Thursday, July 28, 2005 - Posts

SpeechTEK 2005



New York Marriott Marquis, August 1-3

For eleven years, SpeechTEK has been delivering quality, in-depth content focusing on speech technology.

This years line-up looks very impressive, with Intel, IBM, Cisco, Avaya and Microsoft. At last years SpeechTEK in NYC I spotted a dozen Tablet PC's, the most I have ever seen at any conference. Just to meet and listen to James A. Larson, of Intel talk about VoiceXML 2.0 is worth the price of admisson. Today is the last day to register on-line, use the Priority Code EM14  and you'll save $189. 

posted Thursday, July 28, 2005 12:31 PM by paul with 0 Comments

NY SQL UG Meeting

NY Metro SQL Server Users Group

The July 28th Meeting

Graphical Query Plans & Performance Tuning in SQL2000 and SQL2005

Speaker: Steve Kass, SQL Server MVP

In both SQL Server 2000 and 2005, graphical query plans are invaluable for troubleshooting and tuning. Using sample queries on variations of the Northwind sample database, we'll survey as many of the following topics as we can.

  • Using SQL Server 2000 Query Analyzer, we’ll start with a quick overview of graphical plans in 2000 and 2005: general layout, common plan operators, and pop-up details.
  • SQL Server can generate both estimated and actual graphical plans. What additional information is in the actual plan, and what actual plan detail are still not “actual”?
  • Many slow-running queries have a single bottleneck. We’ll see how graphical plans help us identify, understand, and eliminate some common bottlenecks.
  • SQL Server creates and caches parameterized plans for stored procedures and for some ad hoc queries. This usually helps, but not always; graphical plans can help clarify the situation. [New 2005 info]
  • User-defined functions, column-to-column joins, and certain WHERE clauses can lead to inaccurate cost estimates, which can be seen in the graphical plan. [New 2005 info]
  • Graphical plans reveal which new 2005 features (like the analytic functions) provide real performance improvements, and which (like APPLY and recursive queries), do not.
  • In some situations, query cost estimates are very wrong, and a query takes hundreds or thousands of times longer than the optimizer expects. We'll see why, what to do about it, and why it's not a bug. [New 2005 info].
  • Sometimes the estimated execution plan is misleading, and can suggest problems where there are none. You won’t be misled if you’ve seen these situations before.

posted Thursday, July 28, 2005 9:57 AM by paul with 0 Comments

Getting ATOM

Feed Validator

Ever since the the IETF Atom syndication format specification has been declared ready for implementation (more), the demand for Atom 1.0 support by the Feed Validator has been high.

As progress on implementing the Atom 1.0 test cases has reached the point where the feed validator is rarely outright misleading any more, it is time for a wider exposure. At this point, the Feed Validator will accept bug reports on Atom 1.0 support

Particularly be on the look out for the following:

  • Errors being reported as warnings, and vice versa
  • Additional places where warnings would be helpful
  • Unclear or confusing advice

As always, feel free to join us on the feedvalidator-users discussion list.

DasBlog 1.8 will/does produce Atom 1.0 valid documents

 

After a small bit of work today, DasBlog 1.8 will (does) produce valid Atom 1.0 feeds. This is response to the bakededness of Atom 1.0 and Sam's desire to deprecate 0.3. DasBlog won't produce Atom 0.3 anymore (it was always marked under "experimental' anyway) so when you upgrade to 1.8 your Atom feed will upgrade automatically.

We do, as always, produce valid RSS 2.0 and now Atom 1.0. Here's a list of known Atom 1.0 consumers

Atomic RSS

Suppose you’re generating an RSS feed, or you’re thinking about generating an RSS feed, and you’re wondering how Atom fits into the picture. The future of technology is hard to predict, but there’s a good way to hedge your bets. You can generate an RSS feed and, by following a few simple rules, be really sure that there’s a 100%-equivalent Atom 1.0 feed, so that if you’re generating both, they’ll be in sync, and if you need to switch back and forth, it’s just a matter of changing a few strings. Let’s call this future-proofed flavor “Atomic RSS”. It turns out that using Atomic RSS is a Good Thing anyhow, because it will help software in general and news aggregators in particular produce better results.

ATOM is getting traction and gaining momentum, hope to see DotNetJunkies upgrade to ATOM 1.0 real soon!
 
posted Thursday, July 28, 2005 9:23 AM by paul with 0 Comments