Open CASCADE Conference

Pawel's picture

Hello developers,

just like in case of the OCCT forum I wasn't sure if this is the right category to start the discussion here but in the end one could regard a meeting a tool for collaborative work...

This is a cross-post from http://www.opencascade.org/org/forum/thread_23590/?forum=3

The idea is to organize a meeting involving many parties using / working with OCCT such as: the OCC Company, the open source community members, the OCC customers or even representatives of the competing CAD-kernels in order to:
- exchange our experiences gained while working with OCCT,
- support the open source development of the library,
- present our work and achievements and to look for synergies among the users,
- estimate the goals for further development, cooperation etc.

As the original discussion on LinkedIn (http://www.linkedin.com/groups?gid=1546137) got quite long I will not quote it here in order not to discourage you from reading. Just to sum up: The initial conference would very likely be organized by the OCC company and it is possible to hold such a meeting on a regular basis (yearly, for instance).

So would you be interested in attending the proposed event?

Pawel

dbc's picture

Open CASCADE Conference

Thanks to Pawel for his initiative and to those who support it.

We think that this first conference should be intended to the OCCT product development and the OCCT Open Source Community, that is, to those who contribute to OCCT or develop open source projects using or somehow involving OCCT.

In this context, I expect most of our customers and the representatives of CAD kernel market products not really interested by this event.

Nevertheless, we (OCC SAS) would be happy to organize this conference in France, providing that, as already mentioned:

  • we clarify and share the goals and the expectations of such an event,
  • enough people are interested in attending the meeting,
  • and, last but not least, we find the budget for organizing it.

Sorry for being prosaic, but as mentioned by Alexander, not only we (OCC) must find the budget for organizing the conference, but also each attendee should be, as Pawel, ready to cover their traveling and lodging expenses.

Then, for people to commit to attend the conference, I guess we must first materialize possible target topics.
For example:

  • Are there people ready to present their development project, including possible contributions to OCCT, the difficulties they encounter and the way we – OCC - can help them ?
  • Does somebody have a proposition for easing contributions in such a so complex library, both, in terms of process and software design ?
  • Are there people ready to take the responsibility of some developments proposed in our road-map ? I'm especially thinking about code refactoring which does not require any knowledge in the way geometry and topology are implemented in OCCT.
  • Are there people ready to strongly improve some components (example: data exchange) or provide alternatives to existing ones (example: visualization) ?

If this initiative is widely supported, I'm ready to collect your propositions for elaborating the agenda of the conference and organize the event.

Daniel

Pawel's picture

Hello Daniel, thanks for your

Hello Daniel,

thanks for your comments on the conference! I'm glad to see that OCC SAS supports this idea.

Some thoughts on the topic:

- I'm sure finding people willing to present their development including the current challenges connected to the use of OCCT won't be that difficult (let me volunteer at this point).

- A general suggestion on involving external programmers into the development process: I think it would be feasible to designate OCCT experts to guide external programmers and give hints. I'm thinking about the following scenario: Let's assume Bob is an OCCT expert responsible for the Data Exchange module. Alice is an OSC developer and reports a problem with STEP import via Mantis but does not really know where to start looking for a solution and asks for guidance. Bob monitors the Mantis bugtracker - to an extent not affecting his daily work but on regular basis (e.g. once a week) - and analyzes briefly the issue (let's say max 45 min.). After that he gives Alice some hints, for example: consider checking the class Class01, method Class01::Method01 but make sure you check the impact of your changes on Class02 etc. I think that approach would lead to people correcting the bugs and not only reporting them (and implementing workarounds for own needs that are not shared). I guess such a process might relatively quickly reveal who is interested in which module and whether she/he wants to take responsibility for some development proposed in the road-map (refactoring) or completely new functionalities (e.g. visualization, as you mentioned).

- One more thing: Since Alexander mentioned first quarter of 2013 as the possible date for the meeting, I think it would be good to fix the date as soon as possible (there's not much time left). Maybe you could suggest some calendar weeks that would come into consideration from your point of view.

Pawel

Pawel's picture

Open CASCADE Conference

Dear Daniel,

can you maybe deliver some insight on the planned conference? Has it been possible to make any decisions yet?

Thank you!

Best regards
Pawel

alexpena's picture

conference

Hi Pawel,

Thanks for this proposal, this is a great idea.

