. Recently, BugNet also signed an exciting new mass marketing deal with Online Interactive.
BYTE:
With all the free beta software available today, do you feel that people are more willing to put up with more bugs these days?
Brown:
That's a complex question which has to be addressed from a couple of angles. Because of the need to generate revenue and sustain stock prices, there is a lot of pressure on vendors to release products as quickly as possible, and to deal with bugs later. This is a fact of the business, and if anything, it is becoming more prevalent, even though vendors obviously don't want to put out products which are so buggy that they depress sales. Beta is a little different issue, but it also plays into the software vendors' almost maniacal desire to curtail support costs. Beta looks like a dream com
e true to the vendors because they can say up front, "We're not going to support this," and at the same time they derive certain testing benefits from users who use a beta product and inform the company of problems. But for the user, beta software can pose a threat that is real and significant.
BYTE:
Why is beta software a threat?
Brown:
The threat is that installing beta can screw something else up in your system. For example, in a bug story that BugNet recently broke, it was revealed that Internet Explorer 3.0 Beta 2 running under Windows 95 replaces one or more system files with versions which are incompatible with some commercial products that you pay money for. So you install the beta, and all of a sudden, some commercial products may not work as they should. And you go to the vendor you paid money to, and now the vendor may very well say to you, "Our product doesn't work because you are running someone else's beta program, and we don't sup
port beta. You're on your own." Users don't realize that the "free" program that they download may in fact cost them a lot of time and money.
BYTE:
Do you think there will be any change in this situation?
Brown:
This gets into an issue that I think is really societal. We have come to a point where the expectations of computer users are changing, and there is actually conflict between the expectations of the traditional users of PCs, the PC enthusiasts, and the new PC users, normal people who may not want to spend their weekend screwing around with their printer drivers. To make the PC an everyday appliance, this large new group must be satisfied. But these folks expect the darn thing to work when they turn it on, like a toaster or a razor, and of course the PC fails this basic test far too often. So even though the industry as a whole may be doing a better job in the minds of traditional users, it may be doing worse in the minds of the users tha
t the vendors most want to reach. And this, in turn, may be part of why such a shockingly small percentage of PC sales are made to first-time purchasers.
BYTE:
It seems like your site and database have quite a few Win 95 bugs. Is it the mother of all bugs?
BB:
Well this is an interesting question. The number of issues involving bugs or compatibility problems is huge with Windows 95. In our database, the BugMaster database, Windows 95 now makes up over half of it. That means Windows 95 equals everything in Windows 3.x plus all of the Apple and OS/2 and assorted other bugs that we have in there. In terms of bulk or volume of issues, it's more than anything else combined. [laughs]. And that's over the last two years.
But that's only part of the picture. One of the things that BugNet does is we try to provide software vendors the carrot and the stick. We hold the vendors feet in the fire for issues that are causing problems and make them turn
and give their attention and come up with solutions. But at the same time, we want to praise them and give them credit for those situations where they do respond, where they do a good job, or they give it their attention and actually provide their users with a useful workaround or a fix. So one of the things we do is to measure the success rate of all the major vendors, in terms of their success in providing fixes or workarounds that their users bring to them. We do this on an annual and quarterly basis with all the vendors.
We also have taken a look at Windows 95 compared to Windows 3.x. We look at it in terms of the success rate of Microsoft and its allied vendors in providing solutions for this tidal wave of problems which have emerged in connection with Windows 95. And what we see is there are a really large and significant number of issues here which have arisen over the last 12 months, but the success rate in dealing with these issues is at this point a little better than it was with Windows 3.x.
So on the one hand, yes there have been a lot of problems, but on the other hand, Microsoft and its allies are doing a better job of shepherding their users through the minefield and helping them over the fence than was the case before Windows 95. You can't just look at the thousands of issues that have arisen with Windows 95 and its applications and say "Aha." That's only part of the story. The other part of the story is what vendors do in response. I think vendors are doing a somewhat better job than they did before. Whether that means the users are better or worse off depends on the person's individual situation and whether you've got a problem you can't come up with a solution.
BYTE:
What about call avoidance software, where users can go to a database or Web site and find solutions to some of the more generic or mundane bugs, while letting the technical support people concentrate on the more challenging bugs? Do you feel these will provide relief to the poor beleaguered
technical support departments?
BB:
[
laughs
]. Well they certainly are poor and beleaguered. I believe in call avoidance software, we are a vendor of exactly that type of product. In one sense, the BugMaster database is the broadest collection of bug fixes for the PC world. We sell it to individuals and corporations alike with the thought that this system can allow users in many situations to come up with solutions to a problem before they have to waste time, either calling in the case of a corporation, their own IS or IT people, or in the case of an individual, spending 45 minutes on hold waiting for a technical support person to pick up the phone. We are in the business of doing this and I believe it can help. I have sheaves of documents and letters of correspondence who thank us for this service, folks for whom we save an hour of their time. And I know that it can work. The thing that I have not seen proven to my satisfaction yet is the automatic systems. And this is som
ething that's arriving. I don't want to speak against them because it's a great idea and holds tremendous promise, but I haven't actually seen it at work.
