Some thoughts on IBM's "Project Vulcan"
by Volker Weber
Lotus charts out a vision for future versions of their existing portfolio. That is a good thing, and may in part be the answer to a question for a five year roadmap. I see three things in that vision:
- The desire to make the vast portfolio of products work together. Both the stuff you run in your company and the stuff that IBM runs for you as part of LotusLive.
- A new user interface, built on powerful web technology around HTML5.
- Automatic aggregation and filtering of data to relieve the user from too much irrelevant information.
#1 is overdue, and it is the hardest part. It's not going to happen overnight, but IBM is already at work defining new APIs to pave the way. Having said that, IBM is playing catch-up. Msft has been hard at work for a couple of years to build the software and the data centers to make it happen. And they are ready to execute on their plan.
#2 is a huge opportunity to rethink the Eclipse strategy and build something small, fast and nimble. The future client could mostly live in a browser window. The current design prototype works along the lines of Facebook, a very popular UI. IBM will have to see how well this works in a business environment, but the demo that was given to me was very, very impressive and easy to use.
#3 is where IBM can shine. They have the expertise and the background to pull this off.
Overall, a very good move. Some caution is called for. IBM is waking up very late.
Comments
Not having been at Lotusphere, but following from the sidelines, my impression matches yours 100%.
They *can* pull this off. IBM tends to think through implementation very well, and assuming they:
1. Stop
2. Think
3. Plan
4. Dump a TON of resources on it
...they could very well dominate the market. #4 is key. They are waking up late, but if you look at the sheer development effort dumped into N/D for 8.0-8.5.1, they've definitely shown that they can seriously produce when they need to.
So IBM has praised and build an uber fat client that was hardly usable to just throw it away again and build something new around the browser?
It is not that I don't like the idea of Vulcan but IBM has build an astonishing list of fails and experiments over the last decade.
IBM has in way said they are going to dump the Eclipse client. But *I* see a chance to build something much better. Remember how you needed Java applets for the web in R5?
At least they need not worry about backwards compatibility of 8.x plug-ins and composite applications, because nobody actually wrote any.
To me, the presented Vulcan client looked like a rebranded Eclipse/Expeditor client that was running a local XPages engine with mail and calendaring - of course with the known limitations of a web environment (e.g. a HTML editor to create emails, no Notes richtext editor), but I guess they will be closing this gap more and more and add some "rich client" features to the XPages UI to make the local client worth installing it instead of only using a web browser.
Legacy Notes applications will of course still work, but the main technology focus will be XPages app dev.
IBM also announced in one of the XPages sessions that they will add OSGi support to XPages for both client and server. OSGi is the plugin model of Eclipse, so XPages apps will become easily expandable using an extension/extension point model comparable to Eclipse (maybe it's even the same model).
Based on this modular architecture, business partners can add their own extensions to the client. They will have the choice between "write once, run anywhere" using XPages add-ons and Java/Eclipse based tools that only run locally, but with extended capabilities.
I've been trying to find formal training for Notes/Eclipse plug-ins and composite apps. When my training people contacted IBM about it earlier this month they were told I am only the second person in the UK to request it. I figured that may be the writing on the wall. There certainly doesn't seem to be any point in pursuing it now. We had some apps in mind but perhaps we have just been saved by the bell.
All I've seen of Vulcan is this demo on eweek. Vulcan looks impressive even in that demo and I concur that this represents a huge opportunity for IBM but I am left wondering if my company really needs it.
I do too. It just is so contrary to the current strategy that I fear casualties again. And the rich client was a comfortable niche where IBM could well survive a few years without innovation (like they did with R7). The browser is a much more open space where, once you get beyond authentication, you can combine many technologies, products and services. You will either have to do some very ugly things to lock your customers or you have to be really good. The potential to this new model is huge and the barrier for new customers much lower. I start to become excited but IBM has been playing the drums for so many years that there are customers that are in need for complicated fat and heavy software that I started to believe that those really exist.
Why couldn't things co-exist? If XPages is the primary driving force here then the fat client could obviously run things. But simultaneously there could be a "light" (basic?) client -- or just the browser -- that runs Vulcan. No support for legacy Domino views, forms, etc.
Keep in mind that you still have servers, too, user licenses, etc. There's plenty of opportunities for IBM to make money.
Plugins are slightly useful, but personally I wouldn't care if they junked composite apps. I view that as the last bit of nasty junk they created during the Dark Years.
