Monday, January 19, 2004
It was 20 years ago today, Apple taught the world to play...
On 21st January, 1984, the Apple Macintosh first went on sale. Pitched as "the computer for the rest of us", it was radically new, radically different. Apple had a vision of the computer of the future, and put their money where their mouth was. Right from day one, they set themselves up as the adversary, the alternative - in those days the "enemy" was IBM and the IBM PC/DOS combination, and the arcane, awkward, and downright boring standard for business machines. On the other hand this was the era of the 8-bit games computer, and the Mac wasn't one of those either, though it was definitely closer in spirit to these than the IBM. Whatever the ups and downs of Apple over the years, and whatever you think of them (love or hate - there seems to be no in-between), the fact is, they were right - we are all Mac users now. Even if your particular flavour of Mac says "Windows" on the box. To the new generation of computer users who have grown up taking the mouse and windows for granted, it might be hard to understand what all the fuss was about. Back then, Apple had to include a tutorial program with every machine that taught you how to use the mouse, what an icon was, how to move a window. There is no doubt that Apple did create the computer of the future. The astonishing thing is, perhaps, that the Mac is still with us. Reinvented, certainly, but after 20 years you would expect that. Curiously, Mac OS X is not yet as usable and natural as that original Mac, though it's gradually getting there. In part that is due to the very ubiquity and familiarity of the graphical user interface - when the Mac was new every aspect of the GUI had to be explained, and made very obvious, natural, or at least forgiving of its users. Now, everyone "gets" it, so a few corners cut are permissable if it means getting worthwhile heavyweight features out of the door.
Personally, I have always been a Mac user, and in the early days, very much an evangelist for the platform. The GUI was definitely the way of the future - those who doggedly stuck to the IBM PC/DOS were missing the point, stick-in-the-muds, didn't GET it. For me, the Mac was the first machine I could actually get stuff done with. As Apple made error after error and allowed Windows to gradually catch up and eventually overtake, many of us who did get it were pretty dismayed, and gradually the enemy became Microsoft. But let's be clear - Microsoft did get it, in fact right from day 1. They so "got it" that Gates and friends knew they had to have Mac on a PC - they had seen the future and it worked. To many it seems unfair that MS got what Apple should have had, but the blame for this lies squarely with Apple. They had the ball, but fumbled it. It's a damn shame, but there it is. To my mind the success of Microsoft vindicates Apple's vision, and as such there is no point expending energy resenting it. MS's business practices, on the other hand, are another issue altogether.
I still choose Mac over Windows (note - nobody talks about the IBM PC anymore, though of course the modern Intel box is a direct descendant). It's partly a loyalty thing, partly a usability thing (I still feel Windows gets many of the details wrong - using a Windows box is usually an alienating and somewhat frustrating experience for me), and partly a quality thing - I've bought four Macs in my life, and all of them worked well, were built well, and lasted way longer than the average PC. Mac OS X is a thing of greatness. Apple are clearly back from the brink of oblivion and have been for some time. It remains to be seen whether the Mac regains some of the following it deserves - there are some signs that it is doing so, but it's hard to tell. The average computer buyer generally doesn't buy a Mac because there is a strong sheep mentality and pervasive ignorance at large, but so what, it really doesn't matter. People are fond of using car analogies to discuss Windows vs. Mac, and I'd rather drive a Merc than a Ford. The small market share of Mercedes doesn't worry anybody, so why should Mac users worry? They shouldn't. There is plenty of software for the machine, it plays well with others, and does everything that people need it to do. Be proud of being in a minority! You can remain smug and superior (if that's how you take your pleasure) when the next devastating virus knocks out most of the Windows machines in the world and doesn't affect Macs at all.
Where will we be in another 20 years time? Computers are still too hard to use, despite the great leap forward that the Mac gave us. It's unclear what the next leap will be, or even if there will be one at all. I think there does need to be - many of the aspects of computing we take for granted are actually pretty dumb, and as bad in their way as DOS was back in the day. Applications - huh? We shouldn't need to worry about applications, we should instead be focused on our data, documents, or whatever. It's the data I'm interested in, not the particular editor I need to manipulate it. Apple once again tried to move things away from applications with its OpenDoc effort in the mid 90s, but which died a death along with the old Apple (i.e. that before Steve Jobs' second coming). Perhaps it was an idea ahead of its time, and will emerge again, or perhaps they were barking up the wrong tree. Hard to tell. Networking - should "just work" - there is nothing so frustrating and demoralising to the average user than fiddling about with IP addresses and the like. Again, Apple had this sorted with its proprietary AppleTalk networking protocols, but since the rest of the world has condensed around TCP/IP, user transparency has suffered. Zeroconf ("Rendezvous") is addressing this issue, this time as an industry standard rather than a proprietary one. It will need MS to get behind it and adopt it though to make it really take off. Once again, Apple leads but we need to rely on MS to get it adopted widely. These things are not much of a vision of the future, they are just details. Computers should become so transparent, so obvious, so intuitive that we don't even know they are there. We should be able to just get all our stuff done without a second thought. I don't hold out much hope though, since usability of other devices (e.g. video recorders) has not improved greatly in 20 years either. Some things don't evolve much - the telephone has remained essentially unchanged since it was invented. Perhaps the computer of the future - 20, 50, 100 years down the line - will still be essentially the 1984 Macintosh, just a lot faster. One thing I do feel sure of though, is that Apple will still be there, leading from the front.
