I’ve send it to bug-guile, should appear in the issue tracker soon
#74210
On Thu, 7 Nov 2024 17:10:45 +0100
Maxime Devos wrote:
> On Thu, 7 Nov 2024 12:23:08 +0100
> >Maxime Devos wrote:
> >> ‘atexit’ functions are run at ‘exit’. ‘exit’ can be run from signal
> >> handlers (*). Since the hook runs Scheme code, it could do a lot of
> >
even ‘malloc’, so likely I’m looking in
the wrong places.
Best regards,
Maxime Devos
On Thu, 7 Nov 2024 12:23:08 +0100
>Maxime Devos wrote:
>> ‘atexit’ functions are run at ‘exit’. ‘exit’ can be run from signal
>> handlers (*). Since the hook runs Scheme code, it could do a lot of
>> AC-unsafe things, resulting in problems.
>>
>> (*) glibc docum
econd fluid-ref invocation. and=> would avoid that.
Best regards,
Maxime Devos
That seems about right.
Now only needs to wait for someone able to actually merge this, I think.
Best regards,
Maxime Devos
>Where is BREAK defined?
Break is defined by ‘while’:
https://www.gnu.org/software/guile/manual/html_node/while-do.html
/commit/cc0e84cd56df3b07d378f710df39f8822317a2a2
• https://git.sr.ht/~old/guile-parallel/tree/master/item/parallel.scm#L29
• (what’s missing here, is a way to override Guile’s dynamic-wind in a
transparent manner)
Best regards,
Maxime Devos
passing ‘get-char’ you can act on each character, with ‘get-line’
I’m not sure what the difference would be, but apparently it’s not ‘read-line’
(?), if you give it a JSON reading+parsing proedure you iterate over all JSON
objects, with get-u8 you iterate over bytes etc..
Best regards,
Maxime Devos
user might want to set
CLOEXEC or other flags, or uncompress input, …)
Also, documentation is missing.
Best regards,
Maxime Devos
’t heard a good argument
yet for _not_ generalising it.
(*) I.e., while it technically does parse something (extract line from text),
it doesn’t parse the line itself (which in many cases will need to happen), and
‘recognising a line as a line’ is kind of trivial.
Best regards,
Maxime Devos
needs time, and those other parts would remain unanswered.
Also, when I (of my own volition, and in absence of external obligations
competing for time) _choose_ to spend some extra time on writing beyond
necessity, then it’s up to myself whether to consider this extra effort ‘waste’
or not, or
e assuming “\n” is a line delimiter. This is true under Unix
according to the documentation. But it doesn’t say anything about non-Unix
systems.
Best regards,
Maxime Devos
brary. See Venerable Port Interfaces, and See Writing Scheme Values, for
>documentation. Contrary to what R6RS prescribes, neither module automatically
>convers line feed characters to the line terminator of the end-of-line style.
(Please check the last point, seems like something that should be tested
experimentally and I don’t have a Guile setup on this computer.)
(Also, docstrings are to be modified accordingly)
Best regards,
Maxime Devos
.)
Rather, either the used newline in the test needs to be adjusted per-platform,
or the documentation of read-line needs to be adjusted to that \n is always a
newline.
Also, it’s also not a proper “back to the track” reply, since it ignores the
‘generalisation’ component of the track.
Regards,
Maxime Devos
s would be needed), as I
probably mentioned earlier in some other words.
Also, "anyone” includes me. You should stop wasting _my_ time by repeating your
ad nauseum messages -- whether it’s for just repeating your opinion, or
intended for (fallacious) argumentation or persuasion, it is quite ad nauseum.
Regards,
Maxime Devos
Can you stop it with the repetitive claims of “this is just my proposal, I have
freedom to state it, you don’t need to persuade me, etc.”? I’ve heard you the
first, dunno, 5 or so times. Except for the “just” qualifier, I did not
disagree anywhere. For why you shouldn’t do those repetitive claim
nstead of “\n” is
the line feed (I don’t know whether it recognised “\n”, IIRC it doesn’t).
Regards.
Maxime Devos
> Thanks for your patches! Applied now.
