eclipse-clp-users — ECLiPSe Users List

Re: [eclipse-users] (no subject)

[Besnik S. <Bes...@im...>]

Hi,

Look at the bottom link:

http://www.crosscoreop.com/mailman/listinfo/eclipse-users

Cheers,
Besnik



Bertrand Tran a écrit :
> Hi,
>
> Is there any way to stop being in the mailing list?
> Thank you in advance.
>
> Cheers,
>
>
> Tran Bertrand
>
> _________________________________________________________________
> Avec Windows Live OneCare éliminez tous les virus de votre PC ! 
> http://www.windowslive.fr/liveonecare/default.asp
>
>
> _______________________________________________
> ECLiPSe-Users mailing list
> ECL...@cr...
> http://www.crosscoreop.com/mailman/listinfo/eclipse-users
>   

[eclipse-users] (no subject)

[Bertrand T. <ber...@ho...>]

Hi,

Is there any way to stop being in the mailing list?
Thank you in advance.

Cheers,


Tran Bertrand

_________________________________________________________________
Avec Windows Live OneCare éliminez tous les virus de votre PC ! 
http://www.windowslive.fr/liveonecare/default.asp

Re: [eclipse-users] NEWBIE: Eplex installation question

[Kish S. <ki...@ci...>]

brammert wrote:
> I do am using an intel mac, so that probably explains my problems, 
> although I do not know what is causing them. Eclipse itself does seem 
> to start normally but a simple command like halt. does not work either.
You could try compiling the source (which you can download from 
sourceforge) on your Intel Mac -- this would produce native code for 
your machine, but there may be issues with compiling on Intel rather 
PPC, as the porting to the Mac was done before the Intel Macs came out, 
and we have no access to the Intel Macs to compile the code ourselves.

Cheers,

Kish

Re: [eclipse-users] ECLipSe in finance and business

[Andrzej L. <le...@at...>]

I was trying to use Eclipse long ago, but in IC-PARC times there was no
commercial license for Eclipse. Therefore, I am using SICStus, and
taking into account the amount of code developed for SICStus it is
rather unlikely that I switch to Eclipse.

Anyway, I am using Prolog, strictly speaking as CLP(FD) engine for large
commercial applications from supply chain domain.

Cons:

1. One More Language in the enterprise is pain in the butt and is
associated with costs. Not only developers must be familiar with the
language, but implementation services, testing, packaging etc. Not all
of them must know Prolog fluently, but more or less they must know what
is this about,

2. Interface cost. Prolog is embedded into systems written in other
languages, and writing interface (especially bidirectional) is not a
piece of cake. It takes time to develop, in most cases it reduces to the
Common Denominator of Component Technology what means sockets, what
consequently is associated with not-so-great performance, 

3. Practically no programmers, at least loco USA

4. Poor development environments, strictly speaking no IDEs (forget
about EMACS)

Pros:

1. It is much easier and much faster to do constraint engines in Prolog
than in C++, and the difference is dramatic. C++ project was trashed
after we discovered that there is no way to deliver in time and within
budget (even using ... commercial library), and after spending 7 digit
number

2. Prologs come with excellent libraries of global constraints,
substantially outperforming these available in "classical" languages.
This way it is possible to avoid the Discovery of Wheel Syndrome,

Bottom line: in commercial environment the decision of using/not using
specific tools is at least 75% business motivated, the rest is
technical. In my case it required a bit of bravery, including the
possibility of being fired during the process. But the decision paid
back and is paying bask.

A.L.

P.S> By the way, I don't know about other application and company that
is using Prolog commercially...

> -----Original Message-----
> From: 
> eclipse-users-bounces+lewando=att...@cr... 
> [mailto:eclipse-users-bounces+lewando=attglobal.net@crosscoreo
> p.com] On Behalf Of ph...@fr...
> Sent: Wednesday, January 24, 2007 9:44 AM
> To: ECL...@cr...
> Subject: [eclipse-users] ECLipSe in finance and business
> 
> 
> 
> Hello again,
> 
> what would you say are the advantages of using ECLiPSe in 
> accounting, finance and business programs, over C++?
> 
> Why are there so many C++ jobs and so few prolog/ECLipSe jobs?
> 
> Best regards,
> 
> Philippe
> 
> _______________________________________________
> ECLiPSe-Users mailing list
> ECL...@cr... 
> http://www.crosscoreop.com/mailman/listinfo/ec> lipse-users
> 

[eclipse-users] eclipse for eclipse?

