Remember last week's momentary
posting of a book on microsoft.com
about migrating
Notes applications to .NET?
Well, on the assumption that the book lives on, in more discreet
distribution channels, Tom
Duff has written a detailed
article for e-Pro examining the Microsoft book. Duff takes specific
quotes from the book and examines them with a critical eye, and finds that
the story isn't so simple as the 150+ page book would lead the reader to
believe.
Full disclosure -- Tom asked me to review a draft of his article prior
to publication. The final version has some interesting, well-researched
points that I didn't see in that draft, such as:
I'd feel more confident of Microsoft's migration strategy if there were more actual case studies of business successfully implementing this type of change. A highly visible conversion effort is underway at Accenture. This conversion is being managed by a group called Avanade, which is a joint venture between Accenture and Microsoft. Material on Avanade's Web site states that Accenture's application portfolio has been trimmed from 16,000 to 7000 applications. It could be assumed that these are the category 1, 2, and 3 applications that Microsoft converts using this approach. If you assume a three-year conversion period for those remaining applications (2004, 2005, and 2006), you'd have to convert 45 applications a week. You can continue to extrapolate the numbers from there to figure the number of people you'd need to meet this schedule. It appears that their solution to this is to offshore the process to India where you can add lower cost labor to the project and use a higher number of programmers. If that occurs, where do the offshore programmers get the business knowledge inherent in the applications?And with a long life ahead for Notes/Domino, what would be the impetus, anyway?
Any way you look at it, a migration of applications from Notes to .NET is not an easy endeavor and will cost your organization a lot in terms of money, time, effort, and lost opportunity cost when resources are not available for new projects to improve the business.
Link: e-Pro: Converting from Notes to .NET, according to Microsoft >
Post a Comment
- 2
Sagar | 9/3/2004 10:42:49 AM
Same thing is happening at E & Y. They are migrating from Notes to .Net. There is another very very big consulting company like accenture is moving away from Notes in favour of Documentum. Ed might have guessed the company name!
Notes can not handle large content Mgmt requirement which many large consulting companies are facing today so they are moving to more robust content Mgmt platform like Documentum,Vignette,FatWire,DB2 Cm etc.
IBM got ILWCM (aptrix) for us but the truth is ...it sucks .. lots of shortcoming/limitations and poor integration with WPS. DB2 CM has little potential but its too expensive.
- 3
Ed Brill www.edbrill.com | 9/3/2004 2:25:40 PM
I am not sure that it is accurate to say that E&Y is migrating from Notes to .NET.
- 4
Ben Poole http://www.benpoole.com | 9/3/2004 3:44:56 PM
Sagar is talking about one facet of technology here though, to wit: content management. It's probably fair to pick up on the fact that many organisations are moving away from Notes & Domino for content management aimed at the web -- I know my organisation is -- but that doesn't tell the whole story does it? CM is not what Notes & Doino are about, it's just something they can do.
- 5
Michael | 9/3/2004 4:39:01 PM
Documentum and Vignette certainly don't play on the same zone...but they cost A LOT and time to get them running in ENORMOUSLY bigger than the Domino equivalent. If you have millions of documents, you can consider theses stuff, otherwise, the costs associated to them + the flexibility / adaptability you loose compare to Domino is a show stopper.
Not to mention that every ppl having an exp with Documentum consider it's a nightmare to install, maintain and develop with...
- 6
Heini | 9/4/2004 4:54:19 AM
Is it really a good idea to propagate how hard it is to migrate Domino applications (isn't this one of the reasons MS is always bashed for)?
Isn`t this called Vendor lock-in?
Domino is an excellent platform and there are better reasons than this one.
And Notes might not be able to handle as much data as Documentum but it can handle more data than 5 years before (and for Windows and AIX you will have DB/2 quite soon).
And a company that needs 16.000 applications to run their business (even 7000 is a lot) might have bigger problems than Notes or .net. If I read all the comments (from Notes to Oracle, Documentum, Vignette ...)
I really think that I know of a powerful platform. I am quite surprised that there is so little talk about Websphere / Workplace in this context. Why are companies considering .net when Websphere / Workplace / DB/2 <insert your module here> does what they all complaign about?
- 7
Gaston http://www.dblookup.info | 9/4/2004 6:51:47 AM
I agree with Heini, I don't think that telling that migrating away from Domino is a long and difficult process is a good thing to do. It might just make people afraid to come in...
On another part, I have also the feelings that migrating applications from .net or whatever system to domino will take roughly the same amount of time. Usually, problems are not really technical, but if it took your compagny five years to build a complete intranet, it is obvious that you won't change everything in six months.
Gaston
- 8
Brett | 9/4/2004 8:40:05 AM
Surely the real point here is that migration (in either direction and indeed from Notes to Websphere) is a complex process that is likely to take a long time and cost a lot of money and resources. It would be shortsighted and probabl irresponsible for any organisation to migrate from one environment to another to end up with very similar functionality unless there were compelling external drivers. Any organisation making migration sound simple is misleading you in most cases.
