Autodesk and Onshape disagree about cloud technology and focus

I am watching for evolution of cloud CAD, such as OnShape
and A360, they have many important and useful technologies, which are required
to us, but on same time I see profound misunderstanding of reality by the CAD
bosses, as if they are living in they own cloud world.

Let me explain my position.

Using File-database to store drawings is a generic problem (disastrous)
for all CAD-systems, it is extremely uncomfortable, slow and unsafe storage
method. Most companies long time ago (about 10 years) withdraw this method in
favor of modern server database. Modern applications can use files as interchange
format, but not as the main repository. CAD and GIS are the last guardians of
this obsolete technology. This is because such large manufacturers are extremely
inert, they are afraid to make extra breath, just not to offend existing
customers.

This is what I need from cloud technologies:

1)
The project, contains set
of drawings, their components (for example, blocks) and documentations as main
working unit

2)
History of versions and
revisions with ability to rollback each minor changes to each date

3)
Teamwork

4)
Specific permissions for
specific user, even within same team

5)
Server database repository
with the ability to export individual elements of the project to file to
transfer it, for example, by e-mail

Moreover, it is not necessary to implement this on Cloud, it
may be a local server. Some companies (for example military etc.) have strong
security rules, and cannot use global network connections.

It is obvious requirements and, is some way, they are implemented
by OnShape and less by A360 (A360 is more similar to SVN, than to relational
database). However, with these useful things, we get many controversial
features, which are presents as major advance, rather than to solve problems,
that exists more the 20 years.

Why do I need the CAD in browser? Using html 5 and WebGL developers
get access to PC hardware resources, but each such layer makes it more
difficult to control resource usage and slow down performance. Native
application with correct optimization can be much faster on same machine. In
addition, solutions created using modern platform such as .Net or Java are much
easier to maintain in future, without transform the application into a hulking monster
of patches.

Do you want to perform some computations on Cloud? Why? Each
year computers become more powerful. Today for just $1000 you can buy a fairly
fast computer that provides comfortable work even with slow todays CAD. 5-10
years later computer for $1000 will easily provide real time photorealistic
rendering, even for CAD (if you, finally, start using modern technology and
algorithms).

Instead, you invest in the Cloud farms.

Let us try to estimate. Computer for $1000 provides
comfortable work with a single instance of the application. 100 of these
computers, with the total cost of $100 000, will provide a comfortable work for
100 instances of application. Now the question: how much will cost a server
that is capable to execute simultaneously 100 virtual sessions for same
application with same level of comfort as computer for $1000? How much will
cost administration of this server?

Your investments in cloud farms, most likely, will never
return, and most importantly, that people will still buy computers, which would
still be getting more and more powerful.

Do you want to make available desktop cad features for tablet/phone?
I would like to see a man who will seriously work on a 5-inch phone screen
(10-inch tablet screen). On my opinion, maximum that you need for mobile devices
– a good viewer.

Today we have the artifact applications, unchanging for decades,
new features attached to the old frame, and probably you are no longer able to
understand your overloaded patched code. Each year you present new AutoCAD,
Inventor etc., almost same as old, there is no evolution, problems are moving
from one version to another. Now we have multicore CPU, GPU with thousands of
cores, but you are selling application, based on core that has not changed since
the days of single-tasking DOS, and you loudly argue about your vision of the
CAD future.

Take a break from the Clouds. Kill file database format and
make a comfortable modern work with projects, versions and roles, and, most
importantly, teach, finally, your applications to use the hardware resources of
modern multi-core machines, make normal parametric, make 3D direct modelling
with constraints, and make the work with CAD more intuitive – this is your main
task.

With your financial and human resources, you can create a
truly modern system, but if your inertia and laziness will not allow you to
take the right decision – you are doomed. You need to come down from clouds to
the earth and start working.

Congrats! so long and so true… definitely I like it

The onshape is an exclusively cloud based architecture & is working smoothly, with astonishing performance and with embeded collaboration.
As the product is new, the success will depend on the minimum critical mass
of functionality they are going to put in version one of the software and the capacity of onshape to bring in quality /complementary partners.

The openess of the software to existing cad ( for import and translation) is really extensive.
I see some countries and large customers ( germany, japan etc…) will require on premises installations to secure their data and control it.

I believe that onshape will gain market share very fast despite a quite well established set of players. If not, as the technology is really innovative, some players will want to acquire the technology . I would prefer the first option as there is strong innovation capacity behind the creators that could risk to be inhibited with an acquisition.

As far as the market is concerned, the current 2d users ( millions) will find it very simple and easy to adopt. The current 3 D users will be interested as long as it does the work 🙂
Also the price entry point is like less that yearly maintenance of existing software …..