> * Racket CS supports nonatomic allocation via ffi/unsave
I'm almost sure it's a typo: unsaVe -> unsaFe
Gustavo.
El dom, 11 de jul. de 2021 a la(s) 07:07, 'John Clements' via Racket
Developers (racket-dev@googlegroups.com) escribió:
> The release announcement sketch that I have so far is
branches as well.
>
> Sam
>
> On Sat, Oct 26, 2019 at 8:32 AM Gustavo Massaccesi
> wrote:
> >
> > I pushed to the wrong branch in racket/games :(
> >
> > https://github.com/racket/games/tree/19-10-SameTypo
> >
> > Is it possible to remove this
I pushed to the wrong branch in racket/games :(
https://github.com/racket/games/tree/19-10-SameTypo
Is it possible to remove this branch to avoid confusion?
Gustavo
PS: There is another branch in that repo that looks like a similar (old)
error and a few branches in racket/racket. Perhaps it's a
In both cases, I think there will be a license problem. IIUC the license of
libgccjit is GPLv3.
Gustavo
--
You received this message because you are subscribed to the Google Groups
"Racket Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to rac
One posible idea that I think that would be indirectly useful, is to
translate the instructions of Matthew to an example in Travis-CI that
compiles the RacketCS and then run an small amount of tests. This requires
some knowdledge of the quirks of Travis-CI, so I'm not sure if it is newby
friendly e
et type system
> from being used to get access to unsafe operations.
>
> * Racket's internal runtime adds a `scheme_atexit` procedure with
> identical behavior as `atexit` defined in libc.
>
> The following people contributed to this release:
> Alex Knauth, Alexander
Somewhat related ...
I always thought it was strange that mutable pairs can’t be
chaperoned, but boxes and vectors of length 2 can. Moreover, there was
(is?) a plan to replace the implementation of mcons with structs. Does
#:authentic make this more consistent / efficient?
Also, what would happen
It's a good version. thanks
Gustavo
On Mon, Apr 17, 2017 at 10:32 PM, Vincent St-Amour
wrote:
> On Mon, 17 Apr 2017 18:00:42 -0500,
> Gustavo Massaccesi wrote:
>>
>> What about 409321c0?
>> - When raco detects an ambiguous prefix of a command, it shows a list
>&g
a21b33a7
- The bytecode compiler detects more optimization opportunities for
structure operations.
b73e1dfd
[It's too internal. Not worth mentioning.]
What about 409321c0?
- When raco detects an ambiguous prefix of a command, it shows a list
of all the matches.
[I really don't like the redactio
I have been looking at some of the optimizations, and I found a few
interesting differences.
*** Chez may reorder expressions.
(car (list (display 1) (display 2))) ;==> 21#
I think this is ok in RNRS and Chez, but it may cuase a problem for a
program translated directly from Racket. I think th
What about something like:
The bytecode optimizer can reduce more expressions like (equal? x y)
to (eqv? x y) or (eq? x y), in case the type of x or y is known and
the comparison is equivalent for that type.
Gustavo
--
You received this message because you are subscribed to the Google Groups
"
It's fixed. Thanks.
Gustavo
--
You received this message because you are subscribed to the Google Groups
"Racket Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to racket-dev+unsubscr...@googlegroups.com.
To post to this group, send email to r
Hi!
There is no "current" release version at
http://pre-release.racket-lang.org/installers/racket-current-x86_64-linux.sh
. I use it for a few informal automatic test with travis-racket.
Gustavo
On Sun, Jan 15, 2017 at 4:12 PM, Vincent St-Amour
wrote:
> Checklist items for the v6.8 release
>
- c3595c56 9ebfdb54 25dc89a2
I think they are too minor.
- 58333903
I like it, specially because it's the result of many intermediate
steps, like 7c1cb1a2. But it's difficult to explain (specially because
it depends on the knowledge that equal? is much slower than eq?). So I
think it's better to i
I'm not sure that it's so easy. The transformation must ensure that
this program works correctly:
#lang racket
(define-syntax-rule (#%datum . x)
'(2 2 2))
(for ([i 5])
(displayln i))
Gustavo
On Wed, Jan 4, 2017 at 7:17 PM, Matthew Butterick wrote:
>
> On Jan 3, 2017, at 6:07 PM, Robby Findl
It's fine for me.
