Loose cables is an occasional, irreverent look behind the scenes at testing computer products, in particular at IDG's InfoWorld lab in the US. Our insights are gleaned during the long hours spent testing products and even longer hours spent sorting through outrageous vendor claims and press releases. Some of the insights are technical, some are political, and some are just fun.
Being the influential trade journalists we are (and humble too), it's not unusual to see vendors use some of the InfoWorld test results in their marketing materials. And usually that's fine because most vendors get InfoWorld's approval and use the material fairly to back up their own claims with the opinions of InfoWorld analysts.
So we were a bit stunned to open a recent issue of InfoWorld and other US publications, to see an ad touting a company's superiority in its field.
We don't deny that this is a fine product; it came in a close second to a rival product in a recent comparison.
The problem is, the ad doesn't reflect our results fairly. In fact, we're wondering who exactly is working over there that could be guilty of both grossly misrepresenting our results and coming up with such weaselly fine print (which, it turns out, is wrong).
In the biggest and most blatant reworking of our comparison, the ad completely omitted the top performing product from the chart showing the results. Odd, but as far as we can tell, the perpetrator tried to justify the omission with some fine print which might as well have said they were omitting products "with a 'b' in their names".
This company also omitted the score for
another product for no discernible reason. At least this one came in at the bottom of the chart, so dropping it isn't quite as sinister as dropping the actual winner of the comparison.
As if to make up for dropping the lowest scorer, the company helpfully added in a score of zero for another product, which wasn't even in the comparison due to the fact that a new version was imminent, so we didn't think it would be helpful to test the old version in our comparison.
We don't know where they got the "0.0" from - even our most hard-core analysts are pretty sure that the product at least has documentation and a price, and would certainly have got at least a few points somewhere.
It's pretty silly to assume that because we didn't test a product it would have got a terrible score, and it's even more misleading to stick that score on a so-called representation of our results.
So, after taking out two products that weactually tested, and adding in one we didn't, the company has a pretty convincing ad - even if its ethics (and fine print) leave a lot to be desired. (In case you're wondering, the ad hasn't been run in Australia. - Ed.)Although this ad misrepresenting Test Centre results is really blatant, it's certainly not the only instance. Recently, an upset reader pointed out that a vendor was using a fairly negative InfoWorld review of a competitor's product on its Web site.
That wouldn't be so bad, except the review in question said nothing whatsoever about the vendor that was posting it. It was being used strictly to bash a competing product. It shows a lack of professionalism that we find depressing.
Microsoft is doing something right that its competitors just don't get. The declining popularity of Novell's NetWare platform products in favour of the Windows NT juggernaut isn't happening because Microsoft is winning InfoWorld's solution comparisons. In fact, by and large, Novell generally comes out on top in these comparisons.
What Novell and others have consistently failed to do well is mobilise third-party application developers to write products in support of their platforms. Novell has been especially guilty of this with Novell Directory Services (NDS).
Unless Novell can turn this deficiency around, the Redmond-based tractor beam will continue to pull in new customers.
Netscape's openness strategy is a
pretty compelling one. In the intranet product market, Netscape has come up with cross-platform versions of most of the major network service components that are at the heart of every IS manager's nirvana.
Netscape's offerings include open mail and groupware servers tied into a single hierarchical directory and industry-standard security server.
Now that's progress.
If we were the kings of Netscape for a day, we might take that openness strategy out for a spin and see what it could do. The fact is the only thing that's missing is the network OS.
We might announce a new product code-named Netscape OS that is based on a standard high-performance, robust, and stable microkernel Unix. We'd continue to build on Netscape's open platform strategy, adding core integration for CORBA and the Internet Inter-ORB Protocol.
Take it a step further, though. For example,we might bundle some server products suchas Messaging Server, Enterprise Server, and Directory Server to sweeten the deal. In fact, we might throw in SuiteSpot. Let your imagination take it from there. We'll just pose the question, "How far off is the Netscape OS?"
News of the Novell-Netscape deal came on the heels of that thought. And we had to ask ourselves why Netscape would do such a thing. Sure, there's a huge installed base of NetWare, but most NetWare servers will never be Internet or intranet servers; most will just stay in the closet and serve up files.
