Well, much the same observation was true comparing a Commodore 64 motherboard to an Apple ][. The Amiga was made, by necessity, very much in the usual Commodore fashion.. full of custom chips and gate arrays. The original IBM PC was a few Intel interface chips and some hundred TTL parts... it was massive. No integration, no time for integratiion, they tossed it together in something under a year.
Now, that's one reason they cost so much, one reason the motherboard was so huge. But also, a big reason the IBM PC was cloned. Even neglecting software issues, you couldn't easily clone a C64 or an Amiga (yeah, sure, today you can, but it takes a reverse-engineering of the custom chips and a nice FPGA or two). But all the forces lined up behind the PC. It didn't even matter if it was any good, really... the was this pent-up demand. Many people wanted to be in the personal computer business, few were capable of launching something like an Apple or a Commodore or Atari -- there were plenty of attempts, plenty of failures.
With the PC, you had a perfect storm: easy to clone hardware, reasonably easy to clone BIOS, and the OS and all chips you could buy, off-the-shelf essentially, from companies who were not IBM and were really happy to have you as a customer, too. It didn't really matter that early PCs... sucked. Eventually, they fixed big chunks of the architecture, and most of those bad ideas are relegated to a tiny chunk of silicon in the corner of a big system chip, anyway.
Yeah... part of the end-game at Commodore, for those of us who did stick it out in '93 and into '94.. to the bitter end, was still kind of hanging onto the idea that there was a technical solution to the problem of corporate stupidity. There never, ever is. That's why I only made the one film [frogpondmedia.com] about why a company failed. Much of it shot at Randell's old house, in fact (see the extended trailer here [youtube.com]). -- -Dave Haynie
I agree.. this one is pretty cool. Certainly much closer than the Kindle. A couple of things:
* Book Price -- if B&N follow the Amazon pattern, they'll release ebooks at the same time as hardcovers, for less cash. That's a good thing, for certain, though if you're one of those who waits for the paperback, the price might be a bit more. Not that paperbacks are all that cheap these days, compared to the $9.95 "standard" price for many ebooks.
* The e-ink is a must right now for readers. Anything that qualifies as "book" has to be readable on the beach, or it's a no-go. These e-ink systems get better in bright light, worse in the dark... they behave very much like, well, ink on paper. There's no power used to keep the image up, just to change pages... thus, 10-days of reading.
* Also, changing to higher power displays, adding games capability... no! Anyone asking for that stuff ought to look into these things called computers. They already have this stuff. You can already read ebooks on them. Making an ebook reader do this stuff will give you both a bad ebook and a bad computer.
* The ability to lend it makes for a much better "book" model, naturally without the nasty aspect of having to get that book back from the lendee. Libraries ought to be able to tap into this mechanism, too.
* SD expansion and direct PDF support -- kudos there! That's a major issue with the Kindle. What's the point of an eBook reader that can't store all your books. Sure, whatever's left out of the built-in 2GB is sufficient for 1500 novels, but once you talk about illustrations (eg, PDFs, even if converted to monochrome or halftone for space savings), you're getting big. I have over 2GB of PDF datasheets for the current hardware project I'm doing at work... it would be fairly cool to have these all on-tap on a device like this (also makes it tax deductible... hmmm...)
Two big problems still, which really are the same problem. The book model fails if I can't resell the book somehow, if I can't read it on a different ebook reader (though at introduction, B&N is supporting "alternate reader" on way more non-ebook devices than Amazon, and you have to believe Android is going to be on the list soon, given they've already written that code for the Nook already).
The ePub format is a good move.. XML based, world standard, all that jazz. But ePub still supports the option of a DRM, but doesn't specify a DRM. I'm sure they're using some DRM, perhaps a proprietary one for the Nook. My guess is that the books you buy are downloaded keyed to your Nook, and will be an issue to read anywhere else. I would love to be wrong about this. Hopefully, there's more information on just what they're doing here.
I really do want to see a "book" model that really behaves like a book. That's a problem on computer systems... too easy to make it copy, which is not what I'm after... I just don't want to give up the rights normally associated with real books. They could certainly facilitate reselling of ebooks online, just like Amazon does today with used books... that would vanish if B&N ever got out of the business, or a different standard prevailed in some years.
Obviously, this reader will be useful at least for ePub, non-DRMed, for other sources if B&N fails in this can cancels their services... I guess Kindle does that, too, although Amazon's ebooks are proprietary format, and you need Amazon involved to get PDFs on a Kindle.
Well, it's relevant in the same way that 65xx chips have been long relevant.. they're hidden in everything.
At Nomadio, my current company, I helped launch a fairly revolutionary product into the radio controlled hobby market. This was a digital radio, first of its kind, which ultimately became the fastest R/C radio on the market (latency-wise), as well as the only one upgradeable over USB.
One other feature.. it was 2-way. The "receiver" could accept a number of sensors, to monitor temperature, voltage, or motor speed. The motor speed device uses a hall effect sensor, a tiny CPU, and a magnet to measure speed, pretty much anywhere you could hack it in (like the early computer days, most serious RCers are pretty skilled at tweaking their stuff). I originally used a Cypress PSOC, a tiny 8-bit microcontroller with flexible digital and analog function blocks. Unfortunately, that year, Apple was using these to read that silly ring control on the iPod, so all of a sudden, I couldn't get those. I put a junior engineer on the job of finding a replacement... turned out to be a Z80 variation... in the 21rst century. It worked great.
You know, you're right... one reason I decided to write a bunch here.. I figured some Amiga fans would look.
Escom actually did commit to the platform. Not to the extent we all wanted, right off the bad, but in hindsight, I think at least part of that was due to their money problems, which we didn't fully realize at the time. The did establish Amiga Technologies, AT brought in Andy Finkel and me to work with them on what they'd like to do, and what it would take to do that. It was all done at a level that, had it continued, would certainly have been successful... not this kind of quagmired stuff that happened post-Escom. Shame they blew it in the PC market enough to fail so fast.
Amiga people went all places after C= ended. There were a number of Amiga software people on the BeOS team, even more in the BeOS community, at least back in the day. And I know a bunch who got into OS/2 as well. Particularly if you meet someone complaining about all the stupid things, single threading, serialization, etc. in Your Development OS of Choice, you'll win more than lose betting they learned to program correctly on the Amiga.
Oh, and thanks for video link.. I had not seen that . Even the simple fact of an animation genius like Eric coming into his own on the Amiga should make people who don't understand it all think a bit. It's also evidence we were there... art is usually more persistent than technology.
Some of that's actually going away, slowly. It certainly depends on your workflow.
I used to have MIDI, with Bars and Pipes, on the Amiga, which led me to Cakewalk on the PC, and one of those MOTU 8-port MIDI devices, all kinds of stuff. These days, while I could record on the computer, I'm more likely to record on my Fostex portastudio, then bring all the raw tracks in for mixing on the PC. If you're in a big studio, you're going to have external digital ADC/DAC and mixing, so the PC itself needs just some good digital ports.
Curiously, the "PIOS One" project at PIOS/Metabox was exactly the idea of this... a personal computer optimized for audio/video work. Even in 1997, though, it made sense to think most I/O was going to be external. I had a good sound chip (Aureal... another good company that failed, largely do to evildoers from the outside.. I hired some of the Aureal engineers briefly at Metabox USA), with separately regulated and filtered audio power supply. But going beyond four channels, you would hook an external box to the audio expansion port, and bring it in digitally.
Lots of ports, sure. Firewire was critical to video, but that's going away... tapeless is coming on like crazy, and it's great. I bought a little pocket-sized camcorder awhile back, a Sanyo VPC-FH1, which records on SDHC cards. My goal was to reduce wear and tear on the expensive HDV camcorders, but the video quality out this bad boy is crazy.. and it can shoot at 1080/60p, twice the rate of any Blu-Ray mode. But the real key.. flash memory means fast, totally reliable transfer. There are a number of low-end pro cams doing the same things, and even on the high end, lots of people using SxS cards are loading up two SDHC cards into an adapter.
For DSPs, I don't think so... they're just not cost effective. This is the same thing Be realized, going from their original Hobbit+DSP3210/07 prototype to the PPC model. Signal processing on the main PC is really fast these days, and you get to use that power for general purpose computing, not just some specialized bits. There are interesting areas of computing acceleration, but I would look at GPU and FPGA computing, not DSP. If you really like DSP, you could always build one in a system with some kind of FPGA resource. Not just build it, but build a different one optimized for the specific work involved. Both of these run into software issues, too... special support not needed for native signal processing work. So the benefit has to be large.
You definitely want a 64-bit file system, and one tuned to do media well. The last time I built Linux video servers (just over a year ago, eight x86 cores in a 1U rackmount), I got better performance from XFS than ext2 or ext3, which wasn't a shock. I didn't mess around with ZFS, and ext4 was still a work in progress, though they seem to have been moving in the right direction. One nice thing about Linux... all these FS choices. The same thing that optimizes streaming HD video doesn't necessarily optimize a zillion tiny file accesses in a web server.
The best way to implement security in a multimedia environment is simple: don't connect the media network to the internet. Problem solved. Of course, if you're paranoid, go ahead and run the web browser in a sandbox, that's fine.. you don't want the rest of the system slowed down by VM overhead.
Right... it was possible in the 1985, it wasn't so much ten years later (BeOS was first out to developers in 1995). Or MacOS in 1984, if you want to look at something that's still with us.
