Ed Brill

Post a Comment

1. 1  Craig Schumann http://governancefornotes.com | 12/2/2007 10:23:55 AM

   Beyond what I have said on those other threads, I would like to make a point about "RAD". You can do rapid application development in Eclipse, Java, and Composite Apps. RAD isn't about the platform, its about the tools. Domino Designer is a very capable RAD tool. The problem is how RAD is interpreted and used by developers today. It's no longer about quickly developing applications and functionality, and more about being an excuse for being lazy. "I don't need to test because I'm using a RAD tool." or "I can get away with developing in production because Designer is a RAD tool." are things we hear all to often when we talk with prospects.

   I don't think anyone can argue that there are no problems with Designer... but it's not the only thing with problems...

1. 2  David DeWell http://workdomosphere.blogspot.com | 12/2/2007 10:25:57 AM

   I used to think this problem was the lack of individual drive to learn these new technologies... But now as I have worked with all of this new stuff, my main irritations are the following.

   - documentation. - rather lacking these days. And I am not talking about how-tos but more documentation of the code for sdks. Its there - just very limited.

   - tools - my biggest complaint is RAD in this case. I have personally sworn off RAD until someone forces it on me or ibm makes it work all the time with everything

   Other than that, I would complain about marketing. Marketing to developer (and admins) has been an issue even before 8. I go to development sessions for other companies so I can keep my integration skills up but you would be surprised how many sales come out of sessions like these. So you get developers with current knowledge and get new companies interested.

1. 3  LongLiveLotus  | 12/2/2007 10:36:32 AM

   Sounds like IBM have got the message (I broadly agree with the chaps that are blogging for improvements btw) and I'm hoping a few things -

   1. That what is being said about a development focus being needed is no surprise to IBM

   2. Bob Balaban is working his ass off in this space..

   3. Announcements are coming at a major IBM conference in January ;-)

1. 4  Nathan T. Freeman http://nathan.lotus911.com | 12/2/2007 11:14:03 AM

   @2 - "RAD" in your case meaning Rational, right?

1. 5  Bruce Elgort http://ideajam.net | 12/2/2007 11:16:46 AM

   @3

   I hope "Announcements are coming at a major IBM conference in January" translates in to deliverables in 1H of 08. We can only hope.

1. 6  Peter Presnell  | 12/2/2007 12:22:22 PM

   As a long-time Notes developer I am frustrated that Notes 8 was yet another release of Lotus Notes with very little for the Notes application developer. But on this one ocassion I think IBM was justified in focusing its attention on getting the UI right. If it didn't there was a real risk Notes could have gone the same way as Lotus 1-2-3 (anyone remember that product that once dominated the spreadsheet market?). How many Notes developers do we have in shops that run Microsoft Outlook/Exchange for mail???

   But having made the improvements it has, IBM now need to turn its attention to the developers who will create the applications that separates Notes from other products. The non-Notes development world has moved to OOP and SOA. We don't just need a new eclipse IDE, we also need a major overall of the LotusScript language that encourages develoeprs to wrtite better code that can exploit these technologies. I don't want Java. If want a more .Net-like LotusScript language that will allow me to get real applications up and running in hours but exploits the power of things such as Composite applications so that my applications can become a series of components that my users can assemble as it best suits then. When I can do that with the same efficiency I can develop traditional Notes apps then Notes will practically own the small-medium application space in companies with the foresight to have stuck with Notes (or made the switch back)...

1. 7  John Head http://www.johndavidhead.com | 12/2/2007 1:20:30 PM

   @6 As much as I am a LotusScript developer .. the battle lines between Java and .Net are drawn. IBM is a supporter of Java. And while I do not expect IBM to force us to use Java for client side development (I know Nathan, CA being the exception) ... I do not expect to see LotusScript 'get overhaul'. If anything, I am hoping we see IBM push javascript and other languages that are well known and very much the default language for Web 2.0 and the mashup world. I would very much like to see JavaScript run 100% in the client and have access to the entire UI ... and then take that and push it to Symphony. That would allow millions of developers to pick up Notes and Symphony and being writing cool applications and solutions immediately.