That, combined with Novell's marketing problems, should set off alarm bells over at the big 'scape.
Intranet products integrated with NDS is a great idea. Think of where it could have led if Novell had got NDS for NT out the door years ago. You could be using a Netscape Web server, integ-rated with NDS, running on a Microsoftoperating system.
We're glad to see that Microsoft is keeping up its age-old tradition of releasing OSs much later than expected. With the recent announcements that both Memphis (Windows 97? Windows 98?) and NT 5.0 will arrive later than expected, Microsoft is quietly assuring us that there is still order in the universe.
Under the OS/2 hood
This is the last bit about OS/2 for a while. We promise. In a recent discussion with IBM about OS/2 in general, and Directory and Security Services (DSS) in particular, officials told us that IBM's DSS is actually a Distributed Computing Environment (DCE) implementation.
Now this isn't a secret, it's just not well-publicised. DCE doesn't get the press that Novell Directory Services (NDS) or even Network Information Service gets. However, it is a full-featured directory service specification that probably deserves more attention. DCE's problem has been its role as a back-end specification, subject to different implementations by vendors. But that has a plus side, too: DCE is available on a wide variety of platforms.
The cool thing about DSS' reliance on DCE is that it can fit into existing DCE-enabled networks, or it can be the seed for a new DCE environment. What's more, it's been demonstrated that DCE and NDS can be unified. Check out www.citi.umich.edu/tech reports/reports/citi-tr-95-1.ps.Z (this is a gunzipped postscript file for Unix machines) for more information.
The shotgun approach
Have you ever seen a six-year-old play Doom? Parenting tips notwithstanding, it invariably involves a lot of frantic running around and blasting everything - even the walls - with plenty of shotgun shells. Microsoft's Internet strategy is eerily similar.
It even works.
So why didn't the people working on Microsoft Merchant Server take heed? The Mer-chant Server has its own, different scripting language.
What's even stranger is that FrontPage '97 isn't happy about working with pages that use any of those four scriptinglanguages.
Simple scripts usually survive, but even if FrontPage doesn't actually mess up the coding, it wreaks havoc on the code's formatting.
So let's applaud Microsoft for making it possible for us to use four different languages to script pages, but let's hope that next time they take a little more careful aim at the end result and deliver a somewhat more coherent suite of products that leverages each other's strengths in more typical Microsoftian style.
The massive Apple layoffs a few weeks back made us wonder again how the computer industry manages to stay an industry. Imagine if the automobile industry worked the same way. If your company needed a fleet of cars, you could choose between a big manufacturer such as Chevrolet or a smaller one. But you couldn't drive a Chevy on a Ford road. Things would be OK until you heard that Chevrolet isn't doing so well. Suddenly, they stop making new roads for Chevys. If you want to upgrade your fleet, you have to write off the Chevy investment.
Send in the Clones
At the InfoWorld test centre, we've always believed that saving a few hundred dollars by buying no-name, locally assembled clone machines just isn't worth it. For our time and money, we'll take the name brand equivalent every time. Recent news has got us rethinking the value of clones, though.
Consider the hubbub over the cloning of a sheep a few weeks ago. Churches and governments alike are trying to grasp the ethical and moral implications, and we'll leave them to that. We thought we'd take a stab at the business implications, instead.
The first, most obvious target for cloning is Bill Gates. We expect Microsoft has a crash program going (pun intended) to clone its fearless leader.
What does that mean for the rest of the industry?
Every 10 years or so, a new Gates will take over the helm of America's largest monopoly, after having been trained for years by the Gateses before him. Perpetual geekdom, or something to that effect.
Of course, the day will come when Gates' source code is leaked, and competitors will frantically clone Gateses.
Some will no doubt serve as lab rats, in an effort to predict the wily one's next move. Others will actually be employed in management, especially at start-ups.
New companies with serious venture capital could start out with Gates, Ray Ozzie, and Marc Andreessen as their management team.
OK, so maybe that's not such a good idea, but you get the picture.