PICK uses a database for it's directory and file structure (loosely speaking) and Clearcase (a version control/configuration management system from Rational) create what *appears* to be a directory tree with files in it but is just a view into a database. Different users can have different views depending on permissions, versions, filters etc. This is great and powerful stuff. With the avaerage PC having 10k plus files and larger systems having 500k plus files, a hierarchical access mechanism becomes pretty inflexible and inefficient. For old die-hards the files could always be viewed as a directory hierarchy to stay compatible with legacy programs etc.
On my stock standard Gateway p3/600 XP has BSODed a few times, always in the nVidia drivers supplied by Microsoft. I tried the latest detonator drivers that are mean to be faster by they were much more unstable. I'm using an ELSA GeForce2 GTS card.
The crash usually comes when changing graphics modes to run older programs, like a little kids program for my daughter that insists on 256 colour mode (which XP can't do unless in compatibility mode!).
The BSOD looks a bit different (new font used) but still dumps out to disk and when you reboot it wants to send a trouble report to MS.
Also when changing graphics modes between users it got confused and refused to rewrite the background when windows were closed or moved around. I have not allowed my machine to connect to the Net so maybe there are fixes for these probs but I will download the service packs on another machine and copy over the Net. Call me paranoid but I'm not going to let my XP box connect directly as I just don't know what it'll be doing in it's "automatic updates" process.
I did an upgrade from Win2k so maybe that's the prob but I dont want to scrub my drive.
Also runing a really old version of Lemmings (what a great game, brings a tear to my eye to watch my kids getting excited about a game I played on an Amiga 15+ yrs ago!) has sound but no music (midi synth) on my SBLive! card.
Two steps forward and stumbling back...
[sigh]
Elegant Design - but then you have to implement it
on
Software Aesthetics
·
· Score: 1
I usually find that a system, once analysed, comes down to a bueatiful, elegant set of systems and flows that integrate and encapsulate the problem and often provide generalised solutions that alow for future growth and expansion. This usually happens afte r a lot of talking to the end users and pouring over the requirements spec and really understanding the problem.
Then there's a pause. This may take hours or days or weeks (several months in one very large project I managed with several other code architects workingon the job too). In this period to the rest of the world (read: bosses) you're doing prtty much nothing but talking about stuff and having coffee.
Then The Design hits like an epiphany, descending from on high and you just *know* it will work. The final implementation may have to make compromises in the name of speed or resources or the language, but the elegance is always there. You can see this in well written code.
Try to explain this process to a non software person (like a mechanical engineer or a manager)...[sigh]
Re:64-bit architecture - what about HPUX
on
HP Buys Compaq
·
· Score: 1
> [That] makes a "Big 3" of Unix vendors: IBM, Sun, HP/Compaq.
True64 is pretty big in business and defense - it wont go away in a hurry.
Major one you missed is HPUX - it is very big in midrange (read multiprocessor) business systems. These used to be 68k based but are now PA based. The IA64 is meant to be binary compatible and plug in replacements to upgrade this huge market. Not sure of the numbers but I wouldn't be surprised if 20% of HP's 47bn revenue is from HPUX related sales.
DirectX 8 is much much improved over earlier versions. OpenGL always needed extensions to make the best of PC architecture (tri-meshes etc), which was a pain. OpenGL standards change very slowly too, whereas DirectX changes whenever MS wants to add a new feature (once a year or so).
OpenGL is still a contender largely because John Carmack chose it exclusively several yrs back and poo pooed DirectX. More recently he has said DirectX has caught up. The industry tends to listen when JC speaks:-)
Companies often write for PS1 and 2 and other games consoles after the PC version - rarely at exactly the same time and *all* the consoles need specific code. Basically a new platform is a new rewrite (who wants to see a console version of a PC game 1 yr later that's *exactly* the same???). And the MS ppl at E3 said that DirectX 8 code for PC can be recompiled for XBox *but* it'll probably run at 30-50% of it's potential framerate, so you have to recode stuff anyway. This is mainly to make the best of the different nVidia hw compared to the PC (dual Transform and Lighting engines in the XBox).
So DirectX gets you pretty much all the PC market - a good way towards XBox and nowhere on the (very profitable) other consoles - but those are a major rewrite anyway - you don't code a PS2 using OpenGL!!!! It runs like a dog if you do!
On a final note, DirectX 8 was meant to be Fahrenheit (I think the name was) a merger of OpenGL and DirectX, a huge concession by MS after the caning they got from JC and other games developers - but MS and SGI had a falling out (SGI didn't adopt NT fast enough and came out with Linux boxen). In the end MS improved DirectX enough anyway.
It could be with all the reports of security holes in MS Office and back doors etc that one part of the DoD has decided they'd like to run something where they can view and check all the source code themselves to make sure it's secure.
Also, I'm sure the DoD has a lot more than 25k PCs, so this must be just for a small part or division where security is more important than features or compatibility.
Well I've read articles for and against Mr Elz. What I can say is that the Internet untill faily recently in Australia, was firmly controlled by the universities. I remember applying for a.au domain name about 6 years ago - I don't know if it was Mr Elz that I spoke to, but some guy whom I was told handled all.au registrations at the time said (after 3 months trying to track him down) "I'm not going to give your company a domain name as we believe the Internet is for academics and research not for commercial use, so I will never approve your application."!!!
This attitude realy sucked and definitely held back Australia's entry into the commercial domain on the internet, ultimately hurting the development of that type of business here.
On a side issue I notice the govt has decided to ban internet gambling here - except horse races (as they have a mechanism for taxing that). The funny thing was many pokies machines use the internet for monitoring them and the wording of the draft legislation would then ban most pokies machines in Oz which via tax are a big revenue earner for the govt!!! The Prime Minister was heard to quickly assure the gambling industry (non internet!) that the draft would be ammended! He said something about "unforseen loophole" - only he meant they *wanted* a loophole and had not forseen to add it.:-)
Same thing happened to me when offered a job by a very large Silicon Valley unix systems manufacturer (who shall remain nameless).
They had this one paragraph that covered just about every thought I'd have whilst employed by them. I was working on a significant project at the time in my own hours and refused to sign. Their comment was "We thought you might say that, but we thought we'd try it on you anyway"!!!
I got a friend who is a laywer to expand their paragraph into a 10 page document that defined "working for", essentially meaning only when working on jobs under their direction (not just in working hours: what if you work late? and not just in 9-5 hrs: what if you have a sick day and work on your own project at home that day?)
It took two months for them to okay it but they said my final document followed the spirit of their own shorter paragraph anyway.