[Malcolm R. <mal...@cs...>]

Has anyone written a mode for the Eclipse prolog in the IBM eclipse IDE?

Malcolm

Many clever men like you have trusted to civilisation.
Many clever Babylonians, many clever Egyptians,
Many clever men at the end of Rome.
                               - G.K.Chesterton, The Napoleon of  
Notting Hill

Re: [eclipse-users] Conditional jump out of loops

[Kish S. <kis...@cr...>]

On Wednesday 24 January 2007 20:33, Kish Shen wrote:
> On Wednesday 24 January 2007 20:02, Kish Shen wrote:
> > 
> > fromto(_,End,End, end) do .....
> > 

Sorry again, I should have thought more about this. The above will always succeed, as end 
will always unify with End at the end of the first iteration.

The fromto should be in the form:

fromto(continue,_,LoopState, end) 

and for continuing the loop, unify LoopState with continue, and for termination, unify LoopState
with end, e.g.:

(fromto(continue, _,LoopState, end) do
   ....
  (continue(...) -> LoopState = continue ; LoopState = end)
)

where continue(...) is a test to see if you want to continue.

Cheers,

Kish

Re: [eclipse-users] Conditional jump out of loops

[Kish S. <kis...@cr...>]

On Wednesday 24 January 2007 20:02, Kish Shen wrote:
> 
> fromto(_,End,End, end) do .....
> 
> and in your loop body, when you want to terminate, you unify End with end, End = end.
> 
I should have been clearer about this: this will terminate the loop at the end of the current 
`iteration' of the loop, rather than immediately, i.e. all the goals in the loop body will be
executed.

Cheers,

Kish
 

Re: [eclipse-users] Conditional jump out of loops

[Kish S. <kis...@cr...>]

On Tuesday 23 January 2007 23:47, David Tian wrote:
> Hi,
> 
> I want to find out how to jump out of a loop in ECLiPSe and go to the upper
> level of the program. In java, there is 'break'; is there a predicate does the
> same thing in ECLiPse?
> 
> Many thanks,
> David
> 
> _______________________________________________
> ECLiPSe-Users mailing list
> ECL...@cr...
> http://www.crosscoreop.com/mailman/listinfo/eclipse-users
> 
Hi David,

I will assume by `loops' you mean the `logical loops' (do/2) in ECLiPSe. 

To finish such  loops with success, you must meet the termination condition specified in the
loop specifications (i.e. the fromto and friends), e.g.

for(I,1,5) do write(I)

will terminate when I is 5.

If you want to specify a terminating condition that you can set at run-time, you can use a fromto:

fromto(_,End,End, end) do .....

and in your loop body, when you want to terminate, you unify End with end, End = end.

Note that *all* conditions given in the loop specifications must be meet for the loop to end,
so for example, if you use for(I,1,5), you cannot finish early before I == 5 and successfully exit
the loop.

You can always exit a loop early by failing, but you would loose any variable bindings etc. you 
have made.

You can also exit a loop by throwing an exception, as Paulo suggest (block/3 and exit_block/1),
but you should only use this for catching errors, rather as a general way of leaving loops.

Cheers,

Kish

[eclipse-users] ECLipSe in finance and business

[<ph...@fr...>]

Hello again,

what would you say are the advantages of using ECLiPSe in accounting, finance
and business programs, over C++?

Why are there so many C++ jobs and so few prolog/ECLipSe jobs?

Best regards,

Philippe

Re: [eclipse-users] Conditional jump out of loops

[Andrzej L. <le...@at...>]

My solustion is that a) Prolog is not Java, b) In Java, program that
uses break inside a loop can be transformed into program that doesn't
use break, c) There are many languages that don't have break
functionality, and people somehow write programs using these languages,
including loops.

To the original poster: If you feel that you need "break" in the middle
of a "loop" when programming in Prolog, then most likely you are trying
to write Java program in Prolog

A.L.

