通过

Yukon's management tools

  • 项目

As the creator of a few SQL Server-specific tools myself (Sequin, DataPipe, etc.), I have some fairly strong opinions on what type of functionality the tools that come in the white box with SQL Server should have, particularly the GUI apps. At a high level, I think they should:

* be suitable for everyday use by both DBAs and database developers

* compare reasonably well with third-party tools (they might not have as many features, but I don't think they should look anemic next to a product by a company whose entire staff is a fraction of the size of the SQL Server Tools team)

* be better than built-in admin tools shipped free with other DBMS products

* support modern ease-of-use features common in similar tools. For example, they should sport a respectable code editor for editing T-SQL with features on par with those available in many freeware and shareware editors

* be reasonably stable, lean, and fast. Of course, these are all subjective terms, but the prevailing wisdom on them shouldn't be that they're memory hogs or slow or buggy.

* make use of modern user-interface elements common in other Microsoft GUI apps

* be shaped by usability tests and real-world user experiences on par with those a company of Microsoft's means should regularly conduct and learn from

Some of these might seem a bit much, but I would sum them up by saying this: I think a company with Microsoft's resources should be able to produce admin tools for its server products that are first-class, regardless of the fact that they come free in the box. I don't think some guy working out of his garage should be able to build a better GUI admin tool than a whole team of trained engineers at Microsoft can. I hold Microsoft and its legions of the gifted in such high esteem that I expect the very best from them, regardless of whether the app in question is its own product or included in the box with some other.

I've been watching the evolution of the management tools in Yukon for some time, and have come to a pretty firm conclusion about them: I think they've done it. I had some initial doubts about using the VS shell and about combining administration with code development, but, frankly, Management Studio is a joy to use once you get used to it. A long time fan of VS, I felt right at home in Management Studio after I’d used it for a day or two. I thorougly enjoyed having the text editor I'm so used to using for C++ and C# code available for T-SQL. I've written a lot of T-SQL in my time, and I would have loved to have had a tool like Management Studio to have written it in. I admire the extensibility of the interface and the great pains they've obviously gone to in order to balance functionality with sleekness. With so many new features and so much to present to the user, it would have been easy to have created a UI that literally overwhelmed the average DBA -- think: Times Square on a Friday night. But they have worked really hard to present a UI that is balanced, even elegant in some ways, and I think people are going to like it.

One thing I initially had trouble getting past was the resource requirements of the tool. I realize we're still in beta, but they seemed pretty large to me. But, on balance, this really shouldn't be that surprising. With the VS shell lurking beneath the surface, we should expect resource usage on par with a 'real' IDE. No, I don't like using lots of memory or page file space any more than anyone else, but memory and disk are both cheap, and this tool provides some incredible functionality for no more cost than the type of resource use that's common to most modern-day complex GUI apps (and the requisite start-up time that usually comes with such apps). We old fogies have to learn to lighten up a bit and realize that we live in a different world than 1985 or even 1995. It's 2005, and it's not uncommon for complex applications to use what seemed like outrageous amounts of memory and disk just a few short years ago. Whether that's a good or bad thing will have to wait for another day, but, suffice it to say that Management Studio provides an astonishing amount of functionality for the minuscule cost of some disk and memory resources -- resources that have never been cheaper than they are right now.

Now, mind you, I'm not the typical user, and I know that. Some pure DBAs may not like having coding elements in their admin experience. They might not like having the functionality of Query Analyzer and Enterprise Manager effectively combined into one tool. But I think they'll get over it for the most part. As I’ve said, the tool is a joy to use once you get used to it. Users of low-end SKUs such as SQL Express will likely prefer the lightweight tool that comes with that product, and that's fine. There should be room for a lightweight admin tool with limited functionality for that class of customer. But, as for me, I'll be using Management Studio. Enterprise Manager was a formidable tool, but the Tools team was right to reinvent it in this release. What they've created for themselves is a platform that should be viable for years to come. As SQL Server evolves over time, this is a tool that should be able to grow with it. This doesn't mean the tool is perfect (personally, I've never seen perfect software, and I've certainly never come close to creating it), but I do believe it's headed in the right direction, and I will be surprised if people aren't genuinely impressed with it.

Comments