Certainly in Australia I've known of a case where a company tried to get difficult with an ex-employee (taking an idea he developed in his own time to a new employer) and failed before it even got to court. It just couldn't be proved. I imagine if a *lot* of money was involved things might have gone differently.
I remember reading about a guy who did his Phd on this at least 5 years ago. He intended making cubes, each containing a microcontroller, power and motors and connectors (electrical and mechanical) so they could interconnect and execute a generic command given to the whole pile of blocks. The would slide over each other and move by sort of flowing along the ground. I think there was an picture of it too, maybe this was in New Scientist? He only managed to make about 6 cubes in his 4 yr Phd(!), each about 4 inches on a side, made of precision engineered parts. He had hoped to shrink it to 1cm cubed (and eventually much smaller, but this was before people were really thinking of micro mechanical stuff too seriously, and he ran out of time anyway).
The only time I've used netscape's or Outlook's calendars is for To Do lists. These are a major plus when those really busy weeks crop up from time to time and my To Do list extends beyond 30 or so items and I run the risk of forgetting important things.
To Do lists need %age complete and a due by date/time. One thing I haven't seen in ToDo lists that I'd really like is not just a terminal date/time but also how many hrs the task is expected to take. It should then keep tabs of how much work has been done.
Also I'd like a mini-Gantt, so that it can display the tasks I have for the day or week and show free hrs or if I've over committed myself by eg specifying 10 tasks at 2hrs each in a day! This mini-Gantt (or any decent Proj Managment sw, unlike MS Project) must allow you to specify if a task is able to be carried out at the same time as other things or if it is exclusive of anything being done.
This Mini-Gantt should also automatically order tasks in terms of priority (defined when I created the items in the To Do list) and automatically schedule my day, with a pre-defined breaks allowed for morning tea, lunch etc and breaks for mtgs and allow travel time to meetings at known locations. It should summarise the following days events b4 I leave work and should also have that summary as a popup when I first arrive and/or login in the morning.
> > If Echelon is reality (and that's a big "if"), the US is the only player.
> Actually, everything that I've ever seen on Echelon suggests that it's a joint U.S./U.K. project.
Echelon is run by UKUSA (pronounced "you-queue-sah") consisting of the United Kingdom, United States and Australia. The whole Echelon thing came to the fore after an Australian Govt official made comment about a facility in Australia and what it's purpose was.
Two words: scene complexity.
Scene complexity refers to the fact that polygons in a scene overlap each other (from the camera's viewpoint) - ie in an indoor scene there are polygons for the wall in front of you but if there's a room beyond that then there're more polygons that get calculated and put into the frame buffer but are ultimately get overwritten by nearer polygons.
An outdoor scene can have a scene complexity of 3.5 - 5 (or average). So at complexity 5, each pixel is overwritten, on average, 5 times. At PAL resolution you need 50m polys/pixels a sec, *on average*. But there will be times you are suddenly in a position where the complexiy is far higher. For example when standing in a town and looking down a street of buildings. The complexity may be 10 or more (before other culling algorithms reduce the workload). So at reasonable PC screen resolution, of say 1024x768 at 60 Hz (~50m pixels) you'd need 250-500m polys/sec to ensure you have enough headroom. This is why flight simulators etc still don't use PCs! nVidia is doing really well though, in perhaps 4-5 generations (2-4yrs?) they'll have this performance. But then we're talking about 48bit colour (if they listen to John Carmack and others) so we're talking up to 6Gbytes per sec. Of course then we're talking about loading in textures *fast*. In an outdoor scene you might easily have a few hundred megs of textures, turn around quickly (say it takes 6 frames) and the system may have to load maybe 100 meg of textures in 1/10th a second ie 1GB/s of textures. This is a lot of bandwidth PCs and consoles (now and near future) just don't have.
Of course the alternative is to have your game companies spend 2 years fine tuning the geometry of everything you look at so it doesn't run too slow. Which is exactly what *does* happen for now. But how long before we grow tired of viewing painstakingly handcrafted 3D scenes and textures that takes years to tune, rather than more natural and dynamic scenes? And will we be happy with 1024x768 by then?? My guess is there's still a long way to go.
I have one question here: will software really need more and more CPU performance as time goes by?
This is an unbelievably short sighted viewpoint from a Slashdotter! In 30 yrs time I *really* hope my neural tap, eyeball quality rendered (8kx3k, 48bit, 100fps,100Gpoly/sec), speech and vision recognition system doesn't run on a PIII800 or a P8/10k!! It had better use 10k times less power, have 10k times the grunt in 10k times less space and be as far ahead of Pentium anything as todays CPUs are from vacuum tubes. I'd hate to have even a super-duper-duper PC strapped to my head!:-) And that little sucker aint going to be running Windows, Linux or any other OS of today, perhaps a very distant relative. That OS had better be very robust, secure, realtime and distributed, probably something brand new, lending from Unix, Windows and other stuff like QNX and pSOS. UI will be irrelevant. Can any of you honestly say you don't believe this will happen?
I think the main difference btwn a cluster and a single image multi-cpu box is for the programmers. You can't mention "phenomenal inter-module bandwidth" and then dismiss it. That *is* the whole point (well, mostly!). You *can* have a single OS image over a cluster but that's not sensible, so it's not done, rather special libs are used. On a multi CPU box the programmer can choose to ignore the architecture and still get very good speedup (if it's the sort of code that can benefit from multi CPUs). This is because the CPU and to some extent the OS at runtime can manage the modest slowdown in accessing memory in non-local modules. Whereas in a cluster you have to rewrite your code and design it specifically for the architecture and the slowdown for remote memory is unacceptable. I've had code on a single CPU box, recompile it unchanged on a 20 CPU SGI and seen a 18.8x increase in speed. I didn't have to make any changes at all in 20k lines of code (ok, I was lucky too!). You just can't do that in a cluster.
"The best thing about Xbox is that it won't change. Ever. Judging by other consoles, Xbox should have a four or five year run"
Is totally wrong. You'd think they'd realise (especially a bright fellow like Abrash) that things change faster than they used to. Cheap 3D graphics cards today blow away cheap graphics cards of just 2 years ago. By the time the XBox comes out, it's performance will be average, but even if it is ahead of the pack it will fall behind in 6-12 months and be obsolete within 24 months.
I understand MS is considering making the XBox upgradeable. A very wise move, but they'd be *much* better off focusing on making DX8 stable for many years, rather than requiring developers to code to a low level to achieve the advertised performance. But of course, this is just the sort of thing MS has shown no foresight for with DX in the past. A very good reason to program the XBox in OpenGL only.