> -----Original Message-----
> From: Paulo Moura [mailto:pm...@di...] 
> Sent: Wednesday, January 24, 2007 7:51 AM
> To: Andrzej Lewandowski
> Cc: Paulo Moura; 'David Tian'; 'ECLiPSe Users List'
> Subject: Re: [eclipse-users] Conditional jump out of loops
> 
> 
> 
> On 2007/01/24, at 13:38, Andrzej Lewandowski wrote:
> 
> > "Exception Is Not Goto" - this what is being said in 
> Programming 101 
> > course...
> >
> > A.L.
> 
> And given the amount of information in the original message your  
> solution is...? I would say that jumping out of a loop is an  
> exceptional event but that's just me.
> 
> Paulo
> 
> 
> >> -----Original Message-----
> >> From:
> >> eclipse-users-bounces+lewando=att...@cr...
> >> [mailto:eclipse-users-bounces+lewando=attglobal.net@crosscoreo
> >> p.com] On Behalf Of Paulo Moura
> >> Sent: Wednesday, January 24, 2007 5:31 AM
> >> To: David Tian
> >> Cc: ECLiPSe Users List
> >> Subject: Re: [eclipse-users] Conditional jump out of loops
> >>
> >>
> >>
> >> On 2007/01/23, at 23:47, David Tian wrote:
> >>
> >>> I want to find out how to jump out of a loop in ECLiPSe and go to 
> >>> the upper level of the program. In java, there is 
> 'break'; is there 
> >>> a predicate does the
> >>> same thing in ECLiPse?
> >>
> >> Maybe you can use catch/3 at the upper level and throw/1 
> inside your 
> >> loop?
> >>
> >> Cheers,
> >>
> >> Paulo
> >>
> >>
> >> -----------------------------------------------------------------
> >> Paulo Jorge Lopes de Moura
> >> Dep. of Computer Science, University of Beira Interior 6201-001 
> >> Covilhã, Portugal
> >>
> >> Office 4.3  Ext. 3257
> >> Phone: +351 275319891 Fax: +351 275319899
> >> Email: <mailto:pm...@di...>
> >>
> >> Home page: <http://www.di.ubi.pt/~pmoura>
> >> Research: <http://logtalk.org/>
> >> -----------------------------------------------------------------
> >>
> >>
> >>
> >>
> >> _______________________________________________
> >> ECLiPSe-Users mailing list
> >> ECL...@cr... 
> >> http://www.crosscoreop.com/mailman/listinfo/ec> lipse-users
> >>
> >
> >
> 
> -----------------------------------------------------------------
> Paulo Jorge Lopes de Moura
> Dep. of Computer Science, University of Beira Interior
> 6201-001 Covilhã, Portugal
> 
> Office 4.3  Ext. 3257
> Phone: +351 275319891 Fax: +351 275319899
> Email: <mailto:pm...@di...>
> 
> Home page: <http://www.di.ubi.pt/~pmoura>
> Research: <http://logtalk.org/>
> -----------------------------------------------------------------
> 
> 
> 
> 

Re: [eclipse-users] Conditional jump out of loops

[Paulo M. <pm...@di...>]

On 2007/01/24, at 13:38, Andrzej Lewandowski wrote:

> "Exception Is Not Goto" - this what is being said in Programming 101
> course...
>
> A.L.

And given the amount of information in the original message your  
solution is...? I would say that jumping out of a loop is an  
exceptional event but that's just me.

Paulo


>> -----Original Message-----
>> From:
>> eclipse-users-bounces+lewando=att...@cr...
>> [mailto:eclipse-users-bounces+lewando=attglobal.net@crosscoreo
>> p.com] On Behalf Of Paulo Moura
>> Sent: Wednesday, January 24, 2007 5:31 AM
>> To: David Tian
>> Cc: ECLiPSe Users List
>> Subject: Re: [eclipse-users] Conditional jump out of loops
>>
>>
>>
>> On 2007/01/23, at 23:47, David Tian wrote:
>>
>>> I want to find out how to jump out of a loop in ECLiPSe and go to
>>> the upper
>>> level of the program. In java, there is 'break'; is there a
>>> predicate does the
>>> same thing in ECLiPse?
>>
>> Maybe you can use catch/3 at the upper level and throw/1 inside your
>> loop?
>>
>> Cheers,
>>
>> Paulo
>>
>>
>> -----------------------------------------------------------------
>> Paulo Jorge Lopes de Moura
>> Dep. of Computer Science, University of Beira Interior
>> 6201-001 Covilhã, Portugal
>>
>> Office 4.3  Ext. 3257
>> Phone: +351 275319891 Fax: +351 275319899
>> Email: <mailto:pm...@di...>
>>
>> Home page: <http://www.di.ubi.pt/~pmoura>
>> Research: <http://logtalk.org/>
>> -----------------------------------------------------------------
>>
>>
>>
>>
>> _______________________________________________
>> ECLiPSe-Users mailing list
>> ECL...@cr...
>> http://www.crosscoreop.com/mailman/listinfo/ec> lipse-users
>>
>
>

