Sam Gentile's Weblog

	Saturday, October 19, 2002

The .NET COM Interop presentation that I did for the .NET User Group of Greater Boston is now available on my web site as a Powerpoint deck and the sample source code. Warning! The source code is rough, simple and unpolished sample code to illustrate Interop concepts and is not complete nor intended for production use. Update: The presentation that I did for the New Hampshire .NET Users Group is also available on my web site as a Powerpoint deck and the sample code. 1:49:58 PM

	Friday, October 18, 2002

What a night. So I went down to the Wang Auditorium - Boston University Corporate Education Center for our .NET User's Group meeting and got "hijacked" by John Lam and Ingo Rammer stalking the hallways after a day of WinDev. Since I had to give up free pizza to go with them, we all went out to some fantastic barbecue at Smokey Joe's in Nashua. John needed another PC to do nasty CLAW stuff to in prep for his talks tomorrow (today?) so we went back to his hotel room and just spent hours digging into his Runtime Aspect Weaver. I haven't spent much time with this before so I was really amazed at how powerful this stuff is. There is so much potential. I can think of many product uses for this: enforcement of Design by Contract, company standards, logging and tracing, unit test enforcement, and more. Of course, some of the way he implemented this is quite sinister-) and when he weaved onto (into?) System.Windows.Forms.dll, I thought it was the end of my PC-). It still seems to be working. Of course, the three of us shared war stories as well as talking about many things. Ingo has pictures in case my PC does break to blackmail John-) 1:38:43 AM

	Thursday, October 17, 2002

Type Safety in a Loosely Coupled World. In which I'm inspired by Tim Ewald and come to realize that runtime type checking is a very good thing. [sellsbrothers.com: Windows Developer News] Sam amplifies on this showing differences between languages and pointing to his excellent essay Dealing with Diversity. "Until recently, most programming activities were mono-cultures focused around a language centric base.  A much more successful model is emerging.  One in which the service provider has an opportunity to suggest what data types it would prefer.  This gives the intended consumer an opportunity to adapt if it so chooses.  Ultimately the service provider will decide whether or not it can process the request as sent." 2:26:28 PM

Ken Rawlings has been bothered by a few things in .NET recently. I will leave the combustable issue of the size of the runtime aside for now but one point he made I would like to address: Around $6000 upfront software costs for ASP.NET deployment (XP Server plus SQL Server 2000). I honestly don't have any idea where he gets this figure but I don't think its correct. Neither XP Server (whatever that is, I assume you mean .NET Server which isn't even out yet) or SQL Server 2000 is required for ASP.NET development. For that matter, neither is VS.NET. You do need IIS right now (although even that is changing). You can develop .NET on a $99 copy of XP Professional, grab the .NET Framework SDK and Runtime,  and use Notepad. So far, I'm at $99. But I didn't even have to do that. I could have used my old copy of Windows 2000 or Windows 2000 Sever, or even NT4. So my cost would have been $0. ASP.NET applications do not require SQL Server.  If you are talking about data applications, you are talking about a whole different cart of apples. An industrial strength solution for the data enterprise will cost money no matter if its is Oracle, SQL Server or DB2 and that is a whole class of applications. And MySQL doesn't go in that category with its current limitations. This is not ASP.NET development. This is data application developemnt. Even if one did factor in the cost of a development version of SQL Server, it would be nowhere near $6000. Whenever Microsoft comes out with anything, many are quick to leap to the myth added costs, whether they are there or not. Follow-up; Brandon Lee makes the obvious point for me: "Ken Rawlings is bothered by the cost of .NET compared to OpenSource solutions. The comparison seems unfair because the greatest cost is SQL Server vs. MySQL. These are not fairly comparable because SQL Server is a full RDBMS (e.g. transactions) and people in the market for SQL Server will not be comparing it to MySQL. So you really have to compare .NET/SQLServer to something like Linux/Apache/Oracle and the price difference disappears. Ken Rawlings is bothered by the cost of .NET compared to OpenSource solutions. The comparison seems unfair because the greatest cost is SQL Server vs. MySQL. These are not fairly comparable because SQL Server is a full RDBMS (e.g. transactions) and people in the market for SQL Server will not be comparing it to MySQL. So you really have to compare .NET/SQLServer to something like Linux/Apache/Oracle and the price difference disappears. " 12:28:02 PM

Scott Seely, who gave an interesting talk, which may be the first time a Microsoft employee showed Apache Axis alongside .NET, has written a new MSDN article on a timely subject Understanding WS-Security. Of course, the hardest working man in the business has written on this as well as supplyinmg a toolkit. 11:30:34 AM

