>Acis-alliance@cs.columbia.edu>http://lists.cs.columbia.edu/mailman/listinfo/acis-alliance>>>------------------------------------------------>This email has been scanned for viruses and content by the Radan>Computational Webshield Appliances.>>>>>------------------------------------------------>Radan Computational Ltd.>>The information contained in this message and any files transmitted with >it are confidential and intended for the addressee(s) only.  If you have >received this message in error or there are any problems, please notify >the sender immediately.  The unauthorized use, disclosure, copying or >alteration of this message is strictly forbidden.  Note that any views or >opinions presented in this email are solely those of the author and do not >necessarily represent those of Radan Computational Ltd.  The recipient(s) >of this message should check it and any attached files for viruses: Radan >Computational will accept no liability for any damage caused by any virus >transmitted by this email.>>_______________________________________________>Acis-alliance mailing list>Acis-alliance@cs.columbia.edu>http://lists.cs.columbia.edu/mailman/listinfo/acis-allianceFrom alames at sandia.gov  Mon Jun 21 16:32:34 2004From: alames at sandia.gov (Ames, Arlo)Date: Mon Jun 21 16:33:15 2004Subject: [Acis-alliance] Problem with "api_restore_entity_list"Message-ID: <16F54D8E44BBEE4BBBDB5EFB3AEEB86D01C521DC@es06snlnt.sandia.gov>If you are making any direct interface (DI) calls, you must be sure to wrapthem in API_BEGIN/END.  If not, when you make a call to any api_function,such as api_restore_entity_list or api_save_entity_list, ACIS will roll backthe model to the last known state, negatively affecting the model.  It'sparticularly annoying when you're debugging something, and want to save acopy out (to read into some other program, or to look at off-line) and youget nothing in the output file.Also, did you do api_start_modeller, and start up all the requisitesubsystems?  If you don't you can see this kind of behavior when you startdoing things with ACIS.  Certain modeller operations don't cause negativeaffects, so it might not be your first call to ACIS where everything comescrashing down.At 10:18 21.6.2004, Christian Pietersen wrote:>Following on from my previous email.....>>I have been told that any direct api_ calls actually don't need >API_BEGIN/END as all of them do this locally. So I guess I can't really >help... Sorry.>>-----Original Message----->From: acis-alliance-bounces@cs.columbia.edu>[mailto:acis-alliance-bounces@cs.columbia.edu]On Behalf Of Christian >Pietersen>Sent: 21 June 2004 09:06>To: lody24; acis-alliance@cs.columbia.edu>Subject: RE: [Acis-alliance] Problem with "api_restore_entity_list">>>Khalid>>I have had this sort of problem before and the only thing I can suggest >is make sure that you have a bulletin board open, i.e. put >API_BEGIN/END around the api call.>>It might be a problem with the file. Can you read it into other >applications? Can you read it into scheme AIDE?>>Here is a simple code snippet. fp is a FILE* to the open file. As you >can see it's not very complex.>>Regards>Christian>>         API_BEGIN>>         ENTITY_LIST acisList;   // acis list of entities>>         outcome rc = api_restore_entity_list(fp, TRUE, acisList);>         if (!rc.ok())>                 {>                 DEBUG("read sat failed\n");>                 .....>                 .....>                 .....>                 }>>         API_END>>>-----Original Message----->From: acis-alliance-bounces@cs.columbia.edu>[mailto:acis-alliance-bounces@cs.columbia.edu]On Behalf Of lody24>Sent: 19 June 2004 15:52>To: acis-alliance@cs.columbia.edu>Subject: [Acis-alliance] Problem with "api_restore_entity_list">>>Dear all,>I've a problem concerning the function "api_restore_entity_list">it always gives me an error "access violation">so i wonder if anyone got this problem before and>solved it>also if u have a VC++ code that just uses this>function to open sat file it will be great>thanks a lot>>Khalid>>__________________________________________________>Do You Yahoo!?>Tired of spam?  Yahoo! Mail has the best spam protection around >http://mail.yahoo.com _______________________________________________>Acis-alliance mailing list>Acis-alliance@cs.columbia.edu>http://lists.cs.columbia.edu/mailman/listinfo/acis-alliance>>>------------------------------------------------>This email has been scanned for viruses and content by the Radan >Computational Webshield Appliances.>>>>>------------------------------------------------>Radan Computational Ltd.>>The information contained in this message and any files transmitted >with it are confidential and intended for the addressee(s) only.  