-----------------------------------------------------------------
Paulo Jorge Lopes de Moura
Dep. of Computer Science, University of Beira Interior
6201-001 Covilhã, Portugal

Office 4.3  Ext. 3257
Phone: +351 275319891 Fax: +351 275319899
Email: <mailto:pm...@di...>

Home page: <http://www.di.ubi.pt/~pmoura>
Research: <http://logtalk.org/>
-----------------------------------------------------------------

Re: [eclipse-users] Conditional jump out of loops

[Andrzej L. <le...@at...>]

"Exception Is Not Goto" - this what is being said in Programming 101
course...

A.L.

> -----Original Message-----
> From: 
> eclipse-users-bounces+lewando=att...@cr... 
> [mailto:eclipse-users-bounces+lewando=attglobal.net@crosscoreo
> p.com] On Behalf Of Paulo Moura
> Sent: Wednesday, January 24, 2007 5:31 AM
> To: David Tian
> Cc: ECLiPSe Users List
> Subject: Re: [eclipse-users] Conditional jump out of loops
> 
> 
> 
> On 2007/01/23, at 23:47, David Tian wrote:
> 
> > I want to find out how to jump out of a loop in ECLiPSe and go to
> > the upper
> > level of the program. In java, there is 'break'; is there a  
> > predicate does the
> > same thing in ECLiPse?
> 
> Maybe you can use catch/3 at the upper level and throw/1 inside your  
> loop?
> 
> Cheers,
> 
> Paulo
> 
> 
> -----------------------------------------------------------------
> Paulo Jorge Lopes de Moura
> Dep. of Computer Science, University of Beira Interior
> 6201-001 Covilhã, Portugal
> 
> Office 4.3  Ext. 3257
> Phone: +351 275319891 Fax: +351 275319899
> Email: <mailto:pm...@di...>
> 
> Home page: <http://www.di.ubi.pt/~pmoura>
> Research: <http://logtalk.org/>
> -----------------------------------------------------------------
> 
> 
> 
> 
> _______________________________________________
> ECLiPSe-Users mailing list
> ECL...@cr... 
> http://www.crosscoreop.com/mailman/listinfo/ec> lipse-users
> 

Re: [eclipse-users] Conditional jump out of loops

[Paulo M. <pm...@di...>]

On 2007/01/23, at 23:47, David Tian wrote:

> I want to find out how to jump out of a loop in ECLiPSe and go to  
> the upper
> level of the program. In java, there is 'break'; is there a  
> predicate does the
> same thing in ECLiPse?

Maybe you can use catch/3 at the upper level and throw/1 inside your  
loop?

Cheers,

Paulo


-----------------------------------------------------------------
Paulo Jorge Lopes de Moura
Dep. of Computer Science, University of Beira Interior
6201-001 Covilhã, Portugal

Office 4.3  Ext. 3257
Phone: +351 275319891 Fax: +351 275319899
Email: <mailto:pm...@di...>

Home page: <http://www.di.ubi.pt/~pmoura>
Research: <http://logtalk.org/>
-----------------------------------------------------------------

[eclipse-users] Programming with Constraints: an Introduction

[<ph...@fr...>]

Hello,

what do you think of "Programming with Constraints: an Introduction"? Is it a
good beginner's book on CP?

Philippe

[eclipse-users] Conditional jump out of loops

[David T. <D.T...@po...>]

Hi,

I want to find out how to jump out of a loop in ECLiPSe and go to the upper
level of the program. In java, there is 'break'; is there a predicate does the
same thing in ECLiPse?

