Oh I really want a really large sheet that's touch sensitive. I can just see the possibilities.
Get an Architectural sized E sized sheet and you got yourself the right size for almost any diagram. Make it touch sensitive and you can write notes on it, or even make corrections in the middle of meetings. Or even on a building sight. I don't care if it has a lousey refresh rate. 1 or 2 hertz woujld be perfect for this. The possibilities for coloboration and review are enormous. No longer do you have to print out E sized sheets, mark them up, send them back to the drafter and have him correct it. rinse, lather, repeat....
Any type of diagrams/plans would be perfect for this type of item. Post your schedule on the wall, and when the big wigs change the schedule it gets updated automatically.
But the really geeky thing. I'd love a battle map made of this, if it was touch sensitive. You'll be able to tell when a fireball will kill the PCs. You'll be able to figure out who has line of sight. All done with a computer who knows where everything is because the screen is touch sensitive. Hide part of the map, and as the PCs open up the door, reveal that map, no more waiting for the DM to redraw the map.
Maybe I should stop now before I have wet dreams about this stuff.
While I generally agree that its not a good additude to take, there is some basis for it. At least in Canada where the local students pay a lot less then the foreign students because the local students pay taxes. With government cutbacks the universities get less money then they use to for local students, but the money payed by foreign students is money in the bank, they get all of it. When I took Computer Science, there were maybe 5-10% local students, the rest were all from China, and it was a real pain to work on group projects with them, because their english was terrible and my chinese was worse. So there is some resentment from the local students who have to compete with the foreign students, and see that the foreign students get precidence because the Universities get more money for them. And since they come from father away they get preference on dorm rooms forcing the local students to either commute hellishly long distances, or live in hellishly priced housing, which can be up to 10 times the amount for housing on campus. For the first few weeks when I was in university I was at the YMCA till someone dropped out of residence. Like almost everything in the world, it comes down to money, and if a university can get more money from tuition from a foreign student then they will take a foreign student over a local one.
A GSM user can use the same account with as many phones as they wish, switching from one to another in the time it takes to remove the SIM from one phone and slot it in another. That's why most smartphones are GSM - because, frankly, anyone using a smartphone on a cdmaOne network (or, god forbid, a D-AMPS one) will find they're stuck with having to use that phone for all their usage associated with that number.
Actually CDMAOne does have the capability. Its just not used in North America. But it is manditory for all CDMAOne phones in China to have the account information stored on the R-UIM card. The R-UIM card is almost exactly the same as the GSM SIM card. The only difference is the name and the CDMA specific file structure. I have R-UIM cards with both the GSM and CDMA file structure on them, so they can be used in GSM and CDMA phones.
This is mainly a provider issue and not a lacking of the CDMAOne or IS-2000 standard. If you want look at http://www.3gpp2.com/Public_html/specs/CS0023-0.pd f For the relivant standards on the R-UIM card, you'll see that for the most part it points at the GSM standards for how to use the card.
www.3gpp2.com is a good site to look at the standards for IS-2000 and what features are there. I don't know of any of the features you've listed that aren't already in the standards, its just that the providers aren't using them.
What I mean by connected and unconnected isn't based on a connected or unconnected IP protocol. But rather on whether the phone is in the connected or idle state.
I haven't looked through the specification in detail to figure out how long a moble will stay in a connected state when the packets are not flowing. The mobile and base station will decided how which state to go to, depending on demand and how long a station has been idle.
If there's 20 people waiting for a high speed connection and one of the current high speed connections is idle that will probably be moved to another state where it isn't using alot of bandwidth till its needed. Or the channel will be shrunk, or something else can happen. There's lots of things that can change the bandwidth that is used.
Well for a 50 byte packet, if the phone is in the fully connected state it would act like a normal network for 50 bytes. 50 bytes really isn't all that much, so if the server is close it would be maybe 20 ms if using the new 5 ms frames, or it could take 60ms if using 20 ms frames, if its a fair ways way maybe 100-200 ms.
But if you weren't in the connected state you'd see something like this in ideal conditions:
- The mobile would do an access probe with the 50 byte packet 80 ms for the frame to complete, this depends on how large the access preamble is - The mobile would get an ACK back from the base station saying that the access probe packet was received. This will probably take about 40 ms is probably the fastest I've seen. - The mobile would then wait for the returning packet. The frames on the common signalling channel are 20 ms. And if the mobile is lucky it could be withing one or two frames. - That is if the mobile expects a reply packet, if not it will go back into slotted mode, inwhich it only monitors the common signalling channel ever few slots. In most systems this is 2.5 or 5 seconds (roughly), so this will probably be the big portion of your latency.
So you're looking at a minimum of 160ms to 180 ms latency on that packet, and maybe more depending on the network between the base station and the receiving server. I would guess you're looking more along the lines of 500 ms to 1 second.
If you throw in bad radio conditions I think you can get up to 24 seconds before the mobile gives up on its access attempts depending on how the network is configured. And in bad radio conditions you might miss the reply packet the first few times, so add 2.5 or 5 seconds every time it comes through corrupted.
So the rough answer is anywhere between 20-40 ms, to 30-40 seconds.
So my north american mobile phone (Telus mobility in Canada) has a sticker that says Qualcomm 3G CDMA on it and the service is advertised as 1xRTT [telusmobility.com]. You're saying that this was renamed to 3G because they couldn't pull of the 'true' 3G for various reasons?
Basically yes. And Telus is only Rev 0 of CDMA2000, which is only a stepping stone to Rev A which was suppose to have all of the 3G features. I don't remember all the features, but they're mostly funky features to improve capacity and coverage.
CDMA is a confusing world with overlapping acronyms and meanings. Its a pain just trying to figure out what something means from one day to the next. 1xEVDO use to mean 1x Evolution Data Only. I just heard 3rd party that it means 1x Evolution Data Optimised. Because the competting standard to 1xEVDO was 1xEVDV which stands for 1x Evolution Data/Voice. And who wants a data only phone? Its all marketing. And I'm not even sure if 1xEVDO and 1xEVDV are competing anymore but 1xEVDV is the next generation after 1xEVDO.
There's too many number and acroynyms to keep track of, and the meanings keep changing depending on who's talking and which drugs their smoking today.
It depends. In an effort to save on capacity and on power CDMA2000 uses a state machine for its packet connections, determining which access mode the phone is in. Depending on which state the phone is in will determine what the latency will be.
It also depends on packet size. Most of these features come in the next rev of CDMA2000. But this is what happens for the most part:
- If you're using a circuit data mode then you're connected all the time, which costs the carriers and you a hefty chunk of money. This mode is used in all versions of CDMA before IS-2000 rev 0. - If you're using Packet Data, you can be connected with a high speed channel downstream and a low speed channel upstream, or any combination. Depending on the size of the packet the latency can change, that's just standard networking. - But if you're link has been idle for a while your phone may have dropped the channels and gone into another mode. When you send a packet, if its small enough it will go accross the common signalling channel which everyone uses to tell the base station its alive or it wants to make a call. And if the returning packet is small enough then it'll go back accross the the forward common signalling channel. - But if you need to send or receive a large packet, or a stream of packets you need to set up the forward and reverse channels again, and this can take time. Depending on which state the phone is in when it needs to set up the channels this can take as little as half a second, or as much as 3 or 4 seconds before the packets start flowing again.
