pmooney.net

1 - Here's another test for you to try (if you want). Create a new DB with NO template, one form, one view and one field. See which quadrant that ends up in. I'm guessing Quad4 (which is amusing because that DB might be able to be handled by Sharepoint).

2 - Great job, Paul!

I for one really appreciate the work you have put in to this!

3 - Hey Mr. Advocate

Thanks for your interest in this thread, but in this case I do disagree. I am not being unfair. In fact, I have tried my best to keep this as a technical evaluation of the tool at hand (like I did two times previously). MS staff have noted that my evaluations were fair so I think I did a good job. I know there is an overall approach and this tool is the first piece, but this is the tool that is going to be rammed down the throat of my customers, like it has been in the past, and in fact quite recently.

Lets look at the two specific jobs the Analyser does.
1. It tells you what applications are no longer in use.
This does not work. Applications in frequent use are being listed as inactive for over a year in the reports

2. Categorising the apps into quadrants.
The tool has a very simplistic approach to the analysis, which even surprised me. Results are easily manipulated. I agree that if you have one template with lots of databases based off it, and you are happy with the migration of that database, add it into the exclusion list. But, this is very rarely the case.

In relation to the database template name field being used, you write:

“Well, this would be a good point if you were trying to fool the system, but the assumption is not that you would run the Analyzer in a hostile environment where they were trying to fool it. But that is stupid. The people running this tool want to have it work. If they were not sure, they could always copy in the bog standard templates and run the test again, and that is where your discovery that even changing a single field would mess things up becomes a point in the tool's favor.”

I don’t think my assumption is stupid at all! In the competitive environment where this tool is run, the more “QUAD1/2 results, the better”. This is an MS marketing tool. They will be hostile. They want the answers to be positive and “easy” to migrate. In reality, they want the customer to go with Exchange/Outlook and eventually, maybe try to migrate the apps.

I did not want to say that in my evaluation above, because that is my “real world marketing” opinion, and has no place in a technical review.

Again, I am bias to Domino, but I do know Sharepoint/Exchange aswell (and would you believe, like them for what they are). This tool is limited, but my thoughts on it now are that it is unbelievably limited. I have very specific thoughts on how this tool could be written to do a detailed and good job for Microsoft, and it would not be at all like this one. If MS ever want to know, I may even tell them.

This tool is being sold as a big part of the plan. My analysis was technical, that’s all. I don’t believe in knee-jerk Anti-Microsoftism, and in fact, I don’t like it. They make some fine products. This is not one of them.

4 - I am starting to understand (after your comments and others offline) one of the reasons why this tool leaves a different impression with me. Most of my customers are either dedicated Notes/Domino users (Midas) or have already made the decision to move to Exchange/Outlook or to use it for some portion of their business (CoexLinks). I am not on the front lines fighting the wars with those who want to convince companies to migrate one way or the other. From the side of those already having to deal with Exchange/Outlook, this could be used the way I describe. Clearly, it is actually being used as a sales tool more than I recognized. From that point of view, I can understand why the simplistic result bothers you so much (aside from the obvious failings of databases being marked as not having been used when they really are). I still think that you are blaming the tool more than those using it, and that your technical review is colored someewhat by your understanding of how it is being misused. It is a bit like doing a review of Notes without acknowledging that it does anything beyond e-mail, I'm afraid. No matter how people misuse a product, a technical review should probably describe how well it does the job it is purported to do. Reviews of cars shouldn't focus on how well they run over people, although it is perfectly fair game to those who are writing opinion pieces. Technical reviews of weapons should probably not decry their use in crime, but again it is fair game for non-technical opinion pieces. Similarly, decrying this tool for how it will be misused seems a bit strong, even while it is appropriate to point out where it could be misused.

5 - I think some of the criticisms made by the review are slightly unfair but then I also think that for the MS tool to wholly fail to accurately identify the last time a database was used is truely dreadful particularly since it is presented within the Executive Summary - in fact, it really ought to be illegal for MS to do that. Whose fault is it if a company chooses to migrate and blows away a bunch of apps based on that?

Also, the example of a single additional field changing the complexity of migration from QUAD2 to QUAD4 speaks volumes where MS is clearly happy to obfuscate. How does MS get away with it?

Thanks Paul. That was good work.

6 - Well said Ben and fair points. I must say though that I tried to keep bias aside. If this tool did a good job, I would have said it. I am quite happy to point out what I dont like about any product, even Domino. I read the marketing blurb on the tool, and then sought to agree that it could do what they said. Thats all. Everything I looked at in this tool did not meet what is being said about it. If it did. honestly, I would have put that here, regardless. It looks great, and the documentation is excellent, but its functions are lacking.
But, thanks for a differnt perspective. It helps us all do better.
P.

7 - Nice work!

(I paid for the beers, BTW).