Actually, it didn't start with Drexler, it started with Richard Feynman, on Dec 29th 1959 in his now famous lecture "There's Plenty of Room at the Bottom", you can find a copy of this at http://www.zyvex.com/nanotech/feynman.html
From articles I've read on the making of Matrix and watching the extra footage of it's making that comes on the DVD and talking to a guy who worked on the film, I get the following:
Yes they did use a series of still cameras, yes they did then use a computer to interpolate some of the shots.
What also happened is that they used the computer to create a 3D model, based on the data from the different view points. Then they took the images in the shots, used them as textures on the 3D models and voila! they had a 3D model (well, lots actually, one for each cinema film frame). From this they could move the camera *anywhere* and were not constrained to the path of the original cameras. This enabled them to follow all sorts of new camera paths (except where texture data was missing).
This is a major step forward! As someone else mentioned, they also computer rendered the background. On the DVD you can see the wireframe models and incomplete renderings of the subway station.
The issues of 1) right to carry firearms, 2) rights to access whatever I want on the net and 3) freedom of speech are NOT the same thing.
To think they are the same is a kneejerk simplification. Think a bit more before responding. I'm quite happy in my Australian city knowing that my neighbour or the local gangs of kids don't have easy access to automatic weapons.
I'm yet to meet an intelligent Australian or American who believes complete freedom to own whatever armoury people want is a good idea. Most Americans don't realise what a laughing stock the USA is with their 18th Century attitude to firearms which they protect so fiercely whilst having such an appallingly high crime/death rate using weapons!
Perhaps the NRA should push for the Swiss solution of every male between 18-60 *must* own a working firearm and must do a year (or is it 2?) of national service training.
Maybe it's just me, but when I'm in the US (fairly often) I feel the average US citizen is much more worried about being shot that being sexually assaulted by someone who looks at porn on his PC. Like when I was in Washington last year and was told "Whatever you do, don't go to the bad side of town, it's not safe." So I'm educated about the massive crime rate in the nations capital. So please, criticise the recent Net laws, but don't lecture us on firearms. (And I won't mention the appalling racist hierarchy that is openly maintained in the US)
Why is it that the "land of the free" is such a dangerous, unsafe place compared to "close minded" societies like Australia?
I took part in a series of psych studies where they were measuring the rate at which your brain processes information and relating this to IQ. The good news is that they found a direct correlation. There were a lot of different tests, but one that's relevant used a flickering light controlled by computer and you had to say if it was constant or flickering. Over a series of tests they determined the fastest speed at which your brain perceived the light to be flickering.
There's no thinking involved as such in this kind of test: either you see the light flickering or you don't. The bad news is that people of the highest IQ (measured by 6 different kinds of test - it took days!!) could detect flickering down to 16 milliseconds (IQ of 130-140) which is about 62Hz. Lowest IQs tested (90-100) were down around 30ms. The relationship was *roughly* linear.
Hence, next time you're in a room and you're the only one complaining about flickering lights or screens, you might just have to console yourself with the thought that you're the brightest person in the room!
Personally, I had the fastest flicker rate (16ms), I notice flickering lights and screens if I think about it, but don't seem to be adversely affected by fluro lights. Maybe I'm just used to it (my original PC ran at 1024x768 *interlaced* ugg!! I used to think it was horrible to stare at, but strangely, after a few weeks I didn't notice any flicker at all). I'm sure other factors are also involved. (similar to the story of the guy with the upside down glasses, but that's a major digression:-)
By the way, I much, much prefer the new "daylight" fluro tubes you can get, which have a much more natural spectral distribution, making other fluros look yellow or brown by comparison.
I know that sounds like paranoia but consider this: AOL paid out major dollars for Netscape, then said "No we won't use it." Yeah, right, we believe you. They said they didn't want to hurt the relationship with Microsoft that gives them an icon on the Win98 desktop.
So they spend $800m(?) or something just to save little buddy Netscape and with no thought of business advantages - aside from the relicensing of Netscape server products to Sun, which has just recently announced dropping it's own Web server sw in favour of Netscape, which is owned by AOL with whom Sun took out their "I'll scratch my back you scratch mine" deal regarding hw in exchange for advertising. Who bought *that* story? Sun wanting to do some $400m of advertising directed exclusively to AOL clients? Ha ha ha ha [head falls off in embarassment]
So we can assume AOL *will* use Netscape. I recall reading something about the next version of the AOL client being "fully HTML compliant". Their deal with Microsoft would have expired by then no doubt anyway.
Then we have Microsoft's recent loud noises about a new audio standard "a bit like MP3" (Windows Media 4.0) that includes some kind of protection for artists and publishers. Go to the MS web site and click on the Public Enemy link and try to play it. I have the latest version of WinAmp and it loads but gives an error - the protocol is sufficiently different to not be MP3 compliant. With the MS stamp of corporate approval and their marketing, Windows Media format will sweep aside MP3 freeware encoders/players and become the defacto standard. It ships with IE5, Win98, Office2000,Win2000 - it will be on a *lot* of desks. I know that MP3 is wonderfull just the way it is, but MS don't see it that way as they don't control the standard.
So, AOL see this burgeoning market in streaming audio and audio formats as being a big up and coming money maker and MS has a 3-6 month jump on them. They *know* they're going to be changing to Netscape browsers RSN and their users won't be able to access MS's new sound format files easily. What to do. Well when you're as big as AOL you buy the competition to MS and use it as a weapon. I think AOL are on the up and up and this will be an offensive war (ie AOL doesn't look to be sitting back and defending it's territory, they're moving forward, bless their little socks. Personally I hate 'em because the number of friends that got sucked into non standards compliant browsers, emails clients etc whom I have to tailor my email for).
Oh, and don't AOL use Suns, SGI (mainly?) and BSD boxes? What does that tell you about their technical inclination towards what Microsoft would like the world's businesses to run on - NT??
I think we're starting to see some major battle lines being drawn. It's not NT vs Unix. It Microsoft vs Other Companies. Keep a watch on AOL. Oh, and something is going to happen over at IBM, I'm sure of it. I bet you thought they've been kind of quiet lately. I think with the IBM witness at the DOJ case coming up the gloves will start coming off IBM. They're looking very strong financially.
So this isn't NT vs Linux. That's kind of what seems to be the frontline for we Linuxites. But take a step back and see that the HP,DELL,SGI,IBM uptake of Linux is because Linux is a weapon against Microsoft. MS wields NT, Office, Win9x and some other "standards" that flow from that.
