Larry,

I believe that the article described the reality of things at MS in general. In other words, I believe that the 'attitude' is generally common to testers at MS.

However, virtually all of the causes for the decision taken have to do with difficulties within MS and have very little regard for the customer!

Let's take the case where it is not an "admin" installing the subject product... Someone who is not an "admin" likely has far less familiarity with the registry than someone who is. Furthermore I'd guess that someone who is not an "admin" would be more likely to install the subject products in the undesireable sequence. This poor person, the person for whom the tested (but not included) workaround does not address, is the person most in need of a solution.

That MS has designed its products to be so inter-related is really MS' own fault. That MS supports multiple languages the way it does, adding further complexity to an already complex situation, is really MS' own fault.

This was a simple example. The way things are going it may happen faster than anyone thinks that MS is paralysed into being unable to ship any 'fix' simply because the interactions between products is so great that every one of them is affected by any change in any other of them.

It really is NOT a sign of greatness (or macho or virility) that any product is complex, especially when that product is itself designed to be the basis of applications developed by 'users' to perform the REAL WORK of any company.
It really is self-defeating when one's product is so complex, especially with interactions between related componenets/products, the the needs of the users are wholly subservient to the difficulties/costs of making the product properly useable to the intended customers.

How can it be, based on this article, that ANY MS product ships with any bug that was clearly and properly exposed in a prior release??? I'll give you one reason... the "cost of the fix" is judged to outweigh the benefit of developing/installing the fix for the customer base. While business has morphed to make that kind of statement accceptable within its own circles, I can tell you that the majority of people attribute that reasoning as totally invalid and irrational. And, given the clear history that MS does indeed ship next versions with known bugs still in them, it does add strength to arguments that MS ought to publish the list of known bugs for every product it ships. They know its in there and have some rationale for leaving it there yet they choose to keep their customers in the dark. A fine way to treat a customer!

I am one who refuses to accept that there is any justification for shipping any product with known bugs in it. I accept that it can happen but cannot accept that such identified bugs remain hidden from the users of the product in question. That MS has developed such complex interrelated products would be fine and dandy if the customer was totally unaffected. But the customer is not only affected but is well below #1 in the decision matrix for how to handle the issue.

Finally, at least in the eyes of "average users", persistence of same bugs from version to version only ever serves to deteriorate the image of that product in the eyes of its users and that is something that companies used to care about. MS obviously doesn't care all that much, given their demonstrated track record. Yes, they do care, but only when their 'cost to fix' is reasonable in their own eyes.

Jim

>Gerry,
>I believe the opinion is shared by more than one person on one product.
>
>IMO, it's really all about priority. Yes there are bugs that have been left over for years. However, it sounds like there are also work arounds for those bugs that have existed for almost as long. If there is a "relatively straight-forward" work around, then the priority drops. It's that simple.
>
>In a utopian society, we could fix all the bugs found and also implement all the enhancements wanted. This isn't Utopia. Priorities are given and level of effort (LOE) is determined. If the LOE is high for a given bug and the priority is low (which it will be if there is a work around), that bug may exist for several development cycles.
>
>Just my $0.02.
>
>>This represents "one person" on "one product line" (kudos to him) ... and "one bug"; I wouldn't take it as an all pervasive philosophy/attitude across MS.
>>
>>I'm still creating work-arounds for VFP 3.0 bugs that (I guess) MS has deemed "too expensive" to fix (boo hoo).
>>
>>What's next ? ... It will be "un-american" to ask for (more) robust software ?
>>
>>Yep ... let's all sit back now and assume MS knows / does best ...
>>
>>Sorry, Garrett ... I don't ship software with "known" bugs, particularly in subsequent releases; and I won't buy into the concept that it's an acceptable way to run a business ... if you care about your users ...
>>
>>>Joe Bork wrote an interesting defense of why Microsoft would ship a product with known bugs. I recommend it.