yap-users — General Information on Yap

Re: [Yap-users] Fatal YAP Error

[Vitor S. C. <vi...@bi...>]

Hi Oliver,

> 
> It occours after breaking the Program during runtime using Strg-C, then
> entering 
> trace Mode, using Strg-C again while tracing. When your at the commandline
> now and 
> press Strg-C again, yap collapses with the Message
>  [ Fatal YAP Error: likely bug in YAP, segmentation violation exiting.... ] 
> 

Yes, that is a tricky one. I am rewriting the exception handling code,
and hope to fix that and related bugs, but it is quite hard. Any
repeatable script would be much appreciated!

> It occours durring execution when i search a broad tree, without the chance
> of 
> finding what I search for. However, sometimes the search is working and just
> fails in 
> this case, this is the wanted result.
> 

I didn't quite understand. You mean the search is buggy in Yap?

> 
> Thanks for the Support, and thanks for the answers on my last request, it
> was helpful 
> and solved a lot.

Thanks for using Yap!

Cheers,

Vitor

[Yap-users] Fatal YAP Error

[Oliver W. <Ath...@gm...>]

Hello,

I am constantly getting the following Message during Work :

 [ Fatal YAP Error: likely bug in YAP, segmentation violation exiting.... ] 

I tried (and still try) to find a special point where the Error occures, but
it 
varies.
Im using Yap-4.3.20, the Problem existet already in Yap-4.3.19.


It occours after breaking the Program during runtime using Strg-C, then
entering 
trace Mode, using Strg-C again while tracing. When your at the commandline
now and 
press Strg-C again, yap collapses with the Message
 [ Fatal YAP Error: likely bug in YAP, segmentation violation exiting.... ] 

It occours durring execution when i search a broad tree, without the chance
of 
finding what I search for. However, sometimes the search is working and just
fails in 
this case, this is the wanted result.


Thanks for the Support, and thanks for the answers on my last request, it
was helpful 
and solved a lot.
	Oliver Werth


-- 
GMX - Die Kommunikationsplattform im Internet.
http://www.gmx.net

[Yap-users] Playing in a Winter Wonderland

[Vitor S. C. <vi...@bi...>]

Hi,

Yap-4.3.20 is out (at last)! I released sources and WIN32 binaries.

Yap-4.3.20 includes a full rewrite of the module system so that it
will be truly SICStus compatible (at long last). It also includes a
set of improvements on the meta-call. These changes were very
disruptive and it took me a while to get the system back to normal.

The debugger now includes trace. I've been using the debugger myself a
lot ;-), so it should be improving.

There is a new term rewrite library from Erik Alphonse. Thanks!

Last, there are a huge number of fixes. Thanks to Stasinos, Nicos,
Paulo, Henrik, Ines, David, Miguel, James, Ashwin, and Bart for the
many reports.

I believe it will soon be time for a truly stable release 4.4. Before
4.4 I'm planning on changing the catch/throw mechanism and hopefully
on merging the recent work on tabling. Hopefully, these changes will
be much less troublesome ;-).

Enjoy and please keep sending bug reports.

Cheers,

Vitor

Re: [Yap-users] Compatibility with SICStus Prolog

[Vitor S. C. <vi...@bi...>]

Hi Oliver,

> 
> My Main Question is, are there logical Differences between YAP 4.2 and
> SICStus, able 
> to force this outcome, and is there a common solution to the problem ?
> Is there a possibility to use "logical update semantics" in YAP ?
> 

As Paulo said, you should try using

yap_flag(language,sicstus).

This will set logical update semantics to be default. Notice that Yap
does not support everything SICStus supports.

SICStus compatibility has been steadily improving in the
development releases. You may also want to try Yap-4.3.19 and the
soon-to-released Yap-4.3.20.

And, of course, if you have problems you can always contact the list!

Cheers,

Vitor

Re: [Yap-users] Compatibility with SICStus Prolog

[Paulo M. <pm...@no...>]

On Monday, December 17, 2001, at 06:56 PM, Oliver Werth wrote:

> Hello,
>
> I have a quite complex Engine writen in SICStus Prolog, and try runnig=20=

> it
> now with
> YAP 4.2.
> ...
> My Main Question is, are there logical Differences between YAP 4.2 and
> SICStus, able
> to force this outcome, and is there a common solution to the problem ?
> Is there a possibility to use "logical update semantics" in YAP ?

Yes. There is a flag that you can set to choose between "logical update=20=

semantics" and "immediate update semantics". More details in the YAP=20
documentation.

