? > > I believe that this is an issue with Eli's email formatting script, > and not with git itself (the github diff is fine: > http://github.com/plt/racket/commit/a095ebc326aa35b5b19f314080855db5d55ed49b Github seems to do a better job than 'git diff' or 'git show' with these; both of these simply showed me an entire file of lines added and an entire file of lines deleted. John smime.p7s Description: S/MIME cryptographic signature _ For list-related administrative tasks: http://lists.racket-lang.org/listinfo/dev
Re: [racket-dev] rename + changes => bad? (was: Fwd: [plt] Push #21110: master branch updated)
On Fri, Sep 17, 2010 at 1:36 PM, John Clements wrote: > Oof... it looks like doing a rename plus a few changes pretty effectively > hides those changes, though I can dig them out with 'git annotate'. Do we > have a policy (or should we have) a policy stating that you shouldn't put a > rename and modifications to that file in the same commit? I believe that this is an issue with Eli's email formatting script, and not with git itself (the github diff is fine: http://github.com/plt/racket/commit/a095ebc326aa35b5b19f314080855db5d55ed49b ). -- sam th sa...@ccs.neu.edu _ For list-related administrative tasks: http://lists.racket-lang.org/listinfo/dev
[racket-dev] rename + changes => bad? (was: Fwd: [plt] Push #21110: master branch updated)
Oof... it looks like doing a rename plus a few changes pretty effectively hides those changes, though I can dig them out with 'git annotate'. Do we have a policy (or should we have) a policy stating that you shouldn't put a rename and modifications to that file in the same commit? John Begin forwarded message: > From: cleme...@racket-lang.org > Date: September 17, 2010 10:19:37 AM PDT > To: dev@racket-lang.org > Subject: [plt] Push #21110: master branch updated > > clements has updated `master' from 04defa4937 to a095ebc326. > http://git.racket-lang.org/plt/04defa4937..a095ebc326 > > =[ 1 Commits ]== > > Directory summary: > 100.0% collects/test-engine/ > > ~~ > > a095ebc John Clements 2010-09-17 10:18 > : > | added format arg to printf to avoid corner-case printf bugs, rackety changes > : > M collects/test-engine/racket-tests.rkt |2 +- > M collects/test-engine/test-display.scm |2 +- > R collects/test-engine/{test-engine.scm => test-engine.rkt} (99%) > > =[ Overall Diff ]=== > > collects/test-engine/racket-tests.rkt > ~ > --- OLD/collects/test-engine/racket-tests.rkt > +++ NEW/collects/test-engine/racket-tests.rkt > @@ -5,7 +5,7 @@ > scheme/match >lang/private/continuation-mark-key > (only scheme/base for memf findf) > - "test-engine.scm" > + "test-engine.rkt" >"test-info.scm" > ) > > > collects/test-engine/test-display.scm > ~ > --- OLD/collects/test-engine/test-display.scm > +++ NEW/collects/test-engine/test-display.scm > @@ -6,7 +6,7 @@ > framework > string-constants > "test-info.scm" > - "test-engine.scm" > + "test-engine.rkt" >"print.ss" >(except-in deinprogramm/signature/signature signature-violation) ; > clashes with test-engine >deinprogramm/quickcheck/quickcheck) > > *** See above for renames and copies *** smime.p7s Description: S/MIME cryptographic signature _ For list-related administrative tasks: http://lists.racket-lang.org/listinfo/dev
Re: [racket-dev] [plt] Push #21108: master branch updated