Sure each nose-to-nose battle is NT/Linux, ASP/PHP, DCOM/CORBA etc etc, but that's short sighted. This is why I think ppl in the Linux community argue over suits, corporate types and where RedHat is going. This war needs to be fought on all levels. People in the kernel or drivers or wm battle front shouldn't belittle the necessity of the people fighting by corporate alliances, packaging distributions, binary only applications etc. We are all on the same side.
A guess we have to consider why we are on this side and what the goal is. Is MS the enemy? I would tend to think that MS is the *current* enemy because it is the most obvious incarnation of what many people oppose: freedom to choose without bias in a competitive, non-prejudiced sales, marketing and technical "marketplace" (free sw/hw or otherwise).
MS is a bump in the road to this goal. Some people wil see the bump, some will shout out "Lookout!", some will drive around, some will go so far around they'll go off the road into a ditch, some will end up going cross country, some will ride straight over the top. However it happens, the bump will eventually be behind us and won't look so bad in the rear-vision mirror. There will always be more bumps and maybe the wreckage of some crashed cars:-) but remember, that we will never, ever, ever stop. We will always go forwards.
XFS is a lot more than "just" a journaling FS. One of it's other major components is guaranteed I/O rate partitions. When you create these special partitions, the OS tests the disk I/O rate then lets you specify up to the tested rate and then the OS will guarantee you the I/O rate you selected (can be less than the max), nomatter what the machine is doing (ie even under really heavy load, network, UI etc).
This sort of thing is needed when doing uncompressed cinema res images at 24fps (or HDTV) where you need 90-130Mbyte a sec from the disk nomatter what else is going on.
There's a cable channel using 24 odd uncompressed TV res video streams for live delay rebroadcasting (across time zones) using XFS. Works nicely.
Linux doesn't need this right now. Why? Because it's kinda obvious that the whole OS needs to be in on this act, it's not just a FS thing. True, the guaranteed rate stuff can be treated discretely (ie left out). But I think people may be naive when they say, "Yay, use XFS."
I haven't seen what EXT3 promises, but I bet that the current implementation of XFS has fingers going *deep* into IRIX that won't make it a fast retro fit into Linux, compared to EXT3 (unless EXT3 is just conjecture at this time, or only a modest improvement. This'll be why Linus wants a full rewrite: to get greatly improved performance will need a lot of changes on the OS side, and if you're going to go to that effort, you'd better make it worthwhile)
Linux does need a journaling FS and XFS may be the best bet, but it won't happen quickly unless SGI puts some serious resources behind it. For any other effort to pool together enough ppl for long enough to make it happen is just too unlikely for us to just sit around hoping for.
Also, just who has the resources to test large production systems (4+ CPUs) on an OS under test? Corporates, that's who. And they'll contribute their code to Open Source, right? Because...?
This *will* all happen, but I think some of these tougher OS issues will need corporate backing that may have a price current purists will not like but will have to accept (ie less than Open Source code licenses or maybe even (cringe!) binary drivers).
I visited HITL last year to give a talk and chat with the people there. I was lucky enough to have a go with the VRD and I must say it was *very* impressive. These guys have spent a lot of time and money on this project (millions over years) and they've now got it down to a briefcase size.
So I guess the bad news it that it'll be a while before this stuff fits onto a light HMD, they made noises about it being 5 years away, maybe more. Also, they're having trouble pushing the resolution. Given that the human eye has a res of about 10000x6000 (depending on who you talk to) and this display is 640x480 then it only covers a small portion of the visual field. Given 5 or so years maybe we'll have cheap hardware to generate graphics at that res too! That's the frustrating bit.
The good bit is that the colour clarity is excellent and the picture is crystal clear and rock solid. And the extraordinary thing for me is that without my glasses it's still perfectly in focus! Because it writes directly onto the retina, effectively it doesn't use the lens in your eye, so people with very bad vision still see a beautifully clear image. It's kind of odd to take off my glasses and the whole room is a blur (mostly) and then to suddenly see a diamond sharp image looking like a flat panel screen hanging in space about 6 feet away.
There is also some problems with spherical aberration if they try to may the image too wide and getting the scanning components to handle higher (and the right) frequencies. However, *most* of the hard work has been done and it seems to be mostly a clear road of progressive minaturisation untill it is a head mounted "heads up display" (with optional flip down cover for an immersed view).
It doesn't take long to realise that this is the way most computer displays will go. It's not a matter of "if", but "when".
MIPS and PA architecture chips are the architecturally closest *working* chips to Merced that we have today.
I remember all the trouble MIPS had when they rolled out the R10000 chip. Initial performance was not up to spec because early estimates of performance were pretty much correct on the SPECfp numbers, but underestimated *how long* it took to get those numbers. It took a couple of years for the compiler people to wring out the best performance (ok clock speed was off too, but that was not the sole reason, nor was internal CPU wars within MIPS/SGI).
Now the Merced is more complex than the R10000 (and at least the R10K has some *vague* similarity to the R8K, and PA architecture has been around for many years, so these companies had compiler writers experienced in some of the problems they were up against). Intel is starting from scratch here. I'd say when they've done first tapeout and have silicon in their hot little hands, it'll be at least a year before the compilers get close to the performance they hope for.
Meanwhile, IA32 will be up to similar spec and Alpha, PA and MIPS (and SPARC perhaps) will be serious contenders.
Just last week or so MIPS and HP announce they were reviving their CPU development for a further year or so (ie another generation), rather than trusting all to Merced (I assume that means last MIPS or PA in 2003-2005 now). What news did Intel give these guys for them to decide to make such an announcement??
Whilst I think Rob is heading in the right general direction I can see a few problems.
I have to agree with those who say moderators should be able to moderate and post to the same news story.
I only read those stories which interest me (maybe 1 in 3) and these tend to be ones in which I already have an opinion or in which I have some expertise, otherwise I wouldn't be interested in it in the first place. Hence my moderations *and* my comments should be considered acceptable for that particular story.
Secondly, when I read postings, I've taken Rob's advice and set my threshold to +3 and listing in descending order of score. This has reduced my reading time dramatically and is overall a Good Thing.
The trouble is, if I'm made a moderator, then I will only see comments which have already passed a first phase of moderating. Now, considering I'm only reading a few choice postings anyway, I don't mind if I get an extra tab next to each so I can go +1 or -1, but there's no way I'm going to take time to read AC postings.
I imagine a lot of lurker, experienced users like myself will feel the same. If we're added to the ranks of moderators, we will only be moderating the "quality" posts, hence +5 is too low for an upper limit.
