Data Management

Interview with Brian Randell

We caught up with Brian Randell from MCW Technologies at TechEd/Canberra to talk about some of the new features in SQL Server 2005 and it's role as an -Access Killer".

Brian Randell is a senior consultant with MCW Technologies. Brian spends his time between teaching Microsoft .NET technologies to developers for DevelopMentor and consulting with MCW Technologies whose clients include Microsoft, American Honda, DELL, and others. We caught up with Brian at TechEd/Canberra to talk about some of the new features in SQL Server 2005 (formerly code-named -Yukon") and it's role as an -Access Killer"

Builder AU: So what does MCW Technologies do?

Brian Randell: We're certified partner, so in relation to Microsoft we train people on Microsoft technology and we build custom software, which keeps us pretty busy doing work for them. We create white papers, sample applications and we often contribute various pieces to their products. In the past, we have worked on, for example, Visual Interdev 1.0— we wrote some of the wizards that are in that product. For Office 2000, we wrote wizards and add-ins. Right now we are currently working for the SQL Server team on things relating to SQL Server 2000 as well as SQL Server 2005 and we are doing work with the developer division also related to Visual Studio 2005.

We have actually seen a lot of hype and buzz around SQL Server 2005 and some of the new features that have been highlighted in the press, including CLR integration, a new version of reporting services, the express editions, etc. But what are some of the features that you see from a developer's perspective that may not actually get that much press but definitely have the "wow" factor?
I think if you back up, it is not just CLR integration; it is the pervasive use of managed code by the entire product team. Another thing that I should tell developers out there is Microsoft's commitment to managed programming. A lot of people think that it is just marketing and it may all go away. The marketing may have evolved and they are trying to make the messaging clear, but I think that fundamental issue is the Microsoft is committed to the CLR and managed code. Whether it be integrated into SQL Server or actually using the .NET framework to build software that they are shipping to their customers.

Often it will be the perspective that people would say "Microsoft uses C++ but they will never use VB" Well, that is all different now. Microsoft is using managed code to build stuff, so it is not like they are using some super tools and they are only giving their customers something that is not industrial strength. SQL Server is about a billion and a half dollar product unit— so if you are going to bet your billion and a half dollar product unit on something, that shows commitment in my mind. And so right there, I think it is the commitment to the .NET framework would be the biggest thing out there. From there, there are individual features that depending on your perspective are sexy or cool, depending upon what problems you are trying to solve.

With the new express product line, there has been some talk about Microsoft trying to capture the MYSQL market and get entry-level programmers and developers who may have used an open source product— do you think that is the case?
Microsoft is never shy as far as competition goes, but the one thing to look at with SQL Server Express primarily is number one: addressing customer needs. With MSDE, it provided a great solution— it was low cost to entry, zero dollars. You just had to have the appropriate licensing and then we've changed the licensing to where it is available to anyone who is willing to sign the basic license. So we said "What areas of the product were causing pain to the customers?" There was confusion over the governor and now the workflow governor is gone. So database size— 2 gigs, make it 4 gigs. Now let's reduce some of the memory you can use and reduce processor, in other words what value-adds can we do to make customers happy?

We took away the pain points and maybe we cut a few features to the point that we are giving something away for free we can't give it all away, we need you to eventually upgrade to our products. So could you say it is addressing MYSQL? That would be one person's opinion, not necessarily mine. SQL Server Express is about making SQL Server customers happy and preparing them for the Yukon era. And I think that is the biggest thing—whether it is SQL Server CE, SQL Server Express on the desktop or SQL Server 2005 Enterprise Edition for your corporate data centre. That is the main takeaway. If MySQL and the rest of the LAMP community see that is as competition, then that is good for them and good for us, as competition is healthy. But the goal is making existing customers happy, as customers had some issues with MSDE.

Do you see SQL Server Express as an "Access-killer"?
Microsoft traditionally has had a very mixed "data story" and I think SQL Server Express is providing an enterprise solution. Microsoft Access has been primarily targeted towards individual users. While developers have been able to build solutions on it, saying it is a killer... Access has been enhanced heavily in Office 2003 so I wouldn't look at is as a "killer."

I look at it as segmenting the products appropriately. SQL Server Express is where you should be building your foundation applications that will eventually scale into the enterprise. Microsoft Access is for your "mom and pop" end-user, someone who needs the wizards that help you down the path and tight office integration. I think they are more complimentary more than competitive. I think that the issue was that SQL Server Express didn't provide features that Access provided that we needed to have a solid story, so I think it was more fixing the story than anything.

So in terms of other "sexy" features that are in Yukon, one of the areas that developers are really excited about is reporting services that was debuted with SQL Server 2000 in January. So what types of features can developers expect from the Yukon release for reporting services?

I think two of the big ones are going to be the new end-user reporting controls we acquired through ActiveView. There was a Provo, Utah company, small company, 5-user shop that was acquired that will give us a much broader reach. Right now, SQL Server 2000 Reporting Services is a great first release in that reports are easily distributable through the web. That is great, but users often need to slice and dice their data more effectively. So by providing the base report and providing these controls, they will not only be able to access it on the end-user side, but modify the way data is presented. Right now, they are working hard to integrate it into the product and it will be available in the Beta 3 timeframe for customers to try that out.

The other fact is that for developers we will also include an embeddable control that will allow them to ship reports in rich-client applications like Windows forms. So those are the two biggest ones out of the box and there are hundreds of features in reporting services, better integration into the business intelligence studio for developing, better report template support, a full programmable web services interface, etc all built into Yukon.

Looking beyond Yukon, to future releases there have been a number of competitors, including DB2's Stinger and other database platforms that are moving into enhanced and updated storage paradigms, looking at new ways to store structured and unstructured data. Does any of the data storage technologies in Longhorn have a bearing on what they are doing in Yukon today?
Obviously, when we start talking about the future and looking into your crystal ball, it is fuzzy. But I think the best way to look at it is that all the work the team has done in Yukon has provided the framework for future data storage technologies. How that translates to actual product placement and positioning, I can't address, particularly since I am not privy to what code is shared between the two teams. If anything, the long-term goal is how do we optimize the way customers store data. Whether it is on the simple client, to the enterprise. I think Yukon sets the stage to show we have a very good story for storing both standard relational data and the inclusion of the new Varbinarymax data type to really handle BLOB data, really sets the stage to where the foundation we build on the engine can lead to enhancements in things like WinFS, but that is all speculation on my part on where they draw the line.

Editor's Picks

Free Newsletters, In your Inbox