If >you have received this message in error or there are any problems, >please notify the sender immediately.  The unauthorized use, >disclosure, copying or alteration of this message is strictly >forbidden.  Note that any views or opinions presented in this email are >solely those of the author and do not necessarily represent those of >Radan Computational Ltd.  The recipient(s) of this message should check >it and any attached files for viruses: Radan Computational will accept >no liability for any damage caused by any virus transmitted by this >email.>>_______________________________________________>Acis-alliance mailing list>Acis-alliance@cs.columbia.edu >http://lists.cs.columbia.edu/mailman/listinfo/acis-alliance>>>------------------------------------------------>This email has been scanned for viruses and content by the Radan >Computational Webshield Appliances.>>>>>------------------------------------------------>Radan Computational Ltd.>>The information contained in this message and any files transmitted >with>it are confidential and intended for the addressee(s) only.  If you have >received this message in error or there are any problems, please notify >the sender immediately.  The unauthorized use, disclosure, copying or >alteration of this message is strictly forbidden.  Note that any views or >opinions presented in this email are solely those of the author and do not >necessarily represent those of Radan Computational Ltd.  The recipient(s) >of this message should check it and any attached files for viruses: Radan >Computational will accept no liability for any damage caused by any virus >transmitted by this email.>>_______________________________________________>Acis-alliance mailing list>Acis-alliance@cs.columbia.edu >http://lists.cs.columbia.edu/mailman/listinfo/acis-alliance_______________________________________________Acis-alliance mailing listAcis-alliance@cs.columbia.eduhttp://lists.cs.columbia.edu/mailman/listinfo/acis-allianceFrom kyriazis15 at in.gr  Tue Jun 22 08:06:02 2004From: kyriazis15 at in.gr (kyriazis15@in.gr)Date: Tue Jun 22 08:28:49 2004Subject: [Acis-alliance] New User to ACIS R13Message-ID: <0b2f01c45851$49d72bd0$0205030a@internal.ramnet.gr>Could anyone please help me setup the project settingsin Visual C++ 6 for ACIS R13?I cannot exec even the simple examples.From davidkou at hkusua.hku.hk  Tue Jun 22 22:28:47 2004From: davidkou at hkusua.hku.hk (David)Date: Tue Jun 22 22:29:38 2004Subject: [Acis-alliance] Re: New User to ACIS R13Message-ID: <200406230228.i5N2SkpV027258@hkusua.hku.hk>Dear acis-alliance，I hope this helps:[1]Setup two system environment variable: "A3DT" and "ARCH", as stated in $A3DT\docs\r12_acis\ReadMe.htm.[2]Create a new project in VisualStudio, In project settings, set up the following items, a) Change the "Use run-time libray" to "Debug Multuthreaded DLL" or "Multuthreaded DLL", depending on your project settings for "Debug" or "Release" version;b)in the "Preprocessor definitions", append NT_DLL and _AFXDLL to the end;note that NT_DLL can be replaced with NT_NET_DLL etc. to match your ACIS version and VisualStudio version;c) in the "Additional Library path" , add "$A3DT/Lib/$ARCH"to the end;in the "Additional include path", add "$(A3DT)\include" to the end;Generally you are ready to use ACIS now.Best Regards,David2004-06-23>>Could anyone please help me setup the project settings>>in Visual C++ 6 for ACIS R13?>>I cannot exec even the simple examples.From kyriazis15 at in.gr  Thu Jun 24 04:43:53 2004From: kyriazis15 at in.gr (kyriazis15@in.gr)Date: Thu Jun 24 04:44:44 2004Subject: [Acis-alliance] Still unable to exec examplesMessage-ID: <15a001c459c7$60fca2f0$0a05030a@internal.ramnet.gr>I did everything you said, but still cannot run a single example.This is the error message, and it is repeated for several api functions:error LNK2001: unresolved external symbol "__declspec(dllimport) class outcome __cdecl api_stop_modeller(void)" (__imp_?api_stop_modeller@@YA?AVoutcome@@XZ)From akcrawford at gscad.com  Thu Jun 24 10:13:32 2004From: akcrawford at gscad.com (Crawford, Alan)Date: Thu Jun 24 10:14:22 2004Subject: [Acis-alliance] Splitting a FACE by an EDGEMessage-ID: <9BC689DBCB969642B931BF1D87B3E16E7A7464@postman.grad.com>I have two sheets that meet to form a T shape.  I perform a booleannon-regularised subtraction of these two sheets from a thin solid that theycompletely penetrate.  It should create three internal faces, but in somecases I only get two.  The horizontal face of the top of the T is not splitby the edge at the top of the vertical part of the T.  However, the edges ofthis top face are split by the vertices at the ends of this edge.When it fails, the surfaces involved are generally bspline surfaces (ofcourse) and I am assuming it is a tolerance issue.I am trying to find a way to imprint the required edge onto the top face,splitting it into the expected two faces.  I would prefer to avoid manualconstruction of the new topology through the direct interface, but I do notsee any other way.  Does anyone have any suggestions?If I have to use the direct interface, I think I will:- Construct an empty LOOP- Populate this loop with one set of the coedges from the FACE loop,bracketed by the start and end vertices of the edge.  At the same time,remove these edges from the original loop on the face.  Keep track of thenew COEDGE insertion points and directions.   - I know my face has only one loop (no holes)  - I know the single edge cuts all of the way across my face and I do nothave to chain edges to get across- Create the two new COEDGES, using the constructor that inserts the COEDGEinto a loop.- Check and correct the sequence of the COEDGEs around the edge (order ofpartner pointers)- Create the new FACE, using the constructor that takes an existing FACE* toclone and a LOOP*.- Convert the EDGE to a TEDGE (assuming that the original problem was aslight innaccuracy).- Remove and recreate the PCURVES on the body to create the missing PCURVE.I know there are some additional details here, but am I missing somethingmajor?  Or is there a better way anyone else has tried.Thank you for any help you can provide. Alan-------------- next part --------------A non-text attachment was scrubbed...Name: not availableType: application/ms-tnefSize: 2896 bytesDesc: not availableUrl : http://lists.cs.columbia.edu/pipermail/acis-alliance/attachments/20040624/e90084dd/attachment.binFrom alames at sandia.gov  Thu Jun 24 11:14:02 2004From: alames at sandia.gov (Ames, Arlo)Date: Thu Jun 24 11:14:54 2004Subject: [Acis-alliance] RE: Splitting a FACE by an EDGEMessage-ID: <16F54D8E44BBEE4BBBDB5EFB3AEEB86D01C521EC@es06snlnt.sandia.gov>Can you send a copy of the geometry so I can make sure I really understandwhat you're doing?I have sometimes found it advisable to create 3D objects and performordinary booleans, then extract the necessary face from it at the end.  OK,so it's potentially significantly slower, but it's a lot more robust to do3D operations with ACIS than 2D operations.  You can get back some of thespeed by making sure that some of the superfluous faces are well outside theregion where the boolean takes place (in an axis-aligned boxing sense).Face and edge creation is O(n); booleans can be O(n3logn).  If you ensurethat the faces you aren't interested in are by-and-large eliminated withboxing tests, and are, wherever possible, analytic surfaces so boxing isfast, you shouldn't see too much speed degradation.I have had very good luck with api_imprint with sheet bodies.  Make surethey intersect in a non-grazing sort of way.  Attaching attributes to thegeometries lets you label what the results of a given operation are,enabling surgery later if needed.  This kind of approach might be a loteasier to implement than your direct interface-based approach.I have never been satisfied with non-regularized booleans.  Theregularization step was invented in order to take hazy, fuzzy, crummyresults and make them be what you thought you ask for -- they compensate fora lot of nastiness associated with the approximations that show up in CADsystems.  Turn them off and you're asking for it.  Finally, I'm probably the most incorrigible customer Spatial has, because Ido a lot with the direct interface.  Spatial doesn't like it because directinterface calls make it difficult to do certain kinds of maintenance --changing the size of ANY object can break the windows linker.  This ends upmeaning that maintenance activities that involve changing the size ofobjects prevents their being applicable using service packs -- you have towait for the next code release.  But, the direct interface ALWAYS canproduce faster code.  It's just a lot more lines of code, and more to break.Your direct interface pseudocode doesn't look bad.  I'd seriously considerapi_imprint first, though, if possible.  Again, I need to understand yourgeometry.Arlo Amesalames@sandia.gov>  -----Original Message-----> From: 	Crawford, Alan [mailto:akcrawford@gscad.com> <mailto:akcrawford@gscad.com> ] > Sent:	Thursday, June 24, 2004 8:14 AM> To:	ACIS alliance (E-mail)> Subject:	Splitting a FACE by an EDGE> > I have two sheets that meet to form a T shape.  I perform a boolean> non-regularised subtraction of these two sheets from a thin solid that> they completely penetrate.  It should create three internal faces, but in> some cases I only get two.  The horizontal face of the top of the T is not> split by the edge at the top of the vertical part of the T.  However, the> edges of this top face are split by the vertices at the ends of this edge.>