Dezignstuff SolidWorks Blog

Microsoft Azure (link does not work in Opera) is, according to the  Wikipedia entry, “Microsoft infrastructure as a service”; Microsoft Windows on the cloud. Who cares. It turns out that Microsoft is learning a valuable lesson. They are learning that even Microsoft Azure customers who in turn provide cloud services to their customers don’t want Microsoft to host their cloud. So Microsoft is shipping clouds in a shipping container. Big boxy white cloud-colored shipping containers. Full of servers and software to serve up the cloud. The shipping container full of goodies is referred to in this Cnet article as an “appliance”:

As for the Azure appliance, Microsoft says it is too soon to offer many details, but said the option should appeal to businesses and partners that want the benefits of Azure but need physical control of their data for regulatory or other reasons.

It’s almost unimaginable, but even Microsoft customers want physical control of their own hardware, software and data. Doesn’t that sound familiar. I don’t see a way around this for SolidWorks. If they are going to offer a cloud version, they are going to have to offer some sort of an option for either local installs or local clouds. The polls in the previous post shows that there is little support for paying significantly more than customers do now for SWcloud.

So will SolidWorks one day be delivered as a bank of servers in a small shipping container complete with software, where the only input is power and the output is CAD data? The “good ol’ days” of the 1970s computing where the CAD software vendors were also the hardware vendors, and a CAD station cost far more than it does today are on their way back. Apple is trying to return us to that situation – controlling and selling the hardware and the software, and most importantly the prices.

For those who think that resisting the cloud is something “old people” do (and who seem to be implying that I am “old”), it seems to me that “old people” would be the ones wanting to return to the cloud, to relive the glory days, where CAD salesmen lived high on the hog. The idea of the same company controlling the hardware and the software (turnkey CAD vendors like Computervision, Intergraph, Applicon) eventually went extinct. Companies like AutoCAD started specializing in CAD software, and drove the price down. SolidWorks used component technology and drove it down even further. In the 1970s, a CAD station with hardware and software would cost $500,000 or more. Today you get far more power for under $10,000.

Cell phone companies are pulling this same old trick on consumers, and are getting away with it. They control the hardware, the connection service, and the apps, and you are a captive audience once you buy in. CAD vendors are making moves to start cashing in on the same idea. Don’t let them get away with it.

In all of this frenzy about how much of a mistake it would be for SolidWorks to push their tool entirely to the cloud, we might be overlooking situations in which the cloud might be useful. I mentioned some of these early on, but those ideas didn’t get much attention, and certainly no encouragement from SW direct, so maybe I’ll float them again here.

Local Cloud

Let’s just say that as a scaled down proof of concept, SolidWorks decided to make the first release of the cloud something you could install on your own local network.  In the first several releases of SolidWorks, there was a Client/Server Installation. Some people have questioned my recollection, since I have a notoriously bad memory, but that’s because I remember all sorts of useless crap like this. Anyway, check Ricky Jordan’s site, his page on What’s New guides, the 1998 What’s new, on page 1-3:

So. This is something we used to be able to do. It may have been a hangover from techniques on older mainframe computer systems  that shared a processor, and used terminals for display and graphics computing. Bringing back old technology seems to be all the rage, so this would fit right in. I can say that the technique worked, even back then, although there was often a bit of a lag in the display. It worked EXACTLY like a local cloud would work, except that in 2010, a local cloud would also allow distributed computing (running the calculations on more than one computer).

 Loading ...

Distributed computing

Folks in the movie CG industry, as well as SETI (search for extra terrestrial intelligence)  are involved in distributed computing. It doesn’t make sense for all SolidWorks users, but it would for some. Big renderings, rendered animations, big FEA and other types of simulation would be areas that would benefit from this. I’m not talking about computers around the internet, just using spare compute power of machines around your LAN or WAN.

 Loading ...

Serving your own web-based cloud

For some companies, web-based clouds really are a great solution to certain problems. But maybe they still want to control their own data. So they serve up their own cloud from their own machines connected to the web. I’m not talking access to your SW files across the internet, I’m talking about access to your SolidWorks software across the internet. Even currently available Workgroup PDM has the ability to serve files across the internet. You can access the Solidworks Workgroup PDM demo vault at this link: http://pdmdemo.solidworks.com/pdmweb/ (guesta is both username and password).

 Loading ...