Some of that's based on the competition... there was no PC GUI of any major impact in 1985. All personal computers were pretty horrible, too, so true innovation was something people at least took note of. And it was still possible to do that. In modern times, not so much... you pretty much have to be a major chip company to make a significant piece of a modern PC subsystem. No more systems folks like me experimenting with gates or PALs in a lab somewhere.
BeOS was a very good thing, too, but unfortunately, they made a few mistakes. Launching a whole new platform at that time was difficult at best. The BeBox was very cool, but too expensive for anyone but a geek. They could have made a Be motherboard that fit a bog standard PC... but where's the fun in that:-) It's tough, when you're a techie, and evaluate things on the basis of knowing how they should work.. you can see something beautiful, but you have to calm down and realize that most people just want to get their work done and then go yammer about it on Facebook.
Anyway, it seemed reasonable to go software-only, once actual dual processor systems were becoming more common, but that turned out to be a mistake. There was this curious de-valuing of operating systems around 1998-2000... Linux was free, Sun put out lower-end versions of Solaris free, MacOS updates were still free or at least very cheap. Really only Windows was being sold for significant per-copy cash.
Even with Windows, the in-box licensing was a larger business than end-user upgrade. And they successfully blocked the inclusion of any alternative OS... it was a violation of the Windows licensing agreement to build-in a boot manager and allow another OS to run. So, while a few companies wanted BeOS, they didn't want it to the exclusion of Windows. So they went to BeIA, the BeOS for advanced set-top boxes.. but they screwed that up, too. And they were late... I had approached them in 1998 about a set-top-box license, they were sticking strong to the $50 per copy, no hassle price. We wound up using OS/2.. $10 a copy... IBM understood the embedded market. Linux was still not ready for multimedia.
The old tradition of home computing was really about having a computer-centered hobby. You might play games, use a wordprocessor, etc... but the computer itself is the focus. You got Commodore or Amiga magazines, Fish Disks, you joined a user group about that computer, wrote on usenet groups, etc. A PC today is just a tool to help you do something else. I use Altium software for electronics design and Sony for multimedia content creation, and I care a bit about those programs.. I have some loyalty there. But as long as it's a day when Windows isn't biting me in the butt, it's just a means to an end. The OS really doesn't matter. If I could run these under Linux, I probably would, but it's not a big deal either way.
It's pretty clear Microsoft and Apple don't really understand this, either, or at least admit to it. I do recall when the introduction of a new OS was a big, class-A geek event. But, despite Microsoft's creepy videos, does anyone even care if Windows 7 is coming out on Thursday? Sorry, not me.. I'm planning to spend the next 2-3 weeks drooling over the Motorola Droid. Now that's excitement.. finally a pocket computer to make all those iPhone users around me a little jealous. Though I suspect mostly, that'll be along the lines of "wow, you get a signal here...".
The Amiga CLI has nothing to do with BSD... it was adapted from TRIPOS. Sure, we ran some of the same GNU ported shell tools as everyone else, but there was no UNIX in AmigaOS... and that was a good thing in most ways. Believe me, I had been using UNIX since System III in '81 (a summer job at Bell Labs), there was an intention to not copy UNIX. Not a slight against modern UNIX, but it was impossible to do anything remotely multimedia-friendly in an 80s vintage UNIX system. That's why most of the workstations used DSPs if they wanted audio... they could offload the whole audio job to an RTOS on the DSP. The OS couldn't keep up. That was fixed, to an extent, in modern Linux, but back then, no way.
I tried really hard to be a Macintosh fan, but after Apple did their best to kill my Mac Clone making company (PIOS Computer, first shipping 300MHz Mac Clone in the world), I decided I did not have any use for proprietary systems like this, where one guy could just arbitrarily cause hundreds of millions in lost cash, on purpose (Jobs ending the MacOS licensing thing). It's hard enough not doing that to yourself accidentally when you're in the startup game.
Yeah, I was one of the chief guys on the Amiga 3000.. also Greg Berlin and Hedley Davis, and a couple of others on individual chips (Scott Hood did the flickerFixer clone, Jeff Boyer did the original DMA controller.. Greg redesigned it later). And yeah, I'm very proud of the work we did there. It's shame, though, how things ended... there was vastly cooler stuff in the works.
The OS is certainly still usable for hobby purposes, though if that's your main concern, the emulators on any old PC provide an experience dramatically faster than anything we dreamed of in the days of 25MHz 68030s and 68040s. I don't have a problem with folks doing this stuff for fun... in fact, I try to make it out to Commodore shows (yeah, they still have these) when possible.
I just question the expectations anyone ought to have for any kind of commercial future for AmigaOS. And the really sad thing is, had this been open sourced after Amiga Technologies failed or Gateway were through not using it, the situation would have been much, much better today for Amiga fans.
FYI.. I didn't design the "AGA" chips (what we originally called Pandora, and later, AA), just the first computer that used it (Amiga 3000+)... I'm not a chip designer, though I did a couple of gate arrays (the "Buster" chips that drive the Zorro II bus in the A2000 and Zorro III in the A3000, for example). AA was designed by Bob Raible and Victor Andrade. George Robbins did quite a bit of work on the register architecture, and I worked out the signaling for some of the chips together with Bob, to ensure I could actually do something with the chips in an Amiga 3000-based architecture.
I did the Zorro III stuff pretty much on my own... had to keep pace with one of those huge PC industry committees (EISA bus). In retrospect, there some bad decisions in there, but some of that was driven by the idea I wanted this to last another 5-10 years. Had I known the A3000/A4000 was the last one, I might have made things a little simpler. As it turns out, we would have gone to PCI in the next major generation, anyway... they did autoconfig pretty much the same way we did. The main reason for making up way to do things on the Amiga was simple: no one else had done it right yet. When there were working solutions already, we had no problems using them... it's not like we were Apple.
Linux can certainly be fun, though I have done real work using it, too. Unfortunately, some of that's REAL WORK... asynchronous I/O, for example... trivial in AmigaOS, a big pain in the bollocks under Linux. What were they thinking? Ok, it's not quite fair to pick on 70s technology, is it:-)
The state of the art "LED TVs" out now are just LCDs with modulated LED backlights, rather than a single CCFL panel. These deliver much better dynamic contrast, but also take as little as 50% of the power of a CCFL backlit LCD TV... and LCDs were already fairly low power. I calculated some years ago that, based on my average monitor use at my home office, I could pay for new LCD monitors to replace my CRTs in under three years, based largely on power savings. Now, consider that plasmas (the main class of display this legislation will affect) take about 3x as much power as a CRT of the same size... they're crazy power hungry. OLEDs will improve this yet again, for sure, but the power savings are pretty much already here. Plasma's been a failing concept for some years now, and rear project technologies LCoS and DLP are becoming marginalized as consumers overall demand thin screens (LCoS, Sony's SxRD and JVC's D-ILA, are pretty much dead in the consumer market, though they have applications). DLP has already been a fairly low power technology for a television, and they've been moving to LED light sources since 2005, which cuts power consumption another 30% or so. Laser DLPs are coming soon, and claim to use 1/3 the power of similar sized LCD televisions (I would assume that was based on CCFL backlights).
In short, all the cool new TV technologies are already low power. Many of these already have Energy Star certs... this isn't strictly an Californian concern.
SCART had audio and video, inputs and outputs on the same connector. Maybe. It had provisions for composite video, Y/C (split chroma and luma), and RGB... despite the fact that TV works in YUV or YPrPb colorspace, not YUV. And you never quite knew which of the many options were actually implemented by any given TV or other video device. And the connectors are gigantic.. they look like something out of a cheesy 1950's sci-fi movie.
Err... NTSC died awhile ago... it no longer exists in California or the rest of the USA. The only television broadcast in the USA is ATSC, and the color is just dandy, thank-you-very-much. SCART connectors themselves sucked... Dr. Frankenstein would have been embarrassed to use such a huge connector for 21 low-power signals. HDMI, as featured in pretty much all ATSC televisions, is a far better solution.
NTSC is still alive in Canada, Japan, Mexico, South Korea, much of Central and South America, Taiwan, and various other places. Most are switching over to ATSC or ISDB-T in the next few years... some are holding out another 10+.. Mexico isn't dumping NTSC until 2021, I think they're about the last to go. The video layers on ATSC, DVB-HD, and ISDB-T are pretty much the same, MPEG-2 HD video, compressed multichannel audio (AC-3 on ATSC and DVB, with MPEG Layer 2 as an option on DVB, and AAC for ISDB), but ATSC's 8VSB modulation is inferior to the COFDM used for either DVB or ISDB. That affects reception, not the correctly-received a/v stream.
Of course, for most consumer gear, the NTSC color issues were eliminated over two decades ago for any non-broadcast video, by the popularization of Y/C video inputs, via the S-Video connector. Once you have chroma separated from luma, there's no more phase error problem, period. The resulting video looks better than PAL... full vertical color (sure, there's a bit less vertical, but it's still more color), and better field rate. If you really had problems with video games on NTSC systems, the easy fix is simply to recommend using an Y/C connector. Commodore solved that problem in the early 80s with the Y/C output of the Commodore 64, and many video game consoles followed suit.
They also added VIR and GCR reference data sometime back in the early 80s, which is why NTSC televisions since those days have automatically adjusted for phase inaccuracies anyway.
Sure, SCART has useful signalling... you could get RGB or Y/C sometimes (of course, most video wanted to be YPrPb, not RGB, but you can't have everything... ok, there was a rarely used YUV mode in some gear). But you had to worry about which signals were actually implemented and which weren't... it was really hard to get anything more than a guarantee of composite and at least one audio channel.