So depending on which mode you're in, depends on how large the latency is. It also depends on the radio environment. CDMA is fairly good in noisy environments, but if it gets into a really bad area packets will have to be retransmitted several times.
Most carriers who use CDMA now won't go to W-CDMA, they'll go to CDMA-2000 which is a different standard from W-CDMA.
CDMA-2000 is fully backwards compatible with current CDMA systems. If you try to get a 2G CDMA cell phone certified to work with Lucient base stations they will test it for compatibility on their CDMA-2000 base station as well.
The upgrade from IS-95/TSB-74 to IS-2000 is mostly replacing the base stations with new ones, and upgrading the back end. They can also set the upgraded base stations to not operate in 3G mode until the entire 3G network is ready and tuned. And even when they make the cut over to 3G the 2G phones can still use their network. I don't believe you can do this wiith the GSM to W-CDMA transition.
The refit for GSM and TDMA systems is much more costly because of the incompatibilities with W-CDMA. GPRS and UMTS aren't quite so bad, but aren't as good as the upgrad of a CDMA system.
Sprint uses CDMA the last I heard. And no-one is deploying CDMA in a 3xRTT solution right now which was considered true 3G.
Also from my understanding Qualcomm, currently the major/only producer of 3G CDMA chips, is releasing their Rev A chip later this year. So I assume Sprint is using Rev 0 which isn't 3G.
IS-2000 (or CDMA 2000 depending on which marketers you talk to) was set up in phases. Initially Rev 0 and Rev A. Rev 0 was a stepping stone from IS-95B (which is only used in Korea), to 3G. Rev A was suppose to be the full 3G solution.
A few years ago 1xRTT (which Sprint uses) was called 2.5 G. But then came the nice licensing auction in Europe which nearly bankrupted the carriers there for W-CDMA licenses. W-CDMA and 3xRTT require nearly the same licensing because of how much bandwidth they use. 3xRTT uses slightly bit less, but still 3 times more the 1xRTT. 1xRTT can use the same licenses that the carriers in North America and Asia currently have for IS-95. The carriers in North America and Asia mostly scraped plans on going to 3xRTT after the European auction, but the didn't want their systems to be called 2.5G, thus 1xRTT is now called 3G.
So depending on who you talk to Sprint has (or has not) a true 3G system.
I don't think we'll see true 3G anywhere in the world for at least 5 years. W-CDMA which is the competitor to CDMA2000 is still having some technical issues to handle. GPRS is 2.5 G though alot of people call it 3G, its a stepping stone to true 3G.
FWIR 3G is suppose to have 2+ Mbps transfers for stationary wireless stations, and improvements in bandwiths for walking and driving speeds, though I don't remember the numbers. Improved battery time (WCDMA stations haven't fufilled this with stand by times of ~50 hours), and better capacity.
CDMA-2000 1xRTT fufills all these requirements appart from the bandwidth. Rev 0 of CDMA-2000 only allows a maximum of 153 kbps, and Rev A allows 307 kbps. New inititives for 1xRTT do allow for improved bandwidth. They tried 1xEVDO for the world cup, but from what I heard they couldn't pull it off properly and it still needs some work. 1xEVDO is suppose to offer 2.4 Mbps.
From my english classes, every word in a title but a few are always capitalized. It was one of the few lessons that was beat into me in grade 2 when I didn't capitalize properly. From my memory only words like as, and, & the in a title don't get capitalized unless they are at the beginning of the title.
So while capitalizing micron in the title can be confusing, it is proper english, even if the rest of the story and this comment aren't, (or is that ain't?).
I don't have a nokia phone but if you look on the picture near the bottom you see the baseband chip. U0801 is the chip I mean.
It is the VP 22353 PR release of the CDMA+ 200 chip produced by Philips Semiconductors. Nokia doesn't currently use this chip for their baseband.
This is an actual phone, and while Hop-on only has 15-20 people in it. Holley Communications which wrote the CDMA protocol stack software for the chip and the DSP firmware in Vancouver definetly has more then 20 people. Holley also did the Radio, Industrial and MMI design in Dallas. Over the life of the CDMA 200 project there have been around 200 people working on the project between Holley Communications and Phillips Semiconductors.
As for standby time on the phone, it is pretty substancial. I don't know the exact figures, but it is in the 5-6 day range. Which is around average for a CDMA phone.
Not that I can resist tooting our own horn, but the technology behind the phone is solid and well developed. Despite what the slashdot community thinks of Hop-on, the technology behind the phone is solid and developed by a very professional group of people, and some of the most knowledgable people in the CDMA technology world.
And yes the phone can be fabbed inexpensively. Which is one of the benifts for going with the Philips CDMA solution over the Qualcomm set of chips. I do know that the people doing the phone/radio design spend months going over the BOM/design to make it as inexpensive as possible.
And yes there isn't any hook in on that board for an LCD, and that helps with the price on the board, and also increases the standby time. Until you see it for yourself you'll never know how much a simple black and white LCD sucks the juice when you're trying to meet a standby spec.
I'd have to disaree. Going public without a product is stupid I agree. But going public when you're small and you need the capital to expand the company is vital.
For instance if you have a product you're ready to product but you can't get a factory to produce it you go public to pay for the factory.
My company right now is private, and we're looking at going public to pay mainly for development of the next product. Yes we already have one product that we are selling and expecting more. Actually we're looking at the next six months and most of us are scared because it looks like we'll have more major customers then we have resorses to support them, but we don't currently have the money to hire more engineers. And out next product requires one hell of an expensive license from the patent holders. Basically on the order of $50M. We're a small company, we don't have that money hanging around, and VCs aren't going to give us money just for the license so we have to go public.
But going public, just to go public and make millions for the owners, VCs and employees is stupid. You have to have a reason to go public, and a need for all that capital, and then you should really only sell the amount of stock you need to cover the capital and non-capital expenses (salaries and rent and all that), till you can get the contracts and money coming in to cover those expenses. Then if you need more money in the company for later you should have the money from the contracts, or you can offer more shares.
Anyway, I'll probably be modded as off topic. But there are instances where small companies should go public, but most dot-bombs don't fall into this, most were just in it for greed with no real plans or products.
In the case of waring ISPs mentioned in the article, you'll see 95% of the net, you won't see the 5% served by the other ISP.
If you're behind a hacked router then you might see all the net you might not. Most likely you won't see anything.
There's also the case of companies that had IP addresses assigned to them, that keep them unreachable from outside. Depending on their policies you may not see the net.
And if you're on an IP address assigned to an ISP or company that ceases to operate, you'll most likely not see the other 95% of the net.