1. 8  Peter Wilson  | 12/2/2007 3:38:54 PM

   More for developers is certainly needed. However, don't lift your foot off the accelerator on further UI enhancements and performance improvements. They need to keep coming for 8.X, 9.X, 10.X...!

   Pete

1. 9  Mike Brown  | 12/2/2007 4:46:40 PM

   @8

   Couldn't agree more. As a developer, sure I'd like lots of whizzy new stuff to make my life easier (hint: ability to extend standard LotusScript classes, please!).

   But what use will any of that be to us developers if idiot CIOs keep dumping ND because "users think Outlook looks nicer"?

   Cheers,

   - Mike

1. 10  Alan Smith  | 12/2/2007 4:49:55 PM

   And this is the main root for the reason why, I personally, left IBM (Originally Lotus/IBM Software Group) for a Microsoft Consultancy (Avanade).

   IBM has forgotten the Lotus Domino developer community, and also the platform. Microsoft are very much pushing, developing for and enhancing their competing technologies.

1. 11  Alan Smith  | 12/2/2007 4:52:08 PM

   Damnit, wish I could edit :)

   ... Also, the .net platform is an easier transition for the Domino/LS developer than the WSAD environment.

1. 12  Mike Brown  | 12/2/2007 5:51:26 PM

   @ 10/11

   I think I'll take IBM's neglect over Microsoft's royally screwing over its own developers, time after time; not to mention tying them hands-and-feet to a single platform that has had its day.

   Cheers,

   - Mike

1. 13  David Gursky  | 12/2/2007 7:14:36 PM

   Let me begin by saying the Technologist in me is lock step with TIm, Tommy, Nathan, Bob, et. al. As an IDE, Domino simply rocks, and the object based model of the Domino-based database is so much nicer to deal with then the row/column model of SQL databasess

   Now that being said, developers do not buy Domino. Enterprises large and small buy Domino, and business, regardless of size, are not really interested in (re)inventing the wheel.

   Ten years ago, Notes/Domino offered something not on the market -- a strong security model available st a much lower price point. At that time, Domino was a viable option for any of a number of lightweight office automation applications that simply did not exist. Why fill out a training request by hand, copy it, send the original to your boss, who in turn copied it and sent the original somewhere else, when you could have a document in a Notes database that could be stamped with a digital signature and not only routed instantly, but the status made known in real time? This was a no brainer, and companies bought into Domino in droves.

   That model does not apply today. There are Internet-based security models that, while not up to the elegence of Domino, are good enough. Another Ed of great repute (Ed Yourdon) put out a specialty magazine at one time called "American Programmer", and at that time, correctly predicted the threat to US technologists from offshore resources (although I want to say he pegged Brazil, not India as the source of this). So today we have lots of (relatively) inexpensive help available to enterprises that want/need to develop custom applications using tools like php, Python, Ruby, and so on. If they want something REALLY robust, IBM is not going to sell them Domino, they are going to sell them MQ.

   Don't get me wrong. Domino is a great product. I wish IBM would hire a bunch of old Loti to develop a backend office solution to do things like payroll, timekeeping, expense reporting, and so on, but guys, I just don't see it happening.

1. 14  Charles Robinson http://cubert-codepoet.blogspot.com | 12/2/2007 8:49:54 PM

   This meme is repeated rather frequently, and with the same results. The only difference I see so far is there aren't any zealous fanboys jumping to IBM's defense.

1. 15  Andrew Pollack http://www.secondsignal.com | 12/2/2007 8:51:16 PM

   I used to have to fight the "Notes is dead" crowd, but now I seem to have to fight the "You've Killed My Notes" crowd. It’s this weird persecution complex that just won't go away. Hell, even Mr. Lyons (a shudder to invoke his name) recognizes that Notes isn't being killed off. It’s time to move forward.

