Comments on: Google Version of CAD Interoperability Story

By: olegshilovitsky

olegshilovitsky — Mon, 22 Jun 2009 00:28:52 +0000

Laurent, Thanks for your comments! I’d agree 100% of 3D interop is sort of dream. But if we can put practical approach and see different set of apps that involve 3D information – this is something in my view can work for long term. To define these levels and degrees of 3D interop is something customers can drive together with vendors. I think some companies have tried in the past, I’m not sure if it was successful or not. In the end, in each organization there is person who is 3D translation hero… Actually, I think, they can help… Best, Oleg/

By: Laurent

Laurent — Sun, 21 Jun 2009 19:30:36 +0000

Hi,

I do not think that 3D interoperability can work 100%.

There are two aspects of interoperability in CAD files. One aspect is the ability to translate all the entity types in a file, and this is pretty much the same as the problem with Office files. Provided one knows about their existence, if there is a placeholder in the target file, it can usually be translated.

The other one is 3D geometry. And here, there is a problem, because CAD geometry is not a closed set. The problem can be illustrated by a very simple fact: the intersection curve between 2 cubic patches has a degree equal to 324. Since nobody wants to deal with such equations, (and even worse, since they tend to propagate through geometric operations), each CAD company or kernel has designed its own approach to approximate these data and deal with them decently. This implies approximations and more or less algorithmic fine tuning. In the end, there is no perfect approach to deal with these approximated data.

I guess that this second point is the “vulnerable” part Oleg is referring to.

Now, this being said, PLM vendors do not help.

Translating feature history is yet another problem (because, unlike geometry, CAD vendors do not even comply to a common standard, like STEP for 3D geometry).

My 2 cents,

Laurent

By: olegshilovitsky

olegshilovitsky — Sun, 21 Jun 2009 10:28:10 +0000

yml, agree, 100%, but on the other side, I don’t see “translation services” were successful as business for long term in the past. What cloud will change for them? I think they are very vulnerable. Don’t you think so? Best, Oleg/

By: yml

yml — Sat, 20 Jun 2009 21:06:03 +0000

Oleg,
I think this is one of the major lock down that the industry is suffering from. It makes the choice of the PLM system very dependent to the most important CAD vendor in the particular industry.
I tend to believe that offering a translation web service in the “CLOUD” would be a competitive advantage to a new comer on that market.
Regards,
–yml

By: olegshilovitsky

olegshilovitsky — Fri, 19 Jun 2009 12:42:32 +0000

Roberto, I agree with you. I think, interoperability is expensive task, but can be achieved. In competitive word, somebody need to pay for development related to interop. Today, nobody is doing it and we don’t have it. Vendors will do it only in the case they will be interested to acquire competitor’s accounts. -Regards, Oleg

By: olegshilovitsky

olegshilovitsky — Fri, 19 Jun 2009 12:35:46 +0000

Robin, thanks for you comment! Actually, I think Google App improved with quality of Word rendering. For my cases it served 90% of needs. What is good in my view is continues improvement. – Best, Oleg

By: Robin Capper

Robin Capper — Fri, 19 Jun 2009 07:49:52 +0000

Except it looks different in Google and Office, from your screenshots?

The main reason most I know use Office is the sort of translation/layout problems that creep in when sharing documents.

By: Roberto Picco

Roberto Picco — Fri, 19 Jun 2009 05:12:42 +0000

Oleg, I just think that solving the issue of proprietary file formats would make cad migrations a lot easier. Many companies, mine included, wouldn’t even think of changing a cad system fearing their whole CAD history could became useless junk. If CAD files were readable and editable by any applications the competition among CADs would become very tough. My 2 cents.