Slashdot Mirror
PICK uses a database for it's directory and file structure (loosely speaking) and Clearcase (a version control/configuration management system from Rational) create what *appears* to be a directory tree with files in it but is just a view into a database. Different users can have different views depending on permissions, versions, filters etc. This is great and powerful stuff. With the avaerage PC having 10k plus files and larger systems having 500k plus files, a hierarchical access mechanism becomes pretty inflexible and inefficient. For old die-hards the files could always be viewed as a directory hierarchy to stay compatible with legacy programs etc.
On my stock standard Gateway p3/600 XP has BSODed a few times, always in the nVidia drivers supplied by Microsoft. I tried the latest detonator drivers that are mean to be faster by they were much more unstable. I'm using an ELSA GeForce2 GTS card.
The crash usually comes when changing graphics modes to run older programs, like a little kids program for my daughter that insists on 256 colour mode (which XP can't do unless in compatibility mode!).
The BSOD looks a bit different (new font used) but still dumps out to disk and when you reboot it wants to send a trouble report to MS.
Also when changing graphics modes between users it got confused and refused to rewrite the background when windows were closed or moved around. I have not allowed my machine to connect to the Net so maybe there are fixes for these probs but I will download the service packs on another machine and copy over the Net. Call me paranoid but I'm not going to let my XP box connect directly as I just don't know what it'll be doing in it's "automatic updates" process.
I did an upgrade from Win2k so maybe that's the prob but I dont want to scrub my drive.
Also runing a really old version of Lemmings (what a great game, brings a tear to my eye to watch my kids getting excited about a game I played on an Amiga 15+ yrs ago!) has sound but no music (midi synth) on my SBLive! card.
Two steps forward and stumbling back...
[sigh]
I usually find that a system, once analysed, comes down to a bueatiful, elegant set of systems and flows that integrate and encapsulate the problem and often provide generalised solutions that alow for future growth and expansion. This usually happens afte r a lot of talking to the end users and pouring over the requirements spec and really understanding the problem.
Then there's a pause. This may take hours or days or weeks (several months in one very large project I managed with several other code architects workingon the job too). In this period to the rest of the world (read: bosses) you're doing prtty much nothing but talking about stuff and having coffee.
Then The Design hits like an epiphany, descending from on high and you just *know* it will work. The final implementation may have to make compromises in the name of speed or resources or the language, but the elegance is always there. You can see this in well written code.
Try to explain this process to a non software person (like a mechanical engineer or a manager)...[sigh]
> [That] makes a "Big 3" of Unix vendors: IBM, Sun, HP/Compaq.
True64 is pretty big in business and defense - it wont go away in a hurry.
Major one you missed is HPUX - it is very big in midrange (read multiprocessor) business systems. These used to be 68k based but are now PA based. The IA64 is meant to be binary compatible and plug in replacements to upgrade this huge market. Not sure of the numbers but I wouldn't be surprised if 20% of HP's 47bn revenue is from HPUX related sales.
DirectX 8 is much much improved over earlier versions. OpenGL always needed extensions to make the best of PC architecture (tri-meshes etc), which was a pain. OpenGL standards change very slowly too, whereas DirectX changes whenever MS wants to add a new feature (once a year or so).
:-)
OpenGL is still a contender largely because John Carmack chose it exclusively several yrs back and poo pooed DirectX. More recently he has said DirectX has caught up. The industry tends to listen when JC speaks
Companies often write for PS1 and 2 and other games consoles after the PC version - rarely at exactly the same time and *all* the consoles need specific code. Basically a new platform is a new rewrite (who wants to see a console version of a PC game 1 yr later that's *exactly* the same???). And the MS ppl at E3 said that DirectX 8 code for PC can be recompiled for XBox *but* it'll probably run at 30-50% of it's potential framerate, so you have to recode stuff anyway. This is mainly to make the best of the different nVidia hw compared to the PC (dual Transform and Lighting engines in the XBox).
So DirectX gets you pretty much all the PC market - a good way towards XBox and nowhere on the (very profitable) other consoles - but those are a major rewrite anyway - you don't code a PS2 using OpenGL!!!! It runs like a dog if you do!
On a final note, DirectX 8 was meant to be Fahrenheit (I think the name was) a merger of OpenGL and DirectX, a huge concession by MS after the caning they got from JC and other games developers - but MS and SGI had a falling out (SGI didn't adopt NT fast enough and came out with Linux boxen). In the end MS improved DirectX enough anyway.
my 2 cents
It could be with all the reports of security holes in MS Office and back doors etc that one part of the DoD has decided they'd like to run something where they can view and check all the source code themselves to make sure it's secure.
Also, I'm sure the DoD has a lot more than 25k PCs, so this must be just for a small part or division where security is more important than features or compatibility.
Well I've read articles for and against Mr Elz. What I can say is that the Internet untill faily recently in Australia, was firmly controlled by the universities. I remember applying for a .au domain name about 6 years ago - I don't know if it was Mr Elz that I spoke to, but some guy whom I was told handled all .au registrations at the time said (after 3 months trying to track him down) "I'm not going to give your company a domain name as we believe the Internet is for academics and research not for commercial use, so I will never approve your application."!!!
:-)
This attitude realy sucked and definitely held back Australia's entry into the commercial domain on the internet, ultimately hurting the development of that type of business here.
On a side issue I notice the govt has decided to ban internet gambling here - except horse races (as they have a mechanism for taxing that). The funny thing was many pokies machines use the internet for monitoring them and the wording of the draft legislation would then ban most pokies machines in Oz which via tax are a big revenue earner for the govt!!! The Prime Minister was heard to quickly assure the gambling industry (non internet!) that the draft would be ammended! He said something about "unforseen loophole" - only he meant they *wanted* a loophole and had not forseen to add it.
my 2 cents worth
Same thing happened to me when offered a job by a very large Silicon Valley unix systems manufacturer (who shall remain nameless).
They had this one paragraph that covered just about every thought I'd have whilst employed by them. I was working on a significant project at the time in my own hours and refused to sign. Their comment was "We thought you might say that, but we thought we'd try it on you anyway"!!!
I got a friend who is a laywer to expand their paragraph into a 10 page document that defined "working for", essentially meaning only when working on jobs under their direction (not just in working hours: what if you work late? and not just in 9-5 hrs: what if you have a sick day and work on your own project at home that day?)
It took two months for them to okay it but they said my final document followed the spirit of their own shorter paragraph anyway.
Certainly in Australia I've known of a case where a company tried to get difficult with an ex-employee (taking an idea he developed in his own time to a new employer) and failed before it even got to court. It just couldn't be proved. I imagine if a *lot* of money was involved things might have gone differently.