It all depends on the net, the net isn't all that deterministic. Its kind of like a living organism with a personality you have to beat with a large stick. Usually you apply the stick to your neighbours if you're using a cable modem. Or to your ISPs first tier support (who often need a good thrashing, 'No its not my computer your router isn't working').
I haven't gone through the interview process, but I know several people who have. And there's no way that a non-programmer can 'pose as a programmer' and get through the interviews.
My friends that went through the process found it very intense. And while we all dislike the products that get sent out from Microsoft, I don't think anyone will disagree that they hire some of the best programmers out there. You might be able to 'pose as a programmer' like the article says, but there's no way you can pose as a programmer of the caliber that Microsoft hires without being one.
The article seems to read (and I might be reading things into the article) that the terrorists posed as programmers, were hired, made the backdoors and then left. Almost like they walked in with a breifcase and then walked out. This is just crazy, appart from what other people have said with code reviews and other processes to make sure this doesn't happen, the complexity of an OS prevents this.
I work on a relitively small project, the core protocol for a cellphone, and it takes on average 6-10 months of working with the code before you can understand the complexity of it. I hate to think of how long the average programmer at Microsoft goes through to learn the code enough to make backdoors that avoid detection past code reviews.
With an OS as complex as WinXP, I would assume that it would take a good year before anyone knows enough about just one small section to make a backdoor. And that would probably be easy to find. To hide the backdoor it would probably have to cross several sections of the code, you just can't do it with a handful of people posing as programmers. You'd probably have to do the backdoor as a small software project, just to handle the complexities of it.
IMHO, if there's a backdoor in there, then it was put in there with full knowledge of Microsoft. More likely if there were terrorists posing as programmers, they found a backdoor that Microsoft put in there for the government, and not made the backdoor themselves.
I was looking at something similar with using two linux boxes as NATing firewalls and acting as DNS servers. I wonder if this would work for multi-homing with two providers.
What I was thinking on doing was having the two linux boxes setup the same way, having their primary default routes to the connected providers, and a secondary route as the internal address of the other firewall.
So that if their primary default route went down, they'd automatically switch to the other firewall and send all outgoing requests back through the internal network to the other firewall.
These two firewalls would be on completely different subnets and different providers. I wanted to set up my domain name to have the primary DNS on one firewall and the secondary DNS on the second firewall. The IP addresses for DNS lookups on the first firewall would match the IPs provided by the first Provider, and the second DNS for the IPs on the second firewall.
This is where my knowledge goes out the window. I think you can set the expire time for cashing name servers as low as 10 minutes. So the lookups should switch across for name address translation should happen automatically, after ten minutes when the primary link dies.
This doesn't help much with load balancing unfortuantely, though I suppose you could throttle back the DNS (so it doesn't respond) if the primary link is too loaded, or even have it set up to use the IP addresses from the second provider when the link gets saturated.
I don't think you need any support from the ISPs for this, and it should allow you the redundant links.
I don't think the tightly integrated PCs will be the be all, end all like the article says. For consumers and most bussiness uses it might be. But because of the versitility and customization of the PC you'll always see them in their current form.
Some scientfic applications require High Data Rate digital capture cards. You'll never see a chip that allows you to hook the number of inputs into it that some scientists want.
When it comes to capturing this type of data, some scientists are anal, and always wanting more data points at higher resolution. You won't see this type of data being captured on anything put direct to the computers memory/hard drive. You won't see it through gigabit ethernet, firewire or USB. The big word in this sort of application is 'Realtime', I think its mostly a fad that isn't needed in most applications, but everyone wants it, and this requires as high bandwidth as possible.
Other type of equipment in some labs use a custom control interface, often made inhouse. This type of control might move towards USB and Ethernet, but I honestly doubt that.
GPIB isn't what I call a custom interface, but I doubt you'll ever see it based in an integrated box. JTAG interfaces for embedded development probably won't make it into an integrated box. But these interfaces do have ethernet adaptors now. But even with these the direct connect to the computer adapters are still more popular.
There will always be small, important applications that have markets too small to be included in the integrated boxes, but still need to have a computer interface.
You could say this is cirsumstancial evidence but:
A while ago I had to interview a recent immigrant from China for a position at my company. On his resume he had said at his previous job he 'improved' the USB driver code in the Linux Kernel so that it would work better with their systems.
Before the interview I did a grep through the kernel to see if I could find his name or any derivative of the name, I found none. So I asked him about it, and why he wasn't on the contribution list.
His answer (and I'm going from memory), was that since it was done for his employer he wasn't allowed to release it back under the GPL and it was propritary.
From my limited understanding, they do sell their systems (with linux and the mods), and they don't release any source code with them. The system is limited to back end phone networking use, and only in China so it really hasn't come to light.
Take that as you like, while it isn't exactly proof, and I see this as more of a bad coporation then as something 'China specific.' But doing kernel mods and not releasing them back under GPL does happen.
The one thing that I haven't seen mentioned for when assigning priority/severity is repeatability of the bug.
This is where a good test plan and procedures come into play. Where I work if our boards crash when executing a test procedure, but we can't get it repeat we don't file it as a bug, but as a simptom. We of course try several times to reproduce the results that cause the bug.
Alot of projects don't have a good test plan (most I've seen have absolutely no test plans) that defines how to execute the tests.
There are of course good and bad points to having a test plan. The good points is you can repeat a test exactly and reproduce a bug and give the bug produc... er... developers something hard they can use to trace down. Most S1 through S3 bugs should be found with a comprehensive test plan.
The bad point is you might not test the corner cases and the more obscure bugs will be missed. But if you can't reproduce them, then what's the point? You're just going to start locking horns with the developers. It often comes down to alot of ego on both parts. The developer thinks his code is perfect, the tester found a bug.
Reproducability needs to be part of the whole P*/S* assignment. A crash might be fatal, but if you can't reproduce it no matter what you do, are you going to put it in as an S1/P1, and waste alot of time on your part and developers part trying to track it down. That time can be more effectively spent fixing the other bugs, and you also have a good chance that one of these other bugs is causing the crash. Without a repeatable test plan you just don't know.
Filing as a symptom allows you to have the papertrail on the bug so you won't forget about it, but it doesn't waste everyone's time trying to solve something that might take months to solve, especially when you have no frigging clue what caused the bug.
On another note some bugs can be classified as high priority to fix when they aren't all that sevear. Spelling mistakes are often classified as high priority because they're dead easy to fix. And fixing them gives the produce the 'feeling of quality'. Which is total BS but marketing can sell it.
The kind of person you're looking for will be very hard to find. A person who is adaptable and also knows software engineering. I like to believe that I'm a person like that. Actually I'm pretty sure I am the kind of person you are looking for, so I know the job hunt habbits of the type of person you're looking for.
1. You have to be lucky, and find a person who isn't happy with their jobs. The good thing is, is that most large companies still don't understand that Computer Scientists & Engineers, and 'real' software engineers have different needs then your typical factory worker. You might be able to find a high powered, adaptable scientist from one of these larger companies.