Third party based cloud

Ok, just for comparison to the other ideas, I’m interested to see how many folks are still interested in the third party cloud that hosts both your data and your software. There are just too many bad scenarios to add to this one, such as you can’t retrieve your data, your data is encrypted in a database and can only be translated out as a Parasolid or STEP, you can never even delete your own data. If you stop paying for storage, there is no telling what happens to the data…

 Loading ...

People who are trying to make money off of “the next big thing” usually aren’t waiting around for “the next big thing” to happen. They are usually trying to make it happen. And people, especially people hungry for money, are prone to an excess of enthusiasm and a lack of real information from time to time. Don’t go to the grocery store when you’re hungry, and don’t go to a strip club if you’re lonely, don’t take the MENSA test if you’re insecure about your intellectual capabilities. Don’t try to develop a new product if you’re desperate for money.

So how do those little nuggets of wisdom affect you and me, CAD users? At this point in time, they affect us through the cloud. Here’s why, from my point of view.

First analog mobile phone call in 1973 (reenactment 2007)

Life imitates art - Get Smart shoe phone in 1965

The story starts with a cell phone the size of a shoe box. Not women’s shoes, mind you, big manly shoes. Cool thing, but expensive, and not quite portable as a personal device. Primarily owned by traveling businessmen, and carried and powered in the car. Then it progresses to the PDA – maybe a Palm device. Just a portable electronic scheduler and contact keeper. Convenient, expensive, but not connected. Evolve to the Blackberry, aka “Crackberry”. Still primarily in the business applications. This time it is connected to email. Users get addicted to the constant connection. Graduate to a smartphone where all of the business applications of the advanced PDAs are combined with the ubiquity of consumer cell phone.

And a step further to the iPad which uses smartphone software/hardware technology to bridge the gap between portable smart devices that are primarily phones and portable web-connected devices that are primarily computers. With the iPad, we now have portability and data usage in a single device that is not primarily a phone, and it is big enough to actually make data usable.

What’s the difference between an iPad and a traditional computer? To me, its the operating system and how the software is installed. A “real” computer can live without a DVD drive or a rotating hard drive. External USB drives and flash memory can serve those purposes. But on Apple, Windows, Linux, Unix, OS2, DOS, you can install software from sources unaffiliated with the OS writer. With phone-based software, you have to buy each “app” for each new device directly from the company store. I recognize that this is more of a captive business tactic than a technical limitation. But then that’s where this whole thing is headed.

Maybe the difference between a phone and a computer is the use of the word “application” which sounds very technical, and the word “app” which appeals more to the 140 character attention-limited crowd. You can get apps for the phone OSs to do most of the little tasks that traditional computer operating systems do, like screen capture, audio capture, text editing, etc.

Another question that comes to mind here is when do the business-minded tools such as the Blackberry evolve into the entertainment-minded toys such as the iPad? The transition is subtle, yet I think important. SolidWorks itself seems to struggle with the same question. They used to benchmark against other CAD applications, now they benchmark against games. Computers in general have devolved from exclusively high-dollar tools to ubiquitous toys for pre-adolescents.

Do you see how all of this relates to the cloud? All of these mobile devices require data to consume. They don’t have as much internal storage as a “real” computer, so the data has to be delivered from somewhere else. If you are someone hungry to make money, you can make sure the devices require external data, because you can control external data. Anyone who fights copyright battles with folks who do not recognize “rights of ownership” for any kind of data is the big winner in this scenario. Movie companies can control access to their data again. In fact, anyone at all with data on the web (cloud) gains from these mobile devices that can access anything anywhere, especially as the form factor (screen size) gets large enough to actually be meaningful.

The one thing this does not address is content creation. Most mobile users are content consumers, they don’t really create anything. Well, they create images, short movie snippets and text. These are raw materials for real content, not real content in themselves. People who create real content are CAD designers. Writers who use text and images to create a finished piece of writing. There are several other kinds of real content, but these are the kinds that matter to me since those are the kinds of content I create.

First of all, mobile devices are not appropriate form factor for real content creation. I could envision using an iPad to do writing incorporating images, but I don’t think I could envision real CAD work on an iPad. The trend recently is for CAD monitors to get absolutely huge – 30″ or more – not small. CAD laptops are uncomfortably large – 17″ monsters that barely qualify as “portable”. Who even does CAD on a small laptop? I’ve got a 12″ tablet, and I’ve only ever used it for CAD presentations, not real CAD work. So CAD on mobile is going to be limited to viewing, possibly markup. Not real CAD design or data creation.