Many thanks,
David

Re: [eclipse-users] NEWBIE: Eplex installation question

[brammert <bo...@gm...>]

I do am using an intel mac, so that probably explains my problems, 
although I do not know what is causing them. Eclipse itself does seem to 
start normally but a simple command like halt. does not work either.

Kish Shen schreef:
> brammert wrote:
>>
>> On another note, I am also trying to install eclipse on my macbook 
>> using the mac binaries from sourceforge. The installation itslef runs 
>> smoothly, but as soon as I try to load the eplex library in this 
>> distribution it gives the following error
>>
>> [eclipse 1]: lib(eplex).
>> linearize.eco loaded traceable 0 bytes in 0.02 seconds
>> notify_ports.eco loaded traceable 0 bytes in 0.01 seconds
>> hash.eco   loaded traceable 0 bytes in 0.03 seconds
>> constraint_pools.eco loaded traceable 0 bytes in 0.01 seconds
>> loading OSI clpcbc ... dlcompat: Can not open 
>> "/Users/brammert/Eclipse/lib/ppc_macosx/seosiclpcbc.dylib"
>> shared library not found in load("ppc_macosx/seosiclpcbc.dylib")
>> undefined instruction in 'Emulate'
>>
> I am not sure why dlcompat is reporting it can't open the file -- I 
> assume seosiclpcbc.dylib is where ECLipSe think it is? Does it have 
> the right permissions for you to load it?
> The 'shared library not found...' message can be misleading. This 
> message is generated when there is any problem loading the file.
>
> Is your Macbook a PPC Mac, or an Intel Mac? The code is really only 
> for a PPC Mac (it is compiled on a Mac running OS 10.2), although we 
> have reports that it does run on Intel Macs, persumably because some 
> emulation mode, but we have no access to an Intel Mac and cannot test 
> this.
>
> I think I have successfully loaded eplex with the CLP/CBC solver 
> (which is what you are loading) on two PPC Macs: one running OS 10.2, 
> and another running OS 10.4.
>
>
> Cheers,
>
> Kish
>

Re: [eclipse-users] NEWBIE: Eplex installation question

[Kish S. <ki...@ci...>]

brammert wrote:
>
> On another note, I am also trying to install eclipse on my macbook 
> using the mac binaries from sourceforge. The installation itslef runs 
> smoothly, but as soon as I try to load the eplex library in this 
> distribution it gives the following error
>
> [eclipse 1]: lib(eplex).
> linearize.eco loaded traceable 0 bytes in 0.02 seconds
> notify_ports.eco loaded traceable 0 bytes in 0.01 seconds
> hash.eco   loaded traceable 0 bytes in 0.03 seconds
> constraint_pools.eco loaded traceable 0 bytes in 0.01 seconds
> loading OSI clpcbc ... dlcompat: Can not open 
> "/Users/brammert/Eclipse/lib/ppc_macosx/seosiclpcbc.dylib"
> shared library not found in load("ppc_macosx/seosiclpcbc.dylib")
> undefined instruction in 'Emulate'
>
I am not sure why dlcompat is reporting it can't open the file -- I 
assume seosiclpcbc.dylib is where ECLipSe think it is? Does it have the 
right permissions for you to load it?
The 'shared library not found...' message can be misleading. This 
message is generated when there is any problem loading the file.

Is your Macbook a PPC Mac, or an Intel Mac? The code is really only for 
a PPC Mac (it is compiled on a Mac running OS 10.2), although we have 
reports that it does run on Intel Macs, persumably because some 
emulation mode, but we have no access to an Intel Mac and cannot test this.

I think I have successfully loaded eplex with the CLP/CBC solver (which 
is what you are loading) on two PPC Macs: one running OS 10.2, and 
another running OS 10.4.


Cheers,

Kish

Re: [eclipse-users] NEWBIE: Eplex installation question

[brammert <bo...@gm...>]

Hi,

thanks for the response. The fact that the solver is included in the 
eplex package explains everything. What I meant with "same answer" is 
just that in solving a problem that used the eplex functionality, it did 
not seem to matter whether I installed Coin or not. But you're answer 
explains that.