Posted at 09:37 am by GRAHAMUK
Monday, December 08, 2003
iBook screen hinge design
My everyday computer is a 2001 Apple iBook. It's pretty good, a bit slow by today's standards, but does most of what I need it to do very well. It started off life as a 500 MHz machine, but now it's running at 600 MHz, and the internal main bus is up from 66 MHz to a far more useful 100 MHz - in fact it now runs OS X painlessly. The overclock is not the subject of this article - I carried it out quite a while back and had not much of a problem with it - if you're interested, details can be dug up at http://www.xlr8yourmac.com.
One problem the iBook has always had since it was new is that the screen hinges creak terribly - in fact it sometimes feels as if the screen would break, it juddered and graunched so badly. Opening the screen requires a bold and decisive action to avoid this - adjusting it by a small amount was usually tricky, it was often easier to half close it then reopen it to the new position. Others have noticed the same thing - one of the overclock sites I came across also had a modification to the hinges to fix this issue - all it requires is that the friction rollpins are opened up a little bit. On its first take apart, I had a look at my iBook's hinges to see how it all went together. On that occasion I decided against the mod, since it requires quite a bit of dismantling, and I had enough to worry about with the overclock and hard disk upgrade. One thing I did notice though was how the iBook feeds the cables through to the hinged screen - it didn't look too well engineered, but well, I thought, Apple must have lots of experience of this, they know what they're doing and it seems to work, so I'm probably wrong.
9 months later my screen went black. At first I assume some horrible crash has taken place but on close inspection I see that the actual image is still there and working fine - it's the backlight that's gone. I tapped and jiggled a bit, hoping it was just a loose connection, but it wouldn't come back on. Then I closed the lid completely, opened it again and all is fine - until the screen is angled all the way back, at which point it cuts off again. A full closure and reopen will bring it back on. I soon discover there is a critical angle past which the backlight goes off. Easy fix - just don't open it any more than that... except that a week or so of using the machine and my back is aching like buggery, so something has to be done, and anyway there is no way I could accpet that sort of fault. In passing, I enquired of our local Apple repairers what the likely cost would be - a new screen is what they immediately suggested, and like the Canon G3 situation, is both the majority of the cost of a new laptop and total overkill for what must surely be a simple fault. That hinge design was also nagging at me....
I was reluctant to do a full take-apart on the main part of the machine at first - it's not particularly difficult but it's long winded, with about 50 or more tiny screws, and many little bits and bobs that need to be dealt with. Since the fault was in the screen, I decided to start with a that. I'd not opened the screen before, so had to feel my way, but it was quite straightforward (once I'd filed down the only very small hex driver bit I had anywhere near the right size so it fitted properly). In fact it was so straightforward that the main LCD panel nearly fell out - unexpectedly it's only held into its frame by four tiny screws. I quickly identified the inverter board that drives the backlight, and carefully wiggled the wires to see if there was a loose one somewhere. No luck - the fact that the backlight was affected by the screen opening angle was good evidence of a loose connection in the wiring, rather than a fault with the board itself. The fault remained, unaffected by my prodding and poking, so the fault didn't seem to be in this area. Next I turned my attention to the other end of the cable feeding the board. As luck would have it, this is one of the few connectors that can be accessed without taking the main machine apart - you can lift out the keyboard then the memory expansion cover underneath to get at the connector. I disconnected this, wiggled the wires, put it back... no luck, same problem. The cable runs from that connector through the hinge to the inverter board in the screen - disturbance at either end had no effect so the fault had to be in the middle of the cable - and right in the middle of the hinge. A full take-apart was now inevitable, and my concerns about the hinge engineering were seemingly well founded.