So, if you don’t use mobile for CAD, is there any place at all for cloud data storage? Well, some people already have this in a web-based PDM program where they control the web server and access the web through a web portal. Even PDMWorks (SolidWorks Workgroup PDM) has been able to do this for years. So data accessible through the web is no real breakthrough.

The remaining piece to the puzzle is if you would place your application on the cloud. With the answer to the mobile device CAD being for the most part NO, I see very little use for cloud-based CAD. I mean yeah, there will be some folks or some organizations that embrace this because it fits some niche need that they have, but this is not a mainstream capable idea in my estimation. Mobile devices that access networks are a runaway success in the mainstream, but does that mean that they are applicable to CAD in some real way? The internet bubble pop of 2000 was all about basing business plans on unrealistic applications of popular technology. CAD content creation is something that requires hardware that at this point is not portable – large displays. Until tiny and cost effective projectors become ubiquitous, I don’t think this hurdle is one that will be overcome. And even then, you have to ask “just because you CAN, does that mean you SHOULD?”

I think the CAD companies, or one of them anyway, is desperate for profits, and is pushing this cloud concept beyond its natural limits. Cloud for mobile data consumption? Yes, absolutely. Cloud for content creation or data authoring? No. Not yet. There were reasons why CAD users turned away from centralized data storage and time-share applications served by a combined hardware/software vendor when it became possible in the 1980s. Unix workstations became available and obsoleted minicomputers and mainframes. More local control. Why would you let someone else control your hardware, your software AND your data? Why would you hand over 90% of your business operation to an outside party with basically no accountability? I can’t even think of a reason, unless it was as an emergency measure. I don’t want to hear comparisons to Salsafarce.com any more. That is not real data authoring or content creation. I want to hear, if you have it, compelling reasons to use CAD on the Cloud. Now. Put up or shut up. Why are so many subscription dollars being spent on this sort of stuff?

Why are users being threatened with this kind of mentality from Jeff Ray, CEO of SW:

Q: Will the cloud version of SolidWorks be the only version in the future?
Jeff Ray: When the pain of the status quo becomes greater than changing, then they will.

source: Ralph Grabowski’s UpFronteZine

And to emphasize this point, and validate what I’ve been saying here, Bernard Charles, CEO of Dassault Systemes said:

The cloud is an inflection point, as are new form factors like iPod and iPad.

source: Ralph Grabowski’s UpFronteZine

The problem is that I don’t believe that the new form factors or the cloud itself are suitable to the needs of SolidWorks or Dassault Systemes main thrust of business – CAD data authoring. The inflection point does not apply across all possible computer users. There is not necessarily an intersection between mobile computing and CAD data authoring.

Yes sir! No sir! Not that kind of General SolidWorks. Just SolidWorks, generally. You know what I mean. Still, I can’t guarantee that you won’t get fired for reading this, so proceed at your own risk.

Anyway, it has been a while since I posted anything here, and lots of things are happening. Still its summer, so you can excuse me for slacking a bit.

Dassault releases a 2D product?

Yeah, and honestly, who cares? It’s just more in the nasty war between DS and Autodesk. Give away what the other guy sells. Like Microsoft did so long ago. If you really need a 2D tool, check out Dassault’s giveaway product, DraftSight. What does it mean for the former DWG Editor, now known as the SolidWorks 2D editor from Intellicad? Well, who knows, and again, who cares.

If you want to find out more about this, go to Deelip’s blog on the topic. I believe that Deelip comes to the wrong conclusion. DS is not really getting into the 2D market, they are just trying to cut the financial legs out from under Autodesk, and gain access to real 2D users. It’s telling that this product announcement comes not so long after the settlement of the lawsuit over the DWG file type name resulting in SW agreeing to rename the DWG Editor to SolidWorks 2D Editor. That’s really a bit of a misleading name, because the SolidWorks 2D Editor doesn’t edit SolidWorks 2D data at all, only AutoCAD 2D data. Why these infantile companies cannot simply compete on technical merits I will never understand. Playing silly name games only benefits the lawyers, not the users.

SolidWorks 2011 Beta

So how many people haven’t installed 2010 yet? Is a constant onslaught of half-baked new stuff you don’t need really need the best way for SolidWorks to make more money? I rather doubt it, but that’s the way they have chosen. Still, 2011 beta is going to hit on or about July 6, and I’ll be there out of curiosity if  for no other reason. There are new contests, with fabulous prizes. The only prize I hope for is for bugs to not reach sp0, as unrealistic as that sounds.