On another note, I am also trying to install eclipse on my macbook using 
the mac binaries from sourceforge. The installation itslef runs 
smoothly, but as soon as I try to load the eplex library in this 
distribution it gives the following error

[eclipse 1]: lib(eplex).
linearize.eco loaded traceable 0 bytes in 0.02 seconds
notify_ports.eco loaded traceable 0 bytes in 0.01 seconds
hash.eco   loaded traceable 0 bytes in 0.03 seconds
constraint_pools.eco loaded traceable 0 bytes in 0.01 seconds
loading OSI clpcbc ... dlcompat: Can not open 
"/Users/brammert/Eclipse/lib/ppc_macosx/seosiclpcbc.dylib"
shared library not found in load("ppc_macosx/seosiclpcbc.dylib")
undefined instruction in 'Emulate'

*** compilation aborted
        before line 1 in the file 
/Users/brammert/Eclipse/lib/eplex_standalone.eco

*** compilation aborted
        before line 1 in the file /Users/brammert/Eclipse/lib/eplex.eco
Abort

Do you have any idea what the reason for this is? I have already set the 
DYLD_LIBRARY_PATH, so that should not be a problem.

Brammert

Kish Shen schreef:
> Hi,
>
> brammert wrote:
>> I just installed a COIN solver to be used with the EPLEX libraries. 
>> But how can I check if it actually works. Whether Coin is in stalled 
>> or not, running any example from the tutorial gives the same answer.
>>
>> Should I tell EPLEX where to actually find the solver (and if yes how 
>> should I tell it), or should it find the solver itself?
>>
>> _______________________________________________
>> ECLiPSe-Users mailing list
>> ECL...@cr...
>> http://www.crosscoreop.com/mailman/listinfo/eclipse-users
>>   
> You don't need to install any COIN solvers to use the eplex interface, 
> as they are included in the eplex package for the COIN/OSI that you 
> should have downloaded (and if you are using Windows, this should be 
> included in the installer that you run)
>
> If you have only download the eplex solver using COIN, you should not 
> need to do anything else to use it. What messages do you get when you 
> load eplex with the query lib(eplex)?
>
> Can you explain what is this `same answer' you get when you try the 
> examples? If the solver is not correctly loaded, you should not get 
> any answers at all.
>
> Cheers,
>
> Kish Shen
>

Re: [eclipse-users] (no subject)

[Andrzej L. <le...@at...>]

See CPlan here and related publications

http://ai.uwaterloo.ca/~vanbeek/software/software.html

A.L.

> -----Original Message-----
> From: 
> eclipse-users-bounces+lewando=att...@cr... 
> [mailto:eclipse-users-bounces+lewando=attglobal.net@crosscoreo
> p.com] On Behalf Of Malcolm Ryan
> Sent: Tuesday, January 23, 2007 1:11 AM
> To: ecl...@cr...
> Subject: [eclipse-users] (no subject)
> 
> 
> Hello,
> 
> I'm a AI planning researcher by trade, and I am just beginning to  
> looking into CSP as an approach to planning. Are there any example  
> systems which express planning problems as CSPs?
> 
> Malcolm
> 
>       "The act of defending any of the cardinal virtues has today all
>        the exhilaration of a vice."
>                                      - G.K.Chesterton A Defense of  
> Humility
> 
> 
> 
> 
> _______________________________________________
> ECLiPSe-Users mailing list
> ECL...@cr... 
> http://www.crosscoreop.com/mailman/listinfo/ec> lipse-users
> 

Re: [eclipse-users] (no subject)

[<K....@cw...>]

You might look at our papers

Constraint-Based Qualitative Simulation,
Krzysztof R. Apt, Sebastian Brand
http://arxiv.org/abs/cs.AI/0504024

Infinite Qualitative Simulations by Means of Constraint Programming,
Krzysztof R. Apt, Sebastian Brand,
http://arxiv.org/abs/cs.AI/0608017

and their references.