2. Most of the people you're looking for don't post their resume to the internet, or to headhunters, so going through those channels you might not find them. They also don't look in the newspaper want ads. They talk to their friends, and their friends ususually know of a job from a friend of a friend of a friend. The last time I got fed up with my company I wrote an email to a friend, got a responce back and basically got a 40% raise a month later at a different job. I didn't post to the internet, and didn't talk to a headhunter.
3. You have to snag their interest. The only way you're going to get the type of people you seem to want is to snag their interest. A lot of the adaptable scientists are like the hippies of the 60s, they want to change the world. If you offer them a chance they might hop onto it.
4. With the economy of today everyone is skitish about switching jobs. You'll have to offer rock stable stability (not California or Washington rock, they have a habbit of shaking and knocking down buildings). While it is true that the 'real' engineers (not web developers) are still high in demands, they're still nervous about switching jobs because of the economy, if you offer alot but not stability you won't get them.
Still the most important thing is who you know that knows the type of person you're looking for. And getting a reference from someone you know also gives you a better chance at finding a jewel in the rough.
I never really thought of Canada, and especially BC as a place where this was likely to happen first. We don't really seem to have quite the same tendency to get into the whole "Save the Children" hysteria as you hear comming out of the States every so often.
You forget one important thing. The NDP government is going into an election soon. They're doing everything, and I mean everything to buy the voters.
They offered to pay for that kid that needs the liver transplant that can't find a donner in Canada.
They decided to fast track the earthquake improvement in the public schools, the day after they said they weren't, which I believe is the day after the earthquake.
Yes they'll propose legislation to make this a law, but not for any concern for the children, but only to buy more votes, but appearing to be concerned for the children.
Just one last thing: before this latest election here in BC, I knew politicians lied through their teeth, but I had always thought they were decent enough to make then credible or plausible lies. The current government isn't even doing that, their lies are outright lies and almost everyone knows it. So don't expect this legislation to go anywhere before the required election on June, its one of the smaller issues/lies and the only way it'll hit the floor is as something the NDP can use to accuse the liberals of not caring for the children.
Yes I'm cynical, you would be too if you were a high tech worker, living in BC and watching the economy fall behind everyone for the past 10 years.
I remember that story, a bit better. I don't remember the year, but for those interested it goes something like this:
Buchanin (the new arts building at the time) received a bunch of new sculptures that Plant was suppose to put up. Plant are the guys who run the steam plants (yes we have steam tunnels at UBC) and do all the maintenance. Of course plant was taking their sweet time. So Engineering decided to play a prank.
They came up with the most hiddious statues you can think of and then put them inplace. They then started a campaign in the university newpaper, letters to the editor, articles and similar things, about how bad the art students' taste was and how they could make statues just as bad. The arts students fought back saying that the statues were master pieces.
It came to head one day when the engineers went out and started smashing the statues. It all ended when the engineers admitted that they made the statues, and the real ones were still waiting in a basement somewhere for Plant to install them.
The Engineers are UBC have done a similar thing with 'indecent' books at the main library. Planting the books in the library, complaining about all the smut and then having a book burning later on.
A few things in the article leaves me with a few concerns over how expensive it'll be.
The first thing that concerns me is the ranges, 150 mile radius. That's roughly, 70 thousand square miles, how many planes can you get flying in that area? I'm not even going to consider how many you can get if you take into three dimensional space.
With this amount of space to monitor, and all those aircraft, you're going to need two things. The first you need is a fast computer to keep track of all these aircraft. Second you're going to need a damn good radio, and modulation techniques to have broadcasts between all those aircraft without having all the data garbled with collisions.
For the system to be effect you'd have to install it on all aircraft, how long is that going to take? Because if one aircraft, or even a handful of aircraft are flying through the area and only they and the air traffic controllers know about them, the main selling point behind this system isn't really worth while. You still need to rely on ground controllers to tell you who's out there.
For example, if a pilot relies in this system and sees a plane on a collision course, or near collision course, he might turn, into the path of another aircraft that isn't showing up on the system.
For the system to handle 150 mile radius, you need a powerful radio, or large antennas, (both of which you probably won't find in private planes). You'll also need to install it in military aircraft, remember the F-117 that buzzed the airliner a few weeks ago? This of course costs more money and more tax dollars.
I don't see how this system will be as effective as they're promising without it being installed in all aircraft that will be flying through an area. It will help, I'm not denying that, but it will not remove the need of ground controlers and their radars. It won't even lessen their need at all.
I know the point of the article wasn't to mention if the computer would be useful, but to see how far we could go, stating that that would be the limit.
The computer would be interesting for super computers, but never for anything else. I don't see anyone walking around with a nuclear accident waiting to happen. But a super computer would be something else. The specified input and output would work in that enviroment. After all most people who work on super computers don't sit at the console. And if I remember correctly the console is usually burried so deep in the cases that it never sees the light of day unless someone digs into it to help it boot up. But then I've only heard of them, never even seen or worked with a super computer.
Having a laptop like this might be fun though, as a practical joke.
'Hey Joe you're computer's not working right'
'it isn't?'
'it hasn't vaporized the desk yet. You need a hotter computer.'
I've lost alot of faith in BBCs ability to report news. In the past few weeks we've had two reports linked from Slashdot from the BBC that were just pure crap:
Microsoft moving to BC -- Apparently this story came from someone in the BC gov't (not the minister) talking to someone not in Redmond, and not associated with Microsoft about bussiness in BC. Yet the BBC reported it as Microsoft moving to BC.
Hackers hack the space shuttle -- another BBC story. I don't remember anyone claiming that they did this, and the space shuttle is on a closed network, and NASA uses a pretty strong software development process that almost ensures a system that has a low probability of hacking. I believe a story on their development process was on slashdot recently.
This is probably the BBC trying to get more hits on their website so they can get more ad money. And the BBC does like to play hoaxes on the public. I remember hearing of a story in the 70s when they reported (as an April Fool's joke) that the new library in some city in England was built upside down. They pull other hoaxes as well, that's just the one I remember right now.
And besides this falls into one of those wacky inventions. It probably be as popular as those wacky inventions you see in all those black and white news movies. You know the self buttkicking machines, the face guard that rolls down like a shade to protect against grapefruit spray, and all the rest. And where did all these silly inventions go?
The third company should have more then just internet explorer. How is that one going to survive with just IE? After all shipping IE for free is what started this, if they up the price and netscape is still free, who's going to get all the market share? With just IE in this company then this one will fail faster then a Art's major taking a quantum physic exam.
The third company should have IE, and all the Internet products. Like IS, Exchange, and anything else they have dealing with the internet. The other two companies should probably stay the same, one OS, and the other Applications and the rest. For the third company to survive they/NEED/ a product they can sell, and they can't sell IE.
I can actually see a forth company coming out of this, one dealing with hardware. They keyboards and mice are actually quite good. That way you'll have four baby bills: OS, Internet software, General Software, and Hardware. Each of them have products that they can use to make money on.
