Imagine this: Your car fails and you're stranded. Instead of sending a tow truck, your mechanic electronically connects, runs some diagnostics, and enables some alternate circuitry. That fails, but they take additional readings and restart modules in a different sequence and with a bit of diddling (yeah, that's a valid technical term), your car comes to life.
For Hubble Space Telescope fans: NASA has been achieving steady progress toward restoring operation. The 'scope failed on Sept 27 and NASA delayed a Space Shuttle repair mission, which will be the final service call to the Hubble.
They decided to switch over to the "B side" electronics -- a dicey proposition, since they have been idle for 19 years. Imagine finding a radio in your garage that you bought in 1987 and turning it on, and you get the idea. Except that this is a bunch of radios, three processors (including a pre-Pentium Intel '486 and a '386), and a mess of — well, other stuff.
The blow-by-blow description has been interesting to watch. The systems tossed themselves into "Safe Mode," powering down due to detected faults, but these so far have been artifacts. As they learned the causes, they have gotten more and more systems to wake up. Most recently, they found that the safety system was checking for a supply voltage before the monitoring gear had collected the data. The voltage was up, but the monitoring gadget didn't know it yet.
Who doesn't love the inspiring images we get from the Hubble Space Telescope? Alas, it has had troubles and just weeks before the launch of its lastrepair mission, Just two weeks before the next schedule repair mission, the whole machine has failed. NASA has a good record for rescuing victory from the jaws of defeat and the mission is delayed while they revise the strategy.
This interesting report, on the blog of electrical engineer Steve Leibson, gives an engineer's inside view of the troubles.
Why do I dislike Microsoft? They're really a great company, with smart people, and have contributed a great deal. As much as people dislike Bill Gates, I never did. I really admire his charitable work which is precisely what he said he was going to do a decade ago. Their trade practices haven't been pristine but for the most part in recent years, it's been legal and mostly ethical. They're ruthless competitors but that's not, in itself, evil.
I also credit them for fixing the Macintosh! If Microsoft hadn't launched Windows, the Mac would totally suck.
And yet, Microsoft makes me nuts.
This is their biggest offense, in my book. Microsoft pretends to follow industry standards and common practices, while really co-opting, or even sabotaging, them. Under the guise of enhancement, they add "features" in a way that damages competing products. The best example is their web browser. Generation after generation sought to derail web standards, adding elements (such as ActiveX controls) designed to get web developers to write sites that would not work well on other browsers. It is a pleasure now to see them outfoxed.
Another example: Their Exchange mail server uses a Microsoft-proprietary format for attachments and text formatting. Non-Outlook e-mail programs can't read them. There are standard methods for doing this, but Microsoft does not adhere to them. You can configure the server to not do this, but you should not have to.
One more: SharePoint is the most popular collaboration system for companies. It works pretty well if the user has Windows and Microsoft Internet Explorer. Other browsers and systems are supposed to work but it's a hair-pulling experience. As far as I can see, everything they're doing could be done in an industry-standard way that works universally. I can't tell if the Explorer-dependence is deliberate sabotage or incompetence.
It makes me crazy because it is arrogant, anti-user, and predatory.
It's one area where they exploit their leadership in a way that damages
Steve Jobs once said about Microsoft that they "have no taste." Say what you will about Jobs and his arrogance (I would not argue) but he is right. They are successful — and Jobs does not begrudge that, a least in this video — even if their products and strategies are artless. He says, "they don't think of original ideas and they don't bring much culture to their product."
It's obvious to me that engineers design their interfaces and graphics. I recently saw the new release of MS Office 2008 for the Mac. It was demonstrated for me by a Microsoft program manager, who proudly showed me 3D charts in PowerPoint. I was horrified by the amaterish 3D. They had added shadows and highlights that were blocky and unrealistic. The colors were Crayola basic. Labels were improperly centered on shapes. To many business users, I am sure these look fine, but any top-tier company would have them redrawn before using them in any material that goes outside the company.