Drew is back! 11:15:57 AM

	Tuesday, October 15, 2002

Mugshots. Nice to see some faces behind the names i've been meeting online these last 10 days in the Groove Experiments space. Kinda reminded me of the Mugshot discussion on the Grooveforums some time ago. [Jeroen Bekkers' Groove Weblog] This is the group of people I spent most of the two days with. It seemed like we had known each other forever. We also went out to dinner on the last night along with Steve Loughran who is one funny Brit-). His talk was even better than last years (and funnier) and it would do developers good to remember we have to deploy these bloody things. 4:27:11 PM

Too Many Web Services. Here's the "Top 10 Reasons You Know Youve Been Hacking Too Many Web Services..." that Tim and I presented at the Web Services DevCon (which was very fun, btw). [sellsbrothers.com: Windows Developer News]  Ah, Geek humor-) 4:12:00 PM

	Monday, October 14, 2002

Clemens Vasters is probably the most influential and one of the smartest folks I interacted with at the DevCon. He blew my mind a couple of times but one of the biggest revelations he gave me was convincing me of the importance and use of EnterpriseServices in .NET. Now, I knew they were there and I knew it wasn't like COM: dead but still COM+ is so 3 years ago you know? It was Clemens who convinced me that not only is a lot of that stuff useful but some of it is actually extremely vital and important. So thats where I'm going in my research.  The other thing is that Clemens is not even remotely (pun intended) worried about speaking his mind and is extremely smart. So I read with interest today as he skewers Roger Sessions (quite overdue): "Roger Sessions writes about WS-Transactions in his ObjectWatch news letter and the article shows that he shouldn't. First, his "Shootout at the Transaction Corral" has the most confusing lead-in story that I've ever seen for a story about transactions. It starts with how to get breakfast from two places at the same time and how that is a real life coordinated transaction -- it may be so, but why make a "real life analogy" if that by itself is so far fatched that it's losing the whole point." and  "That's a pretty short-sighted statement, because that says that a fortress (I personally prefer the cuter term "fiefdom" coined by the guy who actually came up with this model: Pat Helland) is always implemented as a homogeneous system. Not so: A "fortress" is a system which can very well be implemented as a heterogenous assembly of services implemented on different machines, different OSses and different platforms. If that's so, you will need AT to coordinate local, distributed transactions across, for instance, J2EE and .NET. Web Services are about interop, not the internet!" and: I can tell you what SHOULD happen. IBM, Microsoft, and BEA SHOULD redo their model and make three changes: eliminate the WS-C specification, remove the WS-T dependency on WS-C, and put atomic web service transactions (ATs) where they belong, in the trash. " Well, I think that Roger Session SHOULD try to understand web services, transactions and real world system complexity. Even inside the "fortress", interop counts. 4:56:22 PM

Andres Aguiar's Weblog: "When we looked at the application, we were surprised. The database design was really bad. There were tables with no primary key, referential integrity constraint missing, etc. So, we wrote the article focusing in the database design. You can find it in PDF here (304Kb)"   4:49:04 PM

Fawcette has an article on Dig Into WS-Security with the WSDK, immeditely relevant after Keith Balinger's talk.They also have an article on Use X.509 Certificates with the WSDK. Again, great timing-) 10:28:38 AM

Ingo and I started doing some research into the Rotor implementation of Garbage Collection Saturday night/Sunday morning in my office. We dived down into the assembly (language that is) and found some things. However, we don't have enough data to form any conclusions yet. We willl be continuing this in the next few weeks and wil report what we find. 10:21:03 AM

Sam Ruby: "The Mindreef duo shared a wealth of immediately practical advice. If you ever consider deploying a web service based solution with components from more than one vendor, these are the guys to talk to." The single best thing for me, though, was that there was clearly a very strong shared point of view among most of the speakers as to the primacy of data. If it wasn't for the data, why would we bother writing code? [Dave Seidel :: Wavicle] I agree wuth Dave on the shared view on the primacy of data. Its wonderful to go a geeky development confrence with a lot of different people from different campis if you will and having them all get down to the same something. BTW, I have been holding off on posting anything about MindReef's tool and I didn't give much beta testign time to Beta 1, but after seeing Dave's demo at the conference, I just must get Beta 2 and get going. This tool is fantastic. The WSDL "View" is fantastic taking that jumbled mess and showing me an "object view" among many other features. I was also impressed by Dave's talk and demo. Rairly have I seen such a balanced and fair talk, as they gave equal mention of many other tools. A job well done Dave. 9:24:43 AM

	Sunday, October 13, 2002

12:36:11 PM

My presentation at the DevCon. After the presentation, I had an interesting conversation with Sam Gentile about VSIP and related stuff. [Andres Aguiar's Weblog] Yes, we had a great conversation! This product not only totally rocks but its use of VSIP is beyond what I've seen anywhere else. I am really impressed. Rss-subscribed. 12:32:03 AM