ben lipkowitz a écrit : > > On Mon, 16 Mar 2009, Bryan Bishop wrote: > >> On Mon, Mar 16, 2009 at 12:57 AM, Thomas Paviot >> <thomas.pav...@free.fr> wrote: >>> Nice screenshot! Seems to be a really interesting work. >> >> Thank you. :-) >> >>> According to me, the InteractiveViewer is a kind of pythonOCC RAD >>> environment. If you add features like create a sphere, a line etc., >>> I think >>> it's not in the initial functional scope of this program anymore. >>> Your work >>> is rather a based pythonOCC based CAD demo app. Find a new name for >>> your >>> program and we can include it in the Tools directory, at the same >>> level than >>> CADViewer or InteractiveViewer. >> >> Okay. I'll put some more work into it- perhaps getting it to a usable >> state- and then see what name pops into my head. > > how about "HeeksCAD"? I had a look to HeeksCAD, program that I never tested before today. Bryan's screenshot is clearly the same as Heeks' one.
According to me, it makes absolutely no sense to clone an OpenCascade C++ based app using pythonOCC. Copy is always worse than original, and in the particular case of HeeksCAD, we cannot say that this original program is a way we must follow, or the template we should use. Regarding most of available OS CAD application (HeeksCAD, BRL-CAD, FreeCAD, QCad etc.), I notice that they all try to clone famous commercial 3D CAD programs. That is, in my opinion (what is yours?), a strategy that lead unavoidably to unsucess: it's impossible, for a small team (1?) of developers, to compete with worldwide developmnent teams of DS or PTC, who get a huge amount of business needs from their customers. I have then to precise my goals with pythonOCC development. During the last 10 years, untill 2006, I think that the innovation process/result of commercial CAD vendors was quite poor, mostly for commercial reasons. In a sense, we can make the parallel with the automotive industry: the explosion engine, appeared in the early 20th century, is still the same technology that is used in today's cars. A real innovation in 3D CAD world came in 2006, when SpaceClaim was founded and the SpaceClaim app commercialized. These guys showed that it's still possible to create new products and introduce a new technology that radically changes the way we use CAD softwares. It's still difficult for them to compete with well-established companies, but their product and their technology forced the other actors (think about DS, SW or UGS) to react. pythonOCC can then be seen as a development environment that aims to quickly prototype a new idea. All materials included in the pythonOCC distribution should then: - show the *benefit* of the use of python instead of C++ during the innovation process, - allow new users to quickly learn the API. Bryan's work is however interseting and has to be considered as a kind of "How to build a CAD app using pythonOCC". The requirements are then: - the program as to be as *small* as possible, - the code is clearly explained and widely commented, - focus on one or two (max) tools: for instance, create a sphere or a line, - show the interaction between the viewer and the modeler (i.e. the selection feature): for instance, create a sphere from a point that is selected from the viewer. This part is still missing from the availabe samples (this reminds me that Andy Haywood sent me an update of his InteractiveViewer that I never committed to the repository. Sorry Andy, I will do it). At last, if Bryan wants to use Heeks' material (icons for instance), the best is first to ask for his agreement, and to clearly explain that the objective is *not* to clone his work. And please Bryan, dont't name your app 'BishopCAD' :') Best Regards, Thomas _______________________________________________ Pythonocc-users mailing list Pythonocc-users@gna.org https://mail.gna.org/listinfo/pythonocc-users