And PAL was no great shakes either... the color was a bit better (though it still had red/blue bleeding issues), but at the cost of a delay line and half the vertical color resolution. Lower s/n ratio, and 50 fields per second... used to give me headaches, too much flickering (I had a company based in Germany for six years, Metabox AG, and I designed PAL set-top boxes with SCART connectors... I know of what I speak).
You can download Opera free and try it out, see if you like it.. the cost is very small. It's a completely modern browser... in fact, it's added a number of innovations, such as tabs and bookmark synchronization, well before IE or Firefox had these (and in the latter case, only via plug-in).
Forget the whys... where should I send someone to buy either AmigaOS for the PC (native, not something from the early 1990s running in emulation), or a modern computer with AmigaOS pre-installed? Where's an application I can use to edit high definition video, or make a DVD or Blu-Ray? Since I can do that in Windows, MacOS, and Linux (the latter for free), I think those are completely reasonable standards. This really has nothing to do with the technical merits of AmigaOS, and everything to do with what happens to any OS that has a decade plus of non-development.
The Mac platform was revolutionary... well, the software anyway (the hardware sucked from the get-go, despite the pretty plastic). 20 years later, they have actually fixed the OS (eg, went from a fairly poor low-level design to the Mach kernel and some other decent underpinnings), they have a growing user base, you can actually buy one in a store, there are many modern applications, and the company behind the Mac is making gobs of money. None of those things are true about anything related to the Amiga. And don't get on me for Amiga bashing -- I designed a bunch of them, I love the Amiga. I just hate what happened to it. It has gone nowhere significant since the mid 90s, and things were shakey even near the end, between Commodore's slow death and the year+ it took between that and the sincere attempt to bring things back at Amiga Technologies.
Sure, you may find a few things in AmigaOS that are still better than Windows. You can find that in just about any OS... Windows is an easy target. That doesn't make AmigaOS a useful choice for getting most kinds of real work done today. And it also doesn't remove the fact that the only real platform for AmigaOS today is software emulation of AmigaOS 3.x on a PC, under some other PC OS.
Telling the truth about something is not the same as bashing it; dreaming about what might have been doesn't change what is. But keep in mind... the AmigaOS has been in post Commodore, even post-Commodore/Amiga Technologies neglection longer than it had existed before this. That's a pretty harsh way of looking, but it's the truth. It was 10.5 years from the introduction of the Amiga 1000 (September 1985) to the functional end of Amiga Technologies (March 1986). It's largely been in the hands of lawyers and bozos ever since. Is there anyone really holding their breath for an AmigaOS re-introduction of any kind, much less one that invites a thriving user and developer community? I'd love to see that, but I don't believe in it.
You didn't video editing on an Amiga without serious add-on hardware.
Most folks using Amigas for video were doing analog video, too... digital was barely there at the end. You could use a Genlock for titling, other devices to overlay effects on video, etc. but it was still tape to tape. That wasn't for the feint of heart, and while it was revolutionary for a small number of video professionals, it was only a precursor to today's video revolution, which required digital capabilities.
An Amiga with today's hardware specs would be just like a Macintosh with today's hardware specs: it would be a PC. Guaranteed. Near the end, we were already moving toward using as many commodity parts are possible: PC power supplies, disc drives, etc. Future systems were going to use the PCI bus, and would most likely have been designed CPU-agnostic (look up the "PIOS One" for an example, if there's anything still online... that's the direction I was pushing things before C= failed).
There was a question, back in '93 or so, about the future of desktop CPUs. So we had proponents of the PowerPC, of the PA-RISC, of the Alpha. But by the time the whole Escom/Amiga Technologies adventure was over, things were changing. Shortly thereafter, Apple guaranteed the failure of the PowerPC on the desktop by cancelling Mac Cloning, and it was obviously clear to anyone who was paying proper attention that x86 was the only game in town for this class of computing.
And still the bozos at New Amiga or Hyperion Entertainment or whomever kept their sights on the PowerPC (I do not know specifically where the bozos were, but bozos there were, have no doubts). I know a few of these people, not all of them, but the collective functions as a group of wannabes without proper long term vision. I even tried to hit them in the head with a clue-by-four, even going back to the short tenure of this stuff at Gateway 2000, but there was no help.
The hardware didn't need saving... you need to reinvent computing hardware every five years or so, or it gets too complex to keep advancing. Had Commodore not failed, they would have eventually got out of the graphics chip business, just like Compaq and various other PC companies who once did their own graphics chips stopped. Graphics chips became GPUs, and any one systems company could no more make their own GPU than their own CPU. Everyone who tried either of these failed, in time, unless they built a very strong market well above the level of the personal computer. You could afford to spend $2000 on a CPU for a high-end server or whatever it if went a little faster than the next guy's... you couldn't do that for personal computers. Much less the reality that, without sufficient volume, you couldn't even keep up. The problem Apple so well illustrated with the PowerPC's rise and fall.
The software needed saving, or at least, it would have made things more interesting. There was real opportunity, if they had assembled a team of top notch OS people, like Be did... but that's not really what happened. So, after spending several times as long on AmigaOS 3.x -> 4.x as it took to get from nothing to 1.0 (or beyond), they now have what... a fairly small incremental improvement that runs on... pretty much nothing. We actually did a real engineering analysis of this upgrade in 1995-1996... moving from AmigaOS 3.x to a version, written for CPU-independence, targeted for the PowerPC, with a proper HAL, was about a two year project for the team Amiga Technologies had started assembling. Along the way, that included fixing many of the system's flaws.
When this didn't happen, and AT fell apart, we wound up doing much the same thing at Metabox AG, only this time building a modernized AmigaOS-like OS from scratch... AmigaOS enough that things like Voyager and MUI were easy ports. Took about two years.
The fact that nothing had come of this public Amiga silliness pretty much should drive home what a non-event this is. There's no much of value that can some from this, they're just too far behind.
There are really two needs for a replaceable battery.
The first one is heavy use... every iPhone owner I know (and I know quite a few... the frickin' things are everywhere) charges their phone everyday. And they regularly run out of battery. Of course, when you're using the same device as phone, PDA, and MP3 players, this can happen.
This one's often pretty easily solved... I did it with my Palm Treo, which also served those functions. The Treo sat in a charging station at work, which also served as speakers, if I wanted music without headphones, and as a speakerphone system. But if you don't have a charger around, and you really use the phone as intended for all this stuff, battery life is a real issue.
The big problem is the long-term one... the battery in the iPhone is good for 300-500 charge/discharge cycles before it's capacity is going to freefall into oblivion. Obviously, your mileage may vary, some will last longer than others, based on temperature, specifics of cycling, and just plain luck.
The problem is that the contract runs 730 days. Most iPhone users will have battery issues before their contract runs down. The easy solution is allowing a user-installed replacement... Apple's is, send it back, along with $85.95. They'll install a new battery, and wipe out your iPhone's memory. You get it back in three days.
Well, one good reason: every major cellphone hardware company other Nokia, Apple, and RIM have already announced Android phones, if not yet delivered them. Some large companies, like HTC, have already committed to putting Android on over half their line. The largest carriers now in the USA, Europe, China, and South America have delivered or are in the process of delivering Android phones.
I smell momentum. The big reason this wins is simple: the carriers get to use Android on their terms; they don't have to jump through hoops to please Apple. But Apple's been good... all the information so far suggests that Verizon is actually serious about supporting Android as an open platform, and they have so far been the most closed of the US carriers. Sounds like their new religion is directly driven by Apple now being the most closed company in the smart phone market... and the largest target, despite their sales. Their constant advertising, if nothing else, makes them the obvious target. Everyone knows the iPhone.
Some ARM11 cores support the Thumb-2 instruction set... it was introduced in the ARM1156 core. It's not in the MSM7201A, but it is in all of the iPhone processors, the Samsung S3C6410 and S3C6430. and the nVidia ARM11 chips. All ARMs since the ARM7TDMI support the original 16-bit Thumb instruction set... I'm not sure how important Thumb-2 actually is. A few scaled-down versions, like the Cortex-M3, support only Thumb instruction sets.
Otherwise I agree... the OMAP3 (Cortex-A8) at 600MHz is superior to your run-of-the-mill 800MHz ARM11. NEON, for example, was new for Cortex. You'd have to looking into the SOC details to know more.
I need to know more, NOW! This sounds like the phone I've been waiting for, since my Palm Treo died last year. Pre is pointless on Sprint, and also due to its 8GB memory limit. I won't buy an iPhone, despite my being surrounded by iPhoners... I believe I know better than Steve Jobs which applications I want to be running on such a device. Android was the clear choice, but the phones so far, well, T-Mobile's coverage is as bad as Sprint's... neither reaches my house.
Well, what always bothered me... with all this AI available on the Enterprise, you still have to take a few unshielded blasts from [insert enemy here] before the Captain can give the order to raise the shields. I think I'd want those shields automatically raised by the presence of unknown ships... should happen in femtoseconds, given the power of their computers and sensory gear.
In fairness to Star Wars, the whole premise of Luke trashing the Death Star using The Force rather than his targeting computer required pretty crappy targeting computer technology. In this, they were apparently consistent.
That was a plot element in Babylon 5, too... Earth actually did have pretty good targeting systems, but they couldn't effectively target Minbari ships, at least until the Battle of the Line.
