Are Unfinished Products Now the Norm?
Paul asks: "Long ago when digital synthesizers first became commonly available, I recall a reviewer lamenting how he was getting more and more products to test whose software was unfinished and buggy and would require updates and fixes (this, before the internet allowed easy downloads, would have meant a journey to a specialist repair center). The review also commented how this common problem with computer software was spreading (this was before Windows 95 was out), and asked if it was going to become the norm. These days it seems ubiquitous, with PDAs, digital cameras, PVRs and all manner of complex goods needing after-market firmware fixes often simply to make them have the features promised in the adverts, let alone add enhancements. Are we seeing this spread beyond computers and computer-based products; jokes apart, will we be booting our cars up and installing flash updates every week to prevent computer viruses getting into the control systems? Can anyone comment on any recent purchases where they've been badly let down by missing features, or are still waiting for promised updates even whilst a new model is now on the shelves? How can we make the manufacturers take better responsibility? Apart from reading every review possible before making a purchase, what strategy do you have, or propose, for not being caught out?"
There is no such thing as an "unfinished" product. They're defective out-the-door.
Bill Clinton: Pimp we can believe in. - The Shirt!!!
It's amazing how much effort you can save when you don't take the time to do the job properly. As long as people still buy your product, there's no incentive to actually fix it before it launches.
Software is approaching the complexity of organic life. You know what it means for an organic being to be "finished"?
So what if our software is constantly changing, and is thus "unfinished"? To be finished means it won't improve. Heck, the whole reason for the existence of open source is the "if it's broken, I can fix it" idea.
So, why do we need software to be "finished," anyway?
The DVD recorder has some "issues" with recording to DVD. It's very fancy, otherwise, complete with 6 possible recording inputs and can do slide-shows off USB keys with photos. Nifty specs. It seems that the primary solution is to update the firmware. You would think someone at the factory might have attempted to record video prior to shipping it, alas, they apparently did not. (It is an intermittant bug that causes the audio to progressively lag the video). Hey - it compiles, ship it!. Since the process for updating the firmware seems non-trivial, is riddled with warnings, involves a USB key and I'm lazy - I haven't done it.
Combine this disturbing trend with product reviews that are little more than a regurgitation of the back of the box. (Along with some weird DMCA rules about what can and can't be reviewed on a product esp. vis-a-vis security.) Now you have a situation where you can't even get real reviews of products, and no review is ever "not positive." It's just that some are more positive than others. So, here you are, trying to buy a $500 video camera so you can tape the birth of your fist child and you aren't even really sure that any of them work. On top of that you can't even trust the reviews you read on various sites. I agree with you, this is not a good thing.
Leave the gun, take the cannoli -- Clemenza, The Godfather
That depends entirely on perspective. If entropy is seen as a relative process, software that is not being developed is going to suffer entropy as the world around that software is changing. So, in relative terms, there is software entropy.
Hypothetically, if your current 'perfect OS' software no longer has any development being done, when new storage devices or networking devices become available, that 'perfect OS' is no longer perfect. For this reason, all software will always be 'incomplete' in as much as the world around it changes at an ever increasing pace. Some software is outdated by the time that it is ready for launch as a beta product. For more on that, see the big software projects that some groups around the world have attempted, only to find that on launch it is not capable of dealing with recent changes in the world.
All software will always be no better than beta given that the above is true. This means that for businesses, good enough is as good as perfect as that is as close to perfect as it is likely to ever get.
Sure, there are cases where good enough really isn't; medical equipment, space travel equipment etc. but for the vast majority of software for consumers, beta grade is good enough and thus worth releasing.
Fortunately, some companies release beta software/apps and treat them as such by continuing to improve them before pronouncing the software is out of beta stage. When software is released as final product rather than beta, consumers get upset when they find out it's really only beta that they paid for.
But the point is, yes, software suffers from entropy and atrophy is relative terms.
Support NYCountryLawyer RIAA vs People
However, we routinely produce complicated systems that have excellant reliability. For example, glass displays on aircraft - which are quite common in commercial jets. They have to undergo a much more rigorous level of testing before they can be shipped because the liability to the manufacturer is huge. What's the liability if your Sony cam-corder stops working in the middle of your once-in-a-lifetime round-the-world vacaction, all because of a software glitch? The problem is not with the software, the problem rests partially with the people that make and test the systems, but mostly with the people who hire/fire developers, designers and engineers. They do silly things like higher cheaper, but less qualified engineers. They make marketings's brain-fart of the day the top priority. (I realize we're using the world's cheapest 16 bit micro-controller - but could you write the software in Java with a Gui so we can demo at Java One?) And they do things like sacrifice testing to make schedule. And they're also the ones that do things like set budgets and deadlines.
Leave the gun, take the cannoli -- Clemenza, The Godfather
Posting anonymously, because I review consumer electronic devices for a major web site. It gets depressing. I think I'm one of the few "reviewers" that don't reguritate press releases and/or the specs on the box. I work through each advertised feature and really try it out. I almost always find bugy user-interfaces, features that don't work, and features that are not documented. I used to start these reviews enthusiatically, but over time, I'm gotten more cynical. Today I'm working on a new review and finding the usual problems: Pop-up error messages that are blank except for an "OK" button, security holes big enough to drive a truck through and documented features that plain don't work. And this is with an expensive device that won a major award at an industry trade show. I look at the shiny box with the happy models and I read the glowing quotes from other reviewers and I wonder if they are using the same product I am.
Sigh...
How much does the glass display for an aircraft cost compared to your camera? How much of that cost is testing?
T Money
World Domination with a plastic spoon since 1984
Apart from reading every review possible before making a purchase, what strategy do you have, or propose, for not being caught out?"
Don't buy new products. Seriously, if it is worth buying it will still be on the shelf in six months. Even then I wouldn't buy it until I had read a few *user* reviews, immediately disregarding the top 10%. Check out some forums. Unofficial forums that is, publishers are notorious for nuking negative comments. I do not trust professional reviews. Ever. Even for existing software things can be pretty sketchy for a while. Consider how often Apple manages to botch iTunes, and that's their billion dollar baby. I know it's not what you wanted to hear, but you have to do your due diligence and be patient.
Frankly I don't see this problem going away until it is legislated away. If the bills concerning paid advertisements (i.e. the Sony PSP blog et.al.) have any teeth and clear consumer friendly rules, then reviews might have some value again. Not a lot, but some. Beyond that, liability is the only thing that's going to reign publishers in.
Platform advocacy is like choosing a favorite severely developmentally disabled child.
In the UK, if a product is advertised with certain features and those features either do not exist or do not work, you can return it for a full refund under the Sale of Goods Act. Items must be "fit for purpose" and advertising must be accurate.
That really is the best stratergy. If companies get too many returns, they will realise that their products are not up to scratch and either go out of business or fix them.
BTW, don't be fooled by retailers who claim you can't return things once the packaging is opened. The law appilies to everything, even software and things sold in those stupid "blister" packs you have to destroy to open. Just because the manufacturer made it impossible to find the defect without opening the product doesn't mean you can't return it. Even cars, which loose thousands of pounds in value when you drive them away from the dealers fall under the same law.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC