inefficiency of CELLS=1 - the segfault
>
> OK
>
> > however seems like a silent bug (more likely some optimistic avoidance
> of a
> > NULL check somewhere perhaps). What do you think?
>
> Yes, it is definitely a bug if YOU set the heap to a bad size. As long a
On Wed, Nov 29, 2023 at 08:08:33PM -0800, C K Kashyap wrote:
> Thanks Alex - half a dozen worked in my example :)
Great :)
> I completely get the idea of the inefficiency of CELLS=1 - the segfault
OK
> however seems like a silent bug (more likely some optimistic avoidance of a
>
Thanks Alex - half a dozen worked in my example :)
I completely get the idea of the inefficiency of CELLS=1 - the segfault
however seems like a silent bug (more likely some optimistic avoidance of a
NULL check somewhere perhaps). What do you think?
I say this because, even with CELLS=1, most
Hi Kashyap,
> I attempted to use 1 as the CELLS value in pico.h and immediately ran into
> segfault
Yes, this is not a good idea ;)
CELLS is the number of cells per heap
typedef struct heap {
cell cells[CELLS];
struct heap *next;
} heap;
and PicoLisp allocates as many heaps
Hi Alex,
I attempted to use 1 as the CELLS value in pico.h and immediately ran into
segfault at mark(ApplyArgs); in gc.c. I got around the segfault by simply
doing this -
if(ApplyArgs)mark(ApplyArgs);
if(ApplyBody)mark(ApplyBody);
After that, most things work but I ran into another
On Fri, Jul 15, 2022 at 08:39:03AM +0200, Alexander Burger wrote:
> yes, 'hex' does not verify the input (this was not intended).
>
> But you are right. For consistency with other conversion functions like
> 'format'
> or 'date' it should do so. I will fix.
Done and released.
> ☺/ A!ex
--
Hi frostbyte,
yes, 'hex' does not verify the input (this was not intended).
But you are right. For consistency with other conversion functions like 'format'
or 'date' it should do so. I will fix.
☺/ A!ex
--
UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
(prinl "version: " (version))
(for N (range 0 255)
(let C (char N)
(when (hex C)
(unless (member C (chop "0123456789abcdefABCDEF"))
(prinl C ":" (hex C)) ) ) ) )
(bye)
---
version: 22.6.30
22630
::3
;:4
<:5
=:6
:7
?:8
@:9
W:0
X:1
Y:2
Z:3
[:4
\:5
]:6
^:7
_:8
`:9
That makes sense now. Thanks for the clarification!
Kevin
On Tue, Feb 8, 2022 at 1:03 PM Alexander Burger wrote:
> Hi Kevin,
>
> > I may have encountered a possible bug, unless my understanding of @ is
> > incorrect. It appears the value of @ is not restored in a prog1 bod
Hi Kevin,
> I may have encountered a possible bug, unless my understanding of @ is
> incorrect. It appears the value of @ is not restored in a prog1 body when
> another flow/logic statement occurs as the car of an expression within the
This is right. All flow functions which s
2/8/22 9:15 AM, Kevin Ednalino wrote:
Hello,
I may have encountered a possible bug, unless my understanding of @ is
incorrect. It appears the value of @ is not restored in a prog1 body
when another flow/logic statement occurs as the car of an expression
within the said prog1 body. In both cases, t
Hello,
I may have encountered a possible bug, unless my understanding of @ is
incorrect. It appears the value of @ is not restored in a prog1 body when
another flow/logic statement occurs as the car of an expression within the
said prog1 body. In both cases, the correct result is returned from
Hi all,
as I don't use twitter any more, I think I should announce bug fixes here.
Just found and fixed a bug in reading non-ASCII characters in internal symbols
in pil21.
It was:
: (1 € 2)
-> (1 ⬠2)
: '€
-> â¬
: (char '€)
-> 226
Now (version 21.7.26) it gives:
I see, I had misunderstood the @'s meaning there. Thanks for the
clarification!