Absolutely.. B5 should not be lumped into the gross violations of good storytelling that comprise much of the latter-day Star Trek. Of course, most of B5 was planned in advance by JMS, and while they did have guest writers (including one by Neil Gaiman, "Day of the Dead") and occasionally episodes that didn't really advance the story arc, most of the time, things did move forward. Start Trek's "reset button" at the start of every episode, just about anyway (as long as their nearly endless supply of redshirts held out) was fine for the 1960s, but episodic television has moved beyond that, and for the better.
And one very good thing about B5 -- strangely enough, most characters did not seem to know each and every thing about obscure or even common technology. Just like today. There might be a few experts on board a Starship or a station like B5, but most of the technology would just be used. I think, in the original Star Trek, you had Kirk for that, but I get the impression that every officer in the latter day Treks could, at least when pressed, assemble and disassemble a warp drive, even if it takes Mr. Scott or LaForge to really get it tuned up. Start Trek personnel had an improper relation to the tech behind their tools, and that lead, I suspect, to too many of those deus ex machina solutions.
I'm surprised no one's mentioned the Star Gate series... particularly in the early days, those guys barely understood how to control their primary chunk of tech... not much idea at all how it actually worked (in fact, pretty much no one did... it was probably made by the Heechee or some-such, they just said "the ancients"). One of the main functions of Col. Jack O'Neill was to make anyone shut up when they started babbling tech... pretty clearly aimed at latter day Trek, I think. Sure, the show had its faults, but not those of the TNG and all.
At least 50% of HTC's phones will run Android by the end of next year. Samsung is delivering, and widely rumored to be building a WiMax phone for Sprint. Motorola is betting big on Android, with reported 250 people involved, and many different phone models. More a jumping in, and it's not just LG... Nokia, Panasonic, Sony/Ericsson, Dell, NTT/Domoco, Acer-Garmin, Kyocera, Huawei Technologies, BenQ, Archos, Philips, Asus, Lenovo, Yuhua TeleTec, Haier, QIGI, Kogan Technologies, Vobis Computer, and many others.
Carrier wise, T-Mobile in the USA will be joined by all other major carriers: Sprint/Nextel, Verizon, and finally AT&T, all selling at least one version of an Android phone. In Europe, T-Mobile, Vodaphone, Orange, and Telefonica/O2 have already announced major support for Android. Also NTT DoMoCo in Japan, SK Telecom in South Korea, and China Mobile (eg, the world's largest cellular subscriber base) and China Unicom in China. China Unicom is also supposed to be distributing the iPhone soon. Bharti Airtel is already selling an HTC-based Android phone in India, and Tata DoMoCo just launched a Samsung model. HTC's selling the unlocked "Magic" direct-to-consumer in Russia, as is the local Vobis with their "Highscreen" model. TIM Brazil is releasing an Android phone from Huawei next year, and América Móvil (4th largest carrier in the world... TracPhone in the USA) is releasing the Motorola "Cliq".
So many people in the US think Apple's that huge just because the iPhone has been so successful here, but that success isn't worldwide anywhere near the same level. Apple's trying, but the problem, versus Android, is that the iPhone is always going to be Apple's, and Apple's going to be taking extra fees, just as they do here. With Android, the local companies can customize Android their way... it will be theirs. Traditionally, they've demanded that capability, and had to pay phone providers extra for it. Now it's free, but for the development.
It's also true we've been the first to jump on full fledged Smart Phone systems, while in other parts of the world, they just sell more capable basic phones, so the demand for a smarter phone hasn't been that compelling. I believe an open apps market changes this in ways that we've seen here with the iPhone. But Android is already truly global... and that's it's pretty much already won; even if the numbers are there yet, the momentum already is.
Microsoft is just one company in the hardware business, but it powers hundreds in the software business.
Apple is just one company in the hardware business, and it powers just one company in the software business... as a result, it's a decently large PC company. No HP, but larger than many... but that's also the size of the software market for Apple, and it's unlikely to grow much larger.
And that's the PC business... PDA and Smartphone buyers are easier to sway to another platform. You do need the right kind of openness for the OS to prosper... Symbian never hard that... it was never all that compelling anyway. Palm was, at one time, but like Apple, pretty much limited to one company (ok, it was one, then a couple, then essentially one again).
If it's Apple versus most of the cellular phone market (and it's looking more likely, with Motorola, HTC, Sony/Ericsson, LG, Samsung, Dell, and others), Apple will not remain #1, any more than they did when there were more Apple computers than PCs in the world. Same with the carriers... it's AT&T for iPhone right now, and the much smaller T-Mobile delivering the G1... the much newer G1. So of course the cards are stacked in Apple's favor. But with Android phones moving onto Sprint, Verizon (the largest US carrier), and even AT&T (they're Dell's partner), AT&T won't be leading, and won't just be dependent on the iPhone.
In short, Android is following the same market path that allowed Windows to dominate. Sure, they won't be able to cheat in the way that MS did in the early days, but it won't be necessary... they're own "cheat", open source, is the big win, and one that's far more compelling. Apple could technically complete by opening up, but they'll never do that.. they're Apple... being closed is a big part of the religion.
It's not as if you can't have two (or more) smart phone platforms doing well enough to support healthy markets... again, this is likely to be more volatile than the PC market, just because the current cell phone market is network subsidized, and users are used to replacing units every two years. Apps are certainly one way to enforce loyalty, but smart phones are still such a tiny segment of the cellular market, there are not enough iPhones yet, nor will there be by 2012, for this to be an important enough advantage.
.. over a summer break, I took a course in Science Fiction and Fantasy at Rutgers University (transfers to CMU). It was a pretty decent course, even for a life-long reader of both genres, and not even close to being a "gut course", though I did score a 100% on the final.
One of the principles the professor had in mind -- if you could read and understand the book all on your own, it doesn't belong in the classroom. She approached the class as one might expect of a professor in any other sort of literature class. A couple of the titles we covered: "The Marriages Between Zones Three, Four and Five (Canopus in Argos: Archives)" by Doris Lessing, and "Riddley Walker" by Russell Hoban. I probably would not have selected either, both were interesting and worth a read. High School level, I dunno... you may want to tailor this course more as an introduction to the genres, and offer up more classic than challenging material. I'd make sure to include something modern... you can't go wrong with Neal Stephenson, Stephen Baxter, Neil Gaiman, or William Gibson. I think you have to represent at least some of the more challenging "classic" writers: Clarke, Lovecraft, Verne, Vonnegut, Asimov, Heinlein, Wells, Dick, Bradbury, Ellison... on the mostly Sci-Fi side. For fantasy, Le Guinn, Tolkien, Zelazny, Anthony, Pratchett, etc.
I wouldn't likely offer up more than one book or short story from the same author within such a class, and while I'd like to offer an independent reading project in that context, the likelihood that over 50% of the kids would produce one "Harry Potter" volume or another (I did finally read them, last July) might have me slightly worried for the future of our culture.
One reason I did well on the final... a large part of that was answering the question "what is the difference between Science Fiction and Fantasy". I suspect many readers here would have done well on that final, but this would be a good thing to ask at the end of the course. And perhaps a good indication of how well you did, educating these kids. My answer was certainly vetted by the fine literary folks at Rutgers, and 25-something years later, I am currently working on a novel that attempts to confuse the issue. It's always more fun when you can break things!
Well, there's no particularly good reason to show you and me a mockup of a video card... it more or less looks like the last video card, unless they have announced something special.
On the other hand, assuming the silicon is real and they're actually showing off working chips in some closed PC, it might be some huge mess covered with towers and green wires and all... which means they could still be a chip rev or more away from shipping, even if the functionality has been rendered showable in some way. I've gone to the CES show with a few dozen computers, each of which had to be carefully tuned (with a tweaking tool and freeze spray) to function at all, due to a chip that was just barely working.
Could be they include a mockup simply as intended deception -- they don't care so much what you know about it or don't, and most people don't give a flying fig about the look of the card, they want price, specs, and performance. And a delivery date. However, folks like AMD would dearly love to know that nVidia's behind schedule, not doing as well as they're showing, etc. So making it look like you have working cards (prototype or not... if the card actually works and fits in the space of a regular card, you're much closer than if it's got hundreds of green wires, dead bugs, FPGAs in towers, etc) makes you look more ready. Which is what you want the competition, and to an extent, the consumer, to believe.
In some cases... particularly "new and light" notebooks, pretty much all you're selling is the mechanical design. The motherboard, LCD panel, etc... not exactly much question of those being available from half a dozen manufacturers in China and Taiwan, in any form factor you like. So in this case, they're showing off the look of the thing only, because that's all that actually matters.
And it's quite true, as well, that many companies float "trial balloons" at trade shows. A show like CES isn't for you and me, it's for HP or Dell or Apple to be making deals for new products with Best Buy or Sears or whomever. It's not uncommon for things to be shown off not entirely complete, for the simple fact that, if they're is no interest, there will never be any completed product. CES Winter is right after Christmas for a reason... none of those things are necessarily closer than maybe six months away. Back in my Commodore days, even products we were fully committed to shipping (Commodore 128 and Amiga 1000, at CES Winter 1985, for example) didn't ship until August and September that year, for example.
You just have to realize what the show is for. Sure, there may be companies really trying to cheat, but most of time, that's not actually what's happening. But in a product with maybe a one year design cycle at best, you can't exactly expect to see something fully ready to ship six months before schedule, even if that's when the trade shows are. Even if there's a working product, you can bet software isn't complete, chips may be only partially functional, etc. If they're still doing this AND announcing the product is ready to ship, then you have some fraud to worry about.