Gustavo
On Sun, Oct 16, 2016 at 8:13 PM, Vincent St-Amour
wrote:
> On Sun, 16 Oct 2016 07:59:30 -0500,
> Gustavo Massaccesi wrote:
>>
>> > 397d6041823497a639c09660eca2f0573a58c62e
>> > 8bb79deaa21ebbaba544ec29c6702369a3b23035
>> &
> 397d6041823497a639c09660eca2f0573a58c62e
> 8bb79deaa21ebbaba544ec29c6702369a3b23035
> f8b3ba8253b2b94e5f0b060bc256f651ba73dd48
Are too minor or too internal.
> d4158c2b04b71d193ab32e43396b62b865ec2df6
> dce42313ad3dea250acd1de4c56161b3fd87eca1
> e5e781c4ecffe5494d662e1a43ba3a5dcc2db887
* The R
> Those numbers look like pointers interpreted as fixnums
Yes.
The optimizer thought that the result of (bitwise-and num -2) was a
fixnum, so it changed
(bitwise-ior (bitwise-and num -2) 0)
to
(unsafe-fxior (bitwise-and num -2) 0)
All the pointers are even (actually a multiple of 4 in 32 bi
I´m changing my computer and I got a weird error. I solved this error,
but I'm posting this in case someone gets the same error in the
future.
I copied all the folder with the Racket repository from one computer
to the other instead of using git clone. [Is this a good idea?] [The
same error may ha
I skimmed
https://github.com/mflatt/racket/blob/linklet/racket/src/racket/src/startup.inc
looking for things that look slow.
I noticed that the code uses many keyword procedures with optional
keywords. When all the keyword are used, the application is replaced
with an alternative without keywords
>> > this release in the announcement.
>> >
>> > The list is below. Let me know if anything looks wrong to you.
>> >
>> > Vincent
>> >
>> >
>> >
>> > Adrien Tateno
>> > Alex Knauth
>> > Alexander Shopov
>> > Al
*** 1b54b1c0400dd25110efb90e1da6fda34487f03c and
2bfb851ccc49b3de72e5c5d60b8d337d890a068e
- The optimizer can figure out in more cases that a variable is always
a number, and remove the run time checks and omitable calculations.
*** 3f246dd857c886214ef6766f5c3baf2f50340b06
It's an internal ref
Since the weekend, the installer for the "current" version of the
Release candidate is not available. For example:
http://pre-release.racket-lang.org/installers/racket-current-x86_64-linux-ubuntu-precise.sh
Gustavo
--
You received this message because you are subscribed to the Google Groups
"Ra
:) Just a few minor comments:
It's better to keep the code for strings and bytes in sync, so an
"easy" project is to add the equivalents optimizations for strings.
("easy" because I think that using this as a template, there won't be
any surprise.)
The problem with string/bytes is that too many i
I have a small bugfix for the gui-pkg-manager, but someone that know
more about the code has to review it.
https://github.com/racket/gui-pkg-manager/pull/2
Gustavo
On Fri, Apr 1, 2016 at 2:50 PM, Ryan Culpepper wrote:
> The release process for v6.5 will begin in about a week. If
> you have any
Since this morning, I'm getting this error building racket, when the
script wants to update the packages (in Windows 7)
ssl-connect: connect failed (error:14077410:SSL
routines:SSL23_GET_SERVER_HELLO: sslv3 alert handshake failure)
context...:
C:\Users\[...]\plt\racket\collects\openssl\mzssl
I'm not sure that this was the problem. That day I got a similar
error, and I was using the new repository.
Now I don't get any problem, so whatever it was, it's probably fixed.
Gustavo
On Tue, Jan 26, 2016 at 3:56 AM, Faré wrote:
> The SSL on pkgs.racket-lang.org looks like it's OK after all.
I'm trying to add support in the JIT for primitive-result-arity and
procedure-result-arity. My idea is to copy the implementation of
procedure-arity-includes? and make some modifications.