I remember reading about a guy who did his Phd on this at least 5 years ago. He intended making cubes, each containing a microcontroller, power and motors and connectors (electrical and mechanical) so they could interconnect and execute a generic command given to the whole pile of blocks. The would slide over each other and move by sort of flowing along the ground. I think there was an picture of it too, maybe this was in New Scientist? He only managed to make about 6 cubes in his 4 yr Phd(!), each about 4 inches on a side, made of precision engineered parts. He had hoped to shrink it to 1cm cubed (and eventually much smaller, but this was before people were really thinking of micro mechanical stuff too seriously, and he ran out of time anyway).
Anyone else recall this?
The only time I've used netscape's or Outlook's calendars is for To Do lists. These are a major plus when those really busy weeks crop up from time to time and my To Do list extends beyond 30 or so items and I run the risk of forgetting important things.
To Do lists need %age complete and a due by date/time. One thing I haven't seen in ToDo lists that I'd really like is not just a terminal date/time but also how many hrs the task is expected to take. It should then keep tabs of how much work has been done.
Also I'd like a mini-Gantt, so that it can display the tasks I have for the day or week and show free hrs or if I've over committed myself by eg specifying 10 tasks at 2hrs each in a day! This mini-Gantt (or any decent Proj Managment sw, unlike MS Project) must allow you to specify if a task is able to be carried out at the same time as other things or if it is exclusive of anything being done.
This Mini-Gantt should also automatically order tasks in terms of priority (defined when I created the items in the To Do list) and automatically schedule my day, with a pre-defined breaks allowed for morning tea, lunch etc and breaks for mtgs and allow travel time to meetings at known locations. It should summarise the following days events b4 I leave work and should also have that summary as a popup when I first arrive and/or login in the morning.
> > If Echelon is reality (and that's a big "if"), the US is the only player.
> Actually, everything that I've ever seen on Echelon suggests that it's a joint U.S./U.K. project.
Echelon is run by UKUSA (pronounced "you-queue-sah") consisting of the United Kingdom, United States and Australia. The whole Echelon thing came to the fore after an Australian Govt official made comment about a facility in Australia and what it's purpose was.
Two words: scene complexity. Scene complexity refers to the fact that polygons in a scene overlap each other (from the camera's viewpoint) - ie in an indoor scene there are polygons for the wall in front of you but if there's a room beyond that then there're more polygons that get calculated and put into the frame buffer but are ultimately get overwritten by nearer polygons. An outdoor scene can have a scene complexity of 3.5 - 5 (or average). So at complexity 5, each pixel is overwritten, on average, 5 times. At PAL resolution you need 50m polys/pixels a sec, *on average*. But there will be times you are suddenly in a position where the complexiy is far higher. For example when standing in a town and looking down a street of buildings. The complexity may be 10 or more (before other culling algorithms reduce the workload). So at reasonable PC screen resolution, of say 1024x768 at 60 Hz (~50m pixels) you'd need 250-500m polys/sec to ensure you have enough headroom. This is why flight simulators etc still don't use PCs! nVidia is doing really well though, in perhaps 4-5 generations (2-4yrs?) they'll have this performance. But then we're talking about 48bit colour (if they listen to John Carmack and others) so we're talking up to 6Gbytes per sec. Of course then we're talking about loading in textures *fast*. In an outdoor scene you might easily have a few hundred megs of textures, turn around quickly (say it takes 6 frames) and the system may have to load maybe 100 meg of textures in 1/10th a second ie 1GB/s of textures. This is a lot of bandwidth PCs and consoles (now and near future) just don't have. Of course the alternative is to have your game companies spend 2 years fine tuning the geometry of everything you look at so it doesn't run too slow. Which is exactly what *does* happen for now. But how long before we grow tired of viewing painstakingly handcrafted 3D scenes and textures that takes years to tune, rather than more natural and dynamic scenes? And will we be happy with 1024x768 by then?? My guess is there's still a long way to go.
I have one question here: will software really need more and more CPU performance as time goes by? This is an unbelievably short sighted viewpoint from a Slashdotter! In 30 yrs time I *really* hope my neural tap, eyeball quality rendered (8kx3k, 48bit, 100fps,100Gpoly/sec), speech and vision recognition system doesn't run on a PIII800 or a P8/10k!! It had better use 10k times less power, have 10k times the grunt in 10k times less space and be as far ahead of Pentium anything as todays CPUs are from vacuum tubes. I'd hate to have even a super-duper-duper PC strapped to my head! :-) And that little sucker aint going to be running Windows, Linux or any other OS of today, perhaps a very distant relative. That OS had better be very robust, secure, realtime and distributed, probably something brand new, lending from Unix, Windows and other stuff like QNX and pSOS. UI will be irrelevant. Can any of you honestly say you don't believe this will happen?
I think the main difference btwn a cluster and a single image multi-cpu box is for the programmers. You can't mention "phenomenal inter-module bandwidth" and then dismiss it. That *is* the whole point (well, mostly!). You *can* have a single OS image over a cluster but that's not sensible, so it's not done, rather special libs are used. On a multi CPU box the programmer can choose to ignore the architecture and still get very good speedup (if it's the sort of code that can benefit from multi CPUs). This is because the CPU and to some extent the OS at runtime can manage the modest slowdown in accessing memory in non-local modules. Whereas in a cluster you have to rewrite your code and design it specifically for the architecture and the slowdown for remote memory is unacceptable. I've had code on a single CPU box, recompile it unchanged on a 20 CPU SGI and seen a 18.8x increase in speed. I didn't have to make any changes at all in 20k lines of code (ok, I was lucky too!). You just can't do that in a cluster.
saying:
"The best thing about Xbox is that it won't change. Ever. Judging by other consoles, Xbox should have a four or five year run"
Is totally wrong. You'd think they'd realise (especially a bright fellow like Abrash) that things change faster than they used to. Cheap 3D graphics cards today blow away cheap graphics cards of just 2 years ago. By the time the XBox comes out, it's performance will be average, but even if it is ahead of the pack it will fall behind in 6-12 months and be obsolete within 24 months.
I understand MS is considering making the XBox upgradeable. A very wise move, but they'd be *much* better off focusing on making DX8 stable for many years, rather than requiring developers to code to a low level to achieve the advertised performance. But of course, this is just the sort of thing MS has shown no foresight for with DX in the past. A very good reason to program the XBox in OpenGL only.