Slashdot Mirror
Well, much the same observation was true comparing a Commodore 64 motherboard to an Apple ][. The Amiga was made, by necessity, very much in the usual Commodore fashion.. full of custom chips and gate arrays. The original IBM PC was a few Intel interface chips and some hundred TTL parts... it was massive. No integration, no time for integratiion, they tossed it together in something under a year.
Now, that's one reason they cost so much, one reason the motherboard was so huge. But also, a big reason the IBM PC was cloned. Even neglecting software issues, you couldn't easily clone a C64 or an Amiga (yeah, sure, today you can, but it takes a reverse-engineering of the custom chips and a nice FPGA or two). But all the forces lined up behind the PC. It didn't even matter if it was any good, really... the was this pent-up demand. Many people wanted to be in the personal computer business, few were capable of launching something like an Apple or a Commodore or Atari -- there were plenty of attempts, plenty of failures.
With the PC, you had a perfect storm: easy to clone hardware, reasonably easy to clone BIOS, and the OS and all chips you could buy, off-the-shelf essentially, from companies who were not IBM and were really happy to have you as a customer, too. It didn't really matter that early PCs... sucked. Eventually, they fixed big chunks of the architecture, and most of those bad ideas are relegated to a tiny chunk of silicon in the corner of a big system chip, anyway.
Hi Randell!! See you Saturday!
Yeah... part of the end-game at Commodore, for those of us who did stick it out in '93 and into '94.. to the bitter end, was still kind of hanging onto the idea that there was a technical solution to the problem of corporate stupidity. There never, ever is. That's why I only made the one film [frogpondmedia.com] about why a company failed. Much of it shot at Randell's old house, in fact (see the extended trailer here [youtube.com]).
--
-Dave Haynie
I agree.. this one is pretty cool. Certainly much closer than the Kindle. A couple of things:
* Book Price -- if B&N follow the Amazon pattern, they'll release ebooks at the same time as hardcovers, for less cash. That's a good thing, for certain, though if you're one of those who waits for the paperback, the price might be a bit more. Not that paperbacks are all that cheap these days, compared to the $9.95 "standard" price for many ebooks.
* The e-ink is a must right now for readers. Anything that qualifies as "book" has to be readable on the beach, or it's a no-go. These e-ink systems get better in bright light, worse in the dark... they behave very much like, well, ink on paper. There's no power used to keep the image up, just to change pages... thus, 10-days of reading.
* Also, changing to higher power displays, adding games capability... no! Anyone asking for that stuff ought to look into these things called computers. They already have this stuff. You can already read ebooks on them. Making an ebook reader do this stuff will give you both a bad ebook and a bad computer.
* The ability to lend it makes for a much better "book" model, naturally without the nasty aspect of having to get that book back from the lendee. Libraries ought to be able to tap into this mechanism, too.
* SD expansion and direct PDF support -- kudos there! That's a major issue with the Kindle. What's the point of an eBook reader that can't store all your books. Sure, whatever's left out of the built-in 2GB is sufficient for 1500 novels, but once you talk about illustrations (eg, PDFs, even if converted to monochrome or halftone for space savings), you're getting big. I have over 2GB of PDF datasheets for the current hardware project I'm doing at work... it would be fairly cool to have these all on-tap on a device like this (also makes it tax deductible... hmmm...)
Two big problems still, which really are the same problem. The book model fails if I can't resell the book somehow, if I can't read it on a different ebook reader (though at introduction, B&N is supporting "alternate reader" on way more non-ebook devices than Amazon, and you have to believe Android is going to be on the list soon, given they've already written that code for the Nook already).
The ePub format is a good move.. XML based, world standard, all that jazz. But ePub still supports the option of a DRM, but doesn't specify a DRM. I'm sure they're using some DRM, perhaps a proprietary one for the Nook. My guess is that the books you buy are downloaded keyed to your Nook, and will be an issue to read anywhere else. I would love to be wrong about this. Hopefully, there's more information on just what they're doing here.
I really do want to see a "book" model that really behaves like a book. That's a problem on computer systems... too easy to make it copy, which is not what I'm after... I just don't want to give up the rights normally associated with real books. They could certainly facilitate reselling of ebooks online, just like Amazon does today with used books... that would vanish if B&N ever got out of the business, or a different standard prevailed in some years.
Obviously, this reader will be useful at least for ePub, non-DRMed, for other sources if B&N fails in this can cancels their services... I guess Kindle does that, too, although Amazon's ebooks are proprietary format, and you need Amazon involved to get PDFs on a Kindle.
Well, it's relevant in the same way that 65xx chips have been long relevant.. they're hidden in everything.
At Nomadio, my current company, I helped launch a fairly revolutionary product into the radio controlled hobby market. This was a digital radio, first of its kind, which ultimately became the fastest R/C radio on the market (latency-wise), as well as the only one upgradeable over USB.
One other feature.. it was 2-way. The "receiver" could accept a number of sensors, to monitor temperature, voltage, or motor speed. The motor speed device uses a hall effect sensor, a tiny CPU, and a magnet to measure speed, pretty much anywhere you could hack it in (like the early computer days, most serious RCers are pretty skilled at tweaking their stuff). I originally used a Cypress PSOC, a tiny 8-bit microcontroller with flexible digital and analog function blocks. Unfortunately, that year, Apple was using these to read that silly ring control on the iPod, so all of a sudden, I couldn't get those. I put a junior engineer on the job of finding a replacement... turned out to be a Z80 variation... in the 21rst century. It worked great.
You know, you're right... one reason I decided to write a bunch here.. I figured some Amiga fans would look.
Escom actually did commit to the platform. Not to the extent we all wanted, right off the bad, but in hindsight, I think at least part of that was due to their money problems, which we didn't fully realize at the time. The did establish Amiga Technologies, AT brought in Andy Finkel and me to work with them on what they'd like to do, and what it would take to do that. It was all done at a level that, had it continued, would certainly have been successful... not this kind of quagmired stuff that happened post-Escom. Shame they blew it in the PC market enough to fail so fast.
Amiga people went all places after C= ended. There were a number of Amiga software people on the BeOS team, even more in the BeOS community, at least back in the day. And I know a bunch who got into OS/2 as well. Particularly if you meet someone complaining about all the stupid things, single threading, serialization, etc. in Your Development OS of Choice, you'll win more than lose betting they learned to program correctly on the Amiga.
Oh, and thanks for video link.. I had not seen that . Even the simple fact of an animation genius like Eric coming into his own on the Amiga should make people who don't understand it all think a bit. It's also evidence we were there... art is usually more persistent than technology.
Some of that's actually going away, slowly. It certainly depends on your workflow.
I used to have MIDI, with Bars and Pipes, on the Amiga, which led me to Cakewalk on the PC, and one of those MOTU 8-port MIDI devices, all kinds of stuff. These days, while I could record on the computer, I'm more likely to record on my Fostex portastudio, then bring all the raw tracks in for mixing on the PC. If you're in a big studio, you're going to have external digital ADC/DAC and mixing, so the PC itself needs just some good digital ports.
Curiously, the "PIOS One" project at PIOS/Metabox was exactly the idea of this... a personal computer optimized for audio/video work. Even in 1997, though, it made sense to think most I/O was going to be external. I had a good sound chip (Aureal... another good company that failed, largely do to evildoers from the outside.. I hired some of the Aureal engineers briefly at Metabox USA), with separately regulated and filtered audio power supply. But going beyond four channels, you would hook an external box to the audio expansion port, and bring it in digitally.
Lots of ports, sure. Firewire was critical to video, but that's going away... tapeless is coming on like crazy, and it's great. I bought a little pocket-sized camcorder awhile back, a Sanyo VPC-FH1, which records on SDHC cards. My goal was to reduce wear and tear on the expensive HDV camcorders, but the video quality out this bad boy is crazy.. and it can shoot at 1080/60p, twice the rate of any Blu-Ray mode. But the real key.. flash memory means fast, totally reliable transfer. There are a number of low-end pro cams doing the same things, and even on the high end, lots of people using SxS cards are loading up two SDHC cards into an adapter.
For DSPs, I don't think so... they're just not cost effective. This is the same thing Be realized, going from their original Hobbit+DSP3210/07 prototype to the PPC model. Signal processing on the main PC is really fast these days, and you get to use that power for general purpose computing, not just some specialized bits. There are interesting areas of computing acceleration, but I would look at GPU and FPGA computing, not DSP. If you really like DSP, you could always build one in a system with some kind of FPGA resource. Not just build it, but build a different one optimized for the specific work involved. Both of these run into software issues, too... special support not needed for native signal processing work. So the benefit has to be large.
You definitely want a 64-bit file system, and one tuned to do media well. The last time I built Linux video servers (just over a year ago, eight x86 cores in a 1U rackmount), I got better performance from XFS than ext2 or ext3, which wasn't a shock. I didn't mess around with ZFS, and ext4 was still a work in progress, though they seem to have been moving in the right direction. One nice thing about Linux... all these FS choices. The same thing that optimizes streaming HD video doesn't necessarily optimize a zillion tiny file accesses in a web server.
The best way to implement security in a multimedia environment is simple: don't connect the media network to the internet. Problem solved. Of course, if you're paranoid, go ahead and run the web browser in a sandbox, that's fine.. you don't want the rest of the system slowed down by VM overhead.
Right... it was possible in the 1985, it wasn't so much ten years later (BeOS was first out to developers in 1995). Or MacOS in 1984, if you want to look at something that's still with us.