   Everyone is whining because here comes this insanely powerful new platform -- the Eclipse based client, and nobody can seem to make it do anything yet without a huge amount of effort. It’s a shock to the system for Domino developers. Never before -- not in any release (and I've been certified since version 2) has a ton of new, powerful, sexy features become available in the CLIENT before the DESIGNER was ready to expose them at a high level of abstraction.

   Designer is long in the tooth. Ed says as much in his blog today. Hell, EVERYONE at IBM agrees that's true. It got neglected while IBM made this horrible journey following the false prophets (industry analysts) who proclaimed J2EE as the savior of the industry with record setting consulting revenue for all. Obviously that failed. IBM woke up finally. Well, not so much woke up as was bitch-slapped repeatedly by its own customer base until it grudgingly got out of bed. Regardless, it happened, and major effort is clearly underway on the Notes side.

   The change though, happened in the Client first. Why? The client was changed first imply because that is where the change was needed the most. Customers buy Mail seats (@13 – Mr. Gursky hit this on the head perfectly) and then purchase needed development tools and developers as needed to back up that purchasing decision. In addition, long in the tooth they may be, but the tools DO WORK.

   So today in Notes 8 we have the most powerful new client platform for delivering better looking applications since version 4 added Lotuscript to what was a glorified form tool. The problem is, all that power is only exposed at what amounts to the modern day equivalent of the API level. Sure, it’s not C function calls like you think about in an API, but the level of complexity is roughly on that par by today’s standards.

   What needs to come next – what is coming next if you’ve paid ANY attention whatsoever to the press and the early previews done at Lotusphere 2007 – are the high level tools we’re used to seeing in Designer with the ability to work with, create, manage, and build for these new client capabilities. Personally, I think if you pay attention to the things Bob, Mary Beth, Maureen, and others are talking about, asking about you’ll see that those things are getting a huge focus now.

   If I could set some key strategy for the Notes & Domino platform long term it would be this:

   #1 – Obviously, get a Designer out that Lives in Eclipse. It’s perfectly ok on the DESIGNER side to totally change the look and feel all at once. As long as what it works on doesn’t change, developers are much more tolerant than end users about the way you get there and frankly, the Eclipse development model is much better anyway. Once we’re there, we’ll get real java, javascript, css, html, and maybe even Lotuscript editors with all the modern features you expect.

   #2 – Truly split the Client, Designer, Admin, and Server into totally distinct products which can run side by side or not, and have their own version cycles. That means making the NSF ODS version be the key point of compatibility (e.g. you need a version 8 server to show a version 8 ods based NSF, and a version 8 designer to work with version 8 ods design features, etc.)

   Why split? The product is now too big to be maintained as a single product. By truly splitting into distinct development streams (which internally might have very different splits) you can focus each product on a revision cycle designed to enhance that product, while also upgrading to maintain compatibility with the next ODS version. Obviously the Designer and Client will always be closely coupled from a capability perspective, and the same of course holds true for the Admin client and the Server.

   By the way, @6 -- John, the battle between Java and .NET is so beside the point. What Microsoft fully understands is that the only real selling value of a development environment is that it drives dependence on and sales of your core server software and enterprise platform licenses. VS.NET is about selling the whole Microsoft core suite, not about just being a good development environment. There’s no money in building development tools for their own sake. Ask Borland.