There is nothing evil or badly intentioned about this: I think it's because they're unaware, like the engineer who dresses in plaid and dots and is honestly befuddled when it's pointed out.
Then there are the user interfaces. Microsoft has some of the most advanced user interface labs in the world and a large UI staff. Yet, their interfaces are consistently confusing and convoluted.
MS Office is a case in point. It has tons of functionality, but finding and performing a function is often painful. The worst part is that with each revision, it becomes more painful! Why? Because their UI labs are very good at identifying user complaints, very poor at fixing them. When they find a function people can't do, such as mail merge, they never seen to think that the user interface is fundamentally in need of ground-up redesign. Almost invariably, they think the answer is to add something. They added a mail merge palette. A help system. Then that paper clip animation. Then a wizard. Then a "ribbon." A sequential palette. Now they have a bar thing that has no name that brings up a series of icons that, um, does something. Like a wizard. But they can't call it that because the wizards, um, didn't help.
They seem to toss solutions at the problem, never quite figuring out that the problem is underneath all these wizardly layers.
The project manager who demonstrated Office 2008 for me showed me the wizard-like bar thing they had added for special user operations, like mail merge and templates. It takes up about an inch-high row of the document window. I asked if it could be made vertical. "We thought about that," he said. "Especially with displays tending to be wider. But no, it's horizontal only."
Duh. Even when they know they're making a mistake (and they usually don't), they proceed and make the mistake.
And then, there is the underlying stuff: The code. It's bloated. It's buggy. Year after year, release after release, many bugs remain. It's slow and inefficient. The new Office 2008 for the Mac takes longer to load (a lot longer) and is not noticeably faster than the six-year-old, two-revs-ago version I had.
Someone did an interesting comparison between the open-source web server, Apache, and Microsoft's server, IIS. These "call graphs" hint at the underlying structure and you don't need to be an expert to see the difference.(thanks to Paul Rako's Anablog — which presents his own case against Microsoft — for showing me these.)
Is this typical? Meaningful? I wouldn't know, but I would bet it is.
I Don't Hate Microsoft
I don't hate Microsoft. I certainly don't hate their people. I don't even hate their products or what they have done for the industry. But I hate using their products, I hate being slave to their manipulations and incompetence.
Most of all, I hate it when I have no alternative.
For those in the Bay Area, I would encourage you to attend Maker Faire. May 3-4 at the San Mateo Fairgrounds -- and not to be missed!
I went last year and vowed next time to go both days. It's huge and full of really, really fascinatng exhibits and brilliant ideas. Lots of free stuff.
What is it? It's the heart and soul of do-it-yourself. Hundreds of people who make stuff and love what they do so much that they stand for 12-hour days waiting for people to come by and gawk at their magic. Organizations and museums that have created hands-on demos you can play with.
If you like metal and fire, gears and gadgets, Legos and things that move, honk, squirt, squeak, spark, and belch, you have to go.
In Scientific American: "Designed nearly 150 years ago but never actually built until recently,
the Difference Engine No. 2 designed by Charles Babbage (1791 to 1871)
is a piece of Victorian technology meant to tussle with logarithms and
trigonometry long before the first modern computer. Technophiles have a
rare opportunity beginning May 10 to see one of these devices (only two
exist) on display at the Computer History Museum in Mountain View, Calif."
I think we have a ways to go before machines can translate text. We received this at work.
Note the last sentence, in the P.S.
The customer is asking about samples that did not arrive:
The top of the morning to you. Ex rubs weeks not get to my house samples disposal ( 499923; 499390; 498997; 498983; 498218; 497302; 497294; 497285; 497181; 497179) Whether item what yourselves history ex commissioned sample? Warm-heartedly thank you too those whom received. Readies produces panel LED built of disposal your part. Very I am set on it disposal part whose plannings shopping are not to level pcs/ yer. If you please of correspondent on that deal. If you please quote nr touch to can pay in dollar too ordered disposal. Good day.