Some of that's based on the competition... there was no PC GUI of any major impact in 1985. All personal computers were pretty horrible, too, so true innovation was something people at least took note of. And it was still possible to do that. In modern times, not so much... you pretty much have to be a major chip company to make a significant piece of a modern PC subsystem. No more systems folks like me experimenting with gates or PALs in a lab somewhere.
BeOS was a very good thing, too, but unfortunately, they made a few mistakes. Launching a whole new platform at that time was difficult at best. The BeBox was very cool, but too expensive for anyone but a geek. They could have made a Be motherboard that fit a bog standard PC... but where's the fun in that :-) It's tough, when you're a techie, and evaluate things on the basis of knowing how they should work.. you can see something beautiful, but you have to calm down and realize that most people just want to get their work done and then go yammer about it on Facebook.
Anyway, it seemed reasonable to go software-only, once actual dual processor systems were becoming more common, but that turned out to be a mistake. There was this curious de-valuing of operating systems around 1998-2000... Linux was free, Sun put out lower-end versions of Solaris free, MacOS updates were still free or at least very cheap. Really only Windows was being sold for significant per-copy cash.
Even with Windows, the in-box licensing was a larger business than end-user upgrade. And they successfully blocked the inclusion of any alternative OS... it was a violation of the Windows licensing agreement to build-in a boot manager and allow another OS to run. So, while a few companies wanted BeOS, they didn't want it to the exclusion of Windows. So they went to BeIA, the BeOS for advanced set-top boxes.. but they screwed that up, too. And they were late... I had approached them in 1998 about a set-top-box license, they were sticking strong to the $50 per copy, no hassle price. We wound up using OS/2.. $10 a copy... IBM understood the embedded market. Linux was still not ready for multimedia.
The old tradition of home computing was really about having a computer-centered hobby. You might play games, use a wordprocessor, etc... but the computer itself is the focus. You got Commodore or Amiga magazines, Fish Disks, you joined a user group about that computer, wrote on usenet groups, etc. A PC today is just a tool to help you do something else. I use Altium software for electronics design and Sony for multimedia content creation, and I care a bit about those programs.. I have some loyalty there. But as long as it's a day when Windows isn't biting me in the butt, it's just a means to an end. The OS really doesn't matter. If I could run these under Linux, I probably would, but it's not a big deal either way.
It's pretty clear Microsoft and Apple don't really understand this, either, or at least admit to it. I do recall when the introduction of a new OS was a big, class-A geek event. But, despite Microsoft's creepy videos, does anyone even care if Windows 7 is coming out on Thursday? Sorry, not me.. I'm planning to spend the next 2-3 weeks drooling over the Motorola Droid. Now that's excitement.. finally a pocket computer to make all those iPhone users around me a little jealous. Though I suspect mostly, that'll be along the lines of "wow, you get a signal here...".
The Amiga CLI has nothing to do with BSD... it was adapted from TRIPOS. Sure, we ran some of the same GNU ported shell tools as everyone else, but there was no UNIX in AmigaOS... and that was a good thing in most ways. Believe me, I had been using UNIX since System III in '81 (a summer job at Bell Labs), there was an intention to not copy UNIX. Not a slight against modern UNIX, but it was impossible to do anything remotely multimedia-friendly in an 80s vintage UNIX system. That's why most of the workstations used DSPs if they wanted audio... they could offload the whole audio job to an RTOS on the DSP. The OS couldn't keep up. That was fixed, to an extent, in modern Linux, but back then, no way.
I tried really hard to be a Macintosh fan, but after Apple did their best to kill my Mac Clone making company (PIOS Computer, first shipping 300MHz Mac Clone in the world), I decided I did not have any use for proprietary systems like this, where one guy could just arbitrarily cause hundreds of millions in lost cash, on purpose (Jobs ending the MacOS licensing thing). It's hard enough not doing that to yourself accidentally when you're in the startup game.
Yeah, I was one of the chief guys on the Amiga 3000.. also Greg Berlin and Hedley Davis, and a couple of others on individual chips (Scott Hood did the flickerFixer clone, Jeff Boyer did the original DMA controller.. Greg redesigned it later). And yeah, I'm very proud of the work we did there. It's shame, though, how things ended... there was vastly cooler stuff in the works.
The OS is certainly still usable for hobby purposes, though if that's your main concern, the emulators on any old PC provide an experience dramatically faster than anything we dreamed of in the days of 25MHz 68030s and 68040s. I don't have a problem with folks doing this stuff for fun... in fact, I try to make it out to Commodore shows (yeah, they still have these) when possible.
I just question the expectations anyone ought to have for any kind of commercial future for AmigaOS. And the really sad thing is, had this been open sourced after Amiga Technologies failed or Gateway were through not using it, the situation would have been much, much better today for Amiga fans.
FYI.. I didn't design the "AGA" chips (what we originally called Pandora, and later, AA), just the first computer that used it (Amiga 3000+)... I'm not a chip designer, though I did a couple of gate arrays (the "Buster" chips that drive the Zorro II bus in the A2000 and Zorro III in the A3000, for example). AA was designed by Bob Raible and Victor Andrade. George Robbins did quite a bit of work on the register architecture, and I worked out the signaling for some of the chips together with Bob, to ensure I could actually do something with the chips in an Amiga 3000-based architecture.
I did the Zorro III stuff pretty much on my own... had to keep pace with one of those huge PC industry committees (EISA bus). In retrospect, there some bad decisions in there, but some of that was driven by the idea I wanted this to last another 5-10 years. Had I known the A3000/A4000 was the last one, I might have made things a little simpler. As it turns out, we would have gone to PCI in the next major generation, anyway... they did autoconfig pretty much the same way we did. The main reason for making up way to do things on the Amiga was simple: no one else had done it right yet. When there were working solutions already, we had no problems using them... it's not like we were Apple.
Linux can certainly be fun, though I have done real work using it, too. Unfortunately, some of that's REAL WORK... asynchronous I/O, for example... trivial in AmigaOS, a big pain in the bollocks under Linux. What were they thinking? Ok, it's not quite fair to pick on 70s technology, is it :-)
It's already kind of happening.
The state of the art "LED TVs" out now are just LCDs with modulated LED backlights, rather than a single CCFL panel. These deliver much better dynamic contrast, but also take as little as 50% of the power of a CCFL backlit LCD TV... and LCDs were already fairly low power. I calculated some years ago that, based on my average monitor use at my home office, I could pay for new LCD monitors to replace my CRTs in under three years, based largely on power savings. Now, consider that plasmas (the main class of display this legislation will affect) take about 3x as much power as a CRT of the same size... they're crazy power hungry. OLEDs will improve this yet again, for sure, but the power savings are pretty much already here. Plasma's been a failing concept for some years now, and rear project technologies LCoS and DLP are becoming marginalized as consumers overall demand thin screens (LCoS, Sony's SxRD and JVC's D-ILA, are pretty much dead in the consumer market, though they have applications). DLP has already been a fairly low power technology for a television, and they've been moving to LED light sources since 2005, which cuts power consumption another 30% or so. Laser DLPs are coming soon, and claim to use 1/3 the power of similar sized LCD televisions (I would assume that was based on CCFL backlights).
In short, all the cool new TV technologies are already low power. Many of these already have Energy Star certs... this isn't strictly an Californian concern.
SCART had audio and video, inputs and outputs on the same connector. Maybe. It had provisions for composite video, Y/C (split chroma and luma), and RGB... despite the fact that TV works in YUV or YPrPb colorspace, not YUV. And you never quite knew which of the many options were actually implemented by any given TV or other video device. And the connectors are gigantic.. they look like something out of a cheesy 1950's sci-fi movie.
Err... NTSC died awhile ago... it no longer exists in California or the rest of the USA. The only television broadcast in the USA is ATSC, and the color is just dandy, thank-you-very-much. SCART connectors themselves sucked... Dr. Frankenstein would have been embarrassed to use such a huge connector for 21 low-power signals. HDMI, as featured in pretty much all ATSC televisions, is a far better solution.
NTSC is still alive in Canada, Japan, Mexico, South Korea, much of Central and South America, Taiwan, and various other places. Most are switching over to ATSC or ISDB-T in the next few years... some are holding out another 10+.. Mexico isn't dumping NTSC until 2021, I think they're about the last to go. The video layers on ATSC, DVB-HD, and ISDB-T are pretty much the same, MPEG-2 HD video, compressed multichannel audio (AC-3 on ATSC and DVB, with MPEG Layer 2 as an option on DVB, and AAC for ISDB), but ATSC's 8VSB modulation is inferior to the COFDM used for either DVB or ISDB. That affects reception, not the correctly-received a/v stream.
Of course, for most consumer gear, the NTSC color issues were eliminated over two decades ago for any non-broadcast video, by the popularization of Y/C video inputs, via the S-Video connector. Once you have chroma separated from luma, there's no more phase error problem, period. The resulting video looks better than PAL... full vertical color (sure, there's a bit less vertical, but it's still more color), and better field rate. If you really had problems with video games on NTSC systems, the easy fix is simply to recommend using an Y/C connector. Commodore solved that problem in the early 80s with the Y/C output of the Commodore 64, and many video game consoles followed suit.
They also added VIR and GCR reference data sometime back in the early 80s, which is why NTSC televisions since those days have automatically adjusted for phase inaccuracies anyway.
Sure, SCART has useful signalling... you could get RGB or Y/C sometimes (of course, most video wanted to be YPrPb, not RGB, but you can't have everything... ok, there was a rarely used YUV mode in some gear). But you had to worry about which signals were actually implemented and which weren't... it was really hard to get anything more than a guarantee of composite and at least one audio channel.