   J2EE fails for the same reason. Oh, sure it’s good for some things. Huge amounts of money were poured into it to make sure it was. Java sure didn’t start out that way. Nevertheless, what J2EE failed to do was sell a real platform strategy. By being relentlessly cross platform (in theory), companies couldn’t use it to drive specific server and seat license offerings. Sun spent fortunes on lawyers to prevent anyone (specifically Microsoft) from using an “Embrace and extend” approach. Yet, do you really see anyone creating in house J2EE applications and saying, great, now I can deploy this on Tomcat and WAS tomorrow! It doesn’t really work that way in house. The “Cross Platform” argument was bogus from the start. Meanwhile, J2EE is every bit as hard to write as native compiled C++. It’s too granular and expensive for most in house work. IBM’s response was “Portal”. You don’t have to build it from the ground up, instead you can use Portlets. Only Portlets didn’t do anything by themselves, so IBM decided to tie Portlets together in a framework called “Workplace”. By the time you go there, you were so many layers of complexity deep that it was always doomed to scale and complexity problems. In my considered opinion, it crumbled under the weight of its own relentless insistence on following mythical standards that nobody ever really successfully integrated with outside the scope of the application anyway.

   So no, John, it’s not about .NET vs J2EE – it is still, and always was, about Exchange+SQL Server+Sharepoint+vs.NET vs. Lotus Notes + Domino + Designer. IBM just forgot that for a while.

1. 16  Ed Brill http://www.edbrill.com | 12/2/2007 9:16:50 PM

   @14 I didn't blog this just to repeat an old, tired meme, even if some of the guys I linked to thought they were just venting :)

1. 17  David DeWell http://workdomosphere.blogspot.com | 12/2/2007 10:27:28 PM

   @4 - Yes I mean that bloody piece of horse crap Rational Application Developer. RAD as a concept is over-used quite honestly. Maybe with the right project manager it can truly be RAD.

   RAD is not RAD with a RAD PM.

   That being said - I am at the end of this thread I think. It seems to have gotten a bit muddy now. So I will part ways with this thread now.

1. 18  Patrick Kwinten http://quintessens.wordpress.com | 12/3/2007 2:08:51 AM

   "it's high time that IBM does something for" I guess in the beginning ofthe year 2000 it was already time that IBM should have done more for the LN/D community.

   what should them move this time?

   when looking at the ideas on IdeaJam I see a lot of 'non logic' or missing or 'out dated' designer options that need to be changed or should have been implemented during the years.

   my advice would be, grab the domino designer by it's horns and do a rebuild from scratch (if not too much asked).

   and please remove those 'java applets' and provide us with web 2.0 solutions!

1. 19  Nathan T. Freeman http://nathan.lotus911.com | 12/3/2007 5:26:28 AM

   @15 - Your point number 2 would be an awesome idea. I think it has about a 0.0 percent chance of happening, though. Maybe if Designer was moved to the Rational group, it might have a completely different release cycle. I think most Domino developers find that prospect to be absolutely horrifying, though.

   As an aside, one of the things I hear from time to time is that Designer gets ignored because it's not a profit center. IBM charges for Designer in order to justify staffing it.

   But really, Designer's revenue stream should be the incremental difference between a Notes messaging-only license, and a Notes collaboration license. If the full collaboration license is $10 more, every dime of that revenue comes from the fact that Notes is a development platform, instead of just a messaging platform. And all the development platform value comes from Designer.

   Just a thought.

   @18 - Patrick, what do you think Designer-in-Eclipse is? If the Notes 8.0 client is a re-imagining of the Notes client, then DDE is certainly a re-imagining of Designer.

   And yeah, it's not secret that the web 2.0 stuff is coming. Right now the battle is not over WHETHER it comes, but what IT is. Web 2.0 is not in itself anything, is it?

1. 20  Stuart McIntyre http://collaborationmatters.com | 12/3/2007 7:09:19 AM

   @19 "But really, Designer's revenue stream should be the incremental difference between a Notes messaging-only license, and a Notes collaboration license. If the full collaboration license is $10 more, every dime of that revenue comes from the fact that Notes is a development platform, instead of just a messaging platform. And all the development platform value comes from Designer."

   Bang on the money, Nathan. Or I think you could even argue that Domino Designer should just be included in every Domino licence.