BYTE:
By automated systems, do you mean something that actually goes in to your PC and makes changes for you, instead of just telling you what to do?
BB:
Yes. I have some concern that people may find that this may cause more problems than it fixes, or can even create problems where problems didn't exist. The variety of the hardware/software stew that people are trying to combine is so extreme that it's very hard to create a system that can automatically deal with it. Also, I think these self-help systems may be more useful to expert users. Less experienced users may still require personal assistance from a human.
BYTE:
So sending a novice user out to the Web may not be the best solution. Instead, the answer should be don't have bugs?
BB:
Right, but that's impossible, because the human hand inevitably carries with it imperfections. I think some users or some situations will always require human attention. I just don't think there's any way around it.
BYTE:
Are there any ways to improve customer support, are there any deep dark secrets that people ought to know?
BB:
First, folks should know there are differences in terms of the level of support they give and their success in providing solutions for the problems their users bring to them. I refer to the first quarter of 1996, we did a break out or comparison of the major vendors. Look for example at Corel, which is a vendor primarily of graphics software, with the exception of the suite they recently bought from WordPerfect, and Adobe, which is more or less in the same part of the forest. Our figures showed Corel had about an 85 percent success rate, and Adobe's was under 50 percent for the first quarter. This is not the onl
y criteria that an individual can bear in deciding whether to buy an Adobe or Corel product, but it is a consideration. I think folks should realize that these programs are not cookies that come out of a cookie cutter. And vendors have their own personality, their own priorities, their own staffing situations, and their performance is different.
BYTE:
How do you measure that, the 85 percent versus 50 percent?
BB:
We collect all issues and we determine for each one whether or not an acceptable workaround or fix is available. What we require for an acceptable workaround is essentially a situation where the vendor comes up with a solution that preserves functionality. So an acceptable solution would not be to tell a customer, "The way to avoid this problem with the printer is to not use this printer." The functionality must be preserved in some fundamental fashion to avoid whatever it is that is causing the difficulty. We are the only organization
that I know of that makes any measure of the success of the vendors in this area. And I don't think it's the be all or end all, but I think it is a very significant measure of what's going on in the industry and what kind of experience that a purchaser, whether corporation or individual, may have six months down the road. The other thing that people ought to know is that we feel that people that have paid good money for software, sometimes $500 or $600, should insist on a satisfactory answer, and if you can't get it from the first person, ask for their supervisor or a second-level tech. Make them stay on the phone until you get the thing resolved. I think there's a tendency psychologically for users to come on bended knee to the vendor, as if to the Oracle of Delphi. The basic relationship is one where the vendor is the provider of services and the user is a purchaser who should be able to insist on satisfactory performance from the product they paid money for. We get dozens of bug reports daily and a lot of
them, at least 90 percent of the time, the user turned the bug in, they wrote the vendor e-mail, and they are not getting a response. And one of the reasons vendors may be afraid of BugNet is we will sometimes go to the vendor and shake that tree. This is a sad reflection on the industry, because it means that people that are paying good money can't get minimum service out of the people that are taking their money, and they have to go to someone else like us. I mean, this is part of our business, but sometimes I take a step back and I have to shake my head, because it shouldn't be this way.
BYTE:
What is the gnarliest bug you have heard of?
BB:
I've become a connoisseur of these kinds of things. I respond on a different aesthetic. My favorite is one that BugNet broke is the one in Bill Gates
The Road Ahead
book, which will bomb some machines if you try to run it at more than 256 colors, which is of course the color palette standard of t
hree or four years ago. And then you find that Microsoft doesn't support the thing. But that's kind of a humorous thing. An example of a more serious bug is one that's ongoing right now, is the major incompatibility problem between WordPerfect for Windows 7's address book and Microsoft Exchange. I've gotten a lot of reports of problems with this. And the most recent that we got was that Corel was working to fix this.
But one sophisticated guy got enmeshed in this problem and said he spent 90 minutes with Corel, and the tech rep had him uninstall and then reinstall Exchange as part of the troubleshooting process. That resulted in him no longer being able to run Exchange, at which point the Corel tech rep told him that he was on his own, because Corel does not support Exchange. This is another example of one thing screwing up another thing where the user is left holding the bag. And this is not as it should be. To me, this is symptomatic of the sort of thing that troubles me. It generates ill will, it gen
erates all sorts of negative fallout which the industry as a whole will have to deal with down the road.
For more information, see
http://www.bugnet.com/~bugnet
A Dark Cloud of Bugs