And PAL was no great shakes either... the color was a bit better (though it still had red/blue bleeding issues), but at the cost of a delay line and half the vertical color resolution. Lower s/n ratio, and 50 fields per second... used to give me headaches, too much flickering (I had a company based in Germany for six years, Metabox AG, and I designed PAL set-top boxes with SCART connectors... I know of what I speak).
Again, that's an invalid comparison.
You can download Opera free and try it out, see if you like it.. the cost is very small. It's a completely modern browser... in fact, it's added a number of innovations, such as tabs and bookmark synchronization, well before IE or Firefox had these (and in the latter case, only via plug-in).
Forget the whys... where should I send someone to buy either AmigaOS for the PC (native, not something from the early 1990s running in emulation), or a modern computer with AmigaOS pre-installed? Where's an application I can use to edit high definition video, or make a DVD or Blu-Ray? Since I can do that in Windows, MacOS, and Linux (the latter for free), I think those are completely reasonable standards. This really has nothing to do with the technical merits of AmigaOS, and everything to do with what happens to any OS that has a decade plus of non-development.
You ARE comparing Apples (sic) and oranges here.
The Mac platform was revolutionary... well, the software anyway (the hardware sucked from the get-go, despite the pretty plastic). 20 years later, they have actually fixed the OS (eg, went from a fairly poor low-level design to the Mach kernel and some other decent underpinnings), they have a growing user base, you can actually buy one in a store, there are many modern applications, and the company behind the Mac is making gobs of money. None of those things are true about anything related to the Amiga. And don't get on me for Amiga bashing -- I designed a bunch of them, I love the Amiga. I just hate what happened to it. It has gone nowhere significant since the mid 90s, and things were shakey even near the end, between Commodore's slow death and the year+ it took between that and the sincere attempt to bring things back at Amiga Technologies.
Sure, you may find a few things in AmigaOS that are still better than Windows. You can find that in just about any OS... Windows is an easy target. That doesn't make AmigaOS a useful choice for getting most kinds of real work done today. And it also doesn't remove the fact that the only real platform for AmigaOS today is software emulation of AmigaOS 3.x on a PC, under some other PC OS.
Telling the truth about something is not the same as bashing it; dreaming about what might have been doesn't change what is. But keep in mind... the AmigaOS has been in post Commodore, even post-Commodore/Amiga Technologies neglection longer than it had existed before this. That's a pretty harsh way of looking, but it's the truth. It was 10.5 years from the introduction of the Amiga 1000 (September 1985) to the functional end of Amiga Technologies (March 1986). It's largely been in the hands of lawyers and bozos ever since. Is there anyone really holding their breath for an AmigaOS re-introduction of any kind, much less one that invites a thriving user and developer community? I'd love to see that, but I don't believe in it.
You didn't video editing on an Amiga without serious add-on hardware.
Most folks using Amigas for video were doing analog video, too... digital was barely there at the end. You could use a Genlock for titling, other devices to overlay effects on video, etc. but it was still tape to tape. That wasn't for the feint of heart, and while it was revolutionary for a small number of video professionals, it was only a precursor to today's video revolution, which required digital capabilities.
An Amiga with today's hardware specs would be just like a Macintosh with today's hardware specs: it would be a PC. Guaranteed. Near the end, we were already moving toward using as many commodity parts are possible: PC power supplies, disc drives, etc. Future systems were going to use the PCI bus, and would most likely have been designed CPU-agnostic (look up the "PIOS One" for an example, if there's anything still online... that's the direction I was pushing things before C= failed).
There was a question, back in '93 or so, about the future of desktop CPUs. So we had proponents of the PowerPC, of the PA-RISC, of the Alpha. But by the time the whole Escom/Amiga Technologies adventure was over, things were changing. Shortly thereafter, Apple guaranteed the failure of the PowerPC on the desktop by cancelling Mac Cloning, and it was obviously clear to anyone who was paying proper attention that x86 was the only game in town for this class of computing.
And still the bozos at New Amiga or Hyperion Entertainment or whomever kept their sights on the PowerPC (I do not know specifically where the bozos were, but bozos there were, have no doubts). I know a few of these people, not all of them, but the collective functions as a group of wannabes without proper long term vision. I even tried to hit them in the head with a clue-by-four, even going back to the short tenure of this stuff at Gateway 2000, but there was no help.
The hardware didn't need saving... you need to reinvent computing hardware every five years or so, or it gets too complex to keep advancing. Had Commodore not failed, they would have eventually got out of the graphics chip business, just like Compaq and various other PC companies who once did their own graphics chips stopped. Graphics chips became GPUs, and any one systems company could no more make their own GPU than their own CPU. Everyone who tried either of these failed, in time, unless they built a very strong market well above the level of the personal computer. You could afford to spend $2000 on a CPU for a high-end server or whatever it if went a little faster than the next guy's... you couldn't do that for personal computers. Much less the reality that, without sufficient volume, you couldn't even keep up. The problem Apple so well illustrated with the PowerPC's rise and fall.
The software needed saving, or at least, it would have made things more interesting. There was real opportunity, if they had assembled a team of top notch OS people, like Be did... but that's not really what happened. So, after spending several times as long on AmigaOS 3.x -> 4.x as it took to get from nothing to 1.0 (or beyond), they now have what... a fairly small incremental improvement that runs on... pretty much nothing. We actually did a real engineering analysis of this upgrade in 1995-1996... moving from AmigaOS 3.x to a version, written for CPU-independence, targeted for the PowerPC, with a proper HAL, was about a two year project for the team Amiga Technologies had started assembling. Along the way, that included fixing many of the system's flaws.
When this didn't happen, and AT fell apart, we wound up doing much the same thing at Metabox AG, only this time building a modernized AmigaOS-like OS from scratch... AmigaOS enough that things like Voyager and MUI were easy ports. Took about two years.
The fact that nothing had come of this public Amiga silliness pretty much should drive home what a non-event this is. There's no much of value that can some from this, they're just too far behind.
There are really two needs for a replaceable battery.
The first one is heavy use... every iPhone owner I know (and I know quite a few... the frickin' things are everywhere) charges their phone everyday. And they regularly run out of battery. Of course, when you're using the same device as phone, PDA, and MP3 players, this can happen.
This one's often pretty easily solved... I did it with my Palm Treo, which also served those functions. The Treo sat in a charging station at work, which also served as speakers, if I wanted music without headphones, and as a speakerphone system. But if you don't have a charger around, and you really use the phone as intended for all this stuff, battery life is a real issue.
The big problem is the long-term one... the battery in the iPhone is good for 300-500 charge/discharge cycles before it's capacity is going to freefall into oblivion. Obviously, your mileage may vary, some will last longer than others, based on temperature, specifics of cycling, and just plain luck.
The problem is that the contract runs 730 days. Most iPhone users will have battery issues before their contract runs down. The easy solution is allowing a user-installed replacement... Apple's is, send it back, along with $85.95. They'll install a new battery, and wipe out your iPhone's memory. You get it back in three days.
It takes under a minute to change a Treo battery.
Well, one good reason: every major cellphone hardware company other Nokia, Apple, and RIM have already announced Android phones, if not yet delivered them. Some large companies, like HTC, have already committed to putting Android on over half their line. The largest carriers now in the USA, Europe, China, and South America have delivered or are in the process of delivering Android phones.
I smell momentum. The big reason this wins is simple: the carriers get to use Android on their terms; they don't have to jump through hoops to please Apple. But Apple's been good... all the information so far suggests that Verizon is actually serious about supporting Android as an open platform, and they have so far been the most closed of the US carriers. Sounds like their new religion is directly driven by Apple now being the most closed company in the smart phone market... and the largest target, despite their sales. Their constant advertising, if nothing else, makes them the obvious target. Everyone knows the iPhone.
Some ARM11 cores support the Thumb-2 instruction set... it was introduced in the ARM1156 core. It's not in the MSM7201A, but it is in all of the iPhone processors, the Samsung S3C6410 and S3C6430. and the nVidia ARM11 chips. All ARMs since the ARM7TDMI support the original 16-bit Thumb instruction set... I'm not sure how important Thumb-2 actually is. A few scaled-down versions, like the Cortex-M3, support only Thumb instruction sets.
Otherwise I agree... the OMAP3 (Cortex-A8) at 600MHz is superior to your run-of-the-mill 800MHz ARM11. NEON, for example, was new for Cortex. You'd have to looking into the SOC details to know more.
I need to know more, NOW! This sounds like the phone I've been waiting for, since my Palm Treo died last year. Pre is pointless on Sprint, and also due to its 8GB memory limit. I won't buy an iPhone, despite my being surrounded by iPhoners... I believe I know better than Steve Jobs which applications I want to be running on such a device. Android was the clear choice, but the phones so far, well, T-Mobile's coverage is as bad as Sprint's... neither reaches my house.
Well, what always bothered me... with all this AI available on the Enterprise, you still have to take a few unshielded blasts from [insert enemy here] before the Captain can give the order to raise the shields. I think I'd want those shields automatically raised by the presence of unknown ships... should happen in femtoseconds, given the power of their computers and sensory gear.
In fairness to Star Wars, the whole premise of Luke trashing the Death Star using The Force rather than his targeting computer required pretty crappy targeting computer technology. In this, they were apparently consistent.
That was a plot element in Babylon 5, too... Earth actually did have pretty good targeting systems, but they couldn't effectively target Minbari ships, at least until the Battle of the Line.