   This has long been one of my biggest bugbears, that the sheer power of Domino is excluded from the users/customers that most need to see it demonstrated. We manage to convince an SMB to choose Domino/Notes over Exchange/Outlook so they buy Domino messaging licences as they "only need email right now". Then we cannot help them demonstrate the value in using Domino as an appdev platform as they aren't licenced for designer, and cannot justify the expense of Collaboration licences until they know they will make use of the extra functionality...

1. 21  Andrew Pollack http://www.secondsignal.com | 12/3/2007 7:47:59 AM

   @20 - I hate having to take this position, but in the end I'm forced to recognize that for a great many larger shops, the ability to create code by end users is seen as a big negative. It doesn't create security problems per se, but it exacerbates any that exist and can work against the central IT plan (not always a bad thing, except for Central IT who is paying the money).

   Designer is easily available from demo sites and so on so not distributing to users doesn't really solve the problem. Personally, I'd like to see designer be licensed as part of the collaboration client -- but distributed as a distinct installation kit.

   Further, I sadly find myself in favor of a flag on certificates within the ID file that prevents their use with design elements. In other words, when a certificate is issued to a user, the certificate authority can specify that the certificate may not be used to sign design elements, thus making any code created by users with those ID files run only at the level of "anonymous" (effectively making them nearly useless for development).

1. 22  Mike McP http://www.openntf.org/mPortal | 12/3/2007 8:00:26 AM

   I don't mind a release targeted at end users. End user happiness is what gets apps into the door and keeps them there..it's the horse that comes before the my cart. The UI has been getting a bad rap for so long that it needed a major overhaul, and if it waited much longer most of us developers would be working on other platforms. Just MHO.

   take care,

   Mike

1. 23  Nathan T. Freeman http://nathan.lotus911.com | 12/3/2007 8:01:59 AM

   @21 - "Further, I sadly find myself in favor of a flag on certificates within the ID file that prevents their use with design elements."

   Why sad? I like that idea.

   Actually, I might want to federate some simple agent & view stuff, as well.

   But I'm not advocating end-users getting Designer. Just that the incremental difference between a collab license and a messaging license should be considered to be Designer-generated revenue. I don't think those people should therefore GET Designer.

   As far as some random person just downloading Designer and working on it themselves, I think that's up to a given customer's policy.

   My problem is that as long as IBM charges for Designer, then the Designer revenue stream is considered to be whatever comes in for that particular license. And therefore that team gets a tiny fraction of the resources it deserves. That should change.

1. 24  Andrew Pollack http://www.secondsignal.com | 12/3/2007 8:06:37 AM

   @23 - you're dead on when you say designer's revenue is under and misrepresented when taken as strictly the sales of designer.

   I had never considered that a collaborative license rather than a mail seat was itself proof of designer revenue, but the logic is nearly flawless. Designer, and nothing but designer, drives the price difference and the value difference between a collaborative seat and a mail user seat (especially once you take out quick'r, sametime, etc.. which have their own license streams anyway).

1. 25  Randy Smith http://www.certprimer.com | 12/3/2007 9:48:44 AM

   I think that over the past several years that IBM has abandoned their interest in the Lotus certified community. Both developers and administrators. What a shame, since this was a great opportunity to reach to developers and administrators "working in the trenches". When Grandma Bowen was running the Lotus Certification Program, we knew that there was someone at IBM that was looking out for our best interests. She went to bat for us, making sure that the Lotusphere conference team give us exclusive perks to acknowledge our contribution and loyalty to the brand. Now, those Lotusphere perks are slowly being taken away from us (CLP Discount, CLP Breakfast, certification bootcamps, etc.). Also, the CLP Private forum, which was once a great source for technical, as well as certification-related information, has been dumped into to the IBM Certification Private Site. Have you taken a look at the discussion forum on this site lately? It's little more than a resource for exchanging illegal braindumps so that people can cheat to get their certifications.

1. 26  Axel  | 12/3/2007 10:12:22 AM

   @16: Why j2ee bashing? Look on any job site and you will see that there are lots of demands for Java/J2EE skills. Companies are using it. Are they all fooled by evil consultants or pointy haired bosses? I seem to have more trust in the fact that in the end the more effective solution thrive.