P.S.Pardon me too my english. Beneficiary ex translating program.
A new online service, HealthVault.com, provides a single, web-accessible repository for all your health records. There is much here to like, but — there are a lot of buts.
It has always bothered me that every new doctor or medical service asks the same questions, and each with a separate form. Dentist, specialists, emergency clinics, labs, pharmacies — same questions, again and again. What drugs are you taking? Which of these diseases have you had? Family medical history. It's inconvenient but more important, it compromises one's health care: I'll bet I have answered these 100 times and there are certainly differences, either because things have changed or I don't remember every detail.
This can be important information: A doctor's diagnosis and treatment depend on what the doctor knows about you. And in an urgent situation, or when the patient is not lucid, having accessible medical records can be life and death. I really want better online medical records.
As a technologist, it makes me especially crazy. Everyone else, from the DMV to the grocery store, has an online record system that is better than the medical community's. There are legitimate issues but nothing that can't be solved — and that have already been solved. Every medical office can readily access my credit information already. Clearly we can solve the privacy and networking and access issues.
Some worry about the security issue: Having all your medical data in some centralized database worries people. But do you really think your privacy is more guarded when there are 25 uncontrolled, unencrypted copies of your records, in 25 medical offices all over the county, immediately readable by 100 office staff?
Along comes HealthVault. It's a good start. It keeps all the information for you and your family under what looks like tight security. It holds contact and profile information, medical records, etc. You can upload image files and documents. The sharing facilities seem robust — you can invite someone to view the information and choose what they may see.
Two issues are access and trust. In terms of access, will my doctors accept a HealthVault login as a substitute for their form? Not yet, but they have a few partners who are using the date.
The bigger issue is trust. Note that I did not say "security" — that's a technical issue. It's the human issue of trust that will make or break this, and other, medical automation solutions.
HealthVault is owned by Microsoft. Do you trust Microsoft? I don't share the common distrust of of their corporate intent — I think they mean well and their corporate ethos is that technology is powerful medicine to help people. They also believe that profit and benefit go hand in hand. So I trust their intent.
I am not sure if I trust their technology. They're competent technologists. They certainly are capable of making an accessible, trustworthy, secure system. But will they?
Every Microsoft product is full of defects. Even after years of refinement, glitches remain — sometimes the same issues for year after year.
I worry about bloat. Huge applications mean more opportunity for bugs and functional flaws and Microsoft often seems to not know how to say no to a feature. Worse, they keep adding rather than fixing — a palette, an assistant, a wizard, a "ribbon" — layer upon layer of new things to repair incomprehensible interfaces only a programmer could have designed.
This all affects usability. I am always in awe of what they spend on usability testing and redesign. They seem incapable of simplicity. The continuous repair tells me they know there is a problem but don't know how they are causing it.
I tried to edit my contact information. There is no link to let me do that. I think I know why. I haven't confirmed my identity by replying to the e-mail, but guess what: It's not telling me that. There is no Edit link and they give the user no clue. It's programmer-think: If a capability is not available, don't show it. Would Apple or Amazon or Yahoo have done this?
So, will I use this? Even without my
doctors' buy-in, there is enough functionality for me to get started.
But that's mostly because I am a tech geek. Others would not likely try
And then there is the trust issue.
I expect that the underlying security is fine. But it's like the president of United Airlines once said, "If the tray tables are dirty, the customer has to be wondering about the condition of the engines." As much as I may want this service, I am not going to be uploading my data, given how dirty their tray tables are.
Serious Eats Serious Eats is another blog which plants its stake, and its steak, in the fertile land where food meets science. There is lots of great material here, much of it by J. Kenji Lopez-Alt, or Kenji. If you like FeedMe, you will like Serious Eats.
Cooking for Engineers What do you get when you apply the engineer's mind to the kitchen? Straightforward, practical recipes and tips and a passion for simplifying without sacrificing quality.