TheSwamp
Code Red => AutoLISP (Vanilla / Visual) => Topic started by: DMSS1 on March 03, 2009, 09:25:47 AM
-
I'm looking for Autolisp programming partners if anyone is interested...
(Mediator... I hope i'm in the right spot for this)
I am going to try to develop autolisp software to draw 3D structural & misc steel structures (of any kind) and auto-generate shop drawings from the 3D model(s).
I know this has been done before...
I want to do it differently.
My Vision:
This software will use Auto-Connection technology (the shape sizes and connections auto-update to the design loads).
It's going to be called "No Brain'r Software".
There will be a few different modules governing different types of structures.
Based on the module being used, users will design their product "on the fly".
Change either the 3D model, or any part of the shop drawings, and everything updates accordingly in real-time.
For example, I see this software designing a structural steel building.
As additional floors are added, the shape sizes, and connections automatically update themselves.
If a set of stairs are added, everything pertainate to the stairs is automatically done and anything that needs modifying to accommodate them is done automatically.
Anything can be changed instantly, on-the-fly.
I know this is a challenging project, but I feel it can be done, and fairly quick.
Everyone involved will have an equal share in any profits.
Guidelines and rules for partnership will have to be developed.
A full voting partnership will govern what the objectives are, how it's to be done.
I think groups, working on different projects, my be the way to achieve the long-term goal.
Weekly reports can be posted to the project website, as well as working demo's.
I already have a lot of the technology needed to pull this project off.
I will donate everything I already have to the project.
Together, I hope we can not only make a little money, but give the world a taste of the real power of Autolisp!
Heh! we don't have anything else to do with or spare time!
Thanks,
Dan
-
I dont know structural, but i may be interested. Let me read this post a few more times and think about it.
-
I hate to be a buzzkill, but my conscience dictates I speak up. Based on what you've provide in this loose spec I don't see "it" coming to fruition "fairly quick", though we really don't know what that means. You're basically talking about a reactor based, parametric solid modeler, and while it's certainly do-able (has been done before as you've noted) the notion that it you'll be able to have a robust architecture designed, and solid implementation completed "fairly quick" that will "draw 3D & misc. structures of any kind, auto generate shops drawings, auto connect shapes & connections, and automatically update all derivatives of same" in short order is somewhere between naive and extremely ambitious. Basing the entire thing on AutoLISP, while romantic, may not be the best strategy to fulfill the agenda.
I would suggest you get the entire application to work on paper before you write any code. I personally would be wary to join such a venture until I saw said application map, and witnessed several design iterations (still on paper) to assess the group's methodologies and the project's chances for success.
Having said all that I could be wrong, and it's just my opinion.
Regardless, I sincerely wish you and your partners good luck. There's not doubt it would be an interesting, challenging and educational venture.
/buzzkill
-
MP,
Thanks for your input!
I know you think it may sound impossible to you, but it is your opinion.
Autolisp may not be the easiest way to go... but there are many different ways it can be done with Autolisp.
Quickly, I would defined as "very fast", say 8-10 months.
In order for this project to work, there are a couple of minor things that have to happen.
1) 6-10 min. people working on this project.
2) Project planned to milestones for each week.
3) Dedication of each individual to their portion of the project.
4) Meetings, that govern how the project is being addressed and deployed.
5) Set global variables that are used throughout the whole project, to join each portion.
6) Sound directory and data planning.
Generally, I've found it is very hard to get everyone to agree on any given subject.
Even harder to dictate a project of this scope that everyone involved with is always in agreement on.
I would suggest a small "demo" project to get everyones "feet wet" with.
This would weed-out the ones that don't have the heart to stay on a bigger project.
I think the result of the smaller project, would have to be beneficial to the greater project, so the time is not wasted on something that the greater project needed.
I do have a few ideas about how to take on this task...
-
8-10 months?! ...that's fast! Real fast! May be too much for me.
The first part --''auto-generate shop drawings from the 3D model(s)'' (assuming you mean a 2d rep.)--alone is...well, that's...I dont know what that is.
If you produce a thousand lines a day, assuming that you have a direction to obtain (which is next to impossible most times)...*GACK!*i dont think its gonna work for me.
-
Hi DMSS1. I don't have time for a thorough discussion here but some quick thoughts / points ...
1. I didn't say it was impossible. Please go back and read what I said.
2. You're singing to the choir when it comes to "what can be done with LISP". You'd be hard pressed to find a bigger LISP advocate, but ...
3. Have you fully considered the impacts of a 100% LISP solution? Performance? Security? Upgrade-ability?
4. The application, if using 100% LISP will no doubt have to exploit many ActiveX based LISP functions. Have you explored what the performance / reliability / availability is for said functions under newer versions of AutoCAD, like 2010+?