---* Bill

8 - I wanted to clarify, since the question has been asked. I had no part in the App Analyzer, have not helped in any way for any part of the Red Bull project (although a request was made to use my Midas technology, and I turned it down), and have not even had a chance to run the App Analyzer myself. The preceding comment was purely based on reading Paul's blog and the bit I have gleaned from the web about the App Analyzer.

9 - OK, I'm having just a bit of fun playing devil's advocate, although it is hard in this instance. In any case, I want to put on the other hat for a moment and ask whether you are not being a bit unfair. The Microsoft Application Analyser is supposed to analyze the required tasks to migrate these apps, or to recommend ones that should not be migrated. The Quadrant approach, which you focus on heavily, is not the only evaluation that matters. Certainly, I would take any application written for a different technology and rate the task of rewriting it as difficult, and that is as true going from Sharepoint to Notes as otherwise. The Quadrant approach is supposed to be a first pass. It may be that some large companies have hundreds of databases based directly off the standard Team Room or Document Library. For many companies, touching those templates is verboten, so it makes perfect sense to identify which ones will be "easy migrations".

But the place where I think you are slanting the purpose slightly is when you discuss the ability to modify the AARULES.XML file, and act as if the main reason is to make the results look better. Could I suggest an alternate suggestion. Imagine that the company was DominoDeveloper.net (as unlikely as that may be). They may have literally hundreds of blogs based on Blogsphere. So, part of the purpose of the Analyzer is to identify the task at hand. It may well be that by identifying Blogsphere as an application to be migrated, they can move all those hundreds of applications into the easy quadrant. It may even be the world's biggest bitch to migrate, pardon my French, but once it is done, migrating those applications is a breeze.

Now, you may not believe that that is the intent, but it certainly is what is described as the intent of the Analyzer from what I read, to start a transition plan. An obvious transition plan for DominoDeveloper.net would be to migrate the Blogsphere template, which is identified by the Analyzer as a common template. Some companiesuse fifteen thousand different templates, but a lot have about five really standard templates that account for a whole lot of business. Identifying those templates is not stupid.

Anyway, moving on. You make the comment "What if the template name is the same as the default template name, but it is acutally modified???? Looking up based on a text value in this database property is just not accurate enough for classification." Well, this would be a good point if you were trying to fool the system, but the assumption is not that you would run the Analyzer in a hostile environment where they were trying to fool it. But that is stupid. The people running this tool want to have it work. If they were not sure, they could always copy in the bog standard templates and run the test again, and that is where your discovery that even changing a single field would mess things up becomes a point in the tool's favor.

I could go on, but I want to make the point that while I appreciate the effort you have made, your bias for Domino and against the Analyzer is showing through. Is this a limited tool? Absolutely. Is it as limited as you suggest? That I doubt. Is it easy to migrate applications away from Notes/Domino? No way! Is every company who ever wrote a Notes application still using it? No way to that too. So, it is possible to migrate, and I think we do ourselves and others a disservice by acting like it is always impossible. Then, the first time somebody takes a widely used template and migrates it, and the hundreds of apps that use it, we look like chumps.

Applications can be written on many platforms with many technologies. Some are better suited than others to any particular task. From what I hear, Sharepoint does what it does pretty well. I know that Notes/Domino does what it does pretty well. This Analyzer is probably oversold as a bigger part of an automatic conversion tool than it really is, but it is still probably reasonably useful as part of a transition plan. It may even be that Ed is right and it will convince people that it is really difficult to migrate all those apps. I don't know, but I am growing a bit tired of knee jerk anti-Microsoftism. After all, it wasn't that long ago that I cursed IBM as the big bad meany, and now I live and die by their technologies. I just don't want to forget that this is business, not religion.

10 - Though my comment will look obvious to many, since I work for IBM/Lotus, I have to say : great job. Thorough analysis, hard facts, no fluff. Not that everyone expects anything less from you .
RoB

11 - This is obviously a marketing tool rather than a tech tool, and I do think your review is a little unfair in that it questions the tool's technical credibilty rather than on its ability to do what it was designed to do (ie: confuse and create FUD).

That said, MS probably deserve a little flack for this because they are attempting to position the tool as a technical one.

At any rate, I'm not sure whether automatic analysis is really a good way to go. I think MS would have a better migration tool if they simply created a survey sheet to ask questions of the IT Department. "eg: is this database based on any of the following templates...".

I am eagerly awaiting the IBM tool for analysing the MS Environment for people considering a migration to Domino. I'm particularly keen to see how it handles "MS Other".

12 - Another great effort, well-appreciated by those of us on the front lines. Thank you.

13 - I wonder if Todd Wanke sweats a little prior to posting the tool? Regardless, you have done an excellent job.

For those that want to know about Todd:
http://blogs.technet.com/collabtools/archive/2006/02/08/419038.aspx