   I allways saw this java thing as a longer term and steady learning process. More like investment. And especially since Notes 7, there have been good opportunities which provide real business value to my customers. Another advantage as a developer that I can participate in projects which use bea weblogic, swing, tomcat, websphere, eclipse rcp, etc AND Lotus Domino projects.

   I also find that in the real world that RAD (which does not mean Rational here) thing is much more ambigous and double edged than it sounds in theoretical debates like this. In Java-land there is more discussion about upfront design & architecture, but I am more than happy that the domino guys of my current employer do have the experience and will to invest design & architectural thoughts in our notes databasese. On the other hand all those frameworks, products and tools have made j2ee much more RAD.

1. 27  Dan Sickles  | 12/3/2007 10:15:43 AM

   @7 - See session AD208 - which seems to have disappeared from the view - the link still works:

   https://www-926.ibm.com/events/lotus/lsph2008.nsf/sessionabstract?openform&sessionid=AD208

1. 28  Kevin Mort  | 12/3/2007 10:44:23 AM

   I have always seen ease of development as a huge benefit to the platform. With the use of Eclipse, composite apps etc, I think the danger is that development becomes quite a bit more complex, and thus left only to those who really are focused on development.

   Now, I can't say for sure that this is or will be true, it is however, still something to watch.

   I remember during the Notes 8 beta some of the suggestions on how to change a setting in the mail db involved you needing to go out to the OS into some seven levels of hellatious directories to find a css file to change. Right there I said "whoa!" That was a change not really expected, and might have been a beta thing.

   With that, I believe the message I am trying to get to is that there are various levels of development complexity that need to be considered.

1. 29  Paul Robichaux http://www.robichaux.net/blog | 12/3/2007 12:20:28 PM

   @15: beware #2! Once upon a time, the Exchange client was part of the Exchange team. Then came Outlook. While the Exchange and Outlook teams work hard to keep things in sync, the fact is that they're part of two separate product groups with two separate revenue streams. MS spends an awful lot of time, effort, and money keeping those two teams in sync. The benefits of the split are obvious but getting there is non-trivial.

1. 30  Leo St-Jacques http://www.healthcanada.ca | 12/3/2007 12:27:27 PM

   I am glad to see others feel the same way as I do regarding the lack of progress of the Designer client, little improvement for Notes Domino developers and no real changes to Notes UI for Notes applications.

   I brought up my concerns during the Managed Beta Program training session in Dallas last December. I wanted to see what IBM produced that would extend the experience of our Notes applications for our users. Here at Health Canada we have over 2,000 Notes applications.

   Unfortunately what I got was ND8 Mail and Calendaring overview, ND8 admin server topics, and a 23+slide lesson on how to install, configure and administrate Portal server just to see how Activities work (sigh – too bad Activities is not Domino-driven). It is great that Eclipse is being used with Notes 8 client to enhance the UI, but all of the Notes-based C++ UI is still the same old R5 design, like dialog boxes and @Prompts.

   And we found out that Notes applications in Notes 8 can not fully act on a consumed web service. For example, we need to build an Eclipse plug-in to display data in a data browser format (rows and columns) that was retrieved from a web service. Sigh….

   For me, IBM did a great disservice to Notes Domino and us developers when they decided not to incorporate J2EE into the product. Instead we got two confusing product lines with similar features, which is still confusing today. Example: Quickr 8 (no localization support for Domino version but exist in J2EE version - what happened here??).

   IBM is finally on the right track. I just hope it is not too late.

