Perhaps Outlaw saved money earlier by not spending much on QA'ing the 950.

I think that is a bit of a cheap shot. You know as well as anybody out there that it is simply the nature of anything that involves software that there will be "bugs" uncovered as the product gets more use. What computer program do you have that has never had updates to fix bugs?

What seems to be glaringly missing here is an appreciation that Outlaw DID cover their bases in the design stage by having the software on an EPROM chip. I seem to remember some screaming months ago about the "non-upgradeable" nature of the 950. This was obviously wrong.

Give 'em a break!


------------------
The Soundhound Theater

I am pleased they are fixing some bugs their customers have found. This isn't the first such bug. Maybe Outlaw didn't thorougly test some basic functions in the 950 before releasing it to the public. It's like the hiss problem earlier. Maybe if they had tested the 950 on more home theater configurations with efficient speakers such as you and I have, they would have known about the hiss problem too, before many of their customers (like us) experienced it.

That being said, nobody is perfect and software often has bugs, that's for sure! This does not take anything away from the Outlaws for owning up to the problems and fixing them, and doing it on their dime. But maybe we should not feel too sorry for the Outlaws for spending money on warranty fixes, if the Outlaws saved money by not testing as thoroughly as we might have hoped.

[This message has been edited by Will (edited March 17, 2003).]

You are of course correct, but it must be noted that the incidence of bugs in 'firmware' and personal computer software shouldn't be thought of as comparable. In the case of firmware, the platform is well known (the target hardware) as is the operating environment (no other software installed by users), whereas a large portion of the bugs in personal computer software stem from the diversity of hardware and other software that are involved. This also makes computer software testing much more of a challenge.

To use a case in point, when was the last time your cell phone or microwave needed a firmware update to fix a bug? Ever bring your fuel injection unit in for a bug fix? This sort of thing could happen, but it is very very rare.

Also, the complexity of the firmware of something like the 950, while not trivial, pales in comparison to any significant personal computing application.

Sorry, but in this case I agree - QA should have caught this one - this is one of the 'primary functions' that absolutely should have had a test case.

I always thought a beta test of 5 was incredibly small, too small to be really useful as anything more than a marketing test. The bugs found bear this out. Our last beta was ~2500 users over several months, and at the end the product is very solid, even though that was a small beta test IMO.

I'm glad someone decided to socket rather than solder the EPROM. I hope if opening the case is required (I imagine it will be) there is some provision for those who don't feel comfy doing that sort of thing.

One thing for sure - this has been an opportunity for Outlaw Audio to show what their customer support is made of, and IMO they have passed that test with flying colors.

Nice to know About the way Outlaw takes care of their customers.

Question-- I just recieved my new 950 Thursday, March 13. Has this problem been fixed on it??

------------------

Charlie:

Maybe I should have used preamps like the Rotel 1066 as an example - bugs there too.

Tango:

Since the issue is still being resolved, I'd be pretty sure your 950 does not have a modification.

The way Scott's message read to me, I'm guessing they're working out some retrofit arrangement with a service provider and that owners will NOT be installing their own EPROMS.

I think the complexity is comparable to software because although you control the hardware platform you don't have control over all the software running the individual components. Trying to tie code from different sources together to run one thing in a dynamic situation where everyone is always changing their parts of the code is a nightmare. And my experience as a consultant is just like the Outlaws...my customers don't care that the problem was in the code that one of my sub contractors provided and I had no control over - they paid me - they complain to me (as it should be). Your first pass at testing is thorough, then you start spending all your resources resolving the known bugs while quietly a new bug is introduced by some seemingly irrelevant update. I think it is what's keeping Rolaids in business.

Well done Outlaws, keep up the hard work.

What in Scott's message makes you say that?

I've worked on huge embedded systems that spanned 10 - 30 networked processors and controlled 20+ million dollar machines, and it's not even close QA wise.



Sure you do.



See?

The integrator has the option, at at their discretion, to use or pass on updated libraries. If the integrator uses the updates, then it is their responsibility to make sure no old bugs were reintroduced (regression testing) and all test cases are revalidated. It's all part of an organized QA effort. What I see is a lowering of quality standards across the board driven partly (IMO) by the lowered expectations folks have due to their experience with software, which shouldn't influence them, but does.

[This message has been edited by charlie (edited March 17, 2003).]