Actually, it didn't start with Drexler, it started with Richard Feynman, on Dec 29th 1959 in his now famous lecture "There's Plenty of Room at the Bottom", you can find a copy of this at http://www.zyvex.com/nanotech/feynman.html
From articles I've read on the making of Matrix and watching the extra footage of it's making that comes on the DVD and talking to a guy who worked on the film, I get the following:
Yes they did use a series of still cameras, yes they did then use a computer to interpolate some of the shots.
What also happened is that they used the computer to create a 3D model, based on the data from the different view points. Then they took the images in the shots, used them as textures on the 3D models and voila! they had a 3D model (well, lots actually, one for each cinema film frame). From this they could move the camera *anywhere* and were not constrained to the path of the original cameras. This enabled them to follow all sorts of new camera paths (except where texture data was missing).
This is a major step forward! As someone else mentioned, they also computer rendered the background. On the DVD you can see the wireframe models and incomplete renderings of the subway station.
The issues of 1) right to carry firearms, 2) rights to access whatever I want on the net and 3) freedom of speech are NOT the same thing.
To think they are the same is a kneejerk simplification. Think a bit more before responding. I'm quite happy in my Australian city knowing that my neighbour or the local gangs of kids don't have easy access to automatic weapons.
I'm yet to meet an intelligent Australian or American who believes complete freedom to own whatever armoury people want is a good idea. Most Americans don't realise what a laughing stock the USA is with their 18th Century attitude to firearms which they protect so fiercely whilst having such an appallingly high crime/death rate using weapons!
Perhaps the NRA should push for the Swiss solution of every male between 18-60 *must* own a working firearm and must do a year (or is it 2?) of national service training.
Maybe it's just me, but when I'm in the US (fairly often) I feel the average US citizen is much more worried about being shot that being sexually assaulted by someone who looks at porn on his PC.
Like when I was in Washington last year and was told "Whatever you do, don't go to the bad side of town, it's not safe." So I'm educated about the massive crime rate in the nations capital. So please, criticise the recent Net laws, but don't lecture us on firearms. (And I won't mention the appalling racist hierarchy that is openly maintained in the US)
Why is it that the "land of the free" is such a dangerous, unsafe place compared to "close minded" societies like Australia?
I took part in a series of psych studies where they were measuring the rate at which your brain processes information and relating this to IQ. The good news is that they found a direct correlation. There were a lot of different tests, but one that's relevant used a flickering light controlled by computer and you had to say if it was constant or flickering. Over a series of tests they determined the fastest speed at which your brain perceived the light to be flickering.
:-)
There's no thinking involved as such in this kind of test: either you see the light flickering or you don't. The bad news is that people of the highest IQ (measured by 6 different kinds of test - it took days!!) could detect flickering down to 16 milliseconds (IQ of 130-140) which is about 62Hz. Lowest IQs tested (90-100) were down around 30ms. The relationship was *roughly* linear.
Hence, next time you're in a room and you're the only one complaining about flickering lights or screens, you might just have to console yourself with the thought that you're the brightest person in the room!
Personally, I had the fastest flicker rate (16ms), I notice flickering lights and screens if I think about it, but don't seem to be adversely affected by fluro lights. Maybe I'm just used to it (my original PC ran at 1024x768 *interlaced* ugg!! I used to think it was horrible to stare at, but strangely, after a few weeks I didn't notice any flicker at all). I'm sure other factors are also involved. (similar to the story of the guy with the upside down glasses, but that's a major digression
By the way, I much, much prefer the new "daylight" fluro tubes you can get, which have a much more natural spectral distribution, making other fluros look yellow or brown by comparison.
http://www.caida.org/Tools/Plankton/Images/
It seems AOL is preparing to take on Microsoft.
:-) but remember, that we will never, ever, ever stop. We will always go forwards.
I know that sounds like paranoia but consider this: AOL paid out major dollars for Netscape, then said "No we won't use it." Yeah, right, we believe you. They said they didn't want to hurt the relationship with Microsoft that gives them an icon on the Win98 desktop.
So they spend $800m(?) or something just to save little buddy Netscape and with no thought of business advantages - aside from the relicensing of Netscape server products to Sun, which has just recently announced dropping it's own Web server sw in favour of Netscape, which is owned by AOL with whom Sun took out their "I'll scratch my back you scratch mine" deal regarding hw in exchange for advertising. Who bought *that* story? Sun wanting to do some $400m of advertising directed exclusively to AOL clients? Ha ha ha ha [head falls off in embarassment]
So we can assume AOL *will* use Netscape. I recall reading something about the next version of the AOL client being "fully HTML compliant". Their deal with Microsoft would have expired by then no doubt anyway.
Then we have Microsoft's recent loud noises about a new audio standard "a bit like MP3" (Windows Media 4.0) that includes some kind of protection for artists and publishers. Go to the MS web site and click on the Public Enemy link and try to play it. I have the latest version of WinAmp and it loads but gives an error - the protocol is sufficiently different to not be MP3 compliant. With the MS stamp of corporate approval and their marketing, Windows Media format will sweep aside MP3 freeware encoders/players and become the defacto standard. It ships with IE5, Win98, Office2000,Win2000 - it will be on a *lot* of desks. I know that MP3 is wonderfull just the way it is, but MS don't see it that way as they don't control the standard.
So, AOL see this burgeoning market in streaming audio and audio formats as being a big up and coming money maker and MS has a 3-6 month jump on them. They *know* they're going to be changing to Netscape browsers RSN and their users won't be able to access MS's new sound format files easily. What to do. Well when you're as big as AOL you buy the competition to MS and use it as a weapon. I think AOL are on the up and up and this will be an offensive war (ie AOL doesn't look to be sitting back and defending it's territory, they're moving forward, bless their little socks. Personally I hate 'em because the number of friends that got sucked into non standards compliant browsers, emails clients etc whom I have to tailor my email for).
Oh, and don't AOL use Suns, SGI (mainly?) and BSD boxes? What does that tell you about their technical inclination towards what Microsoft would like the world's businesses to run on - NT??
I think we're starting to see some major battle lines being drawn. It's not NT vs Unix. It Microsoft vs Other Companies. Keep a watch on AOL. Oh, and something is going to happen over at IBM, I'm sure of it. I bet you thought they've been kind of quiet lately. I think with the IBM witness at the DOJ case coming up the gloves will start coming off IBM. They're looking very strong financially.
So this isn't NT vs Linux. That's kind of what seems to be the frontline for we Linuxites. But take a step back and see that the HP,DELL,SGI,IBM uptake of Linux is because Linux is a weapon against Microsoft. MS wields NT, Office, Win9x and some other "standards" that flow from that.