1. 31  Andrew Pollack http://www.secondsignal.com | 12/3/2007 4:45:14 PM

   @26 - Why bash J2EE? Because its a solution in search of a business model. Ok, as an architecture the in-memory back end objects add a lot of value in terms of ability to process transactions in volumes way over what a web server might otherwise hope for. That is the one thing it really brings to the table, and 99% of the rest of just overhead. Java failed at the desktop, and at the browser client side. It failed completely. J2EE was invented as a way to keep Java useful. It worked, I suppose, but the things that make it work have nothing to do with it being Java, nothing to do with cross platform (as I said, nobody is building apps first and then deciding what platform to deploy them on), and its only revenue stream to vendors is services -- and that's due to the extreme complexity of the server platforms and the deployment requirements.

   Sure, there are lots of jobs for you if you learn it. Good pay too -- because the complexity limits the competition, and also because it takes a metric butt-load of code to get anything done. How is that good for the customer?

1. 32  John Rowland  | 12/4/2007 3:13:21 AM

   Tom Yager -- no MS lapdog -- makes comments about the release of VS2008. It has some interesting comments about how MS courts developers. Especially notes are the efforts to draw developers into Company-sponsored blog communities. { Link }

1. 33  Nathan T. Freeman http://nathan.lotus911.com | 12/4/2007 6:27:42 AM

   @31 - "nobody is building apps first and then deciding what platform to deploy them on"

   No, but they are deploying them on a platform, then deciding later to redeploy them on a different platform. Which is just as valid a reason to pick a flexible architecture.

1. 34  Andrew Pollack http://www.secondsignal.com | 12/4/2007 6:46:29 AM

   @33 - I just don't see that (redeployment of J2EE apps on competing server platforms) happening much with the exception of using the open source platforms as testbeds or startup platforms with the long term plan in mind for possible upgrade. The numbers are hard to get, but I'd be very surprised to see any kind of volume in the enterprise j2ee server market of customers moving from one platform to another.

1. 35  Ed Brill http://www.edbrill.com | 12/4/2007 7:44:36 AM

   @32 thanks - some good ideas in there

1. 36  Kerr  | 12/4/2007 8:43:12 AM

   @34. It's true that you are not going to see an installed app being moved, lock stock, from one platform to another. While that is often touted as a reason to use java it's mostly a red herring. But that doesn't make the JVM a bad idea and it doesn't mean that platform independence isn't a feature that is used every day, by many many developers.

   What you do see is people using skills, techniques, tooling, libraries and frameworks that are automatically available on a huge range of platforms. There are extreme edge cases were something doesn't work, but it's really not that often and is normally where someone is relying on behaviour from one JVM that is not part of the spec.

   Every implementation of a java enterprise stack grows the market and encorages competition. EJB<3 is generally considered to be a miss step, but the competitive nature of the platform meant that a better solution sprang up and was taken on board, removing complexity. This came from the java enterprise community rather from the classic spec designers.

   There are a lot of things wrong with java, but there is an awful lot good in there as well and it doesn't have to be complicated. Anyone know when WAS 7 is out ;)

1. 37  Andrew Pollack http://www.thenorth.com/apblog | 12/4/2007 9:40:04 PM

   @36 - Sure, Java is proof that if you put enough effort and money into something, you can turn crap into a useful tool. It becomes useful because people are using it. There are libraries out there because there are libraries out there.

   I work with some C and C++ code that has a Makefile which compiles under Win32, OSX, Linux (32 and 64 bit), BSD, Solaris, etc.

   Sure, the JVM is an interesting idea. Its the same idea that separating hardware from code via BIOS was. Its the same idea that separating BIOS from code with OS was. its another level of abstraction as operating system complexity has increased.

   Domino does the same thing. When you write for Domino, you're writing for a cross-platform, sandboxed, secured, run time environment. It just happens to offer a level of abstraction far deeper (aka with a much higher level of simplicity) than either the CLR or the JVM.

1. 38  Jimmy Bracco http://www.lotus911.com | 12/4/2007 10:12:36 PM

   Ummmm,

   hello???

   parallel streams of development mean anything to anyone?

   If I ran my business by concentrating on one area 5 years after there is a problem there, and then move towards the next problem area 5 years later, I would be in the Landfill business, not Collaboration software.