Regards,

Paulo


-----------------------------------------------------------
Paulo Jorge Lopes de Moura
Dep. of Informatics                   Office 4.3  Ext. 3257
University of Beira Interior          Phone: +351 275319700
6201-001 Covilh=E3                      Fax:   +351 275319732
Portugal

mailto:pm...@no...
http://www.ci.uc.pt/logtalk/pmoura.html
-----------------------------------------------------------

[Yap-users] Compatibility with SICStus Prolog

[Oliver W. <Ath...@gm...>]

Hello,

I have a quite complex Engine writen in SICStus Prolog, and try runnig it
now with 
YAP 4.2.

In SICStus the Engine was working fine. 
I already changed the different predicates in the library to the YAP
predicates and 
fixed the small syntax differences. The Engine, including all its librarys,
and 
included Modules or Databases is compiling fine, and all of the simple
predicates do 
work. 

However, using the complex functions of the Engine it seems to hook up in an
infinite 
Loop. While in Sicstus the procedure took about 0.23 seconds, I run it for
over an 
hour without success. 
During tracing I wasn't yet able to find the reason for an infinite Loop,
but I keep 
on searching. I am aware of the "immediate update semantics" of YAP and I
not yet 
figured out if this is the problem.

My Main Question is, are there logical Differences between YAP 4.2 and
SICStus, able 
to force this outcome, and is there a common solution to the problem ?
Is there a possibility to use "logical update semantics" in YAP ?

Thanks, for bothering

Oliver Werth

-- 
GMX - Die Kommunikationsplattform im Internet.
http://www.gmx.net

[Yap-users] Module changes

[Vitor S. C. <vi...@bi...>]

Hi,

I've just commited a huge number of changes to the module sytem. The
goal is to address incompatibilities with the Quintus/SICStus in a
principled way, instead of just hacking around it.

The changes affect *all* built-ins that deal with predicates, that
is, a lot! Also, there are some changes to the CVS tree for
CLPQR. I'll work on testing and fixing the new bugs, please don't
despair if things are a bit iffy for a while.

Thanks for understanding,

Vitor

Re: [Yap-users] compiling projects

[Vitor S. C. <vi...@bi...>]

Hi Erick,

>=20
> I have both prolog code and external predicates written in C++.
>=20

ok.

> I would like to know if there is a way to produce one single executable
> file with yap 4.3.19. Otherwise, could someone tell me a convenient way
> to compile my project?
>=20

Yap is compiled as follows: first we create a libYap.a library with
the files in the C directory. Then we link this library with the
console/yap.c file. yap.c sets up the system and then calls the
top-level.

To change the top-level you likely want to hack over yap.c. yap.c
essentially initialises everything and then calls the predicate $live
over and over again. You can change your top-level by editing $live.

To extend Yap with new builtins one solution is to add files in the C
directory. In this case you will have functions of the form

/* returns either TRUE or FALSE */
static Int
p_whatever(void)
{
  /* get input args */
  Term t1 =3D Deref(ARG1);
  Term t2 =3D Deref(ARG2);
  /* check possible typing errors */
  if (IsVarTerm(t1)) {
   /* set up an error */
   Error(INSTANTIATION_ERROR, t1, "my message";
   /* fail */
   return(FALSE);
  }
  /* do something useful and place the result in term tf */
  tf =3D something_really_nice(t1, t2);
  /* we're done */
  return(unify(tf,ARG3));
}

at the end of the file you need to add your new predicates, which is
quite simple:

/* this is not static */
InitMyStuff()
{=20
  /* first argument is the name, second is arity, third is C-function,
last is flags (0 is safe) */
  InitCPred("whatever", 3, p_whatever, 0);
}

last, you need to call InitMyStuff from somewhere. A good place is
C/stdpreds.c in the function

InitCPreds();

> Thanks in advance for your help,
> Regards,
> Erick.
>=20

Hope this helps and feel free to ask more questions,

Regards,

V=EDtor

[Yap-users] compiling projects

[Erick A. <Eri...@lr...>]

Hello,

I have both prolog code and external predicates written in C++.

I would like to know if there is a way to produce one single executable
file with yap 4.3.19. Otherwise, could someone tell me a convenient way
to compile my project?

Thanks in advance for your help,
Regards,
Erick.

[Yap-users] Meta-call changes

[Vitor S. C. <vi...@bi...>]

Hi Folks,

I just updated the CVS version with the new faster, and much buggier,
meta-call. There were a huge number of changes throughout the system,
and debugging still goes on, but I think it's time to release the
system to the masses.

BTW, I'll be spending the next year in Wisconsin, where I hope to do a
lot of development on Yap ;-).

Cheers,

V=EDtor