A careful 45 minutes or so later, the machine is in pieces. This time I have to dismantle the hinges altogether so I can get at the cable in its entirety. This involves unscrewing the hinges from the main chassis, and also, from within the screen shell, unscrewing two screws which hold the hinge cover in place. This is the metal-looking thing that you can actually see joining the screen to the main unit. Once done, the cable run is fully exposed. What happens is that a bunch of wires (about 5 or 6) is scrunched into a tight bundle, wrapped in a metallised cloth sleeve and tightly wrapped with cloth adhesive tape to form a cylindrical cable. This then passes in front of the hinge, turns a tight 90° bend to go through to the screen, and then spreads out again to link to the board. The bunched cable is trapped tightly between the hinge itself and various openings in the plastic shell and the hinge cover - it is more or less unable to move and in addition is strongly chafed by the opening in the main shell as the lid is operated. It is a recipe for failure. In addition, the fact that the bunch is unable to move means that over a short distance of this cable, say about 10mm, the full accommodation of the changing angle (about 120° total) must be met within this length of wiring. Again, this is certain to fail after a while - think about twisting a thin piece of wire backwards and forwards until it breaks. If a longer run had been allowed for accommodation of the angle, things would be better - but 30mm or so would be a minimum to guarantee reliability.
Sure enough on exposing the wires within this bunch, one was found to be broken clean in two. The others looked OK despite the strain they must be under. I quickly fixed the break and bound up the cable again. I looked for ways to alleviate the twisting problem, but there is so little room - a little bit extra could have been allowed for in the hinge design, but it wasn't, so things are very very tight. Still, it looks nice, so that's OK then.... From now on I'll be moving the screen far less than I used to. Reassambled, everything is now fine (and I finally fixed the creaky hinges) - but I wonder how long it will keep on working. More worrying is that through the other hinge, there is a similarly tightly wound film ribbon cable carrying the image data to the screen. This cable is extremely fine, and if it broke, would be just about impossible to fix. No doubt a replacement isn't listed as a service part. So, in conclusion, while I'm a great admirer of the industrial design of Jonathan Ive, who designed the iBook, I feel that some better engineering in the hinge area would benefit the machine greatly. It would only take a few extra millimetres of room within the hinge itself, and some attention to where the cable goes as it passes along the front of the hinge - the key factor is that the cable should not be trapped tightly here, but allowed to move freely to accommodate the range of screen angles over a reasonable length of the cable.
Posted at 01:56 pm by GRAHAMUK
Thursday, November 13, 2003
Another saltwater damage tale
The Canon story (see 10/11/03) reminds me of the time, about 1979, I worked for a company making two-way radios, called Dymar Electronics - now long forgotten and passed into history, sadly. During my apprenticeship there I worked for a time in the repair department. In those days we repaired right down to the individual component level, no computer diagnostics, and very rarely replacement of entire assemblies. Instead we relied of proper knowledge of electronics, a little deductive reasoning, and persistence!
One day a set came in - an 883 portable if my memory serves - that was owned by an oil-rig operator, and had been found by an inspection diver lying at the bottom of the North Sea having been dropped off the side of the rig by accident six months previously. Naturally it was in a pretty poor state! Being the college boy it was assigned to me as a "challenge". I'm pretty sure that everyone expected it to be simply written off and replaced, but hey, my 17 year-old labour was cheap and they probably thought it would shut me up! The damage was fairly extensive, with significant corrosion of the metal shell of the set, and the two circuit boards were heavily salted. The Ni-Cd battery was definitely beyond recovery. I took the set up to the production area and got them to run it through the post-soldering cleaning tanks - huge vats filled with hot steaming trichloroethane - and then set to work. The receiver was basically functional, requiring little more than a general retune and setup. The transmitter worked but had very low power. This was traced to a blown output transistor, which was duly replaced. The speaker/microphone had to be replaced, as did the handset connector. One tuneup later, it worked perfectly. Total time, about two hours, total cost, maybe a few pounds. A couple of new covers and a battery from the stores and the owner got their set back for a very reasonable charge.
The upshot was that the company greatly impressed the customer, who had sent the thing back as a joke, it turned out. Whether a coincidence or not I don't know but we got a large order from them the following year. Maybe "things were different in my day", but Canon could learn from this.
Posted at 07:57 pm by GRAHAMUK
Monday, November 10, 2003
Servicing the Canon G3 digital camera
OK, first rant. Now, I own a Canon G3 digital camera. It's just a brilliant device, I'm completely satisfied with it in every way and would recommend it to anyone. In fact I carry it with me most places I go, and that's where this little tale of woe begins. A few months back, I was up on Australia's Gold Coast, taking a walk along the clifftops and admiring the rather wild sea below. Going close to the edge to get a better look and a picture, a sudden freak wave suddenly washed straight up the cliff-face and rained down on me with some force - I got out of there as quick as I could and started to count my blessings that I hadn't been swept away! Drying off, I noticed that the G3 was now sitting in a few millimetres of salt water in its carry case. I quickly emptied it out and dried it off as best I could, also removing the battery "just in case".
Later, I tried to operate the camera and discovered that it was working rather oddly - perhaps not unexpected. The lense extension/retraction motor would work intermittently and the camera would not focus properly. Then it would work. Then it wouldn't. After a while it kept reporting a "lens" error and shutting down. In short, it was buggered.