Regardless of the question whether to generalise, there was also the thing
about the test testing for returning #true even though this behaviour isn’t
documented anywhere. Also, there was the thing about needing to verify whether
(ice-9 rdelim) always rec
t’s a similar
kind of nonsense I don’t want to deal with either (the generalisation Is
trivial so nothing needs to be shown, what you are claiming is different from
what you are doing, and I did no cheap talking), so consider yourself added to
the blocklist.
Bye,
Maxime Devos
>Any opinions on what is best: Having a define-method* or having the
>functionality in define-method itself?
You can’t unify define-method with define-method* without making some arbitrary
choices in some special cases (the same applies to define-method* too actually,
and also to define-method
doesn’t.
A variant with untyped keyword arguments:
(define-method (f (a ) (b ))
(pk 'positional))
(define-method (f (#:key foo))
(pk 'optional-keyword foo))
(f #:foo 'bar)
Who should win? Both are a quite specific match.
Best regards,
Maxime Devos
use scm_i_ttyname.
Also, the explanation in the commit message belongs in a (code) comment, so
people don’t have to track down the commit that introduced the changes.
Best regards,
Maxime Devos
).
(For a non-malloc example, see the bug report (+ patch) about GOOPS methods and
,trace.)
Best regards,
Maxime Devos
’t think it can be shortened: it’s for Guile (so a SCM_ prefix for
namespacing), it’s for syscalls (so _SYSCALL suffix), and it’s a variant for
locking/unlocking (so LOCKED in the middle). Nothing to remove there, I’d think.
Best Regards,
Maxime Devos
t regards,
Maxime Devos
e.
How is fixing ttyname outside the scope of fixing ttyname? Previously, you said
that:
>[context: ttyname]
>In any case, I'm happy to fix all the problems if we can be confident […]
Best regards,
Maxime Devos
ing else that simply
takes a long while. During this time, the scm_i_misc_mutex is locked, so other
threads that want to use ttyname can’t progress for a long duration!
The variant of the SCM_SYSCALL loop I provided, avoids these issues.
Best regards,
Maxime Devos
.
(In case of unclarity and to be really sure to not copy too much, strncpy
exists where n=length of array excluding space reserved for terminating zero,
and things could be initialised to \0.)
Best regards,
Maxime Devos
are here about not exceeding the nominal maximum for its own
sake, rather it’s just to have a sufficiently large buffer.
Best regards,
Maxime Devos
xist, and hence
many of the procedures in that module still make sense, including
program-arities. In particular, arity information is still recorded (in
.guile.arities), so it should be made / kept accessible somewhere.
Best regards,
Maxime Devos
mplementation exists for
bootstrapping the Scheme Scheme interpreter, so it’s of less interest to get
arity information in there.)
Best regards,
Maxime Devos.
0 (with some error margin to deal with
inexactness). I'm pretty sure it's in radians.
(Usually I'd mention something about documentation here, but IIRC
Guile-Lib has some set-up where it copies docstrings.)
Best regards, Maxime Devos
:
For which θ is θ/π and tan(θ) rational?
Assume t=tan(θ)∈ℚ. Note that sin (2θ)=(2t)/(1+t^2)∈ℚ. Since sin(2θ) is
rational and 2θ/π is rational, then 2θ is an integer multiple of π/2.
So, θ is an integer multiple of π/4.
(For the other direction, this can be verified with a lookup table.)
Best regards,
Maxime Devos
On 16/02/2025 11:31, Maxime Devos wrote:
Some others you can add:
* tangent
* arcsine
* arccosine
* arctan
One more thing: arctan2. Can be implemented in terms of arctan, but
getting the sign right is inconvenient.
ception, or (arctan +infinity) being inexact 90.0, this
wouldn't work as-is.
+ (< (sqrt (expt (- value estimate) 2)) margin))
That's simply (< (abs (- value estimate)) margin), no? Or does Guile not
have 'abs'?
Best regards,
Maxime Devos
On 16/02/2025 12:19, Maxime Devos wrote:
I'm going to look into what's the case for 'tan'. For 'tan',
tan(pi/4)=tan(45°) is rational as well - I'm going to look into
whether the 'S' of 'tan' is only the integer multiples pi/4, or
whether
301 - 338 of 338 matches
Mail list logo
