Letters from You

• Integrity Compromised
  
• Xing Praised
  
• StringBuffer Prefered
  
• Dohhh!
  
• Eye Strained
  
• Interface Defended
  
• Molly Recommended
  
• Radio Hijacked
  
• Gap Bridged
  
• Write to Us
  

Integrity Compromised

Your magazine severely compromised its journalistic integrity in the article "Data Mining on the Web" in the January issue. The author should have stated from the outset that he is the CTO at one of the companies surveyed. Isn't it a coincidence that the data-mining algorithm he specifically recommends above the others just happens to be the one in his company's product? All and all, the article blurred the lines between information and advertising, which is a scary trend.

Best,
Ted Shab

Ted,

From time to time Web Techniques invites CEOs, CTOs, and other industry leaders to author first-person articles. These articles focus on the products and technologies of the companies that the leaders are involved with. As with any article, whether feature, review, or column, we work with the author to identify bias and compensate editorially. You will find that we often print articles written by leaders in competing companies and that we identify our feature authors—and the companies they work for—with a biography at the end of each article.

Amit Asaravala
Executive Editor

Xing Praised

Thanks for your many great articles—usually very meaty and right to the point. [Regarding the February "Webmaster's Domain" column] Audiocatalyst 2.0 (they may have a newer version) from xing is my favorite ripper-coder. It's an all-in-one, but in the table on page 16, xing was listed as an encoder. I know it's hard to keep up with all the software out there, especially the shareware, but this product is excellent, with many customizable features—worth a look!

Thanks,
Cory Riley
Intel SMD Webmaster
[email protected]

StringBuffer Prefered

Al Williams' February "Java@Work" column ("The Postman Always Rings Twice") mentions "one small compromise in its code" when referring to an additional ampersand at the end of the constructed URL created by the Poster object. Al mentions two possible solutions for this, but I suggest this simple (if inelegant) fix:

while ( enum.hasMoreElements() )
{
    // this will add the appropriate number of ampersands
    if ( buf.length() > 0 )
        buf.append( "&" );
    // etc., etc.
    buf.append( item + "=" + val );
}

Also, one thing Al forgets to mention in his description of using StringBuffer over String is the speed enhancement. Not only do you save process space when it comes to garbage collection, but every time you use the immutable String, as in his example:

"a = a + b;"

to make another String, here's essentially what happens:

a = new StringBuffer( a ).append( b ).toString();

String has to be converted to StringBuffer (then back again) to handle the change. This may not cause much overhead if your application has only a few Strings, but it's not rocket science to imagine the exponential drain on larger apps. To use the buzzword du jour: Disintermediate—use StringBuffer.

Michael Kmiec
[email protected]

Dear Michael,

Thanks for your interest in "Java@Work." There are many possible solutions to the ampersand problem and your way doesn't strike me as any less elegant than any other way.

And, of course, as you point out StringBuffer is more efficient than String in several ways. This is a good example in which you shouldn't have to know how things work, but in real life knowing the internal details can make a big difference.

Regards,
Al Williams

Dohhh!

Hi!

Good article [see "Webmaster's Domain," February 2000]! Couple of things, though. I was very curious how you came up with the numbers on page 16, column 1:

"Standard stereo CD audio produces two channels of 16/bit data at 44.1KHz, which works out to 14,000 kbps."

Isn't this:

(2 channels) * (44,100 samples/sec) * (16 bits/channel) = 
1,411,200 bits/sec => 1,411.2 kbps =~ 1,400 kpbs?

Also, just as a side note from personal experience, when using low bit rates (for example, 56 kbps or less) I found that if you sample a CD at lower speed the resulting MP3 will be of much more tolerable quality. (I don't know if Grip can change its sampling rate.) For example, if your target bit rate is 56 kbps, then sampling at 44.1kHz will produce nasty quantization effects. But sampling at 22kHz sounds much cleaner, though you lose the higher frequencies. I suppose if you get really inclined, you can take the 22kHz sampling and equalize it to lower the volumes of the midrange frequencies.

Elliott Lee
EIF Group

[email protected]

Hi,

Yes, it looks like you caught a typo. The uncompressed rate is 1.4Mbps, about ten times the rate of good-sounding MP3 audio. Thank you, too, for the information on lower sampling rates. The version of the lame MP3 encoder that I experimented with at the time I wrote the article didn't support lower sampling rates, but it does now. You're quite right that the nasty quantization artifacts (which appear as high-frequency buzzes) disappear at lower sampling rates.

