1. Strange behavior in sequence handling
- Posted by Henri.Goffin at sbs.be
May 07, 2001
Hello everybody,
I'm a newcomer to Euphoria but I immediately found it interesting enough to
spend the few bucks that entitle you as a registered supporter. BTW are there
Euphoria T-shirts or something?
Now let's come to my first contribution (ahem) to this forum.
Can somebody try the following and tell me if he/she gets the same weird
behavior:
sequence L
with trace
trace(1)
-- OK
L = {'a', 'b', 'z'}
L = {L, 1, 2}
L = {L, 4, 5}
-- probably OK all along the line
-- Weird
L = {'a', 'b', 'z'}
L[1] = L -- Still OK
L[1][1] = L -- Wow! Is this LISP or Prolog or something?
trace(0)
It might be that I missed a subtlety of the language (I'm a newbie after all),
but I would rather suspect that Euphoria has trapped itself in some of its
optimizations in the internal representation of sequences, somewhere deep down in
its guts.
Anyway have a nice day, all of you!
Henri from Belgium
PS: I apologize for my pidgin English. My mother tongue is French. So let's all
talk Euphorian.
2. Re: Strange behavior in sequence handling
Henri.Goffin at sbs.be wrote:
> ...
sequence L
with trace
trace(1)
-- OK
L = {'a', 'b', 'z'}
L = {L, 1, 2}
L = {L, 4, 5}
-- probably OK all along the line> -- Weird
> L = {'a', 'b', 'z'}
> L[1] = L -- Still OK
> L[1][1] = L -- Wow! Is this LISP or Prolog or something?
> trace(0)
>
Henri,
I agree, seems you found a bug (which is really seldom now)! To much
recursion!
Have a nice day, Rolf
3. Re: Strange behavior in sequence handling
----- Original Message -----
From: <Henri.Goffin at sbs.be>
To: "EUforum" <EUforum at topica.com>
Subject: Strange behavior in sequence handling
Hi Henri,
>
> I'm a newcomer to Euphoria but I immediately found it interesting enough
to spend the few bucks that entitle you as a registered supporter. BTW are
there Euphoria T-shirts or something?
Now there's an idea. Maybe coffee mugs, and mouse pads too 
> Can somebody try the following and tell me if he/she gets the same weird
behavior:
>
> sequence L
> with trace
> trace(1)
> -- OK
> L = {'a', 'b', 'z'}
> L = {L, 1, 2}
> L = {L, 4, 5}
> -- probably OK all along the line
>
> -- Weird
> L = {'a', 'b', 'z'}
> L[1] = L -- Still OK
> L[1][1] = L -- Wow! Is this LISP or Prolog or something?
> trace(0)
I sure do. I'd say you have discovered a bug in Euphoria. Of course, it
could be just an undocumented feature.
------
Derek Parnell
Melbourne, Australia
"To finish a job quickly, go slower."
4. Re: Strange behavior in sequence handling
>I sure do. I'd say you have discovered a bug in Euphoria. Of course, it
>could be just an undocumented feature.
>
Yeah, BTW when are you documenting all the 'features' in win32lib?
Graeme
5. Re: Strange behavior in sequence handling
Hi Henri,
> L[1] = L -- Still OK
> L[1][1] = L -- Wow! Is this LISP or Prolog or something?
> trace(0)
Eu outputs CauseWay crash-file with message,
if ? L, or ? L[1], or ? L[1][1]
but ? length(L) or ? L[2] or ? L[3] are OK.
And what does LISP or Prolog do in this case ?
Just interesting to know, I am not
familiar with these languages.
Regards,
Igor Kachan
kinz at peterlink.ru
6. Re: Strange behavior in sequence handling
- Posted by aku at inbox.as
May 07, 2001
Very Error:
sequence L
L = {'a', 'b', 'z'}
L[1] = L
L[1][1] = L
?L
In EX.EXE:
{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{
{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{
{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{
{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{
{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{
CauseWay DOS Extender v3.45 Copyright 1992-99 Michael Devore.
All rights reserved.
Exception: 0E, Error code: 0006
EAX=FFD1AB60 EBX=00000000 ECX=00000042 EDX=C600077B ESI=00000197
EDI=0000007B EBP=8380EC4C ESP=8380EC1C EIP=837FCCCF EFL=00010286
CS=018F-7C822000 DS=0197-7C822000 ES=0197-7C822000
FS=0000-xxxxxxxx GS=019F-xxxxxxxx SS=0197-7C822000
CR0=00000000 CR2=00000000 CR3=00000000 TR=0000
Info flags=00008018
Writing CW.ERR file....