5. This application will be GUI intensive, are you going to use DCL? OpenDCL? ObjectDCL? DLL based dialogs? What are the impacts of using the chosen platform with regards to expressiveness, performance, security ad infinitum?
6. As stated previously I think the 8 month time frame is ambitious for the lofty plans for the application. Have you factored in all facets of a typical development cycle, including testing, documentation?
7. Have you considered model size issues / limits? Co-existence issues?
8. Have you explored the various way the 3D objects can be modeled, e.g. meshes, faces, solids etc. and impacts to file sizes, dwg integrity / fragility, modify-ability etc?
9. Have you explored the myriad of ways the data will be stored, xdata, dictionaries, xdictionaries, external databases yada?
10. A demo, which can illuminate some of the above would be useful / intersting, and sounds like a very good idea.
11. I may sound like a foe but I'm not. I'm trying to share my expertise and experience, which spans many decades, multiple engineering disciplines and includes parametric modeling.
12. If it emerges that I wish to be involved, either directly or in some advisory capacity, I can provide a private and secure project workspace.
PS: Why will people want to invest their time in effort on your project? What is your expertise? What is your experience? What is your real name? Etc.
Sincerely,
Michael.
-
That's ok 7,
Didn't say it would be easy... (it's not a difficult as you think!)
Wouldn't have brought it up if I didn't already have formulated a good plan!
Michael,
The only thing that might hurt the project is the newer releases of autocad getting away from autolisp.
If they maintain autolisp and keep adding to the available tools, it is still worth considering.
Security covered!
Reliability? To be seen.
Size? It doesn't mater about the size! I doubt it will get bigger than 2-3 gigs.
I've got to go for now, more latter!
-
• The only thing that might hurt the project is the newer releases of autocad getting away from autolisp.
• If they maintain autolisp and keep adding to the available tools, it is still worth considering.
• Security covered!
• Reliability? To be seen.
Not a very compelling / convincing response to be honest.
• Size? It doesn't mater about the size! I doubt it will get bigger than 2-3 gigs.
Sorry, I think you're mistaken. I've seen civil / structural models easily 5 times larger than this, and it can become an issue with regards to performance, stability and integrity.
-
The maintenance alone would be a nightmare!
Im horrible at 3d stuff.
Marks got a grasp on the private and secure workspace thing too; i can set something up as well if need be.
That time frame is FAST, FAST, FAST in my opinion. I dont think i could keep up.
Docs are super important to me; i spend more time on my Docs then i do my code.
I'll keep an eye on this thread for sure, but maybe you can keep me in mind if you need someone to empty garbage cans or something.
-
I remember back when 2000i came out our reseller was telling us AutoDESK would NOT support AutoLISP in the next release, and for a fee, he could port all our AutoLISP over to VBA.
:lmao:
-
Dan,
I'm a pretty good structural steel detailer ...
I'm also a reasonably competent AutoLisp programmer.
I've also used several, and been involed with the development of a few detailing packages.
Personal opinion:
AutoLisp is a great language for customisation and minor development projects.
AutoLisp is not the language to use for a package of this nature and complexity.
You are committing to costs in the region of a minimum 1.5 million ( arguably double that) for a product that has a limited market and which will require a different set of specifications for each of the countries you could potentially target.
At least consider a language that has a class structure ...
///
That being said, I wish you good fortune.
-
In my opinion (I detail steel for a living, and began programming in AutoLISP in the late 1990s), you are wasting your time.
Reasons:
1. Autodesk is very unlikely to provide much, if any, added capability to AutoLISP
Please see this thread for specifics:
http://discussion.autodesk.com/forums/thread.jspa?threadID=715071&tstart=0
Please note particularly the comments of Mr. Tanzillo with respect to APIs for AutoCAD 3D solids.
2. More to the point, to do this properly, IMO, requires entity capabilities which are not presently provided by native AutoCAD entities.
As one example, extrude a polyline to represent a rolled section, then punch a hole, or several, in the extrusion. Guess what: you can no longer grip-stretch the solid.
Specialized software for structural steel/miscellaneous iron modeling & detailing within AutoCAD overcomes these rather severe limitations by creating custom entities designed specifially for the purpose. As of now, it is only possible to construct custom entities using ARX. It *may* become possible to do so at some time in the future using *.NET languages (uncertain), but I believe it is safe to state categorically that this will *never* be possible using AutoLISP.
3. Existing steel modeling/detailing packages for AutoCAD (I know of three, including one product owned by Autodesk) are available for < US$10,000.
4. In at least one instance, the 3rd party product is itself programmable via COM (which means, via AutoLISP).
See here for examples:
http://sites.google.com/site/mclisp/prosteelisp
AFAIK, it is not possible to do this with AutoLISP as the base coding language.