Krzysztof Apt

> Hello,
>
> I'm a AI planning researcher by trade, and I am just beginning to
> looking into CSP as an approach to planning. Are there any example
> systems which express planning problems as CSPs?
>
> Malcolm
>
>       "The act of defending any of the cardinal virtues has today all
>        the exhilaration of a vice."
>                                      - G.K.Chesterton A Defense of
> Humility
>
>
>
>
> _______________________________________________
> ECLiPSe-Users mailing list
> ECL...@cr...
> http://www.crosscoreop.com/mailman/listinfo/eclipse-users
>

[eclipse-users] (no subject)

[Malcolm R. <mal...@cs...>]

Hello,

I'm a AI planning researcher by trade, and I am just beginning to  
looking into CSP as an approach to planning. Are there any example  
systems which express planning problems as CSPs?

Malcolm

      "The act of defending any of the cardinal virtues has today all
       the exhilaration of a vice."
                                     - G.K.Chesterton A Defense of  
Humility

Re: [eclipse-users] NEWBIE: Eplex installation question

[Kish S. <ki...@ci...>]

Hi,

brammert wrote:
> I just installed a COIN solver to be used with the EPLEX libraries. But 
> how can I check if it actually works. Whether Coin is in stalled or not, 
> running any example from the tutorial gives the same answer.
>
> Should I tell EPLEX where to actually find the solver (and if yes how 
> should I tell it), or should it find the solver itself?
>
> _______________________________________________
> ECLiPSe-Users mailing list
> ECL...@cr...
> http://www.crosscoreop.com/mailman/listinfo/eclipse-users
>   
You don't need to install any COIN solvers to use the eplex interface, 
as they are included in the eplex package for the COIN/OSI that you 
should have downloaded (and if you are using Windows, this should be 
included in the installer that you run)

If you have only download the eplex solver using COIN, you should not 
need to do anything else to use it. What messages do you get when you 
load eplex with the query lib(eplex)?

Can you explain what is this `same answer' you get when you try the 
examples? If the solver is not correctly loaded, you should not get any 
answers at all.

Cheers,

Kish Shen

[eclipse-users] NEWBIE: Eplex installation question

[brammert <bo...@gm...>]

I just installed a COIN solver to be used with the EPLEX libraries. But 
how can I check if it actually works. Whether Coin is in stalled or not, 
running any example from the tutorial gives the same answer.

Should I tell EPLEX where to actually find the solver (and if yes how 
should I tell it), or should it find the solver itself?

Re: [eclipse-users] Can't run the java interface QuickTest

[Joachim S. <jsc...@ci...>]

Matthew Ford wrote:
> Hey,
> After compiling QuickTest.java, with the command javac -classpath 
> /Users/matt/eclipse/lib/eclipse.jar: QuickTest.java
> when I subsequently try and run the quick test I get a Null Pointer 
> Exception,


Hi Matthew,

this is because you are running the PowerPC version of ECLiPSe (ppc_macosx)
on an Intel Mac (presumably in some emulated mode).  This isn't really
tested, and probably not very efficient, but your earlier mail seems to
suggest that the basic functionality works.

I have fixed the Java-Interface problem for build 5.10#43.

The proper solution is of course to have a native MacOS/Intel
version.  This needs to be built by someone who has such a machine,
but we can help, of course.


-- Joachim

Re: [eclipse-users] Endless loop

[Marco G. <mar...@un...>]

Philippe de Rochambeau wrote:
> Hello,
> 
> eclipse seems to go into endless loop when you type
> 
> X = f(X), Y = f(Y), X = Y.

This is something "you should not write" according to Prolog semantics :-)

In theory, X=f(X) should simply fail, because a variable cannot unify 
with a term containing the variable itself. But checking at every 
unification that the resulting term does not contain cycles is quite 
expensive, so most Prolog systems do not perform the check (by default).

In ECLiPSe, you can switch the occur-check on by setting the flag 
occur_check:

[eclipse 1]: set_flag(occur_check,on).

Yes (0.00s cpu)
[eclipse 2]: X = f(X), Y = f(Y), X = Y.

No (0.00s cpu)

So, you should write X=f(X) only if you have the occur-check activated; 
otherwise you know that you may get inconsistent results.

Cheers,
Marco

-- 
Marco Gavanelli, Ph.D.
Computer Science Division
Dipartimento di Ingegneria
University of Ferrara
Via Saragat 1 - 44100 Ferrara (Italy)
Tel  +39-0532-97-4833
Fax  +39-0532-97-4870
http://www.ing.unife.it/docenti/MarcoGavanelli/