OT: Linux popularity

I go into a lot of engineering departments, and talk to a lot of IT people, it's my job. I see mostly microsoft, of course, a lot of Novell, some Unix, few Macs, and no Linux. I hear of Linux being used sometimes on an invisible server, but never where end users, including mechanical engineers, have to touch it.

For SW to port to Linux, Linux is going to have to come out of the IT closet, literally. Linux right now is for server administrators, geeks and general establishment fighting rebels who don't care what it is as long as it's not embraced by the masses.

If you were to put Linux on your mechanical engineering resume, most of the people reading it wouldn't even know what it was.

Here's some stuff that's more than a bit biased toward MS, but at least gives an alternate point of view, and does have some good points. It basically says that the Linux crowd looks at MS much more closely than they look at Linux.

"P." wrote in news:1100742896.994178.47340 @f14g2000cwb.googlegroups.com:

We've talked about this before Matt, and my position on the issue has changed considerably.

Heh. I used Linux for *years* as my primary home desktop OS, and recently wiped it in favor of Windows, because that's where today's CAD/CAM apps are. I seriously doubt Linux will ever gain any traction in the CAD/CAM market, not for reasons mentioned above, but because the legacy UNIX apps (UG / ProE / Catia, etc) are dying a slow but sure death in favor of Windows popularity. For example, PTC is dropping IRIX (SGI) in their next release of Wildfire, and one by one the rest are sure to follow.

Heck, look at Solaris 10. Sun's giving it away starting in January in a last ditch effort to make some money selling service for it. You just don't give away "3,000 engineering years and an investment of more than $500 million in research and development" [*] of something that has a chance of turning a profit for you.

Death isn't purty'. :-/

I even ditched Linux in favor of the more mature FreeBSD on the server. Don't have any Linux media anymore either. No sense in hanging on to stuff that's never going to get used again.

Solidworks on Linux? Never going to happen, IMHO.

Got a link? That's a shame. For a long time the nicest place for an Engineer was in front of a blue or purple box with a cube on it.

Regards,

Indeed. :-(

Haven't had the opportunity myself. Yet, that is.

Most of the major CAD/CAM systems still run very well on various flavors of UNIX.

Guys vision is were it's at. I can assure the death you are describing for linux is the same for Solidworks if you think about it.

The operating system is becoming a comodity. I assume the only way that Linux will win is if the model that Microsoft uses fails. Think Explorer Vs Netscape and then Vs Firefox.

The same is the case with Cad. In the last few years I have been seing products pop up out of nowhere and suddenly seem almost complete. A product that took a small team a few years to develop is way ahead of products that took decades before them. The reason is they are standing on the shoulders of past developers and solutions that have become generic.

So to get back on topic. Solidworks will become one of many generic 3D products in 5 years time. The most critical wanted feuature is the one it must avoid like the plague. Intercompatibility ... Imagine buying a $80 product that can read Solidworks files/save Solidworks files and you will have trouble telling the two products apart.

Can't be done you say ????

So unless it can lead by doing what otheres avoid, it will be one of the others. So intercompatibility is give away your file format. Work with other operating systems, abstraction is the key (I could be wrong :-)

Giorgis

Matt,

I didn't think you led such a cloistered life. When I was working for a VAR here in Indiana I ran into several engineering departments that had to make SW work with Linux as the file server. I am sure you must have run into some in New York. I have been using it for a file server for years. Much easier to maintain that Windows and faster too. In fact I take the Linux box for granted because once setup it just runs and runs without reboot or glitch. The last time it was shut down was to add a hard drive. I compare that with the place I used to work that had at least 3-4 server crashes a week plus a weekly shutdown for maintenance.

I also noticed that all the articles you quoted were from 2000, almost four and a half years old. Hardly current information by software standards.

But the article I posted is about China, an area in which 99.999% of the people on this group have little if any first hand knowledge. However, China is growing at such a feverish pace, especially in engineering and manufacturing that it will soon rival America as being the driver of technology. That is why I posted the article. If China continues this growth and still feels left out of the mainstream they will simply do their own thing and from what I see that will be towards their own flavor of Linux. For example they already have their own WiFi standard that is propriety to China, Inc. and it leaves the American companies in the dust. When it comes to software, growth and sales are where it is at, not support and maintenance. When was the last time SW touted its growth in maintenance support contracts at SWW? So if CAD sales in China are to grow like their consumption of so many other resources then SW better take note or half the world will be using ShanghaiWorks.

I don't remember the link, but I actually saw Microsoft point people to a link that touted Linux. The link said that implementing Linux was four times the work of upgrading Microsoft server. If you read the article you came away with the thought that if Linux was just four times more difficult than running through a Windows upgrade that involves sticking a CD in the server and waiting a half hour, then why not do it? After all they were comparing an upgrade to a full change and rollout. The rest of the article talked about how many people where looking at Linux and its robustness. There is a reason IBM, MSC, Noran Engineering and many others are pushing/supporting Linux for engineering.

Check out : "Microsoft and Dassault Systèmes Form Strategic Alliance"

I actually heard John McEleney speak on this item a week ago. The big part is that the Linux Desktop OS is not in big use. On the server side he all for it. But the common OS just isn't there.

I got the feeling that they are keeping an eye on it and if the OS starts making a push they will be ready for it. They won't make the same mistake Pro-E made.

