Threads
Do you see a double standard in the way software products are evaluated? Do non-Microsoft products gain a halo effect simply by not being from Microsoft? Conversely, do good Microsoft products get disparaged for no good reason other than the fact that they are from Microsoft? Which do you think is the best browser, and why? Discuss it in LangaLetter threads.
Recent Columns
Open-Source Split?
Tilting At Windmills New Rules In The Domain Name Game End Of The Intel Era? LangaLetter Archives
Bio
Fred Langa is a senior consulting editor and columnist for Windows Magazine. Fred's free weekly newsletter is available via subscribe@langa.com. You can contact him at fred@langa.com or via his website at http://www.langa.com.

The Software Double Standard

By Fred Langa

Netscape is driving me crazy. The new releases of Navigator (4.08) and Communicator (4.51) have retained several long-standing bugs that get in the way of complex Web pages. (One of these bugs is so old it traces its roots back to the earliest days of JavaScript!) Other browser makers (including Opera and Microsoft) have worked around these problems, but Netscape has not.

It's driving me nuts for two reasons, and the first is very personal: I've literally wasted days trying to program around the bugs in Netscape's browsers. You see, I'm working on "BrowserTune 2000," the next iteration in the BrowserTune series, which is sponsored by Windows magazine.

The current iteration is BrowserTune98; it's a series of browser-neutral Web pages that tests for free some 300 browser features and functions so you can see exactly what any browser supports--or fails to support. It's hugely popular, serving more than a million and a half tests a month.

The new version, BrowserTune 2000 (BT2K for short), will offer a "fast test" option that will automatically test a limited subset of essential browser features; a full-blown manual test will still be available as a separate item.

The automated testing is handled by simple, basic client-side JavaScripts. For maximum compatibility, the JavaScripts don't take any shortcuts or do anything fancy: it's all pedestrian, straightforward, explicit, inline, simple code.

But Navigator has had a bug since JavaScript first appeared: Complex scripts cannot be placed within the table-data (TD) tags of a table. It's not an insurmountable problem--Internet Explorer, for example, has no problem with code of arbitrary complexity placed anywhere you want inside tables. It's just that Netscape has never gotten around to fixing the bug.

This is a particular problem for BT2K because all of CMP's pages are table-based. There's no good way to avoid the tables, so the only option is to use a long, complicated, pain-in-the-butt "document.write" trick to fool the browser into not realizing it's working inside a table. As long as you hide the TD tags from the browser in this way, Navigator/Communicator is happy.

Similarly, Netscape is equally finicky about font tags inside HTML "divisions." For example, even the latest versions of Netscape browsers will crash and burn when they try to run the following simple code:

[div id="crashandburn" style="position:absolute"]
[TABLE WIDTH=300]
[TR] [TD] Netscape browsers never get this far.[/font][/TD][/TR]
[/TABLE]
[/div]


The meaningless unpaired tag is the problem.

Believe me, I've heard all the "purity of code" arguments defending the placement of tags, and yes, there should be a font tag in there. If Netscape simply generated an error message, that would be fine. If it even garbled the display or exhibited weird font behavior, it might be OK. But instead, that little extra font tag actually causes instant browser death. Navigator/Communicator generates an illegal operation/invalid page fault: The browser stops cold and immediately shuts down.

Is this necessary behavior brought about by some fanatical devotion to pure and perfect HTML? No, it's just a bug. For example, Opera (generally regarded as the "purest" of the browsers) has no trouble at all with the code--it just ignores the meaningless font tag. Likewise, Microsoft IE just ignores the tag and renders the page perfectly. Only Netscape crashes and burns.

There's lots more, but you get the idea.

OK, all software products have bugs and design flaws. It's inevitable. With most products, we take bugs (even the nasty ones like Netscape's) in stride.

For example, Opera (as of version 3.51) has loads of bugs and weird behaviors. It's maturing fast, but chokes and stumbles on many pages that give Netscape and IE no trouble at all. Still, it's a well-regarded browser, despite its shortcomings.

Netscape browsers are still well-regarded despite their flaws--some of which have remained unfixed for so long they're practically fossils. But the presence of these bugs hasn't dampened Netscape fans' affection for the browser.

Microsoft IE has its share of bugs and flaky behaviors. But it correctly displays any page that Opera can display; and with the exception of some of Netscape's proprietary and non-World Wide Web Consortium-compliant tags, will also display just about anything a Netscape browser can display. In fact, after coding hundreds and hundreds of pages for all three major browsers, I find IE is by far the most flexible and robust of the pack.

But IE is reviled by many. "It's buggy," some say--although the competing browsers are at least as buggy, and are actually worse. "It's bloated," others cry--as if the 14-plus Mbyte Communicator is a model of software svelteness. "It's a slug," still others cry, although IE is perceptibly faster than, say, Communicator 4.51 in side-by-side tests.

A reasoned, objective approach might look at strengths and weaknesses of each browser in turn, and leave it at that. But there's a significant number of people who hold Microsoft to an unrealistic standard of software perfection to which they would never hold any other company.

And that's the second part of what's, err, bugging me: Why do Opera's and Netscape's browsers get excused for their shortcomings, while Microsoft's gets nailed? Why is there a double standard in judging software quality?

Is it Microsoft's allegedly bad business practice? But that has nothing to do with software quality per se, and judged on their own merits, with the same standards used to judge other software, some of Microsoft's products are actually quite good--even best-of-class. I think IE is one of them.

Maybe--just maybe--some Microsoft products earned their popularity by simply being better than what else is available.

But some regard statements like that as heresy. Bugs from Netscape, Opera, Red Hat, or Apple--or almost anyone except Microsoft--are just part of the package and no big deal. But bugs from Microsoft are a glaring example of shoddy workmanship, lazy monopolistic practices, lousy programming, and an expression of the vast evil that lurks in the suburbs of Seattle. After all, all Microsoft products stink, right?

Give me a break.

Do you see a double standard in the way software products are evaluated? Do non-Microsoft products gain a halo effect simply by not being from Microsoft? Conversely, do good Microsoft products get disparaged for no good reason other than the fact that they are from Microsoft? Which do you think is the best browser, and why? Join in!

---