Slashdot Mirror
Oh I really want a really large sheet that's touch sensitive. I can just see the possibilities.
Get an Architectural sized E sized sheet and you got yourself the right size for almost any diagram. Make it touch sensitive and you can write notes on it, or even make corrections in the middle of meetings. Or even on a building sight. I don't care if it has a lousey refresh rate. 1 or 2 hertz woujld be perfect for this. The possibilities for coloboration and review are enormous. No longer do you have to print out E sized sheets, mark them up, send them back to the drafter and have him correct it. rinse, lather, repeat....
Any type of diagrams/plans would be perfect for this type of item. Post your schedule on the wall, and when the big wigs change the schedule it gets updated automatically.
But the really geeky thing. I'd love a battle map made of this, if it was touch sensitive. You'll be able to tell when a fireball will kill the PCs. You'll be able to figure out who has line of sight. All done with a computer who knows where everything is because the screen is touch sensitive. Hide part of the map, and as the PCs open up the door, reveal that map, no more waiting for the DM to redraw the map.
Maybe I should stop now before I have wet dreams about this stuff.
While I generally agree that its not a good additude to take, there is some basis for it. At least in Canada where the local students pay a lot less then the foreign students because the local students pay taxes.
With government cutbacks the universities get less money then they use to for local students, but the money payed by foreign students is money in the bank, they get all of it.
When I took Computer Science, there were maybe 5-10% local students, the rest were all from China, and it was a real pain to work on group projects with them, because their english was terrible and my chinese was worse.
So there is some resentment from the local students who have to compete with the foreign students, and see that the foreign students get precidence because the Universities get more money for them. And since they come from father away they get preference on dorm rooms forcing the local students to either commute hellishly long distances, or live in hellishly priced housing, which can be up to 10 times the amount for housing on campus. For the first few weeks when I was in university I was at the YMCA till someone dropped out of residence.
Like almost everything in the world, it comes down to money, and if a university can get more money from tuition from a foreign student then they will take a foreign student over a local one.
A GSM user can use the same account with as many phones as they wish, switching from one to another in the time it takes to remove the SIM from one phone and slot it in another. That's why most smartphones are GSM - because, frankly, anyone using a smartphone on a cdmaOne network (or, god forbid, a D-AMPS one) will find they're stuck with having to use that phone for all their usage associated with that number.
Actually CDMAOne does have the capability. Its just not used in North America. But it is manditory for all CDMAOne phones in China to have the account information stored on the R-UIM card. The R-UIM card is almost exactly the same as the GSM SIM card. The only difference is the name and the CDMA specific file structure. I have R-UIM cards with both the GSM and CDMA file structure on them, so they can be used in GSM and CDMA phones.
This is mainly a provider issue and not a lacking of the CDMAOne or IS-2000 standard. If you want look at http://www.3gpp2.com/Public_html/specs/CS0023-0.p
For the relivant standards on the R-UIM card, you'll see that for the most part it points at the GSM standards for how to use the card.
www.3gpp2.com is a good site to look at the standards for IS-2000 and what features are there. I don't know of any of the features you've listed that aren't already in the standards, its just that the providers aren't using them.
What I mean by connected and unconnected isn't based on a connected or unconnected IP protocol. But rather on whether the phone is in the connected or idle state.
I haven't looked through the specification in detail to figure out how long a moble will stay in a connected state when the packets are not flowing. The mobile and base station will decided how which state to go to, depending on demand and how long a station has been idle.
If there's 20 people waiting for a high speed connection and one of the current high speed connections is idle that will probably be moved to another state where it isn't using alot of bandwidth till its needed. Or the channel will be shrunk, or something else can happen. There's lots of things that can change the bandwidth that is used.
Well for a 50 byte packet, if the phone is in the fully connected state it would act like a normal network for 50 bytes. 50 bytes really isn't all that much, so if the server is close it would be maybe 20 ms if using the new 5 ms frames, or it could take 60ms if using 20 ms frames, if its a fair ways way maybe 100-200 ms.
But if you weren't in the connected state you'd see something like this in ideal conditions:
- The mobile would do an access probe with the 50 byte packet 80 ms for the frame to complete, this depends on how large the access preamble is
- The mobile would get an ACK back from the base station saying that the access probe packet was received. This will probably take about 40 ms is probably the fastest I've seen.
- The mobile would then wait for the returning packet. The frames on the common signalling channel are 20 ms. And if the mobile is lucky it could be withing one or two frames.
- That is if the mobile expects a reply packet, if not it will go back into slotted mode, inwhich it only monitors the common signalling channel ever few slots. In most systems this is 2.5 or 5 seconds (roughly), so this will probably be the big portion of your latency.
So you're looking at a minimum of 160ms to 180 ms latency on that packet, and maybe more depending on the network between the base station and the receiving server. I would guess you're looking more along the lines of 500 ms to 1 second.
If you throw in bad radio conditions I think you can get up to 24 seconds before the mobile gives up on its access attempts depending on how the network is configured. And in bad radio conditions you might miss the reply packet the first few times, so add 2.5 or 5 seconds every time it comes through corrupted.
So the rough answer is anywhere between 20-40 ms, to 30-40 seconds.
So my north american mobile phone (Telus mobility in Canada) has a sticker that says Qualcomm 3G CDMA on it and the service is advertised as 1xRTT [telusmobility.com]. You're saying that this was renamed to 3G because they couldn't pull of the 'true' 3G for various reasons?
Basically yes. And Telus is only Rev 0 of CDMA2000, which is only a stepping stone to Rev A which was suppose to have all of the 3G features. I don't remember all the features, but they're mostly funky features to improve capacity and coverage.
CDMA is a confusing world with overlapping acronyms and meanings. Its a pain just trying to figure out what something means from one day to the next. 1xEVDO use to mean 1x Evolution Data Only. I just heard 3rd party that it means 1x Evolution Data Optimised. Because the competting standard to 1xEVDO was 1xEVDV which stands for 1x Evolution Data/Voice. And who wants a data only phone? Its all marketing. And I'm not even sure if 1xEVDO and 1xEVDV are competing anymore but 1xEVDV is the next generation after 1xEVDO.
There's too many number and acroynyms to keep track of, and the meanings keep changing depending on who's talking and which drugs their smoking today.
It depends. In an effort to save on capacity and on power CDMA2000 uses a state machine for its packet connections, determining which access mode the phone is in. Depending on which state the phone is in will determine what the latency will be.
It also depends on packet size. Most of these features come in the next rev of CDMA2000. But this is what happens for the most part:
- If you're using a circuit data mode then you're connected all the time, which costs the carriers and you a hefty chunk of money. This mode is used in all versions of CDMA before IS-2000 rev 0.
- If you're using Packet Data, you can be connected with a high speed channel downstream and a low speed channel upstream, or any combination. Depending on the size of the packet the latency can change, that's just standard networking.