Service Pack 4.0 released

This is bigger news than it sounds. If you check out the release notes, there are tons of bug fixes in sp4. Several pertained directly to some of my customers.

Head in the clouds

I was working in my cloud-based blog software, WordPress, the other day, when notice of a new version came down. I suddenly realized that one of the main selling points that CAD in the cloud proponents are touting is probably a sham. One particular claim is that on the cloud, there are no versions, you are simply running whatever software is installed on the cloud that particular day. Just like Facebook.

But what about WordPress? WordPress is not as big as Facebook, but it is certainly one of the most widely used web apps for maintaining blogs. This blog is written using WordPress. The thing is that a lot of independent developers develop plugins for WordPress, stuff like things that allow readers to upload files, or provide statistics or filter out spam, and their software may only be compatible with certain versions of WordPress. When you update WordPress, sometimes you have to update the plugin as well. And, guess what? WordPress does not update itself. You have to initiate it, and it certainly matters what version you are running on. And Facebook? Facebook is a rather loose ship, and I’m sure it matters every bit as much to applications that run within Facebook which version it is running with, but they try to keep this transparent to the user. And apps that run with Facebook aren’t exactly well known for quality and stability, which may be in part due to version incompatibilities.

So if I run WordPress on the cloud, what’s the big deal? Well, first, I have direct access to the files on the server because I own the domain. I don’t maintain the hardware, but I do maintain the software and through FTP I can add and remove files directly. The direct control makes all the difference. If I could host my own SolidWorks in the cloud, and the cloud could be on my domain or on a local server right in my office and only served across the network, that would be a value to me. But on someone else’s server across the internet? I don’t have any need for that, and doubt strongly that many people actually do. The people who are the biggest advocates of this are people who have something to sell, not actual users.

I do believe that some people will be willing to pay for the SW on someone else’s cloud, but it will not be the majority of users. And the reason for this has little to do with age. Some folks claim that anyone who disagrees with them is just an old fart. Which is strange, since old farts are the ones who remember previous attempts to push paying customers to take delivery of CAD software over the internet. Plus, I’m not really that old. And even if I am past the age where I’m able to accept the cloud, why do I blog on the cloud, and bank on the cloud when possible?

I think there is a real difference between blogs and banks and CAD. The blog is all about getting information out there, and being seen. That’s why its on the web. The bank, well, the bank’s data belongs to the bank. And whether I access it or not, it’s already available. So I might as well benefit from it.

But CAD data, well, it’s just so big. And it takes so long to compute. And, quite frankly, unless I am sharing data (which I generally do not) there is not even any reason for me to have the data on my local network, much less someone else’s network. I’ve long been a proponent of solving problems with the least possible complexity. Anyone who thinks the internet is less complex than a local computer or even a local network is simply not being honest with anyone. So my beef against CAD in the cloud is the complexity of the solution, and the lack of alternatives if something goes wrong. Not my age.

Update update

This should be a separate post, but I don’t want to give this that much exposure, so it’s just tacked on here at the end of the cad in the cloud. Anyway, Spatial, the folks who bring you the ACIS geometry kernel, are now offering a really simplistic CAD viewer. (gotta use IE7/32) It does slightly more than just view, you can also remove faces and add fillets. I suppose these tools have been added to prove the actual geometry kernel is being run across the web, to distinguish it from a straight viewer. Is it more than just a toy? Not right now, but you can be sure this is not all that is coming. You can also be sure that a viewer is a long way from a real CAD tool.

If you want to read more about it, and from someone who is more impressed with it than I am, again, go to Deelip’s blog.

The timing of an announcement like this seems lined up to take some of the punch out of the Synchronous Technology 3 announcements coming from the Siemens PLM event.

The new Spatial offering will allow developers to license the Catia V6 geometry kernel as well as the cloud tools to develop 3D cloud  applications for end users. It seems uncharacteristic for Catia to license anything at all, but now that SolidWorks is getting a V6 make over, they might as well let the rest of the world do it too. but I think the really important thing here is that they are licensing their cloud tools. If the customers are (D)assaulted on all sides with cad in the cloud from many different vendors, they may be more likely to succumb. Or a rising tide floats all ships. However you want to look at it. They are trying to make some money off of other vendors who think they can sell cad in the cloud.

