[News.eclipse.foundation] Re: IBM, Sun, NB, and Eclipse ?

[Ed Merks <Ed.Merks@xxxxxxxxx>]

Vlad,

Comments below.

Vlad Varnica wrote:

Ed,

I agree with the fact that the eclipse foundation doesn't adopt open source projects and simply hosts them. What I don't agree is that this information is not really public and that many companies believe that each project is under Eclipse Umbrella and protection.

Not really public?  Isn't that like being a little bit pregnant?  What's implied by this Umbrella of protection?  A 1-800 support service?   Certainly the technology is IP clean and ready for commercial consumption, but buyer ought to read all the licenses, terms, and conditions, which I think are quite clear on this matter.

This is today breaking our traditional ISV business model which is to build value on the top of open source frameworks such as EMF and not to provide out of box fully professional tools in order to help large corporations to reduce their software investments.

So all the Eclipse members building products on top of Eclipse are broken?  I don't see the evidence for that.  Definitely Eclipse requires that the projects provide extensible technology so that there is the possibility to provide value add.  But in the end, there is no limit to how good or complete the out-of-the-box experience for a project is allowed to be.  JDT for example is world class.  It would be hard to build a product that improves on it.  That's okay but we should have a double standard for other projects?

Our problem is in the last three years many open source tools used the EMF with GMF framework in order to directly compete with existing vendors by adding many professional features.

A problem indeed.  Better work hard to keep ahead.  Ride the wave of change or be swept away by it.

The result is that the monetized modeling plugin market is impossible today

More challenging I agree.  Few things are impossible though

and that for example we can't pay high level consultant such you Ed, or even add free contributors members to this group because the generated revenue is too small. This is why today Eclipse is non monetized market and I hope this will finally change.

I wouldn't hold my breath.  The waves of changes don't tend to stop.

Omondo is using two plugins which are EMF and GEF. We refuse to use GMF.

You make EMF sound so small and light weight.  I like that. :-)

Now be sure that we will not contribute to any EMF project because we don't agree with the global project technical orientation which is more to compete UML that provide real value and because other open source projects would reuse it and provide it for free to our end users.

I'd suggest you never complain about the quality of things others have invested in heavily and you use for free.  EMF was used to build things like XSD and UML.  If it competes with them, perhaps that's a short coming in UML and XSD.

It is therefore breaking our existing market.

Find a new one.

There are over 5,000 bugs inside EMF as soon as you try to develop an advanced and professional solution.

That's the most nonsensical thing you've said so far.  My list shows not a single unresolved defect, only requests for more goodness.  It's hard to buy that level of quality, service, and support from a vendor.

Omondo did pay the price to fix them and we will certainly not give them to our open source competitors today.

Likely you'll take them to the big bit bucket in the sky then.  Of course if you've directly modified the base EMF code and you are shipping those modifications in your products without sharing the source you are in violation of the terms of the EPL.  I'd not discuss such things on this forum were I you...

Sorry about that we like to play the rules but not to be stupid.

It sounds like you aren't playing by the rules, which would be quite stupid to admit in public.

Concerning GMF it is different because we just don't see any interest to use it.

So don't.

GMF is helping to reduce development cost when building an UML tool

No, it's more general than that.  It's for building a graphical editor for any model, with UML being just one instance.

and this is today used by all the open source projects.

Neither UML nor GMF are used by all the open source projects.  EMF comes close though. :-P

We consider that only native integration to EMF and GEF can generate technological value and adding and intermediate stage in the middle is major problem for scalability, live synchronization or for full project life cycle modeling keeping the same UML Id.

Yet IBM builds its UML products on this stack. I imagine you must compete very well with them then...

It is not possible today to keep all these advanced features because of all these transformation stages. Except to reduce time to market development in order to compete tools there is no other interest in GMF.

Why bring it up then?  There must be a lot of projects you're not interested in...  I'll bet you don't like Papyrus much and no doubt Ecore Tools isn't high on your list either.  I'm rather fond of Ecore Tools personally.

Concerning GEF this project is just fantastic and don't really need additional support.

My previous post was concerning the change that IBM acquisition have on the market. For me the professional open source tool market providers is dead because the real need is to build value on the top of existing frameworks and not to try to reduce large corporations software expenses.

We see many examples now of groups of consumers getting together to collaborate on building technology that they can reuse more cheaply than buying it from a vendor, and thereby also avoid vendor lock-in.  That will happen regardless of whether IBM buys Sun.  I don't see a connection between the two.

If EMF and GMF goal are only to compete with UML or existing tool vendors then be sure that these projects will not find many support from larger corporations and will sooner or later stop supporting them.

You've got a very narrow view on what EMF is all about.  Even if what you said were true, your conclusion is still wrong: large corporations would happily use free UML tools.  That might be very unfortunate for Omondo and would even be unfortunate for IBM which derives a great deal of revenue from it's UML tools.

It is still time to change this EMF and GMF none sense strategy and be more positive.

Don't hold your breath.  You aren't the arbiter of what's positive and what's not.  In fact, I set the strategy for EMF and that strategy has nothing to do with competing with UML tools.  The goal is the provide the world's best build data integration platform upon which anyone can build any kind of cool application.  Some of those applications might compete with you and some might be ones you use to build your ISV technology more expediently.

We need stable frameworks to add value on the top of them

Like EMF.

and not hundred of new open sources projects trying to compete with us.

Too bad, you get those too.  The sword has two edges.

Eclipse foundation has given too much marketing to these projects and made end users believe that they don't need anymore traditional ISV vendors.

You can't put a lid on competition, so be prepared for it.

I will never agree with this vision and financial institution which are currently helping IBM to acquire Sun have the same vision.

You don't need to agree with it.  You don't even get involved enough to help influence the direction.  The guy in the driver's seat controls where the car goes.  The guy in the back seat can only move his lips and hope that effects the changes he desires.

WE NEED TO MONETIZED ECLIPSE.

So be innovative. If you don't, someone with a different vision will.

Vlad,
Omondo