On Tue, May 18, 2021 at 12:03 AM Alexander Burger
wrote:
> On Mon, May 17, 2021 at 04:53:50PM -0500, polifemo wrote:
> > So, running:
> > (task -5000 (/ @ 5) (tty (println 'a)))
> >
> > does not save in *Run my
On Mon, May 17, 2021 at 04:53:50PM -0500, polifemo wrote:
> So, running:
> (task -5000 (/ @ 5) (tty (println 'a)))
>
> does not save in *Run my expected:
> (-5000 -1000 (tty (println 'a)))
>
> but instead it saves this:
> (-5000 NIL (tty (println 'a)))
>
> my guess is that `(eval (++ Prg) 1)`
I'm not sure if this is really a bug, since the behaviour I expected only
runs in a version of 'task I'm refactoring right now. But I'll try anyways.
So, running:
(task -5000 (/ @ 5) (tty (println 'a)))
does not save in *Run my expected:
(-5000 -1000 (tty (println 'a)))
but instead it saves
oh! I get it! thanks for clearing that up.
On Fri, May 14, 2021, 00:40 Alexander Burger wrote:
> On Fri, May 14, 2021 at 07:16:27AM +0200, Alexander Burger wrote:
> > In Vip, this is useful for example to set the correct namespace:
> >
> >$ vip --symbols llvm src/main.l +
> >
> > (note the
On Fri, May 14, 2021 at 07:16:27AM +0200, Alexander Burger wrote:
> In Vip, this is useful for example to set the correct namespace:
>
>$ vip --symbols llvm src/main.l +
>
> (note the double '-' as the function is '-symbols'), or to do other useful
> things:
>
>$ vip -"trace 'foo"
On Thu, May 13, 2021 at 05:40:39PM -0500, polifemo wrote:
> I realized the source of my previous loading problem:
>
> `((pre? "-" S) (load S))` this is the offending line: it checks if the
> argument string starts with a "-" to load it, but then it does not get rid
> of it, so it looks for the
this is the corrected function:
(bye
(if
(vip~vi ## [+ | +[]] [+[]] ..
(make
(while (opt)
(let S @
(cond
((pre? "+" S)
(link
(cons
I realized the source of my previous loading problem:
(bye
(if
(vip~vi
(make
(while (opt)
(let S @
(cond
((pre? "+" S)
(link
(cons
Ah cool. I’m not familiar with the android bit so wasn’t sure how the UI
works.
On Mon, Jun 22, 2020 at 01:48 Alexander Burger wrote:
> On Mon, Jun 22, 2020 at 05:27:37AM +, Shaughan Lavine wrote:
> > Does something have to be allowed for forms.l to work? I'm getting
> > '*Get [::1] not
On Mon, Jun 22, 2020 at 05:27:37AM +, Shaughan Lavine wrote:
> Does something have to be allowed for forms.l to work? I'm getting
> '*Get [::1] not allowed'.
That's right. Some global variables used in HTTP transactions need to be
allowed. But this is done automatically in the first few lines
Yep. Got it. That's about what I'd settled on---though without thinking about
the security implications.
Does something have to be allowed for forms.l to work? I'm getting
'*Get [::1] not allowed'.
And I'm still stumped about '"@doc/form/refD.html" -- Open error: Permission
denied' without
Hi all,
> OK, thanks, It isn't worth trying to modify the reader. I'll just add the
> decimal point while processing the input.
It is not a good idea to use the Lisp reader in a GUI. Imagine what a malicious
user can do with read macros! She could enter things like
`(call "sh" "-c" "rm -r *").
Original Message
>> On Jun 21, 2020, 12:16 PM, John Duncan < duncan.j...@gmail.com> wrote:
>>
>>> Not a bug, this is the design of the reader (symToNum). Can you make input
>>> include the decimal point?
>>>
>>> On Sun, Jun 21, 2020 a
rom ProtonMail mobile
>
>
>
> Original Message
> On Jun 21, 2020, 12:16 PM, John Duncan < duncan.j...@gmail.com> wrote:
>
>
> Not a bug, this is the design of the reader (symToNum). Can you make input
> include the decimal point?
>
> On Sun, Jun 21, 2
Original Message
On Jun 21, 2020, 12:16 PM, John Duncan wrote:
> Not a bug, this is the design of the reader (symToNum). Can you make input
> include the decimal point?