I’m sure the Really Big CAD companies hear the protests from users, and they are developing this stuff anyway. So far Solid Edge seems to be the one that is not getting caught up in it. I guess they have enough to worry about with sorting out the last band wagon they jumped on.

The cloud discussion on this blog has been wide ranging, and has adjusted in tone somewhat after some quotes from Messrs. Ray and Charles (sorry, not intentional bit of name mixing…) were made public. From what they are saying, it looks as if they are poised to make a monumental mistake. The biggest mistake they could make at this point would be to kill off, or begin the end of the lifecycle of the locally installed version of SolidWorks. While that’s what they seemed to have said, I can’t believe that they would do that.

I do see a lot of benefits of a cloud implementation of SW, but as the only implementation, I have to say I think it would fail miserably, and SW would lose a lot of customers. To be clear, the comment from Jeff Ray saying that when the pain of using a version that does not get updated becomes too much, slow-to-change customers will move to the cloud version is disingenuous. I would be willing to bet that people who do not move to a cloud implementation do not do it for a reason, and will be as likely to move to another brand software package altogether as to move to a completely new look and feel (and learning curve) of SolidWorks on the cloud. Surely people who are paid a lot of money to grow rather than destroy a company/product are going to see this as well. While they seem to be saying that the future is completely in the clouds, at the same time, there is no way they really intend to kill off the local install.

The new version of SW (speculation warning) will likely be based on the Catia V6 geometry kernel as well as the Enovia V6 backbone (Enovia V6 was confirmed at SWWorld, Catia V6 was not), which together would allow… well, you’d better read the DS site about what V6 allows, and Enovia V6… or a summary here.

What if… now, I’m going back to some earlier conjecture here… what if the current manifestation of SolidWorks disappeared, and in its place we get a choice between SW served up over a web server farm AND the same cloud version of SW that you could install locally on either a single machine or your own local server farm. Let’s face it. The arguments for stability and data loss being much better on the web are pure bunk. If you can do it across the web, you can do it in your own office. That’s just a little bit of SW preying on the lack of information that customers have at this point. Did you know that even without a new version, SolidWorks 2010 has the capability to automatically save your work as you go? did you also know that this capability has been there for over a decade, but most users don’t use it? Did you also know that 3D Instant Website is cloud-based collaboration that SolidWorks has had for again about a decade, but people don’t know about it and don’t use it?

So here are the polls:

 Loading ...

 Loading ...

 Loading ...

I know I’ve done this before, but with all of the new speculation and information I wanna do it again. (three polls)

 Loading ...

 Loading ...

 Loading ...

I’ve burnt out on this topic long ago, but I feel like I have to summarize it. There has been a lot of good information that readers have come up with, and I just want to bring it all together.

Advantages of using the cloud:

• You can use SW anywhere you have a web connection
• With a mobile card (cell phone) you can connect to the internet from anywhere you have cell phone access
• There is no longer the need for separate desktop and laptop machines or the portability/performance compromise between them
• Distributed computing for multi-threaded tasks
• You don’t have to worry about local installation issues being part of the stability problem
• You don’t have to buy fancy hardware any more
• Local system maintenance is not as crucial as with local installs
• Sharing and visualizing SW data with non-SW users/owners will become much easier
• (advantage DS:) on the cloud, no more piracy – all users are paying users

Disadvantages of using the cloud:

• You have to be on the web to use SW
• SW is not sufficiently multi-threaded in general modeling to take advantage of distributed computing
• If there are technical problems, you are not in control, and you aren’t in close contact with the person in control (reseller->solidworks->cloud provider->support person/division)
• The fancy hardware that you still have to buy for other purposes doesn’t give you any advantage in the cloud, and you don’t get the advantage of not needing to buy that fancy hardware
• Some third party non-partner add-ins will almost certainly not be offered in the cloud
• Once you stop paying, you no longer have access to the software unless you have a local install
• It has been shown many times that cloud applications are vulnerable to criminal activity. Identity theft is a huge source of crime, and using the fact that online banking is so prevalent as a way of supporting the safety of cloud data is a bit disingenuous because there is also a lot of  identity theft through online banking and other cloud activities.
• Governmental agencies and contractors have policies against storing data in places where security is not directly under that agency’s control.
• Internet standards compliance has not been SW’s strong suit. How many of us use browsers that still don’t display the SW sites correctly?