Sure, a parameter is fine. And since we're going to have a paramter, then defaulting to the old way seems best. Thanks! Robby On Fri, Sep 17, 2010 at 12:02 PM, Casey Klein wrote: > On Fri, Sep 17, 2010 at 11:52 AM, Robby Findler > wrote: >> On Fri, Sep 17, 2010 at 11:50 AM, Casey Klein >> wrote: >>> On Fri, Sep 17, 2010 at 11:39 AM, wrote: robby has updated `master' from c6fc7137ee to ce211ac364. http://git.racket-lang.org/plt/c6fc7137ee..ce211ac364 =[ 1 Commits ]== Directory summary: 10.1% collects/redex/private/ 83.2% collects/redex/tests/bmps-macosx/ 6.6% collects/redex/tests/ ~~ ce211ac Robby Findler 2010-09-17 11:39 : | adjusted metafunction application rendering so that ellipses do not | get commas put in front of them : >>> >>> FWIW, if I were typesetting this example by hand, I would have done it >>> the old way >>> >>> rdups[| x_1, x_2, ..., x_1, x_3, ... |] >>> >>> not the new way >>> >>> rdups[| x_1, x_2 ..., x_1, x_3 ... |] >>> >>> but maybe I'm alone in that preference. >> >> I had thought that for a while, but some examples I'm working with >> today get really confusing when you do that and just seem ugly. The >> ellipses really isn't a separate item in the argument list; it is an >> operator on the thing that comes before and the lack of a paren >> emphasizes this properly. >> >> IMO. >> > > In that case, maybe there should be a parameter? (I'll be the one to > add it if you want.) > > This rendering seems like a departure from convention. The first two > books I pulled off my shelf (_Semantics of Programming Languages_ and > _Invitation to Discrete Mathematics_) do it the old way. > _ For list-related administrative tasks: http://lists.racket-lang.org/listinfo/dev
Re: [racket-dev] [plt] Push #21108: master branch updated
On Fri, Sep 17, 2010 at 11:52 AM, Robby Findler wrote: > On Fri, Sep 17, 2010 at 11:50 AM, Casey Klein > wrote: >> On Fri, Sep 17, 2010 at 11:39 AM, wrote: >>> robby has updated `master' from c6fc7137ee to ce211ac364. >>> http://git.racket-lang.org/plt/c6fc7137ee..ce211ac364 >>> >>> =[ 1 Commits ]== >>> >>> Directory summary: >>> 10.1% collects/redex/private/ >>> 83.2% collects/redex/tests/bmps-macosx/ >>> 6.6% collects/redex/tests/ >>> >>> ~~ >>> >>> ce211ac Robby Findler 2010-09-17 11:39 >>> : >>> | adjusted metafunction application rendering so that ellipses do not >>> | get commas put in front of them >>> : >> >> FWIW, if I were typesetting this example by hand, I would have done it >> the old way >> >> rdups[| x_1, x_2, ..., x_1, x_3, ... |] >> >> not the new way >> >> rdups[| x_1, x_2 ..., x_1, x_3 ... |] >> >> but maybe I'm alone in that preference. > > I had thought that for a while, but some examples I'm working with > today get really confusing when you do that and just seem ugly. The > ellipses really isn't a separate item in the argument list; it is an > operator on the thing that comes before and the lack of a paren > emphasizes this properly. > > IMO. > In that case, maybe there should be a parameter? (I'll be the one to add it if you want.) This rendering seems like a departure from convention. The first two books I pulled off my shelf (_Semantics of Programming Languages_ and _Invitation to Discrete Mathematics_) do it the old way. _ For list-related administrative tasks: http://lists.racket-lang.org/listinfo/dev
Re: [racket-dev] [plt] Push #21108: master branch updated