- But if you're link has been idle for a while your phone may have dropped the channels and gone into another mode. When you send a packet, if its small enough it will go accross the common signalling channel which everyone uses to tell the base station its alive or it wants to make a call. And if the returning packet is small enough then it'll go back accross the the forward common signalling channel.
- But if you need to send or receive a large packet, or a stream of packets you need to set up the forward and reverse channels again, and this can take time. Depending on which state the phone is in when it needs to set up the channels this can take as little as half a second, or as much as 3 or 4 seconds before the packets start flowing again.
So depending on which mode you're in, depends on how large the latency is. It also depends on the radio environment. CDMA is fairly good in noisy environments, but if it gets into a really bad area packets will have to be retransmitted several times.
Most carriers who use CDMA now won't go to W-CDMA, they'll go to CDMA-2000 which is a different standard from W-CDMA.
CDMA-2000 is fully backwards compatible with current CDMA systems. If you try to get a 2G CDMA cell phone certified to work with Lucient base stations they will test it for compatibility on their CDMA-2000 base station as well.
The upgrade from IS-95/TSB-74 to IS-2000 is mostly replacing the base stations with new ones, and upgrading the back end. They can also set the upgraded base stations to not operate in 3G mode until the entire 3G network is ready and tuned. And even when they make the cut over to 3G the 2G phones can still use their network. I don't believe you can do this wiith the GSM to W-CDMA transition.
The refit for GSM and TDMA systems is much more costly because of the incompatibilities with W-CDMA. GPRS and UMTS aren't quite so bad, but aren't as good as the upgrad of a CDMA system.
Sprint uses CDMA the last I heard. And no-one is deploying CDMA in a 3xRTT solution right now which was considered true 3G.
Also from my understanding Qualcomm, currently the major/only producer of 3G CDMA chips, is releasing their Rev A chip later this year. So I assume Sprint is using Rev 0 which isn't 3G.
IS-2000 (or CDMA 2000 depending on which marketers you talk to) was set up in phases. Initially Rev 0 and Rev A. Rev 0 was a stepping stone from IS-95B (which is only used in Korea), to 3G. Rev A was suppose to be the full 3G solution.
A few years ago 1xRTT (which Sprint uses) was called 2.5 G. But then came the nice licensing auction in Europe which nearly bankrupted the carriers there for W-CDMA licenses. W-CDMA and 3xRTT require nearly the same licensing because of how much bandwidth they use. 3xRTT uses slightly bit less, but still 3 times more the 1xRTT. 1xRTT can use the same licenses that the carriers in North America and Asia currently have for IS-95. The carriers in North America and Asia mostly scraped plans on going to 3xRTT after the European auction, but the didn't want their systems to be called 2.5G, thus 1xRTT is now called 3G.
So depending on who you talk to Sprint has (or has not) a true 3G system.
I don't think we'll see true 3G anywhere in the world for at least 5 years. W-CDMA which is the competitor to CDMA2000 is still having some technical issues to handle. GPRS is 2.5 G though alot of people call it 3G, its a stepping stone to true 3G.
FWIR 3G is suppose to have 2+ Mbps transfers for stationary wireless stations, and improvements in bandwiths for walking and driving speeds, though I don't remember the numbers. Improved battery time (WCDMA stations haven't fufilled this with stand by times of ~50 hours), and better capacity.
CDMA-2000 1xRTT fufills all these requirements appart from the bandwidth. Rev 0 of CDMA-2000 only allows a maximum of 153 kbps, and Rev A allows 307 kbps. New inititives for 1xRTT do allow for improved bandwidth. They tried 1xEVDO for the world cup, but from what I heard they couldn't pull it off properly and it still needs some work. 1xEVDO is suppose to offer 2.4 Mbps.
Ummm, not to defend the editors, but its a title.
From my english classes, every word in a title but a few are always capitalized. It was one of the few lessons that was beat into me in grade 2 when I didn't capitalize properly. From my memory only words like as, and, & the in a title don't get capitalized unless they are at the beginning of the title.
So while capitalizing micron in the title can be confusing, it is proper english, even if the rest of the story and this comment aren't, (or is that ain't?).
I don't have a nokia phone but if you look on the picture near the bottom you see the baseband chip. U0801 is the chip I mean.
It is the VP 22353 PR release of the CDMA+ 200 chip produced by Philips Semiconductors. Nokia doesn't currently use this chip for their baseband.
This is an actual phone, and while Hop-on only has 15-20 people in it. Holley Communications which wrote the CDMA protocol stack software for the chip and the DSP firmware in Vancouver definetly has more then 20 people. Holley also did the Radio, Industrial and MMI design in Dallas. Over the life of the CDMA 200 project there have been around 200 people working on the project between Holley Communications and Phillips Semiconductors.
As for standby time on the phone, it is pretty substancial. I don't know the exact figures, but it is in the 5-6 day range. Which is around average for a CDMA phone.
Not that I can resist tooting our own horn, but the technology behind the phone is solid and well developed. Despite what the slashdot community thinks of Hop-on, the technology behind the phone is solid and developed by a very professional group of people, and some of the most knowledgable people in the CDMA technology world.
And yes the phone can be fabbed inexpensively. Which is one of the benifts for going with the Philips CDMA solution over the Qualcomm set of chips. I do know that the people doing the phone/radio design spend months going over the BOM/design to make it as inexpensive as possible.
And yes there isn't any hook in on that board for an LCD, and that helps with the price on the board, and also increases the standby time. Until you see it for yourself you'll never know how much a simple black and white LCD sucks the juice when you're trying to meet a standby spec.
I'd have to disaree. Going public without a product is stupid I agree. But going public when you're small and you need the capital to expand the company is vital.
For instance if you have a product you're ready to product but you can't get a factory to produce it you go public to pay for the factory.
My company right now is private, and we're looking at going public to pay mainly for development of the next product. Yes we already have one product that we are selling and expecting more. Actually we're looking at the next six months and most of us are scared because it looks like we'll have more major customers then we have resorses to support them, but we don't currently have the money to hire more engineers. And out next product requires one hell of an expensive license from the patent holders. Basically on the order of $50M. We're a small company, we don't have that money hanging around, and VCs aren't going to give us money just for the license so we have to go public.
But going public, just to go public and make millions for the owners, VCs and employees is stupid. You have to have a reason to go public, and a need for all that capital, and then you should really only sell the amount of stock you need to cover the capital and non-capital expenses (salaries and rent and all that), till you can get the contracts and money coming in to cover those expenses. Then if you need more money in the company for later you should have the money from the contracts, or you can offer more shares.
Anyway, I'll probably be modded as off topic. But there are instances where small companies should go public, but most dot-bombs don't fall into this, most were just in it for greed with no real plans or products.
In some cases yes, in some cases no.
In the case of waring ISPs mentioned in the article, you'll see 95% of the net, you won't see the 5% served by the other ISP.
If you're behind a hacked router then you might see all the net you might not. Most likely you won't see anything.
There's also the case of companies that had IP addresses assigned to them, that keep them unreachable from outside. Depending on their policies you may not see the net.