Things that could go either way depending on how SW implements it:

• Long term licensing: has the potential to eliminate purchase cost, but why would SolidWorks leave that money on the table for technology that is going to cost more than just the base software?
• Short term licensing: if I just need an extra license for 3 days, will that be cost effective? Is there a one-time buy-in charge for licensing deals?
• Very short term licensing: if I just need a SW add-in for 3 hours, will that be cost effective?
• Third party gold partners (like Delcam for SolidWorks) may not be offered on the cloud
• It remains to be seen what happens to your data once you stop paying for cloud. can you get it back? does someone else now own it?
• Can you delete your data off the cloud? even after you quit paying?
• Will you be able to select from multiple versions of SW for compatibility with non-cloud using contractors/customers/etc?
• Within major releases, can you also opt to use a different service pack if the latest sp has bugs you can’t live with?

Please add more to the list. I’ll add them here.

Read this.

First, please read this article from the Electronic Frontier Foundation. They are talking about the Autodesk vs Vernor case where Autodesk tried to prevent Vernor from selling old copies of AutoCad on eBay. This might sound like the wrong end of the pirates issue I posted about recently, but its not. Vernor was not selling pirated software, it was legitimate software in the box with the disks and everything. It wasn’t some hack copy with a cracked registration. The court (and all CAD users) sided with Vernor. The points about selling used books, movies, games, music CDs are all exactly the same as software. CAD corporations should not be able to distinguish their product as any different from these other forms of electronic media.

Anyway, the pot I want to stir here, is that now that the Adesk v Vernor case has gone in favor of the end user, the corporation must find another way around that. Licensing in the cloud is going to be brutal. It really is going to be the old “you don’t own it, you license it” thing. If they can get a lot of people to buy into this way of doing things, they will have more and more customers locked in. There will be no such thing as “legacy” data in the cloud. You keep paying or you say good bye.

We’ve talked a lot about some of these other cloud issues. Let’s talk about licensing. What do you think would be a fair from the user’s point of view for cloud licensing? It seems clear that SolidWorks will own the application, but who will own the data you create? This idea needs more scrutiny. What do you think?

Read this ZD net article saying that the cloud must be the future because criminals use it. It’s this kind of thing that shows I’m an amateur at this kind of thing. I would think that if you combine criminals and security vulnerabilities, you’d come up with something bad rather than something good. Oh, so much to learn…

With all of the talk about clouds recently, and the interesting coincidence that John McEleney, the former CEO of SolidWorks, is now the CEO of CloudSwitch, I thought now might be a good time to talk to John. I don’t have any special access to Mr. McEleney, but we were acquainted before he became SW’s CEO. I went to dinner with him once in his role as a sales VP (if my memory serves correctly) at the Anchor Bar in Buffalo, NY back in the late 90′s. The Anchor Bar is the birth place of Buffalo wings. John also got his bachelors degree in mechanical engineering from the University of Rochester, and I worked at SolidWorks resellers around Rochester for several years.

Here is a summary of what CloudSwitch does, right from the CloudSwitch website:

CloudSwitch enables enterprises to move their existing applications to the right cloud computing environment—securely, simply and without changes. With CloudSwitch, applications remain tightly integrated with enterprise data center tools and policies, and can be moved easily between different cloud environments and back into the data center based on the requirements of the business. CloudSwitch protects enterprises from the complexity, risks and potential lock-in of cloud computing, freeing them to leverage the cloud’s advantages in cost and business agility.

If you are following this cloud business closely, you will want to watch this YouTube video to get a better sense of what CloudSwitch is, and how it could potentially play into SolidWorks or even your own company’s plans to move applications to the cloud.

This picture shows John McEleney at a table full of users at SWW 2007 Super Bowl party signing a copy of SolidWorks 1995 that Richard Hall had with him. The other users shown are (left to right) Devon Sowell, Rob Rodriguez, Richard Hall and Ricky Jordan. John was the type of SW CEO that users could relate to. He had a mechanical engineering degree and a manufacturing engineering degree, in addition to the obligatory MBA. He can obviously think at a higher level, but also knows the difference between counterbore and countersink. This is why he can relate to users in a way that other SW CEOs might not.