Herman Mayfarth
www.tktn.com
-
I wrote a big, long speech about the subject.
Then I realized, I'm wasting my time here.
Think what you want, complain as you will... about all the the reason why I can't be done!
I will do it myself... as always!
I would have liked to have had some help, it's the reason I came here, to ask for help.
Autolisp has a really bad reputation in the programming world!
In my opinion, the reason is, there are some many programs out there for free.
So good, some bad, but you can't complain if it's free.
Many write .lsp, but not many can write on this level...
Because of all the free stuff, people expect it to be free.
It's a lot of work, but then there's nothing else I would rather do...
Thanks again...
-
I wrote a big, long speech about the subject.
Post it.
I'm wasting my time here.
No you're not. You just gained the advice of many seasoned programmers, most of which are highly qualified to discuss LISP, and it's applicably to large scale software development in particular.
Think what you want, complain as you will... about all the the reason why I can't be done!
No one said it can't be done, however it was pointed out why LISP shouldn't be your first choice and that your estimates are very conservative, detrimentally so.
I will do it myself... as always!
Good luck.
I would have liked to have had some help, it's the reason I came here, to ask for help. Autolisp has a really bad reputation in the programming world!
In what context?
In my opinion, the reason is, there are some many programs out there for free.
No. The reason LISP may lack credibility in some contexts is not the language, nor free programs per se.
It's because any monkey can knock out some LISP and hang a "I are a programmer now" sign on his door. This is often the case with scripting languages because there's no cost to purchase compilers etc. and the apps written tend to be simplistic, not typically requiring one to possess a computer science degree (or equivalent experience).
Many write .lsp, but not many can write on this level ...
I would hazard a guess that most folks that responded in this thread could write at your level or better.
Because of all the free stuff, people expect it to be free.
LISP has no monopoly in this phenomenon.
Thanks again ...
My pleasure.
PS: You strike me as the kind of guy who would curse someone for preventing them from walking into traffic rather than thanking them.
(http://www.theswamp.org/screens/mp/wave.gif)
-
Many write .lsp, but not many can write on this level...
at what level ?
I've had a look at your site .. I was writing that sort of stuff 15 years ago.
-
I can sell some macro
carry
See site
http://sites.google.com/site/mclisp/prosteelisp
Sure hope you're joking.
You're going to sell code that provides automation for Bentley Prosteel objects, code that's largely a translation from the vba code examples they provide in their documentation? Furthermore, you're suggesting that our friend Dan base his Strut-o-matic application on an existing commercial application, in other words a consumer would have to purchase a license for Bentley Prosteel and Struct-o-matic, assuming Bentley would grant him permission?
Am I missing something here?
Anyone? Anyone? Bueller?
:D
-
indeed, I feel violated in a troll kind of way
-
joke
good !
'cause 'we' have a few old ones that don't get used any more. :)
http://www.theswamp.org/index.php?topic=17458.msg210350#msg210350
.. Speaking of VLisp with Prosteel .. it may be necessary when AutoDesk scrap VBa.
.. .. or perhaps Bentley will finally add a .net API.
-
.. Speaking of VLisp with Prosteel .. it may be necessary when AutoDesk scrap VBa.
.. .. or perhaps Bentley will finally add a .net API.
FYI -- there are indications Bentley is moving away from AutoCAD in their efforts to achieve CADD engine agnosticy. Read up on OpenPlant.
-
Just out of curiosity, how does lisp handle complex linear algebra such as matrices and vectors and the operations on them?
Even if it does handle it fine, how would it perform given that the proposed app is to be dynamic/automatic/parametric and would involve many 100's or 1000's of these calc's with a few simple mod's or additions?
The OP hasn't mentioned his level of understanding 3d geometry and its manipulation in an 'efficient' manner, this really is key, if that's even just a little shaky the project will eventually stumble or even fall.
Yes, on paper it's do-able but in reality it's a lot to ask of one mere mortal and a scripting language, best of luck.
-
I can write 3d a program kernel.
Reading and record in dxf will be used only.
Without use vla -*!!!
For example:
Select of sides, edges, vertex in 3dSolid (grread)
Creation 3dSolid (entmakex)
Moving (entmode)
Rotation (entmode)
Mirroring (entmode)
Offset (entmode)
Union (entmode)
Subtraction (entmode)
Intersection (entmode)
Transformation USC (entmode)
Moving of sides (entmode)
Moving of edges (entmode)
Moving of tops (entmode)
Change of radius's faces (entmode)
Change of initial region for Extruded 3dSolid (entmode)
Purge for 3dSolid (entmode)
Possibly, more many various functions. :-)
---------------------
PS. But I, cannot do similar work free of charge. This very big task...
-
So ... did this thing get to market and I missed its debut? (http://www.theswamp.org/screens/mp/oh.gif)