I might suggest at first to piggy back on an existing conference, this way we get the greatest effect without the effort of organizing a conference.

we could all submit a paper to a conference such as siggraph,ecaade,cadfutures,caadria, and meet during conference.

while we can all write individual papers, we could also collaborate on similar topics and submit joint papers.

we could then organize to meet for dinner/pub during the conference nights, which logistically is manageable.

this could serve as a test bed for the possibility of creating a pure opencascade conference.

if it does emerge, I am sure it will be a very interesting group.

Best,

Alex

Pawel's picture

conference

Hi Alex,

thanks for commenting and supporting the idea!

Can you suggest a conference where the OCCT paper could fit? If we had this we could see if there were any volunteers wishing to participate in the event / cooperate on the paper.

Some time ago I started a poll (http://www.linkedin.com/groupItem?view=&gid=1546137&type=member&item=143...) on linkedin in order to see if people were ready to participate. There was not that much feedback but the tendency was to meet somewhere in Europe - that would be the option I would also prefer myself. Would that be an option for you?

Pawel

alexpena's picture

Hi Pawel, Ill check the poll.

Hi Pawel,

Ill check the poll. There are many conferences where OCC can fit.

Although it may sound strange, Architects(as in brick and mortar) have been interested in the intersection between computation and fabrication, and how we bridge design and manufacturing. since they now have more access to styling features in design authoring packages (Nurbs, subdivision surfaces), but they are having a hard time translating those designs into constructible designs that can be "described" using either traditional documentation "projective geometry" or more Numerically Controlled Machining processes. this field is being termed "Architectural Geometry" or "Rationalization" and there are many opportunities for anyone with experience in OCC to contribute to this field.

I met Jelle Feringa for example during one of the ecaade conferences, and it was a great opportunity to catch-up.

Where it becomes more useful is if anyone here has Mathematical Knowledge so that they can encode their knowledge as mathematical notations so that it is not just a technological description but also a mixture of theory, for example much of the theory behind BREP has somehow been buried in the midst of Geometry Kernels, but in fact this knowledge could potentially solve many problems related to how we analyze geometry for fabrication, for example a really useful command like splitting a surface by using Gaussian curvature as the criteria could help architects reduce cost of double curved surfaces by translating higher curvature components into lower curvature ones, since cost of manufacturing normally decreases as you go from highly curved,double curved, single curved, and then flat, in that order, Many tools even advanced ones like(CATIA) dont have these features built-in, you have to develop plugins using CAA or VBA, so there is still keen interest in the industry to solve these "Problem-Solving" problems which have only recently emerged with the adoption of Nurbs packages like Rhino/CATIA/Alias or even with the appropriation of Animation Packages like Maya/3ds Max/XSI/Houdini/MudBox.

So I welcome anyone who is not in the AEC industry and has advanced knowledge of Geometry at the analytically level and BREP level to contribute to these conferences and Debates occurring around how architects are reinventing they way they translate designs into fabrication, Frank Gehry started the debate in the 1990s when he adopted CATIA for the fabrication of the Barcelona Fish and subsequently the Guggenheim in Bilbao , 23 years later the industry is finally catching up, but the same problems he was facing with translating his design intentions into fabrication intentions are re-emerging as other architects who did not have access to these tools now have access to low cost tools that enable them to design complex surfaces.

We are struggling with encoding material behavior in the models, for example OCC's fair-curves and Batten-curves are great examples of tools that could be used to simulate how wood bends in lofting, trick is new materials emerge very quickly and architects sometimes need to encode the behavior of these materials into the models, so the models have to be not only representations of form and shape but increasingly representations of the behavior in materials, so that they reduce the uncertainties of matching the design intention with how the assembly is going to perform once built on site(rework increases cost and litigation).

If you come from the Aerospace and Shipbuilding industries this is your bread and butter, and architects are Hungry for this kind of knowledge and expertise, since buildings are becoming ever more expressive and complex.

here is a list of conferences where you could contribute:

Industry Based(where people show software with possible applications and sometimes how their software was used on a complex project):

there is this new conference in Paris started by Prof. Helmut Pottman, http://aag12.architecturalgeometry.at/ Advances in Architectural Geometry, it happened last year in Paris.

also https://www.design-modelling-symposium.de/ digital design modelling in Berlin.

Research Based(Application is not necessarily the imperative):
these conferences have been meeting for more than 20 years around the topic of Computers in Architectural Research.

-ACADIA: http://acadia.org/conferences
-CAADRIA:http://www.caadria2013.org/
-CAAD FUTURES: http://cf2013.tongji-caup.org/
-ECAADE:http://www.ecaade.org/

Hope to see any of you on these conferences.

Best,

Alex

www.opencascade.com

Copyright 2011-2017
OPEN CASCADE SAS
Contact us