CauseWay error 09 : Unrecoverable exception. Program terminated.
----------------------------------------
In EXW.exe:
EXW caused an invalid page fault in
module EXW.EXE at 0177:0040cd63.
Registers:
EAX=00000001 CS=0177 EIP=0040cd63 EFLGS=00010287
EBX=800b0196 SS=017f ESP=0055fffc EBP=0056fc46
ECX=00580cf8 DS=017f ESI=00000043 FS=267f
EDX=0041fce8 ES=017f EDI=00580cb0 GS=0000
Bytes at CS:EIP:
e8 a8 fe ff ff 83 3d 90 3d 42 00 ff 0f 84 98 00
Stack dump:
>
H> I'm a newcomer to Euphoria but I immediately found it interesting enough to
spend the few bucks that entitle you as a
registered supporter. BTW are there Euphoria T-shirts or something?
H> Now let's come to my first contribution (ahem) to this forum.
H> Can somebody try the following and tell me if he/she gets the same weird
behavior:
H> sequence L
H> with trace
H> trace(1)
H> -- OK
H> L = {'a', 'b', 'z'}
H> L = {L, 1, 2}
H> L = {L, 4, 5}
H> -- probably OK all along the line
H> -- Weird
H> L = {'a', 'b', 'z'}
H> L[1] = L -- Still OK
H> L[1][1] = L -- Wow! Is this LISP or Prolog or something?
H> trace(0)
H> It might be that I missed a subtlety of the language (I'm a newbie after
all), but I would rather suspect that Euphoria
has trapped itself in some of its optimizations in the internal
H> representation of sequences, somewhere deep down in its guts.
H> Anyway have a nice day, all of you!
H> Henri from Belgium
H> PS: I apologize for my pidgin English. My mother tongue is French. So let's
all talk Euphorian.
7. Re: Strange behavior in sequence handling
On Mon, 07 May 2001, Igor Kachan wrote:
> And what does LISP or Prolog do in this case ?
>
> Just interesting to know, I am not
> familiar with these languages.
>
(list)
((uses)(lots)(of)
(((paranthesis))))
((to)(make)(code)(clearer))
I don't think this will execute, but one never knows....
Regards,
Irv
8. Re: Strange behavior in sequence handling
On Mon, 07 May 2001, aku at inbox.as wrote:
> sequence L
> L = {'a', 'b', 'z'}
> L[1] = L
> L[1][1] = L
> ?L
>
> In EX.EXE:
> {{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{
> {{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{
> {{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{
> {{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{
> {{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{
> CauseWay DOS Extender v3.45 Copyright 1992-99 Michael Devore.
For what it's worth, in Linux, this prints {{{{{ continually, and the memory
usage quickly climbs into the 100 meg range. At which point, I shut down the
program (call me chicken). Anyway, if you want to shut down a server, or
maybe stress test a system, this would be an excellent program to run.
Regards,
Irv
9. Re: Strange behavior in sequence handling
Henri Goffin writes:
> L[1][1] = L -- Wow! Is this LISP or Prolog or something?
It's a bug. Thanks for reporting it.
It affects the translator too.
I'll fix the interpreter for version 2.3.
The translator beta-2 is almost ready, but
if it's an easy fix I'll do it right away.
If not, I'll do it for the translator 2.2 official release.
Regards,
Rob Craig
Rapid Deployment Software
http://www.RapidEuphoria.com
10. Re: Strange behavior in sequence handling
----- Original Message -----
From: <graemeburke at CROSSWINDS.NET>
To: "EUforum" <EUforum at topica.com>
Subject: Re: Strange behavior in sequence handling
> Yeah, BTW when are you documenting all the 'features' in win32lib?
Between 2:37am and 5:54pm on Wednesday, May the 9th, God willing.
Nah, just kidding. Actually I'm expecting some help from people who ask me
these sort of questions. Are you offering to help by some chance? In the
meantime, its now 3:13am Tuesday morning and I've got a lot of work still to
do, so it might be awhile off yet before its completely documented.
------
Derek Parnell
Melbourne, Australia
"To finish a job quickly, go slower."