Also, all you Linux experts already know this but, if you are going to throw SolidWorks parts and Assemblies out on a Linux server it sounds like PDM is a must because if SolidWorks crashes on an assembly it doesn't know what to do with the ~tilda files that windows creates and you can't open the assembly until all the ~tilda files are removed. And in a multi-user enviroment......

Ken M

I don't have that problem with a crash. My files are on a Linux based box and after a crash, the ~tilde files sometimes disappear and sometime they don't. Either way, I can always open up any of the non~tilde files without deleting the ~tilde equivalent first.

About anything can work as a file server . Except, perhaps, MS ..... that usually has problems I gather.

And what platforms must the kernel work on?

Consider the effects of common kernels, like ParaSolid.

The computing world is consolidating, it seems to me. Apple/Next looked at the Unix/Linux versus Microsoft and decided it would be better to align with the .Nix world. Hence, now, software written for Unix/Linux is easy to run also on the Mac OSX. That has helped reinvigorate Unix a bit.

Mac Servers are getting serious respect in IT departments and even in some of the heavy lifting with Mac clusters in acedemia.

I would surely enjoy being able to run Solidworks on my Macs, even if I had to move to the PC to finish the job later.

In today's world, I really do NOT understand why corporate users are not driving OS manufacturers toward a common .Nix standard and letting the user pick his favorite flavor of user interface. I realize MS will not like it, but what it would do is place an emphasis on service, features, reliability and support, and then a company or individual would not be locked into a "dud" company.

I am sorry to say that I have never viewed MS's OS as anything I could call "great". It may not be a "dud", but I don't use it when I don't have to do so. I am just not as productive as on the Mac.

Bo

There is one major contributing factor to the unpopularity of linux on the desktop, and that is the fact that none of the mainstream application developers have products for linux. Decision makers of Companies, small and large, rarely purchase products without knowing what they are getting. You can't evaluate something that does not exist.

I have been using Solidworks since 1998 and I can say this, with no uncertainty. If there was a version of solidworks for Linux or Mac OS X, I would immediately begin the process of switching platforms.

I will tell you my reasoning. I used to have a completely Microsoft server setup and was spending quite large percentage of my time maintaining it. I changed the layout to replace my Exchange my server with a Linux server running Sendmail. The linux server had zero downtime in its first year of service. Where as the Exchange server had monthly maintenance and usually a semi-monthly restart to keep it running. Also, the performance of the Linux system is much more robust under heavy loads. Now, if I can bring that to my CAD workstations which get restarted at least once a day, if not more, then I could be that much more productive and that much more competetive in today's market. Considering that the average time to get back into design after a restart is about 5-10 minutes, this is a significant slice of time. Also, considering the fact that Microsoft addresses the problem of system instability by improving the bootup time, I don't think the computing world is in competent hands.

Rob Muncer, Owner

CGR Technologies, Inc.

Palatine, IL 60067

snipped-for-privacy@cgrtech.com

The whole point of this thread is that a lot of decision makers are going to be coming on line who simply do not hold these truths to be self evident. In other words old wisdom will fly out the window. They won't have preconceived notions and they will have a large resivoir of experienced Linux users to tap. As we speak China Inc. is writing its own version of Linux. MicroSoft and Linux are on a much more even playing field in China. And China is the next big growth market.

We used to have a lot of trouble with this, I was constantly deleting all the ~tilda files off my Solaris file server running SAMBA. But one of the Solidworks users came up with the fact that if the person who crashed just immediatly opens back up that assembly and exits correctly, its all dealt with just fine.

The simple facts are that if MS wanted to ensure only legitimate copies of MS exist, it would do it. It is quite easy in fact. The problem is that if they push for that, a HUUUGE number of people will move to Linux. In a world of high bandwidth and cheap storage you can download XP and Office in mean hours or less. All Microsoft is doing is preventing the inevitable, that is critical mass for Linux. For now large companies pretend to go for Linux and get huge discounts from Microsoft.

So that is point one.

The question now is: Is Linux upto scratch ? Well the answer in my opinion is Yes. You see MS is in the same trap SolidWorks is. Put new features or Fix old bugs. Well new features get you customers. Once you have them, you lock them in by ensuring your files are proprietary. Can use SolidWorks files with ProE. Do you wish you could. Well you're not getting it. So it's just market forces.

Microsoft XP is actually very stable, the blue screen of death is pretty much in the imagination of the Linux enthusiasts now. (at least on the desktop) But at the end of the day, Linux can afford to rid it self from bugs, add features and bug test for nothing when MS has to spend billions and introduce stupid features like "clippy" hoping to stay alive. Linux can avoid new features. If you want them, Linux is quite modular, you take the risk.

PS: There is one real factor that MS is gearing up for in my opinion ... legal challenges to every patent it owns. And it owns tens of thousands. I am sure the use of the letter "e" in emails is patented by MS or somebody MS can fund.

Giorgis

preconceived

huge discounts >from Microsoft.

But is this true of China?

proprietary.

How does the China market view "proprietary". China Inc. takes a very dim view of proprietary unless they hold the cards. China is still a state run government. If the government can't control something it won't succeed. It is viewed as not good for the people.

I just read an ad in Fortune that IBM paid for. The add was for Linux. Kind of makes you wonder why Bill forged a bond with Charles. Historically IBM and Dassault have been joined at the hip. Could it be that Gates sees a danger in Dassault being conviced by IBM to port Catia and its PLM solutions to Linux (which after all isn't a big stretch given Dassault's experience with Unix.) In an oblique way, Gates in my opinion is showing us that LInux is a real threat, a real competitor and therefore a real alternative.