I let the camera dry out for a week, but no joy. I took it to my local friendly camera dealer who said yes, they could send it off to Canon in Sydney for analysis and repair, but they also warned me that Canon were notoriously slow getting around to doing the work, so not to expect it back anytime soon. Also, it was fairly expensive to ship it there and back and there was a considerable charge just for them to look at it. Well, the prospect of having to buy a whole new camera didn't appeal, so I went ahead.
About 8 weeks went by, then I got a call from the camera shop. They'd received the report from Canon - the main optical assembly and main PCB needed replacement, the work would cost AU$899. They also offered a replacement new camera for $1,099. I'd just seen the G3 advertised at a discount camera centre for $999, so this didn't seem very generous. I said not to bother, please would they return the camera unrepaired. Two weeks later I had it back. Inserting the battery and powering it up, it worked perfectly.... for a few minutes. Then the "lens" error message flashed up again and it shut down. Back to square one.
Or was it? The fact that it worked once demonstrated that no permanent damage had occurred to the unit - it was capable of operating properly, but something was preventing it. I've seen what salt-water can do to an piece of electronics - it ain't pretty, but then again corrosion is usually in the form of a semi-conducting salt that is easily cleaned off. I figured that's what must have happened to the camera, and knowing that the amount of water that got in there was probably quite small, I figured I'd have a go fixing it myself. Nothing to lose after all, I was resigned to buying a new camera anyway.
Carefully removing first the front, then the back cover of the body, I could see that the bottom part of the chassis showed minor signs of saltwater ingress (discolouration of the steel chassis), but it was only a few millimetres deep. The water had not touched the majority of the innards. In passing I was impressed at both the design and the build quality of the unit, a triumph of miniaturisation. The main board is at the back of the camera, and there are three connectors at the bottom edge of the board, connecting film-wire (flexible circuits) to other parts of the camera. One of these was lower than the other two and in the "wet zone" of the corrosion. Carefully unlocking the connector and releasing the film wire showed that the tiny fingers of solder were bridged by salty corrosion. Taking a pencil eraser I carefully rubbed it off, using an eyeglass to check it was completely clean. I reassembled the connector and powered up the camera, covers off. Success - it worked perfectly, and still does now, weeks later. The fault has been fully rectified.
I'm very happy to have my camera back, but at the same time pissed off that the repair dept. at Canon Sydney are such a bunch of inept losers. They knew it was salt water damage, the extent of which was easily seen by the telltale "tide mark" on the chassis. Given the symptoms of the fault, the cause was pretty easy to guess at. So glibly opting for replacement of the two largest and most expensive parts of the camera without proper investigation is to my mind a failure of good service. I'm not even sure they took the covers off the camera - they probably plug it into an external diagnostic tool and blindly do what it says. I can imagine that in many cases of damage to a camera there is little can be done other than replace major assemblies, but this case wasn't like that. The fact they couldn't tell the difference tells me there is something wrong somewhere with Canon's training, procedures or attitude. Incidentally the whole repair took me half an hour and I'm not even familiar with dismantling the camera, so there would be little merit in claiming that a more thorough investigation would be too time consuming. It's just lazy.
Posted at 10:38 am by GRAHAMUK
Well, I thought I'd have a go at this blog thing. Not that I expect anyone to read it, I mean, how can one advertise these tiny little backwater websites? Anyway, let's give it a go.
I'm an electronics engineer by trade and training, and engineering topics are what I like to talk about - or rant about, if I'm being more honest. Often these rants/discussions need an outlet other than my long-suffering partner, though I have to say she is very patient with me. Hence those rants will often find their way here. They usually fall into one of two categories - good/bad design, and good/bad service. So expect lots of examples of those. The other purpose is to share ideas and thoughts about the various projects I work on, or have worked on in the past. This might be interesting to somebody, and if I ever get any readers, getting feedback will be helpful to me too.
Some basic info up front. I'm 41, educated as an electronics engineer and spent the first 12 or so years of my working life designing hardware. Then I migrated into software and spent the next 16 or so years designing and writing code. Mostly this has been on the Mac, with some other stuff thrown in too - embedded, DOS PC, but next to no Windoze. I've also always had an interest in all things mechanical, and have always done my own tinkering with cars. I owned and restored a classic 1974 droopsnoot Firenza - something of a love/hate relationship with design that one! I'm also a small-time pilot, having gained my PPL in 1999, flying 172s for fun. Last year (2002) I emigrated from England to Australia, living in the country, so lots of scope for all sorts of interesting projects here - solar and wind power pencilled in among them.
I hope you'll find this interesting! --Graham
Posted at 09:59 am by GRAHAMUK