And if you're on an IP address assigned to an ISP or company that ceases to operate, you'll most likely not see the other 95% of the net.
It all depends on the net, the net isn't all that deterministic. Its kind of like a living organism with a personality you have to beat with a large stick. Usually you apply the stick to your neighbours if you're using a cable modem. Or to your ISPs first tier support (who often need a good thrashing, 'No its not my computer your router isn't working').
I haven't gone through the interview process, but I know several people who have. And there's no way that a non-programmer can 'pose as a programmer' and get through the interviews.
My friends that went through the process found it very intense. And while we all dislike the products that get sent out from Microsoft, I don't think anyone will disagree that they hire some of the best programmers out there. You might be able to 'pose as a programmer' like the article says, but there's no way you can pose as a programmer of the caliber that Microsoft hires without being one.
The article seems to read (and I might be reading things into the article) that the terrorists posed as programmers, were hired, made the backdoors and then left. Almost like they walked in with a breifcase and then walked out. This is just crazy, appart from what other people have said with code reviews and other processes to make sure this doesn't happen, the complexity of an OS prevents this.
I work on a relitively small project, the core protocol for a cellphone, and it takes on average 6-10 months of working with the code before you can understand the complexity of it. I hate to think of how long the average programmer at Microsoft goes through to learn the code enough to make backdoors that avoid detection past code reviews.
With an OS as complex as WinXP, I would assume that it would take a good year before anyone knows enough about just one small section to make a backdoor. And that would probably be easy to find. To hide the backdoor it would probably have to cross several sections of the code, you just can't do it with a handful of people posing as programmers. You'd probably have to do the backdoor as a small software project, just to handle the complexities of it.
IMHO, if there's a backdoor in there, then it was put in there with full knowledge of Microsoft. More likely if there were terrorists posing as programmers, they found a backdoor that Microsoft put in there for the government, and not made the backdoor themselves.
I was looking at something similar with using two linux boxes as NATing firewalls and acting as DNS servers. I wonder if this would work for multi-homing with two providers.
What I was thinking on doing was having the two linux boxes setup the same way, having their primary default routes to the connected providers, and a secondary route as the internal address of the other firewall.
So that if their primary default route went down, they'd automatically switch to the other firewall and send all outgoing requests back through the internal network to the other firewall.
These two firewalls would be on completely different subnets and different providers. I wanted to set up my domain name to have the primary DNS on one firewall and the secondary DNS on the second firewall. The IP addresses for DNS lookups on the first firewall would match the IPs provided by the first Provider, and the second DNS for the IPs on the second firewall.
This is where my knowledge goes out the window. I think you can set the expire time for cashing name servers as low as 10 minutes. So the lookups should switch across for name address translation should happen automatically, after ten minutes when the primary link dies.
This doesn't help much with load balancing unfortuantely, though I suppose you could throttle back the DNS (so it doesn't respond) if the primary link is too loaded, or even have it set up to use the IP addresses from the second provider when the link gets saturated.
I don't think you need any support from the ISPs for this, and it should allow you the redundant links.
I don't think the tightly integrated PCs will be the be all, end all like the article says. For consumers and most bussiness uses it might be. But because of the versitility and customization of the PC you'll always see them in their current form.
Some scientfic applications require High Data Rate digital capture cards. You'll never see a chip that allows you to hook the number of inputs into it that some scientists want.
When it comes to capturing this type of data, some scientists are anal, and always wanting more data points at higher resolution. You won't see this type of data being captured on anything put direct to the computers memory/hard drive. You won't see it through gigabit ethernet, firewire or USB. The big word in this sort of application is 'Realtime', I think its mostly a fad that isn't needed in most applications, but everyone wants it, and this requires as high bandwidth as possible.
Other type of equipment in some labs use a custom control interface, often made inhouse. This type of control might move towards USB and Ethernet, but I honestly doubt that.
GPIB isn't what I call a custom interface, but I doubt you'll ever see it based in an integrated box. JTAG interfaces for embedded development probably won't make it into an integrated box. But these interfaces do have ethernet adaptors now. But even with these the direct connect to the computer adapters are still more popular.
There will always be small, important applications that have markets too small to be included in the integrated boxes, but still need to have a computer interface.
You could say this is cirsumstancial evidence but:
A while ago I had to interview a recent immigrant from China for a position at my company. On his resume he had said at his previous job he 'improved' the USB driver code in the Linux Kernel so that it would work better with their systems.
Before the interview I did a grep through the kernel to see if I could find his name or any derivative of the name, I found none. So I asked him about it, and why he wasn't on the contribution list.
His answer (and I'm going from memory), was that since it was done for his employer he wasn't allowed to release it back under the GPL and it was propritary.
From my limited understanding, they do sell their systems (with linux and the mods), and they don't release any source code with them. The system is limited to back end phone networking use, and only in China so it really hasn't come to light.
Take that as you like, while it isn't exactly proof, and I see this as more of a bad coporation then as something 'China specific.' But doing kernel mods and not releasing them back under GPL does happen.
The one thing that I haven't seen mentioned for when assigning priority/severity is repeatability of the bug.
This is where a good test plan and procedures come into play. Where I work if our boards crash when executing a test procedure, but we can't get it repeat we don't file it as a bug, but as a simptom. We of course try several times to reproduce the results that cause the bug.
Alot of projects don't have a good test plan (most I've seen have absolutely no test plans) that defines how to execute the tests.
There are of course good and bad points to having a test plan. The good points is you can repeat a test exactly and reproduce a bug and give the bug produc... er... developers something hard they can use to trace down. Most S1 through S3 bugs should be found with a comprehensive test plan.
The bad point is you might not test the corner cases and the more obscure bugs will be missed. But if you can't reproduce them, then what's the point? You're just going to start locking horns with the developers. It often comes down to alot of ego on both parts. The developer thinks his code is perfect, the tester found a bug.
Reproducability needs to be part of the whole P*/S* assignment. A crash might be fatal, but if you can't reproduce it no matter what you do, are you going to put it in as an S1/P1, and waste alot of time on your part and developers part trying to track it down. That time can be more effectively spent fixing the other bugs, and you also have a good chance that one of these other bugs is causing the crash. Without a repeatable test plan you just don't know.
Filing as a symptom allows you to have the papertrail on the bug so you won't forget about it, but it doesn't waste everyone's time trying to solve something that might take months to solve, especially when you have no frigging clue what caused the bug.
On another note some bugs can be classified as high priority to fix when they aren't all that sevear. Spelling mistakes are often classified as high priority because they're dead easy to fix. And fixing them gives the produce the 'feeling of quality'. Which is total BS but marketing can sell it.
The kind of person you're looking for will be very hard to find. A person who is adaptable and also knows software engineering. I like to believe that I'm a person like that. Actually I'm pretty sure I am the kind of person you are looking for, so I know the job hunt habbits of the type of person you're looking for.