Absolutely.. B5 should not be lumped into the gross violations of good storytelling that comprise much of the latter-day Star Trek. Of course, most of B5 was planned in advance by JMS, and while they did have guest writers (including one by Neil Gaiman, "Day of the Dead") and occasionally episodes that didn't really advance the story arc, most of the time, things did move forward. Start Trek's "reset button" at the start of every episode, just about anyway (as long as their nearly endless supply of redshirts held out) was fine for the 1960s, but episodic television has moved beyond that, and for the better.
And one very good thing about B5 -- strangely enough, most characters did not seem to know each and every thing about obscure or even common technology. Just like today. There might be a few experts on board a Starship or a station like B5, but most of the technology would just be used. I think, in the original Star Trek, you had Kirk for that, but I get the impression that every officer in the latter day Treks could, at least when pressed, assemble and disassemble a warp drive, even if it takes Mr. Scott or LaForge to really get it tuned up. Start Trek personnel had an improper relation to the tech behind their tools, and that lead, I suspect, to too many of those deus ex machina solutions.
I'm surprised no one's mentioned the Star Gate series... particularly in the early days, those guys barely understood how to control their primary chunk of tech... not much idea at all how it actually worked (in fact, pretty much no one did... it was probably made by the Heechee or some-such, they just said "the ancients"). One of the main functions of Col. Jack O'Neill was to make anyone shut up when they started babbling tech... pretty clearly aimed at latter day Trek, I think. Sure, the show had its faults, but not those of the TNG and all.
At least 50% of HTC's phones will run Android by the end of next year. Samsung is delivering, and widely rumored to be building a WiMax phone for Sprint. Motorola is betting big on Android, with reported 250 people involved, and many different phone models. More a jumping in, and it's not just LG... Nokia, Panasonic, Sony/Ericsson, Dell, NTT/Domoco, Acer-Garmin, Kyocera, Huawei Technologies, BenQ, Archos, Philips, Asus, Lenovo, Yuhua TeleTec, Haier, QIGI, Kogan Technologies, Vobis Computer, and many others.
Carrier wise, T-Mobile in the USA will be joined by all other major carriers: Sprint/Nextel, Verizon, and finally AT&T, all selling at least one version of an Android phone. In Europe, T-Mobile, Vodaphone, Orange, and Telefonica/O2 have already announced major support for Android. Also NTT DoMoCo in Japan, SK Telecom in South Korea, and China Mobile (eg, the world's largest cellular subscriber base) and China Unicom in China. China Unicom is also supposed to be distributing the iPhone soon. Bharti Airtel is already selling an HTC-based Android phone in India, and Tata DoMoCo just launched a Samsung model. HTC's selling the unlocked "Magic" direct-to-consumer in Russia, as is the local Vobis with their "Highscreen" model. TIM Brazil is releasing an Android phone from Huawei next year, and América Móvil (4th largest carrier in the world... TracPhone in the USA) is releasing the Motorola "Cliq".
So many people in the US think Apple's that huge just because the iPhone has been so successful here, but that success isn't worldwide anywhere near the same level. Apple's trying, but the problem, versus Android, is that the iPhone is always going to be Apple's, and Apple's going to be taking extra fees, just as they do here. With Android, the local companies can customize Android their way... it will be theirs. Traditionally, they've demanded that capability, and had to pay phone providers extra for it. Now it's free, but for the development.
It's also true we've been the first to jump on full fledged Smart Phone systems, while in other parts of the world, they just sell more capable basic phones, so the demand for a smarter phone hasn't been that compelling. I believe an open apps market changes this in ways that we've seen here with the iPhone. But Android is already truly global... and that's it's pretty much already won; even if the numbers are there yet, the momentum already is.
Microsoft is just one company in the hardware business, but it powers hundreds in the software business.
Apple is just one company in the hardware business, and it powers just one company in the software business... as a result, it's a decently large PC company. No HP, but larger than many... but that's also the size of the software market for Apple, and it's unlikely to grow much larger.
And that's the PC business... PDA and Smartphone buyers are easier to sway to another platform. You do need the right kind of openness for the OS to prosper... Symbian never hard that... it was never all that compelling anyway. Palm was, at one time, but like Apple, pretty much limited to one company (ok, it was one, then a couple, then essentially one again).
If it's Apple versus most of the cellular phone market (and it's looking more likely, with Motorola, HTC, Sony/Ericsson, LG, Samsung, Dell, and others), Apple will not remain #1, any more than they did when there were more Apple computers than PCs in the world. Same with the carriers... it's AT&T for iPhone right now, and the much smaller T-Mobile delivering the G1... the much newer G1. So of course the cards are stacked in Apple's favor. But with Android phones moving onto Sprint, Verizon (the largest US carrier), and even AT&T (they're Dell's partner), AT&T won't be leading, and won't just be dependent on the iPhone.
In short, Android is following the same market path that allowed Windows to dominate. Sure, they won't be able to cheat in the way that MS did in the early days, but it won't be necessary... they're own "cheat", open source, is the big win, and one that's far more compelling. Apple could technically complete by opening up, but they'll never do that.. they're Apple... being closed is a big part of the religion.
It's not as if you can't have two (or more) smart phone platforms doing well enough to support healthy markets... again, this is likely to be more volatile than the PC market, just because the current cell phone market is network subsidized, and users are used to replacing units every two years. Apps are certainly one way to enforce loyalty, but smart phones are still such a tiny segment of the cellular market, there are not enough iPhones yet, nor will there be by 2012, for this to be an important enough advantage.
.. over a summer break, I took a course in Science Fiction and Fantasy at Rutgers University (transfers to CMU). It was a pretty decent course, even for a life-long reader of both genres, and not even close to being a "gut course", though I did score a 100% on the final.
One of the principles the professor had in mind -- if you could read and understand the book all on your own, it doesn't belong in the classroom. She approached the class as one might expect of a professor in any other sort of literature class. A couple of the titles we covered: "The Marriages Between Zones Three, Four and Five (Canopus in Argos: Archives)" by Doris Lessing, and "Riddley Walker" by Russell Hoban. I probably would not have selected either, both were interesting and worth a read. High School level, I dunno... you may want to tailor this course more as an introduction to the genres, and offer up more classic than challenging material. I'd make sure to include something modern... you can't go wrong with Neal Stephenson, Stephen Baxter, Neil Gaiman, or William Gibson. I think you have to represent at least some of the more challenging "classic" writers: Clarke, Lovecraft, Verne, Vonnegut, Asimov, Heinlein, Wells, Dick, Bradbury, Ellison... on the mostly Sci-Fi side. For fantasy, Le Guinn, Tolkien, Zelazny, Anthony, Pratchett, etc.
I wouldn't likely offer up more than one book or short story from the same author within such a class, and while I'd like to offer an independent reading project in that context, the likelihood that over 50% of the kids would produce one "Harry Potter" volume or another (I did finally read them, last July) might have me slightly worried for the future of our culture.
One reason I did well on the final... a large part of that was answering the question "what is the difference between Science Fiction and Fantasy". I suspect many readers here would have done well on that final, but this would be a good thing to ask at the end of the course. And perhaps a good indication of how well you did, educating these kids. My answer was certainly vetted by the fine literary folks at Rutgers, and 25-something years later, I am currently working on a novel that attempts to confuse the issue. It's always more fun when you can break things!
Well, there's no particularly good reason to show you and me a mockup of a video card... it more or less looks like the last video card, unless they have announced something special.
On the other hand, assuming the silicon is real and they're actually showing off working chips in some closed PC, it might be some huge mess covered with towers and green wires and all... which means they could still be a chip rev or more away from shipping, even if the functionality has been rendered showable in some way. I've gone to the CES show with a few dozen computers, each of which had to be carefully tuned (with a tweaking tool and freeze spray) to function at all, due to a chip that was just barely working.
Could be they include a mockup simply as intended deception -- they don't care so much what you know about it or don't, and most people don't give a flying fig about the look of the card, they want price, specs, and performance. And a delivery date. However, folks like AMD would dearly love to know that nVidia's behind schedule, not doing as well as they're showing, etc. So making it look like you have working cards (prototype or not... if the card actually works and fits in the space of a regular card, you're much closer than if it's got hundreds of green wires, dead bugs, FPGAs in towers, etc) makes you look more ready. Which is what you want the competition, and to an extent, the consumer, to believe.
In some cases... particularly "new and light" notebooks, pretty much all you're selling is the mechanical design. The motherboard, LCD panel, etc... not exactly much question of those being available from half a dozen manufacturers in China and Taiwan, in any form factor you like. So in this case, they're showing off the look of the thing only, because that's all that actually matters.
And it's quite true, as well, that many companies float "trial balloons" at trade shows. A show like CES isn't for you and me, it's for HP or Dell or Apple to be making deals for new products with Best Buy or Sears or whomever. It's not uncommon for things to be shown off not entirely complete, for the simple fact that, if they're is no interest, there will never be any completed product. CES Winter is right after Christmas for a reason... none of those things are necessarily closer than maybe six months away. Back in my Commodore days, even products we were fully committed to shipping (Commodore 128 and Amiga 1000, at CES Winter 1985, for example) didn't ship until August and September that year, for example.
You just have to realize what the show is for. Sure, there may be companies really trying to cheat, but most of time, that's not actually what's happening. But in a product with maybe a one year design cycle at best, you can't exactly expect to see something fully ready to ship six months before schedule, even if that's when the trade shows are. Even if there's a working product, you can bet software isn't complete, chips may be only partially functional, etc. If they're still doing this AND announcing the product is ready to ship, then you have some fraud to worry about.