Re: [Scilab-users] metanet for scilab 6.0.1 ?
Le 21/09/2018 à 14:27, philippe a écrit : Hi, is there any metanet user here ? I would like to install metanet in scilab 6.0.1, I tried to compile master branch from scilab-forge but it fails , see : https://antispam.utc.fr/proxy/1/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/forge.scilab.org/index.php/p/metanet/issues/1772/#preview metanet has been bugged from a long time but problems where solved recently in scilab 5.5.* so it's a shame to see it again bugged :-\ Best regards, Philippe ___ users mailing list users@lists.scilab.org https://antispam.utc.fr/proxy/1/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/lists.scilab.org/mailman/listinfo/users Hello Philippe, After looking at the final lines of the log, I have the feeling that the low-level gateway still uses the stack. This could explains why it does not compile with Scilab 6.0.1. S. -- Stéphane Mottelet Ingénieur de recherche EA 4297 Transformations Intégrées de la Matière Renouvelable Département Génie des Procédés Industriels Sorbonne Universités - Université de Technologie de Compiègne CS 60319, 60203 Compiègne cedex Tel : +33(0)344234688 http://www.utc.fr/~mottelet ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users
Re: [Scilab-users] metanet for scilab 6.0.1 ?
Le 21/09/2018 à 14:31, Stéphane Mottelet a écrit : > > After looking at the final lines of the log, I have the feeling that the > low-level gateway still uses the stack. This could explains why it does > not compile with Scilab 6.0.1. yes that's what I get, all fortran files in : metanet-master/sci_gateway/fortran/ have the line : include 'stack.h' but the toolbox call sci_malloc.h in many other files, which the new api for scilab6 (introduced since scilab 5.3?). So do you think that removing the line "include 'stack.h' " from all *.f files while solve the problem ? Best regards Philippe ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users
Re: [Scilab-users] metanet for scilab 6.0.1 ?
Le 21/09/2018 à 16:54, philippe a écrit : Le 21/09/2018 à 14:31, Stéphane Mottelet a écrit : After looking at the final lines of the log, I have the feeling that the low-level gateway still uses the stack. This could explains why it does not compile with Scilab 6.0.1. yes that's what I get, all fortran files in : metanet-master/sci_gateway/fortran/ have the line : include 'stack.h' but the toolbox call sci_malloc.h in many other files, which the new api for scilab6 (introduced since scilab 5.3?). So do you think that removing the line "include 'stack.h' " from all *.f files while solve the problem ? No. Fortran gateways disappeared with Scilab 6 S. Best regards Philippe ___ users mailing list users@lists.scilab.org https://antispam.utc.fr/proxy/1/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/lists.scilab.org/mailman/listinfo/users -- Stéphane Mottelet Ingénieur de recherche EA 4297 Transformations Intégrées de la Matière Renouvelable Département Génie des Procédés Industriels Sorbonne Universités - Université de Technologie de Compiègne CS 60319, 60203 Compiègne cedex Tel : +33(0)344234688 http://www.utc.fr/~mottelet ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users
Re: [Scilab-users] metanet for scilab 6.0.1 ?
Hi, I use the NARVAL package in scilab 6. I think it is basically the same as the old metanet stuff. It is not supported directly though, and I believe I had to do some hacking to get it working. Cheers, Arvid -- Arvid Rosén VP Research & Development m: +46 73 25 12347 p: + 46 13 21 16 23 softube.com On 2018-09-21, 16:54, "users on behalf of philippe" wrote: Le 21/09/2018 à 14:31, Stéphane Mottelet a écrit : > > After looking at the final lines of the log, I have the feeling that the > low-level gateway still uses the stack. This could explains why it does > not compile with Scilab 6.0.1. yes that's what I get, all fortran files in : metanet-master/sci_gateway/fortran/ have the line : include 'stack.h' but the toolbox call sci_malloc.h in many other files, which the new api for scilab6 (introduced since scilab 5.3?). So do you think that removing the line "include 'stack.h' " from all *.f files while solve the problem ? Best regards Philippe ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users
Re: [Scilab-users] metanet for scilab 6.0.1 ?
Le 21/09/2018 à 17:12, Stéphane Mottelet a écrit : Le 21/09/2018 à 16:54, philippe a écrit : Le 21/09/2018 à 14:31, Stéphane Mottelet a écrit : After looking at the final lines of the log, I have the feeling that the low-level gateway still uses the stack. This could explains why it does not compile with Scilab 6.0.1. yes that's what I get, all fortran files in : metanet-master/sci_gateway/fortran/ have the line : include 'stack.h' but the toolbox call sci_malloc.h in many other files, which the new api for scilab6 (introduced since scilab 5.3?). So do you think that removing the line "include 'stack.h' " from all *.f files while solve the problem ? No. Fortran gateways disappeared with Scilab 6 You mean Fortran gateways are no longer supported in 6.x? Is there any plan to reintroduce this feature? Antoine S. Best regards Philippe ___ users mailing list users@lists.scilab.org https://antispam.utc.fr/proxy/1/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/lists.scilab.org/mailman/listinfo/users -- +++ Antoine Monmayrant LAAS - CNRS 7 avenue du Colonel Roche BP 54200 31031 TOULOUSE Cedex 4 FRANCE Tel:+33 5 61 33 64 59 email : antoine.monmayr...@laas.fr permanent email : antoine.monmayr...@polytechnique.org +++ ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users
Re: [Scilab-users] metanet for scilab 6.0.1 ?