>
> On Sun, Jun 21, 2020 at 14:34 Shaughan Lavine
> <[shaug...@nometaphysics.org](mailto:shaugha
Not a bug, this is the design of the reader (symToNum). Can you make input
include the decimal point?
On Sun, Jun 21, 2020 at 14:34 Shaughan Lavine
wrote:
> Isn't this a bug?
> ---
> :(scl 2)
> :212
> 212
> :212.
> 21200
> -
> Shouldn't 212 be equal to 212.
Isn't this a bug?
---
:(scl 2)
:212
212
:212.
21200
-
Shouldn't 212 be equal to 212. ?
If not, how do I force a "." after an integer input by a user? Do I really have
to use a string input and look for a "."?
Sent from ProtonMail mobile
Hi Alfonso,
> I've noticed that the "rand" function doesn't work with negative arguments:
>
> Version 20.3.16
>
> : (rand -10 10)
>
> -> 1152921504606846966
>
> Version 19.12.28
>
> : (rand -10 10)
>
> -> -5
>
> I don't know in which ex
Hello,
I've noticed that the "rand" function doesn't work with negative arguments:
Version 20.3.16
: (rand -10 10)
-> 1152921504606846966
Version 19.12.28
: (rand -10 10)
-> -5
I don't know in which exact version the bug appeared.
Regards,
Alfonso V.
--
UNSUBSCRIBE:
On Wed, Jan 02, 2019 at 07:50:44PM +0900, Jean-Christophe Helary wrote:
> I'm currently working on documenting the wiki source syntax *by* using the
Cool!
> wiki source syntax. I spent most of yesterday afternoon doing that. There was
> this item that was not working and I found one minor glitch
> On Jan 2, 2019, at 19:07, Alexander Burger wrote:
>
> On Wed, Jan 02, 2019 at 06:48:09PM +0900, Jean-Christophe Helary wrote:
>> I am not seeing the change here.
>>
>> Can you confirm that you changed
>>
#main > ul li {
>>
>> to
>>
#main > ul > li {
>
> Yes, I copy/pasted from
On Wed, Jan 02, 2019 at 06:48:09PM +0900, Jean-Christophe Helary wrote:
> I am not seeing the change here.
>
> Can you confirm that you changed
>
> >> #main > ul li {
>
> to
>
> >> #main > ul > li {
Yes, I copy/pasted from your mail. A browser cache issue perhaps?
☺/ A!ex
--
UNSUBSCRIBE:
Alex,
I am not seeing the change here.
Can you confirm that you changed
>> #main > ul li {
to
>> #main > ul > li {
?
there should be a ">" between "ul" and "li".
Thank you.
Jean-Christophe
> On Jan 2, 2019, at 15:48, Alexander Burger wrote:
>
> On Tue, Jan 01, 2019 at 05:44:24PM
On Tue, Jan 01, 2019 at 05:44:24PM +0900, Jean-Christophe Helary wrote:
> It looks like the culprit is the wiki.css file (L86):
>
> #main > ul li {
>list-style-type: square;
> }
>
> That selector applies the "list-style-type: square;" to *all* elements
> under a element (including nested
+{
-{item}
-{
*{-{nested item}
-{nested item}}}
-{item}}
renders normally the ordered/unordered nesting,
but
*{
-{item}
-{
+{-{nested item}
-{nested item}}}
-{item}}
sets all the
Hi Thorsten,
> ,
> | (=0 'any) -> 0 | NIL
> | Returns 0 when any is a number with value zero. See also n0, lt0,
> | le0, ge0, ge0 and =1.
> `
>
> Should probably ge0 and gt0 here ...
Indeed! Thanks! Will be fixed in the next release :)
♪♫ Alex
--
UNSUBSCRIBE:
and opening it as a bug-with-patch on the Debian BTS, I figured it is
simpler and faster to post about it here - so if anyone else finds the time,
and if the fix really is that simple, it can help all users more quickly.
--
Rowan Thorpe
PGP fingerprint:
BB0A 0787 C0EE BDD8 7F97 3D30 49F2 13A5 265D CCBD
Hi all,
yesterday Mike Pechkin found a bug in 64-bit bignum multiplication.