Sure each nose-to-nose battle is NT/Linux, ASP/PHP, DCOM/CORBA etc etc, but that's short sighted. This is why I think ppl in the Linux community argue over suits, corporate types and where RedHat is going. This war needs to be fought on all levels. People in the kernel or drivers or wm battle front shouldn't belittle the necessity of the people fighting by corporate alliances, packaging distributions, binary only applications etc. We are all on the same side.
A guess we have to consider why we are on this side and what the goal is. Is MS the enemy? I would tend to think that MS is the *current* enemy because it is the most obvious incarnation of what many people oppose: freedom to choose without bias in a competitive, non-prejudiced sales, marketing and technical "marketplace" (free sw/hw or otherwise).
MS is a bump in the road to this goal. Some people wil see the bump, some will shout out "Lookout!", some will drive around, some will go so far around they'll go off the road into a ditch, some will end up going cross country, some will ride straight over the top. However it happens, the bump will eventually be behind us and won't look so bad in the rear-vision mirror. There will always be more bumps and maybe the wreckage of some crashed cars
My 2.5c worth.
cheers
Michael Snoswell
XFS is a lot more than "just" a journaling FS. One of it's other major components is guaranteed I/O rate partitions. When you create these special partitions, the OS tests the disk I/O rate then lets you specify up to the tested rate and then the OS will guarantee you the I/O rate you selected (can be less than the max), nomatter what the machine is doing (ie even under really heavy load, network, UI etc).
This sort of thing is needed when doing uncompressed cinema res images at 24fps (or HDTV) where you need 90-130Mbyte a sec from the disk nomatter what else is going on.
There's a cable channel using 24 odd uncompressed TV res video streams for live delay rebroadcasting (across time zones) using XFS. Works nicely.
Linux doesn't need this right now. Why? Because it's kinda obvious that the whole OS needs to be in on this act, it's not just a FS thing. True, the guaranteed rate stuff can be treated discretely (ie left out). But I think people may be naive when they say, "Yay, use XFS."
I haven't seen what EXT3 promises, but I bet that the current implementation of XFS has fingers going *deep* into IRIX that won't make it a fast retro fit into Linux, compared to EXT3 (unless EXT3 is just conjecture at this time, or only a modest improvement. This'll be why Linus wants a full rewrite: to get greatly improved performance will need a lot of changes on the OS side, and if you're going to go to that effort, you'd better make it worthwhile)
Linux does need a journaling FS and XFS may be the best bet, but it won't happen quickly unless SGI puts some serious resources behind it. For any other effort to pool together enough ppl for long enough to make it happen is just too unlikely for us to just sit around hoping for.
Also, just who has the resources to test large production systems (4+ CPUs) on an OS under test? Corporates, that's who. And they'll contribute their code to Open Source, right? Because...?
This *will* all happen, but I think some of these tougher OS issues will need corporate backing that may have a price current purists will not like but will have to accept (ie less than Open Source code licenses or maybe even (cringe!) binary drivers).
My 2 cents worth.
I visited HITL last year to give a talk and chat with the people there. I was lucky enough to have a go with the VRD and I must say it was *very* impressive. These guys have spent a lot of time and money on this project (millions over years) and they've now got it down to a briefcase size.
So I guess the bad news it that it'll be a while before this stuff fits onto a light HMD, they made noises about it being 5 years away, maybe more. Also, they're having trouble pushing the resolution. Given that the human eye has a res of about 10000x6000 (depending on who you talk to) and this display is 640x480 then it only covers a small portion of the visual field. Given 5 or so years maybe we'll have cheap hardware to generate graphics at that res too! That's the frustrating bit.
The good bit is that the colour clarity is excellent and the picture is crystal clear and rock solid. And the extraordinary thing for me is that without my glasses it's still perfectly in focus! Because it writes directly onto the retina, effectively it doesn't use the lens in your eye, so people with very bad vision still see a beautifully clear image. It's kind of odd to take off my glasses and the whole room is a blur (mostly) and then to suddenly see a diamond sharp image looking like a flat panel screen hanging in space about 6 feet away.
There is also some problems with spherical aberration if they try to may the image too wide and getting the scanning components to handle higher (and the right) frequencies. However, *most* of the hard work has been done and it seems to be mostly a clear road of progressive minaturisation untill it is a head mounted "heads up display" (with optional flip down cover for an immersed view).
It doesn't take long to realise that this is the way most computer displays will go. It's not a matter of "if", but "when".
cheers
MIPS and PA architecture chips are the architecturally closest *working* chips to Merced that we have today.
I remember all the trouble MIPS had when they rolled out the R10000 chip. Initial performance was not up to spec because early estimates of performance were pretty much correct on the SPECfp numbers, but underestimated *how long* it took to get those numbers. It took a couple of years for the compiler people to wring out the best performance (ok clock speed was off too, but that was not the sole reason, nor was internal CPU wars within MIPS/SGI).
Now the Merced is more complex than the R10000 (and at least the R10K has some *vague* similarity to the R8K, and PA architecture has been around for many years, so these companies had compiler writers experienced in some of the problems they were up against). Intel is starting from scratch here. I'd say when they've done first tapeout and have silicon in their hot little hands, it'll be at least a year before the compilers get close to the performance they hope for.
Meanwhile, IA32 will be up to similar spec and Alpha, PA and MIPS (and SPARC perhaps) will be serious contenders.
Just last week or so MIPS and HP announce they were reviving their CPU development for a further year or so (ie another generation), rather than trusting all to Merced (I assume that means last MIPS or PA in 2003-2005 now). What news did Intel give these guys for them to decide to make such an announcement??
cheers
Michael Snoswell
Whilst I think Rob is heading in the right general direction I can see a few problems.
I have to agree with those who say moderators should be able to moderate and post to the same news story.
I only read those stories which interest me (maybe 1 in 3) and these tend to be ones in which I already have an opinion or in which I have some expertise, otherwise I wouldn't be interested in it in the first place. Hence my moderations *and* my comments should be considered acceptable for that particular story.
Secondly, when I read postings, I've taken Rob's advice and set my threshold to +3 and listing in descending order of score. This has reduced my reading time dramatically and is overall a Good Thing.
The trouble is, if I'm made a moderator, then I will only see comments which have already passed a first phase of moderating. Now, considering I'm only reading a few choice postings anyway, I don't mind if I get an extra tab next to each so I can go +1 or -1, but there's no way I'm going to take time to read AC postings.
I imagine a lot of lurker, experienced users like myself will feel the same. If we're added to the ranks of moderators, we will only be moderating the "quality" posts, hence +5 is too low for an upper limit.
cheers
Michael Snoswell