The CloudSwitch product is aimed at enterprises that make use of data centers. This is not necessarily a one-to-one relationship with what Solidworks is hoping to do, but Mr. McEleney has a lot of insight into the how and why of cloud computing in general, some of which can give us a glimpse into what we can expect to experience with SW in the cloud. His company doesn’t really deal with software end users, they deal with the corporate entity implementing the cloud software, and moving (switching) from a local cloud which may be on all the time, but only needed on a part time basis, to a much larger and much more flexible hosted cloud with the ability to turn it on or off as needed.

Anyway, I caught John with some open time, and he was just a firehose of information for about 45 minutes. I took 3 pages of notes, but should have had 3 times that easily. What follows is a bit of a paraphrase of some things he had to say about himself personally, the SolidWorks community, and about this business of cloud.

John started with a bit of an explanation of why he left SolidWorks in the first place. Jon Hirschtick served as CEO for about 6 years before McEleney took over for the next 6 years between 2001 and 2007. Then he had some personal issues to take care of, and decided to step down. In short order he had a wedding, the birth of a child and the death of first his father then his mother, all of which made this period quite demanding for him.

After some time, as you might expect from any ambitious and motivated person, John started feeling a little restless, eager to get back into the fray of business life. He began talking to investors in the Boston area, and started noticing some things about some of the business proposals that passed his desk. His experience at SolidWorks taught him some things about what computer users wanted: faster computers and better graphics. But these things weren’t really core to the customer’s business, they were secondary at best. He recognized the need for what he calls a platform shift. The emergence of the internet itself is a platform shift, like moving CAD from mainframes to Unix to Windows. Moving applications to the cloud is the latest platform shift.

John also noticed that “every” company that was using cloud type applications was using Amazon’s Elastic Compute Cloud (EC2) rather than building their own server structure. This was because of the flexibility of Amazon’s cloud to match the compute need with the business need.

Of course I had to ask if there was any sort of connection between him joining a cloud company, and SolidWorks announcing that they had been working on cloud applications for 3 years, and he said, quote, “Absolute coincidence”. I’ll take him at his word for that. Cloud is an emerging market force, and it makes sense that SolidWorks is looking at it, and it also makes sense that there are startups in that discipline that need leaders. In fact, John rather echoed the claims of Jeff Ray and Bernard Charles that SolidWorks has a “fiduciary and strategic obligation” to be looking into this platform shift. The fact that this is a platform shift will “only be obvious to the masses in hindsight”. How long will this shift take? John recognizes that you may never get everybody off of the old way of doing things, but most people will switch over the course of several years, with the usual curve (as with first board to CAD and then 2D to 3D) bringing early adopters first, followed by increasingly conservative businesses and individuals. John says “There are no homogeneous solutions anywhere”, by which he means that cloud will coexist with other application delivery methods for a long time if not indefinitely.

When I asked John for an example of a customer that can benefit from using a hosted cloud such as the EC2, he cited one unnamed company that uses Siebel as their CRM (customer/contact resource management) system. When they want to test and roll out a new version of the system, they have to have a parallel install, and thus a second large bank of servers to handle the testing. It is so much easier to use a scalable hosted solution to do this rather than setting up a local server farm to do the testing.

McEleney gave a great analogy that helped put cloud computing in perspective. John’s wife is from Buffalo, NY, and she might want to go and  visit sometime. She might want to stay a couple of days, or maybe longer, who knows. When she gets there, she needs to have a car. She could fly to Buffalo, buy and insure a car, then sell it when she leaves, but that’s wasteful and inconvenient if shes just there a couple of days. She might take a taxi, but that might also turn out to be too expensive if she wants to go to Niagara Falls, or to Canada, or to the Finger Lakes wine country. Or what if she wants to go somewhere and take the whole family of about 15 people? Other options include leasing and rental, and between the four options, buy, lease, rent, taxi, different scenarios might call for different solutions, possibly even hybrid solutions. That’s where the great value of the cloud starts to shine. It gives you immediate and more cost effective ways to quickly scale up or down without a huge investment or waste.

He gave another example of a company called Animoto, that takes music files and image files from Facebook, and makes a quick video from them. They started with a couple of servers in the closet of an apartment, but that was prone to all sorts of local problems, like cat hair in the fan and his girlfriend hitting the wrong light switch. After moving to the Amazon server farm, they went from say 2o hits an hour to a million hits a day over a very short time frame. There was no way they could have physically set up the necessary 3500 servers that fast, or had the room for them in their apartment. Being able to move into the cloud allowed for the business to grow explosively.