It is a really stupid bug, returning the wrong register value when the
first argument is a bignum and the second argument is zero. Only in this
special combination. For example
(* 1448194831602515360 0
)
nevertheless try this gcc.l code, when I have the time.
Thanks to all,
Denis
From: jon.klei...@usit.uio.no
To: picolisp@software-lab.de
Subject: Re: PicoLisp on mac osx : bug
Date: Wed, 20 Nov 2013 08:49:10 +0100
Hi Denis,
It's not clear to me exactly
Good news and thanks for the information for the font-menu.
Denis
From: jon.klei...@usit.uio.no
To: picolisp@software-lab.de
Subject: Re: PicoLisp on mac osx : bug
Date: Thu, 21 Nov 2013 09:31:50 +0100
Hi Denis,
Thanks a lot! This works very fine
@software-lab.de
Subject: Re: PicoLisp on mac osx : bug
Date: Wed, 20 Nov 2013 08:49:10 +0100
Hi Denis,
It's not clear to me exactly what you did to get OpenGL working with
pil32 on Mac.
On 19-11-13 16:05 , Denis Fourt wrote:
Hello Alexander,
Yes, I installed pil32 and Jon Kleiser's library. I
: PicoLisp on mac osx : bug
Date: Wed, 20 Nov 2013 08:49:10 +0100
Hi Denis,
It's not clear to me exactly what you did to get OpenGL working with
pil32 on Mac.
On 19-11-13 16:05 , Denis Fourt wrote:
Hello Alexander,
Yes, I installed pil32 and Jon Kleiser's library. I built everything
a
proper fix (unless the real 64 bits for mac osx will be soon available).
Thanks,
Denis
From: a...@software-lab.de
To: picolisp@software-lab.de
Subject: Re: PicoLisp on mac osx : bug
Date: Tue, 19 Nov 2013 08:27:51 +0100
Hello Denis,
I have
Hi Denis,
1) The changes I made are surely not perfect, they are likely to cause
some troubles on the other OSs or other (older?) versions of mac os
x. So I have just explained them here, because they could be useful
for another user or to correctly fix the makefile (sorry I cannot help
, but it would be nice if someone else could issue a
proper fix (unless the real 64 bits for mac osx will be soon available).
Thanks,
Denis
From: a...@software-lab.de
To: picolisp@software-lab.de
Subject: Re: PicoLisp on mac osx : bug
Date: Tue, 19 Nov 2013 08:27:51
Hello,
I downloaded and installed Picolisp 32 bits on mac os 10.6. Everything went all
right. Then I tried to install the library for openGL and I had to modify the
makefile to build a 32 bits lib.
**DYNAMIC-LIB-FLAGS = -m32 -dynamiclib -undefined dynamic_lookup
-framework OpenGL
Hello,
Ok, second try
I downloaded and installed Picolisp 32 bits on mac os 10.6. Everything went all
right. Then I tried to install the library for openGL and I had to modify the
makefile to build a 32 bits lib.
**
DYNAMIC-LIB-FLAGS = -m32 -dynamiclib -undefined dynamic_lookup
Hello Denis,
I have no Mac myself, but let me ask for understanding the problem.
I downloaded and installed Picolisp 32 bits on mac os 10.6. Everything
went all right. Then I tried to install the library for openGL and I had
to modify the makefile to build a 32 bits lib.
I assume that you
Sorry if I am posting this to the wrong place (if so, please let me
know where is the right place). I am trying to report a bug in the
latest version of the picolisp-wiki, and can't find where to do that
(there is one page on the wiki which appears unused).
Anyway, in case this is the right place
Hi Rowan,
Sorry if I am posting this to the wrong place (if so, please let me
know where is the right place). I am trying to report a bug in the
latest version of the picolisp-wiki, and can't find where to do that
This list is a good place to report bugs. The wiki is more intended to
post
Hi Thorsten,
hmm... I just downloaded the new (??) version and it still crashes when
I type (env) directly after start-up ...