In the JIT, procedure-arity-includes? only has a special case for
scheme_native_closure_type and scheme_prim_t
te from multiple snapshot builds, instead of the
>> step that creates a site for a single build. Unless Ryan creates the
>> links manually (or through his own scripts), someone would need to
>> adjust the general build script (in "distro-build") to add that.
>>
&
> out, to prepare for the release? (The release branch is a relatively
> short-lived thing that exists only in the run up to a release.)
>
> Robby
>
> On Sat, Oct 10, 2015 at 10:01 AM, Gustavo Massaccesi
> wrote:
>> I have setup a copy of one little-schemer [1] project
I have setup a copy of one little-schemer [1] project, with some
tweaks to runs the test in travis [2]. The 6.2 version had a bug that
broke their code, that was fixed in 6.2.1 [3]. So I thought it would
be nice to test it to be sure that the new version doesn't breaks
their code. (This looks easie
I went to the old "obsolete" repository webpge
http://git.racket-lang.org/plt and it's not updated since 3 days (as
expected?).
I think it would be a good idea to put there a message with a link to
the new repository.
Gustavo
On Wed, Sep 16, 2015 at 5:15 PM, Sam Tobin-Hochstadt
wrote:
> Summary
[I should take a week to read this more carefully, but I'll ask now anyway ...]
Can you explain (again) the difference between aclock and sclock? Does
`box` increases aclock but don't increase sclock? Does `set-box!`
increase both? Is it possible to increase sclock without increasing
aclock?
Gust
I think you'd like to delete the MzCOM.tmp_Release_x64.vcproj file.
Gustavo
On Mon, Aug 17, 2015 at 1:16 PM, wrote:
> mflatt has updated `master' from 641c56b6e9 to 1b493f2146.
> http://git.racket-lang.org/plt/641c56b6e9..1b493f2146
>
> =[ One Commit ]=
Slightly off-topic: I didn't notice 'original-for-check-syntax. A few
weeks ago I had problems with a missing arrows in DrRacket, and I
searched for 'disappeared-use and I got
http://docs.racket-lang.org/reference/stxprops.html?q=disappeared-use#%28idx._%28gentag._188._%28lib._scribblings%2Freferen
> | For reasons that are not clear, the new expander triggered
> | the problem through an existing test.
I think the problem was made evident by the previous commit "fix
optimizer's single-use tracking, especially on inlining" (
https://github.com/plt/racket/commit/bfc2b27d653de50ce0e111962f59e1a3
On Tue, Jul 14, 2015 at 10:48 AM, John Clements
wrote:
>
>> On Jul 11, 2015, at 9:46 PM, Gustavo Massaccesi wrote:
>>
>> For a Racket version, it would be very helpful to have a special case
>> for tail calls.
>
> Can you tell me what you mean by this? I’m n
For a Racket version, it would be very helpful to have a special case
for tail calls.
(The support of continuations looks difficult, but they are no very
used in beginners programs)
Gustavo
On Sat, Jul 11, 2015 at 12:28 PM, 'John Clements' via Racket
Developers wrote:
>
>> On Jul 7, 2015, at 5
After the recent change, DrDr sends all the messages of the rebuids to
the latest pusher. But Most of the times the changes are due to
modifications in the packages, not in the main repository.
I think that the it should send the messages only after the first rebuild.
(My latest commit is only a
Sometimes, I have no internet connectivity.Is there an option for
"make" to rebuild Racket from the source without trying to update the
packages?
Gustavo
--
You received this message because you are subscribed to the Google Groups
"Racket Developers" group.
To unsubscribe from this group and st
Just for fun. Don't try this at home:
#lang racket
(define-syntax-rule (quasiquote x) (reverse 'x))
`(1 2 3 4)
Gustavo
--
You received this message because you are subscribed to the Google Groups
"Racket Developers" group.
To unsubscribe from this group and stop receiving emails from it, send
I'm still not sure that this is related, but in the old system two
equal consecutive marks cancel. But if you compose two markers, the
"composed" mark doesn't cancel when it's applied twice. For example:
;---
#lang racket
(define A (make-syntax-introducer))
(define B (make-syntax-introducer))
(def
42 matches
Mail list logo