Cost for servers like this is charged on the order of 10cents per server hour. So If you have a complex computation that takes 10 servers 1 hour, that would cost $1. Figures are meant to be approximate, not exact. And there is no guarantee that SolidWorks will bill things like that either. John suggested that some of his beta customers envisioned payments like minutes on a cellular plan, to allow them to budget. I believe SolidWorks is still investigating the whole cost issue. How much to charge, and what sort of granularity to offer for billing or payment plans.

John did have a few other comments to add. He said a couple of times “the cloud is real”, I suppose to emphasize that it is not a passing fad or a plot to extort more money from customers. He was careful to add, however that he thought it was overhyped, but it was still real. He recognizes that there are a lot of real and legitimate questions that people have, some of which have good technical answers, and some don’t. He compared the early cloud to early product design, where CAD vendors created some inadvertent consequences. For example, large data sets and disparate locations required PDM applications, which became a market unto itself, and growing now into PLM. PDM wasn’t an intentional plot to make more money, it was a solution to some problems created by CAD.

He recognized that some smaller vendors might not need a cloud solution now, and that possibly a hybrid solution can allow interaction between cloud and non-cloud applications. He cited a triangle that shows the cloud implementation as a combination of Software as a Service, Platform as a Service, and Infrastructure as a Service :

He used the term “machiavellian” several times, by which I assume he meant that some naysayers propose that the cloud is really an underhanded scheme just to make more money, and that business people are forcing it where there seems to be no technical reason. John believes this application of technology solves many actual end user problems, although he acknowledges it does create others.

My overall take on John’s comments are that he strongly believes in the cloud (how could you not as a CEO of a cloud company), and that while his business is focussed on businesses offering cloud applications rather than end users, he sees gains in efficiency and stability through the cloud. He acknowledges that there are no homogeneous systems anywhere, and due to late adopters, naysayers and some possibly incompatible scenarios (such as military), some sort of a hybrid state may need to exist. Where this all settles out, time will tell some years into the future.

In parting, John wanted me to communicate to the SolidWorks community that he still loves the industry and the people, and wished to encourage us to keep challenging ourselves, our tools, and the people who make the tools. He also related a fun story. His 11 year old son said that he had two dreams. One of them was to some day go to SolidWorks World, and the other was to some day go to Mac World. With this cloud business, and platform compatibility through the cloud, those two events may have more in common in the future than they have in the past. Anyway, John said that he may be looking to attend a future SolidWorks World event. I would certainly be one to welcome him.

Overall, in this interview, Mr. McEleney was focussed on delivering cloud services to other businesses who in turn deliver software to end users. He was not commenting directly on application of CAD in specific to the cloud, and said that he was not involved in the development of SolidWorks cloud applications. He is by necessity an optimist when it comes to the cloud, and while he acknowledges some of the unanswered questions and difficulties about applying SolidWorks specifically to the cloud, as a businessman he is not letting those hurdles stand in the way of developing what at least some customers will find useful.

So, while I have no doubt that the cloud will march forward regardless of how many reasons people can find not to implement it, I believe that SolidWorks users still need to play a little devil’s advocate and bring up the questions. Someone may find a clever solution to them.

In closing, I want to thank John for the time he spent with me today on the phone. I think this conversation benefited my understanding of some of the underlying issues, and hope that it did the same for some readers.

Read the Cnet article, and the original article quoting Larry Ellison.

This is a great quote, but it sounds a little sarcastic:

Said Ellison: “Everything’s called cloud now. If you’re in the data center, it’s a private cloud. There’s nothing left but cloud computing. People say I’m against cloud computing–how can I be against cloud computing when that’s all there is?”

Is this cool? Yes, it seems like it will be useful for a certain range of work, but certainly not all work. It may be something I would like to use with customers who don’t have CAD.

But here’s the funny thing. Or funny to me, anyway. SolidWorks already has this thing called 3D Instant Website that people already don’t use. It exists now, and does probably 70% of what his new thing does. The big difference here is storage of actual models in the new stuff. Did I mention that no one uses 3DIW, even though you don’t have to pay anything for it if you have SW Professional.

I personally like, or liked 3DIW. But I couldn’t get people to use it. Doing the same thing and then putting your files up there on an unknown server doesn’t sound to me like a big improvement. I’m just curious why if it failed the first time, why they want to call it new and try essentially the same thing again (and possibly charge you extra for it?!?!)