Sorry, the fix was not complete. :(
Though it worked for me yesterday,
$ bin/picolisp
: (env)
- NIL
I noticed that it did _not_ work in a debug break:
Alexander Burger a...@software-lab.de writes:
Hi Alex,
hmm... I just downloaded the new (??) version and it still crashes when
I type (env) directly after start-up ...
Sorry, the fix was not complete. :(
Though it worked for me yesterday,
$ bin/picolisp
: (env)
- NIL
Now
Hi Alex,
in the newest testing version on 64bit Archlinux PicoLisp terminates
when I call (env) without args. A bug?
,---
| : (env)
| Speicherzugriffsfehler
| $
`---
--
cheers,
Thorsten
--
UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject
Hi Thorsten,
in the newest testing version on 64bit Archlinux PicoLisp terminates
when I call (env) without args. A bug?
Yes, indeed! 'env' crashed when called at the top level (i.e. without
any existing bindings). Same for 'trail' and 'up'. The bug was
introduced recently by a check for debug
Hi Jon,
This is not a new bug. I'm using (if ( N 2) since 30 years for such
simple benchmark purposes, so I didn't care to change. I don't remember
if I made this up by myself, of if I found it somewhere.
I suggest you add a comment in misc/fibo.l that explains this, in case
You
'expr' converts a SUBR function (i.e. a function written in C or
assembly which evaluates all its arguments) to an EXPR (a Lisp level
function).
OK, so it's not built for FSUBR conversion.
Unfortunately, it doesn't work with FSUBR functions (C or asm functions
_not_ evaluating all
.
This is not a new bug. I'm using (if ( N 2) since 30 years for such
simple benchmark purposes, so I didn't care to change. I don't remember
if I made this up by myself, of if I found it somewhere.
In any case it is a matter of definition, i.e. where you start counting
the fibonacci numbers. If you
On Fri, Oct 26, 2012 at 05:42:14PM +0200, Alexander Burger wrote:
In any case it is a matter of definition, i.e. where you start counting
the fibonacci numbers. If you omit the initial zero, you get a proper
sequence starting with index '1' (as is the convention in PicoLisp):
: (for N 7
Hi!
I noticed the 'daemon function does a fine job distinguishing builtin
C functions from lisp functions. However, it redefines the function to
evaluate all arguments, rather than letting the function decide for
itself. Here is an example transcript:
: (setq a 3)
- 3
: (while (gt0 (dec 'a))
Hi Mansur,
Now I see another strange behaviour. My steps (CentOS 6.2 x64):
...
Then I load http://wiki.mysite:5000 in Firefox, log in
(admin:admin), navigate to Articles Essays in the left menu.
I see New document 'Articles', This document doesn't exist yet.
Seems ok. I press Create
Hi Mansur,
What I see is that '+Doc object loses it's 'nm property after any
try to edit. This is visible also in the maintenance gui: Name field
is always resets to empty state with Document hint, when I tab to
body field
You are right! I can reproduce it. It was a bug in pil64:
: (fold
Hi Mansur,
On Wed, Feb 29, 2012 at 11:33:52AM +0100, Alexander Burger wrote:
Please try once more, after downloading the new picoLisp.tgz.
Sorry, but the fix this morning was not completely correct. So I had to
upload yet another testing release.
Cheers,
- Alex
--
UNSUBSCRIBE:
Hi all,
idForm -- Undefined
?
While this was not really a bug in the Wiki, there _was_ another bug.
José Romero triggered it unintentionally yesterday, and I could easily
reproduce it. It crashed the Wiki when you tried to delete a Document
(via the Delete button), because of a bug
On Tue, 28 Feb 2012 13:01:31 +0100
Alexander Burger a...@software-lab.de wrote:
Hi Mansur,
I've just installed fresh PL release and wiki from this link:
http://software-lab.de/wiki.tgz
OK, great.
When I'm logged in and navigate to Role administration, I see this
error in the
regards,
Mansur
Hi all,
idForm -- Undefined
?
While this was not really a bug in the Wiki, there _was_ another bug.
José Romero triggered it unintentionally yesterday, and I could easily
reproduce it. It crashed the Wiki when you tried to delete a Document
(via the Delete button), because
Looks like the example JSON functions choke on empty objects. I made a
quick and dirty fix, but I'm not sure if I introduced extra bugs.