Hi, Le 24/09/2018 à 11:20, Arvid Rosén a écrit : > Hi, > > I use the NARVAL package in scilab 6. I think it is basically the same as the > old metanet stuff. It is not supported directly though, and I believe I had > to do some hacking to get it working. > > Yes I remember a presentation of NARVAL, but I would like to refactorize a lot of old scilab code created with metanet for scilab4. This is possible with few work with a new version of metanet, but not with NARVAL. Best regards, Philippe. ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users
Re: [Scilab-users] metanet for scilab 6.0.1 ?
Le 24/09/2018 à 13:41, amonm...@laas.fr a écrit : Le 21/09/2018 à 17:12, Stéphane Mottelet a écrit : Le 21/09/2018 à 16:54, philippe a écrit : Le 21/09/2018 à 14:31, Stéphane Mottelet a écrit : After looking at the final lines of the log, I have the feeling that the low-level gateway still uses the stack. This could explains why it does not compile with Scilab 6.0.1. yes that's what I get, all fortran files in : metanet-master/sci_gateway/fortran/ have the line : include 'stack.h' but the toolbox call sci_malloc.h in many other files, which the new api for scilab6 (introduced since scilab 5.3?). So do you think that removing the line "include 'stack.h' " from all *.f files while solve the problem ? No. Fortran gateways disappeared with Scilab 6 You mean Fortran gateways are no longer supported in 6.x? Is there any plan to reintroduce this feature? No. The "stack" is definitively gone. However, this is just a gateway problem. You can still use Fortran code, and actually there are a lot of Scilab functions which still rely on sources written in this language BUT gateways have to be coded with the "new Scilab API", either the "transitionnal" C API which appeared with Scilab 5.2.0, and preferably the new C++ API of Scilab 6. But I know that migrating API for a stuff like Metanet is not an easy task. As far as I am concerned, I have to rewrite the gateway for the FSQP optimization code of A. Tits and until now I didn't write one line of code... S. Antoine S. Best regards Philippe ___ users mailing list users@lists.scilab.org https://antispam.utc.fr/proxy/2/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/antispam.utc.fr/proxy/1/c3RlcGhhbmUubW90dGVsZXRAdXRjLmZy/lists.scilab.org/mailman/listinfo/users -- Stéphane Mottelet Ingénieur de recherche EA 4297 Transformations Intégrées de la Matière Renouvelable Département Génie des Procédés Industriels Sorbonne Universités - Université de Technologie de Compiègne CS 60319, 60203 Compiègne cedex Tel : +33(0)344234688 http://www.utc.fr/~mottelet ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users
Re: [Scilab-users] metanet for scilab 6.0.1 ?
Hi, Yeah, I know the pain. We had to do exactly that when metanet was dropped from being part of scilab. Actually, it wasn’t quite as bad as I first thought, but NARVAL changed a couple of times initially, which was annoying. However, the functionality (that we use) is exactly the same as metanat. Cheers, Arvid -- Arvid Rosén VP Research and Development Softube AB S:t Larsgatan 9D S-582 24 Linköping Sweden Phone: (+46) 13 21 16 23 Mobile: (+46) 73 25 12 34 7 Mail: ar...@softube.se URL: www.softube.se From: users on behalf of philippe Sent: Wednesday, September 26, 2018 9:08:42 AM To: users@lists.scilab.org Subject: Re: [Scilab-users] metanet for scilab 6.0.1 ? Hi, Le 24/09/2018 à 11:20, Arvid Rosén a écrit : > Hi, > > I use the NARVAL package in scilab 6. I think it is basically the same as the > old metanet stuff. It is not supported directly though, and I believe I had > to do some hacking to get it working. > > Yes I remember a presentation of NARVAL, but I would like to refactorize a lot of old scilab code created with metanet for scilab4. This is possible with few work with a new version of metanet, but not with NARVAL. Best regards, Philippe. ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users
Re: [Scilab-users] metanet for scilab 6.0.1 ?
Le 26/09/2018 à 17:11, Arvid Rosén a écrit : > Hi, > > Yeah, I know the pain. We had to do exactly that when metanet was > dropped from being part of scilab. Actually, it wasn’t quite as bad as I > first thought, but NARVAL changed a couple of times initially, which was > annoying. However, the functionality (that we use) is exactly the same > as metanat. there was another solution : using scicoslab to run old metanet code! Up to 2014 it worked for me . Please don't kill me ! Philippe ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users