- 9
Ben Rose http://www.jaffacake.net/bensblog.nsf | 9/4/2004 1:16:47 PM
We have documentum and have many problems with it. Not sure if it's the product or the way it's been installed. It was certainly not specified correctly hardware wise which made the costs look much cheaper up front.
On the migration side of things, I never did understand why I can copy a contact from Notes and paste into Outlook, but not the other way...now that's vendor lock-in!
- 10
Nathan T. Freeman | 9/6/2004 3:41:11 AM
Heini makes an excellent point. Do I really want to take advice on IT management from a company boasting "16,000 applications?" I hadn't thought about that one before, and it really should jump out and bite you in the face. When you have one deployed, custom application for every ten employees, perhaps your corporate development strategy could stand a good long look.
I'd also be curious whether it would be feasible to build and globally deploy 16,000 applications in, what, 10 years?, on any other platform. We're talking 1600 apps a year, or, on a conservative estimate, 5 new apps deployed PER DAY! Think there were any standards in place or any code review for systems churned out at that rate?
- 11
| 9/6/2004 8:41:53 AM
If I start to get to know a larger organization and they have Domino, I nearly expect, that they have a corporate policy to move away from Domino towards J2EE or .NET.
This was the case in the 2 of 3 larger organizations I have started contact with in the last 12 month.
Domino is actually the most effective lock-in platform that I can imagine. This has to do with the non-layered architcture (UI, Database, Business logic) all mixed in one ("easy-to-build") salsa.
There is nothing to be proud of regarding this lock-in mess.
I wish Microsoft, Accenture and all other companies who are struggling with away-from-Domino conversion all the best in their difficult task.
If there are good conversion tools, companies might stop freezing new Domino development. This is what actually happens, after companies realize that Domino apps create a BIG lock-in.
... and the Marketing Manager who allways plays Mr. Nice Defender of the world against "evil" Microsoft does portray those lock-in effects as a good thing.
Axel
- 12
Nathan T. Freeman | 9/6/2004 9:00:11 AM
Yes, Domino is a "lock-in" platform in terms of non-switchable development. I hate to drop the big secret, but so is everything else. It's not like .NET doesn't lock-in corporate developers to one approach.
I've never been able to understand why people say Domino doesn't layer UI and database. Of course it does. If you've ever written a form formula, or used a computer subform, or written a condition in a view column, you've done exactly that. Just because they're stored in the same file facility (and that assumes you've never used a single-copy template, either) doesn't mean they aren't layered.
Business logic is another matter, but anyone who claims that .NET or J2EE layer business logic by their design is full of crap. It's layered by CODING PRACTICE, and it's not hard to do the same thing in Domino. I've been doing it since R4. Most shops don't, because they refused to take Domino development as a serious effort, and throw people with Excel training at it instead.
- 13
Ed Brill www.edbrill.com | 9/6/2004 9:11:51 AM
Axel - welcome back, I guess. You might have missed it that I'm actually in sales now, but whatever.
The point is, migration for migration's sake never makes sense for an IT organization. IBM offers Lotus customers a long-term roadmap for those Notes applications -- both stand-alone and/or integrated into other infrastructure (Portals, WebSphere applications, IBM Workplace client technology, etc.). In the next Domino release, a relational database will become an optional backend...making Notes/Domino data more accessible than ever. And Domino 7 will also allow application functionality to be exposed as Web Services. I don't see how you could label those kinds of efforts as "lock-in".
Anyway, Tom's point (if I may assert what I drew from his article) is that spending money to migrate existing apps from one platform to another is not a good use of IT investments. It is done sometimes in the context of strategic application platform adoption, or for other reasons -- but migration for migration's sake? Why? What's the real reaturn on investment?
- 14
| 9/7/2004 8:06:01 AM
I am hearing different reasons for migration.
(faster RDBMS connections, simpler development model, simpler browser based solutions so that people can work from home where replication is not an option and Notes access isn't allowed, better printing, all this stuff).
I am exagerating quite a bit. Domino was and is still a great plattform for what I call typical business type of functionality (workflow, email, security, integration to backend systems, etc.).
But others have significantly catched up.
Unfortunatedly this is an emotional business. So not all migration policies might be rational. Some from my pov, are (simpler development model, better rdbms integration, etc.).
DB2 as optional backend should make migration simpler, I agree. But I don't know enough about it, yet (is everything stored as simple data type and not as blob, etc.).
Webservices makes a system more open, but there are extra performance costs attached with it (this serialize/de-serialize stuff needs ressources).
For example from Websphere, Domino is not the simplest platform to integrate (can't include Domino in a 2-phase-commit transaction, I agree with Bob Balaban who says that people to "take care" when accessing Domino from Java. A lot of people simply doesn't care about .recycle().
I met an Avanade principal a couple years ago at an alumni reception (ironically, it was the evening of an all day Lotus event, so I was able to kill 2 birds with 1 stone). It was clear from our discussion that the move by Accenture was clearly a political move because of the Avanade joint venture with MS and that it had nothing to do with MS being a better technology. But you only pick these things up in one on one conversations, not in company puff pieces or web sites.