Diff follows:
##
--- json.orig.l 2012-01-27 07:34:25.0 -0300
+++ json.l
On Fri, 27 Jan 2012 12:55:29 +0100
Alexander Burger a...@software-lab.de wrote:
Hi José,
Looks like the example JSON functions choke on empty objects. I
made a
Oops. Right.
If it's OK I'll edit the task page, buggy code is ugly code!
Indeed. Please do so. Many thanks!
Done!
Hi José,
Hmmm, it still doesn't work quite well, it chokes on negative
Oops again ;-)
numbers, no idea how to fix that. :(
We can simply add the - character to the atom-specifiers:
(de readJson ()
(case (read -_)
({
...
Cheers,
- Alex
--
UNSUBSCRIBE:
On Fri, Jan 27, 2012 at 01:45:12PM +0100, Alexander Burger wrote:
We can simply add the - character to the atom-specifiers:
Sorry, forget that. These are only for symbols :(
Let me check more thoroughly.
Cheers,
- Alex
--
UNSUBSCRIBE: mailto:picolisp@software-lab.de?subject=Unsubscribe
On Fri, 27 Jan 2012 13:58:50 +0100
Alexander Burger a...@software-lab.de wrote:
On Fri, Jan 27, 2012 at 01:48:41PM +0100, Alexander Burger wrote:
Let me check more thoroughly.
How about that?
(de readJson ()
(case (read _)
...
(T
(let X @
Hello, I'm new to the list and Picolisp.
I tried this, and obtained a segfault:
: ('(1 2) 6)
Violación de segmento
But, if i try:
: ('(a b c) 6)
- NIL
I mostly understand why, but it was a surprise that Picolisp responded with
a segfault instead of an error.
Does this means that Picolisp
On Tue, 10 Nov 2009, Javier wrote:
Hello, I'm new to the list and Picolisp.
I tried this, and obtained a segfault:
: ('(1 2) 6)
Violación de segmento
But, if i try:
: ('(a b c) 6)
- NIL
I mostly understand why, but it was a surprise that Picolisp responded with a
segfault instead of an
Hi TC,
sorry, I messed something up. You are of course right!
: ('(1 2) 6)
Violación de segmento
...
I bet you know what happens here; you are trying to use a number as a
variable, which is illegal - crash (besides it does not make sense
anyway)
It is correct what you said. The
On Wed, Apr 22, 2009 at 10:07:54AM +0100, Tomas Hlavaty wrote:
Are there other special symbols in the 32bit PicoLisp that behave
like NIL?
Not that I can think of any. Except for the mentioned example where
symbols look like numbers.
Cheers,
- Alex
--
UNSUBSCRIBE:
Hi Alex,
The question is: What is the correct way?
I am writing my own sexp reader which is safe against malicious input,
i.e. it does not call eval (the escape chars like ` are not
understood).
: (intern 123)
- 123
: (+ @ 7)
123 -- Number expected
I thought I could use
Hi Alex,
is there a function to look up an interned symbol?
'intern' does not seem to be doing that:
: (intern NIL)
- NIL
: (pack NIL)
- NIL
: (pack (intern NIL))
- NIL
:
This behaviour seems a bit strange to me. I would expect (pack
(intern NIL)) to return NIL...
Thank you,
Tomas
--
Hi Tomas,
leaves single space in front of all cookie names except the first one.
...
(cons (htArg (clip (car L))) (htArg (cadr L))) )
Thanks! I didn't notice that, as I don't use cookies in applications,
and obiously never tested with more than one :-)
Cheers,
- Alex
Hi Alex,
in file @lib/http.l
(de _htHead ()
..
(setq *Cookies
(mapcar
'((L)
(setq L (split L =))
(cons (htArg (car L)) (htArg (cadr L))) )
(split @X ;) ) ) )
leaves single
Hi Henrik,
You're trying to access unassigned memory in your examples, C would crash if
you went out of bounds on an array too. The only viable options you have in
Thanks for the comment.
Let me add an explanation why this access to unassigned memory happens
in the first place.
In the call
82 matches
Mail list logo