On Fri, Sep 17, 2010 at 11:50 AM, Casey Klein wrote: > On Fri, Sep 17, 2010 at 11:39 AM, wrote: >> robby has updated `master' from c6fc7137ee to ce211ac364. >> http://git.racket-lang.org/plt/c6fc7137ee..ce211ac364 >> >> =[ 1 Commits ]== >> >> Directory summary: >> 10.1% collects/redex/private/ >> 83.2% collects/redex/tests/bmps-macosx/ >> 6.6% collects/redex/tests/ >> >> ~~ >> >> ce211ac Robby Findler 2010-09-17 11:39 >> : >> | adjusted metafunction application rendering so that ellipses do not >> | get commas put in front of them >> : > > FWIW, if I were typesetting this example by hand, I would have done it > the old way > > rdups[| x_1, x_2, ..., x_1, x_3, ... |] > > not the new way > > rdups[| x_1, x_2 ..., x_1, x_3 ... |] > > but maybe I'm alone in that preference. I had thought that for a while, but some examples I'm working with today get really confusing when you do that and just seem ugly. The ellipses really isn't a separate item in the argument list; it is an operator on the thing that comes before and the lack of a paren emphasizes this properly. IMO. Robby _ For list-related administrative tasks: http://lists.racket-lang.org/listinfo/dev
Re: [racket-dev] [plt] Push #21108: master branch updated
On Fri, Sep 17, 2010 at 11:39 AM, wrote: > robby has updated `master' from c6fc7137ee to ce211ac364. > http://git.racket-lang.org/plt/c6fc7137ee..ce211ac364 > > =[ 1 Commits ]== > > Directory summary: > 10.1% collects/redex/private/ > 83.2% collects/redex/tests/bmps-macosx/ > 6.6% collects/redex/tests/ > > ~~ > > ce211ac Robby Findler 2010-09-17 11:39 > : > | adjusted metafunction application rendering so that ellipses do not > | get commas put in front of them > : FWIW, if I were typesetting this example by hand, I would have done it the old way rdups[| x_1, x_2, ..., x_1, x_3, ... |] not the new way rdups[| x_1, x_2 ..., x_1, x_3 ... |] but maybe I'm alone in that preference. > M collects/redex/private/core-layout.rkt | 10 +++--- > M collects/redex/tests/bitmap-test.rkt | 12 > A collects/redex/tests/bmps-macosx/rdups.png > > =[ Overall Diff ]=== > > collects/redex/private/core-layout.rkt > ~~ > --- OLD/collects/redex/private/core-layout.rkt > +++ NEW/collects/redex/private/core-layout.rkt > @@ -258,9 +258,13 @@ > (just-after (close-white-square-bracket) last)))] > [(null? (cddr lst)) > (cons (car lst) (loop (cdr lst)))] > - [else (list* (car lst) > - (just-after (basic-text "," (default-style)) > (car lst)) > - (loop (cdr lst)))] > + [else > + (if (eq? '... (lw-e (cadr lst))) > + (cons (car lst) > + (loop (cdr lst))) > + (list* (car lst) > + (just-after (basic-text "," (default-style)) (car > lst)) > + (loop (cdr lst] > > (define (just-before what lw) > (build-lw (if (symbol? what) > > collects/redex/tests/bitmap-test.rkt > > --- OLD/collects/redex/tests/bitmap-test.rkt > +++ NEW/collects/redex/tests/bitmap-test.rkt > @@ -187,6 +187,18 @@ > ;; make sure two metafunctions simultaneously rewritten line up properly > (test (render-metafunctions S T TL) "metafunctions-multiple.png") > > +;; make sure that the ellipses don't have commas before them. > +(define-metafunction lang > + rdups : x ... -> (x ...) > + [(rdups x_1 x_2 ... x_1 x_3 ...) > + (rdups x_2 ... x_1 x_3 ...)] > + [(rdups x_1 x_2 ...) > + (x_1 x_3 ...) > + (where (x_3 ...) (rdups x_2 ...))] > + [(rdups) ()]) > + > +(test (render-metafunction rdups) "rdups.png") > + > ;; Non-terminal superscripts > (test (render-lw lang (to-lw (x_^abcdef x_q^abcdef))) > "superscripts.png") > > collects/redex/tests/bmps-macosx/rdups.png > ~~ > Binary files /dev/null and NEW/collects/redex/tests/bmps-macosx/rdups.png > differ > _ For list-related administrative tasks: http://lists.racket-lang.org/listinfo/dev
Re: [racket-dev] [plt] Push #21102: master branch updated
Since you asked... probably the string-constants library should not be used for those error messages, since none of the other error messages are translated, but since they are already being used it seems fine to continue (at least to me). Robby _ For list-related administrative tasks: http://lists.racket-lang.org/listinfo/dev