1. You have to be lucky, and find a person who isn't happy with their jobs. The good thing is, is that most large companies still don't understand that Computer Scientists & Engineers, and 'real' software engineers have different needs then your typical factory worker. You might be able to find a high powered, adaptable scientist from one of these larger companies.
2. Most of the people you're looking for don't post their resume to the internet, or to headhunters, so going through those channels you might not find them. They also don't look in the newspaper want ads. They talk to their friends, and their friends ususually know of a job from a friend of a friend of a friend. The last time I got fed up with my company I wrote an email to a friend, got a responce back and basically got a 40% raise a month later at a different job. I didn't post to the internet, and didn't talk to a headhunter.
3. You have to snag their interest. The only way you're going to get the type of people you seem to want is to snag their interest. A lot of the adaptable scientists are like the hippies of the 60s, they want to change the world. If you offer them a chance they might hop onto it.
4. With the economy of today everyone is skitish about switching jobs. You'll have to offer rock stable stability (not California or Washington rock, they have a habbit of shaking and knocking down buildings). While it is true that the 'real' engineers (not web developers) are still high in demands, they're still nervous about switching jobs because of the economy, if you offer alot but not stability you won't get them.
Still the most important thing is who you know that knows the type of person you're looking for. And getting a reference from someone you know also gives you a better chance at finding a jewel in the rough.
You forget one important thing. The NDP government is going into an election soon. They're doing everything, and I mean everything to buy the voters.
They offered to pay for that kid that needs the liver transplant that can't find a donner in Canada.
They decided to fast track the earthquake improvement in the public schools, the day after they said they weren't, which I believe is the day after the earthquake.
Yes they'll propose legislation to make this a law, but not for any concern for the children, but only to buy more votes, but appearing to be concerned for the children.
Just one last thing: before this latest election here in BC, I knew politicians lied through their teeth, but I had always thought they were decent enough to make then credible or plausible lies. The current government isn't even doing that, their lies are outright lies and almost everyone knows it. So don't expect this legislation to go anywhere before the required election on June, its one of the smaller issues/lies and the only way it'll hit the floor is as something the NDP can use to accuse the liberals of not caring for the children.
Yes I'm cynical, you would be too if you were a high tech worker, living in BC and watching the economy fall behind everyone for the past 10 years.
I remember that story, a bit better. I don't remember the year, but for those interested it goes something like this:
Buchanin (the new arts building at the time) received a bunch of new sculptures that Plant was suppose to put up. Plant are the guys who run the steam plants (yes we have steam tunnels at UBC) and do all the maintenance. Of course plant was taking their sweet time. So Engineering decided to play a prank.
They came up with the most hiddious statues you can think of and then put them inplace. They then started a campaign in the university newpaper, letters to the editor, articles and similar things, about how bad the art students' taste was and how they could make statues just as bad. The arts students fought back saying that the statues were master pieces.
It came to head one day when the engineers went out and started smashing the statues. It all ended when the engineers admitted that they made the statues, and the real ones were still waiting in a basement somewhere for Plant to install them.
The Engineers are UBC have done a similar thing with 'indecent' books at the main library. Planting the books in the library, complaining about all the smut and then having a book burning later on.
A few things in the article leaves me with a few concerns over how expensive it'll be.
The first thing that concerns me is the ranges, 150 mile radius. That's roughly, 70 thousand square miles, how many planes can you get flying in that area? I'm not even going to consider how many you can get if you take into three dimensional space.
With this amount of space to monitor, and all those aircraft, you're going to need two things. The first you need is a fast computer to keep track of all these aircraft. Second you're going to need a damn good radio, and modulation techniques to have broadcasts between all those aircraft without having all the data garbled with collisions.
For the system to be effect you'd have to install it on all aircraft, how long is that going to take? Because if one aircraft, or even a handful of aircraft are flying through the area and only they and the air traffic controllers know about them, the main selling point behind this system isn't really worth while. You still need to rely on ground controllers to tell you who's out there.
For example, if a pilot relies in this system and sees a plane on a collision course, or near collision course, he might turn, into the path of another aircraft that isn't showing up on the system.
For the system to handle 150 mile radius, you need a powerful radio, or large antennas, (both of which you probably won't find in private planes). You'll also need to install it in military aircraft, remember the F-117 that buzzed the airliner a few weeks ago? This of course costs more money and more tax dollars.
I don't see how this system will be as effective as they're promising without it being installed in all aircraft that will be flying through an area. It will help, I'm not denying that, but it will not remove the need of ground controlers and their radars. It won't even lessen their need at all.
I know the point of the article wasn't to mention if the computer would be useful, but to see how far we could go, stating that that would be the limit.
The computer would be interesting for super computers, but never for anything else. I don't see anyone walking around with a nuclear accident waiting to happen. But a super computer would be something else. The specified input and output would work in that enviroment. After all most people who work on super computers don't sit at the console. And if I remember correctly the console is usually burried so deep in the cases that it never sees the light of day unless someone digs into it to help it boot up. But then I've only heard of them, never even seen or worked with a super computer.
Having a laptop like this might be fun though, as a practical joke.
'Hey Joe you're computer's not working right'
'it isn't?'
'it hasn't vaporized the desk yet. You need a hotter computer.'
I've lost alot of faith in BBCs ability to report news. In the past few weeks we've had two reports linked from Slashdot from the BBC that were just pure crap:
Microsoft moving to BC -- Apparently this story came from someone in the BC gov't (not the minister) talking to someone not in Redmond, and not associated with Microsoft about bussiness in BC. Yet the BBC reported it as Microsoft moving to BC.
Hackers hack the space shuttle -- another BBC story. I don't remember anyone claiming that they did this, and the space shuttle is on a closed network, and NASA uses a pretty strong software development process that almost ensures a system that has a low probability of hacking. I believe a story on their development process was on slashdot recently.
This is probably the BBC trying to get more hits on their website so they can get more ad money. And the BBC does like to play hoaxes on the public. I remember hearing of a story in the 70s when they reported (as an April Fool's joke) that the new library in some city in England was built upside down. They pull other hoaxes as well, that's just the one I remember right now.
And besides this falls into one of those wacky inventions. It probably be as popular as those wacky inventions you see in all those black and white news movies. You know the self buttkicking machines, the face guard that rolls down like a shade to protect against grapefruit spray, and all the rest. And where did all these silly inventions go?
The third company should have more then just internet explorer. How is that one going to survive with just IE? After all shipping IE for free is what started this, if they up the price and netscape is still free, who's going to get all the market share? With just IE in this company then this one will fail faster then a Art's major taking a quantum physic exam.
/NEED/ a product they can sell, and they can't sell IE.
The third company should have IE, and all the Internet products. Like IS, Exchange, and anything else they have dealing with the internet. The other two companies should probably stay the same, one OS, and the other Applications and the rest. For the third company to survive they
I can actually see a forth company coming out of this, one dealing with hardware. They keyboards and mice are actually quite good. That way you'll have four baby bills: OS, Internet software, General Software, and Hardware. Each of them have products that they can use to make money on.
Just my 2 cents.