@Erik
Of course things could co-exist and I expect this to happen. Currently you are forced to use the fat client because the basic client is not allowed to innovate. If you give choice I am skeptical that the fat client, as it currently is, has a chance to survive. The current client has just moved too far and with its new role as an XPages container IBM themselves is fading it out.
Now if IBM would have taken the old client, made it prettier, implemented full Javascript and CSS support years ago (that all existed in parts but never got completed). Sometimes I think it would been much easier to bring a full working browser into the old client than now to find ways to move old Notes components into the browser.
XPages actually run in an instance of Mozilla XULRunner, a lightweight wrapper around the Gecko engine that powers Firefox. If things are truly going toward XPages the leap to the browser is a short one.
Most Notes shops are mail-only. A browser client might be all they ever need. And you can add in more services on the the server instead of messy plugins and composite apps.
IBM says they want to do Notes, browser and mobile. For many customers, Notes won't be necessary. Which then makes even Domino expendable.
@Henning - I definitely agree. The current client has splintered too far, too fast as a result of the "catch up" that needed to be played. I wouldn't be shocked to see pieces of it phased out, or at least non-futured, like layers, outlines, layout regions, (and in a way, Lotusscript) etc. have been.
@Vowe - I wouldn't say Domino would become expendable, rather that its role would change. *Something* would still have to handle XPages delivery, authentication, etc. and it would make sense to continue that on Domino since that's where all the work has been. And you'd definitely still need Designer. The template and replication models would still be relevant and useful, and NSF is actually a *very* good container for Vulcan-style activities, especially considering the built-in security model.
I could see a scenario where Domino installs with a "lightweight" mode, similar to a Foundations server, where a bunch of the old server tasks aren't running and the NAB is lean-and-mean.
Domino's role could change, but I bet it will still be around.
Certainly you would need a server. But it could be based on different middleware. :-)
@Vowe - Ah, I think I misunderstood - you mean Domino would be expendable for the customer.
I don't necessarily disagree there, and in fact I think long-term it might be the case that the majority of customers end up running things in the cloud anyway. Would Domino be "gone" then? I guess that depends on how you look at it.
Regarding "Most Notes shops are mail-only. A browser client might be all they ever need." They have a browser client today -- how is Vulcan making things different there?
I'll be curious to see how this plays out in SMB. SMB will eat up something like Vulcan, especially if the server is either in the cloud or installs very easily. The won't want composite apps, plugins, or legacy Notes dev, but they will want all the portfolio pieces and XPages.
@Vowe - I have long said that Domino will eventually be absorbed by Websphere. I keep getting told I'm wrong, but I keep seeing it happening. As Domino's services become more loosely coupled it makes it easier to move those services to other platforms. It will be interesting to see how it gets spun as it progresses.
@Erik - I have worked at SMB's since 1995 and I can tell you from experience there is a lot of comfort in the control from owning something end to end, which makes the cloud is a scary place for them. With Blackberry, Google and Amazon outages happening regularly it's going to take a while before the cloud becomes stable enough for SMB's to fully embrace it.
"I've been trying to find formal training for Notes/Eclipse plug-ins and composite apps. When my training people contacted IBM about it earlier this month they were told I am only the second person in the UK to request it. "
Amazing. But then again, not. The Eclipse RCP was obviously such a stupid idea from the get-go; it made me realise that if that was how they saw a future for Notes, then it truly was dead. I jumped ship and stopped using Notes for anything other than an email client.
R.I.P. Notes
@Bernard,
well if you view the Notes Client only as an vehicle for Email then you may be right, otoh the Eclipse version contains Sametime, Symphony, links into Lotus Connections and other products.
i.e. I'v got a plugin in my Notes Client for my company voice mail, very slick.
Could all of this have been implemented in different technology (such as the Fat Client)?
Probably, but as things are the Notes RCP Client investment
DID expand the reach of the Notes Client helping to reinvigorate the platform and infuse new technology.
Did the market embrace Workplace/Expeditor?
Hell no, but a lot of its technology is now very commonplace.
(Osgi, Eclipse eg)
If you broaden your perspective to other Eclipse RCP based tools, then hell no you are completly wrong.
The Eclipse Rich Client Platform was a brilliant idea!
Take a look at http://en.wikipedia.org/wiki/List_of_Eclipse-based_software,http://www.eclipse.org/community/rcpcp.php and http://www.eclipse.org/community/rcpos.php