Lincoln

Eye Strained

Let me first of all express my appreciation to Lincoln for his excellent article "Humming Along with the Web" in the February issue of Web Techniques. I've always appreciated his articles, but this one was particularly timely for me.

Now to the purpose of this email. I'm one of those that keep past issues of magazines until they are long past the obsolete category. Because of this, I exercise my right to "reproduce for research purposes" by copying an article of interest and placing it in an appropriate folder. Because of the numerous URLs listed in this article, I wanted to place a copy of this in the folder, but found that due to the colors chosen for the "Online" box, not all of the information would come through on a copy. The contrast differential between the red background and the black text was not enough to bring anything through regardless of the copier mode and contrast selections. (Although I'm not that "up in years" (yet!), I also find that this type of color selection renders the text difficult to read under non-optimum lighting conditions. I feel that this is an issue that should be addressed in Publishing 101, but what do these young whippersnappers know?) I know that you like the brightly colored backgrounds to catch the eye's attention, but would it be possible to use a different color choice? For example, I think a shade of pink or even a light green would have worked just as well and rendered the text more readable—and reproducible.

Other than my knit-picking, thanks for a great magazine. I look forward to each issue.

Dave Pfaltzgraff
Staff Engineer, Patapsco Designs

Dave,

The February issue was the last one published in that style before the redesign of our magazine. Beginning in March, our new look should solve the product online-box readability issue. Let us know what you think!

Cliff Scorso

Art Director

Interface Defended

Bob,

I enjoyed reading Molly Holzschlag's February 2000 "Integrated Design" column, and always like differing opinions. I must differ with her on using the Excite home page as an example of bad design. I have used it as my main portal for two years now. I use it for the very reason that Molly derides it (that is, "its rich interface"). It presents the most information in the most concise format, and allows unusual customization options for its users.

Having been a software designer for 20 years, I have often found that good screen and Web-page design, very much like beauty, is in the eye of the beholder.

In some cases, I believe that the best approach is to have a menu of customization choices, so that different users can select differing formats according to their own preferences. This is particularly true for Web portals that are heavily viewed, such as Excite.

It is always interesting to hear different opinions on what is good design.

Thanks,
Neal Miller
Miller Systems, Inc.
[email protected]

Molly Recommended

I have been reading Web Techniques for a while now and I just wanted to tell you how much I enjoy it. Web Techniques is the only magazine I read from cover to cover. I am looking forward to the new look it's going to have.

And I wanted to tell Molly how much I enjoy her column. It's just great!

I have absolutely no bad comments about it. After reading each issue's "Home Page," "Integrated Design" is the first thing I read. I take notes about the stuff Molly writes and I put it in practice. I have to say that I wouldn't be the same Web designer without her knowledge. Keep up the excellent work.

Piero Rosito
Digicom, GT

Radio Hijacked

Bob Kaehms' "Home Page" column in the December 1999 issue made many good points about radio's similarities to the Web. Here's another one:

It was not uncommon in the early days of radio for stations to be hijacked. Literally. A competitor would actually overpower a station's frequency and broadcast its own message over someone else's frequency.

Chuck Williams
www.sme.org

Gap Bridged

I've been reading your magazine for a few years and find it the most insightful and useful publication available for interactive marketing-development. You continue to write about timely and consistently significant topics. Web Techniques manages to bridge the gap between technical innovation and real-world implementation.

January's issue was particularly valuable as Dan Greening helped define the tangible value of site data analysis. The article acknowledges the potential of dynamic site data, and simplifies the thought process of its management, which is critical to understand when designing the system. The next step is to define methods of mining the aggregate data available from integrated legacy and ERP data systems.

I also enjoyed Molly Holzschlag's perspective on the project management and development process. The association of media production and software development is one I've evangelized and implemented for years. It is essential for the development team, internal management, and the client to understand this.

Stone Korshak
Director, Special Projects
OgilvyInteractive WW

---

Write to Us

We want to hear from you! Please send your thoughts to [email protected], or via snail mail to Letters to the Editor, Web Techniques, 600 Harrison St., San Francisco, CA 94107. Letters may be edited for length or clarity.