Top 3 is our 3 favorite local shops, Joey's, Caz's and Stoyle's. Not in our preferred order of course (Stoyle's is best), but Google can't be expected to know that (yet).
Now I just need a "Download location to in-car navigation system" button.
Back in university we used to play Wolf3D (and later Doom) in small groups, one guy playing and the rest watching. We had one guy who used to give out advice like that; we took to calling him Coach. His most sage piece of advice: "Don't get hurt!"
Interesting to compare these guys to Truckstop.net. Currently have rolled out 420 or so hotspots in 5 months, with plans to get around 3000 locations. Of course, their typical customers are truck drivers (hence the name), but anyone with with wireless and proximity to a location can sign up. Also, they charge the user directly, instead of charging the location. Still, once you sign up you can use any Truckstop hotspot at all, which can be pretty useful if you are travelling.
You can get to a nifty map on their page that shows all current and planned locations.
Re:Or, if this doesn't interest you
on
Paranoia
·
· Score: 1
Paranoia was awesome. Some of the weapons were hysterical. There was one in particular I remember, the Nuclear Hand Grenade (I can't recall if this was official or one of my friends made it up). Had specs something like this: Max throwing distance: 10m Blast radius: 50m
I can't comment on XMLTV's quality vs Tivo, but I can tell you that it is better (more accurate, better data) than the Guide+ feature built into my RCA TV (automatically aquires guide data from the cable signal somehow). I used Guide+ for quite some time after I bought my TV, and it was really great, particularly because I wasn't even aware of the feature when I bought the TV, I only discovered it after I got home. But since then I have built myself a MythTV box, so now I'm using XMLTV's guide data (which gets its data from clickTV since I'm in Canada). Why one is better I cannot say, since you'd figure all guide data originally stems from a single source, but I guess this is not the case.
I had to deal with this for the web GUI for one of our products. The solution has gone through a few revisions. The graphs consisted of two pie charts and a bar chart.
Version 1: images generated on the server (perl script using GD). This worked well, but I came to dislike how it had to download a new image every time you refreshed the page. Some of our customers were on pretty slow links. Also, we wanted the next version to work on Windows (v1 was on linux), and the perl/GD thing was going to be a pain. Our product is a web cache, so obviously I'm a bandwidth conservationist, and the bandwidth needed to download the images over and over offended me.:)
Version 2: images generated on the client using a java applet (purchased...it was only $50 or so). Nice, because the images are generated client-side, but the applet took awhile to download on slow links, and the startup time is BRUTAL. I came to hate this solution.
Version 2.1: pie charts done with Flash, bar charts done with DHTML/javascript. This has been working great! Flash starts up instantly, and almost everyone already has it. The numbers for the graphs are just included in the HTML code, and Flash/javascript uses those numbers to generate graphs. There are 2 numbers for each pie chart, and 60 numbers for the bar graph. Being text, they take up very little bandwidth.
But Joe Programmer works for The Company. Therefore, acting as a representative of The Company, he can quite legally release the stuff under GPL.
Now, whether or not he was supposed to do that or not is an internal matter at The Company. If he was not supposed to, Joe will probably find himself looking for a new job really fast. But I believe that doesn't give the company the ability to revoke the GPL release. Its released, its under GPL, sorry about your luck. Next time hire employees that are more trustworthy.
I learned this the other day from an article at Tom's Hardware. In retrospect it makes logical sense but I don't think it would have occured to me. We're sorta trained to think faster == more performance.
Anyways, what the article discovers is that you'll get BEST performance when memory speed == FSB speed. In benchmarks they find that a Athlon 3000+ (333Mhz FSB) with DDR333 is faster than the 3000+ with DDR400 (or DDR444). So, mental note, when shopping for a system, don't bother paying extra for that faster RAM, just get whatever matches your FSB.
True, in some cases. Not, however, in my case. I would LOVE to buy a Tivo. I'd do it today. Tivo won't sell to me because I live in Canada. However, I do have an RCA TV, with Guide+ (which is great!). So this RCA PVR with Guide+ would be just the ticket for me. Sure, it doesn't have the Tivo suggestions, which I think would be cool, but its better than having nothing.
Thats slightly misleading. A well provisioned web cache will save approximately 25% of your bandwidth. You'll see a document hit rate of 40-50%, and a byte savings of about 20-30% (average 45% and 25% respectively). I'm a cache vendor, I see these stats all the time on customers servers. Its really interesting actually, the size doesn't seem to matter (T1, T3 or more) the numbers are almost always identical. Some people don't seem to believe it, but I'll say it straight out: A web cache WILL reduce bandwidth usage on a pipe by 25% (well, 25% of web traffic, that is. For most ISPs and companies, thats the vast majority of traffic on the pipe).
I just went through this. My WinME system was starting to get goofy (don't laugh, I had really good stability with WinME, but lots of people seem to have trouble). Anyways, I decided to give XP a try. I've still got 40 days to activate, haven't decided wether to stay or not. My main concern was boot time. XP is *slightly* slower than WinME (I can live with it). Win2k was bad.
Anyways. For those that reinstall a lot: Do you tend to erase C:\windows and reinstall or just install on top of the existing install? It'd be nice to just re-install on top so you don't have to reinstall all your apps again, but I wonder if you still get that Just Installed(tm) speed feeling you get from a real virgin install. Experiences? I've never had a chance to experiment with it myself.
Not directly related, as there were no fires involved (that we know of), but a good story none the less.
Around 1990 or so (think "386's are brand new) I was working in a computer store. At this time motherboards didn't have I/O on them, it was all on a multi I/O ISA card, including the real-time clock w/battery. Well, the maker of the cards we used switched from a rechargable battery on this card to a long life lithium battery. However, they neglected to remove the recharging circuit from the design. Apparently it is not good to try to recharge lithium batteries, as we started getting frantic phone calls from our customers.
Me: "Hello, can I help you?" Customer: My new computer just exploded!! Me: You mean it crashed? Customer: No, it EXPLODED! There was a loud BANG and sparks came out the front!!
Sure enough, they brought the systems in and we inspected them. The battery had disintegrated. In one case the floppy cable had been pretty much lying right on top of it. The cable was all black and scorched and had little bits of metal shrapnel embedded in it.
Fortunetly it didn't do much real damage. We just replaced the card, cleaned all the little metal bits out and sent them on their way.
Needless to say we did a quick recall on the others! (Only about 5 systems went out like that.)
Have you ever told a newbie to "not be afraid" and "you can't damage the computer no matter what you do"? Man, I felt sorry for these people!
Re:Quake was supposed to do this...
on
P2P Roaming Chat
·
· Score: 1
I actually implemented this in Quake2. By hacking around with the game source, I turned the exit of each level into a "portal" to another server. I actually had it working. I could connect to Server1, run through e1m1 (or whatever the offical map name was, I forget) and hit the exit. Server1 would tell my client to connect to Server2, where I would spawn in at the beginning of e1m2, complete with all my stats/weapons/etc from when I exited the previous level.
The warp was interesting. I repurposed the savegame/loadgame code to dump the players current status to disk, then I transmitted the file to other server via a TCP connection, and used the loadgame code on the other server to restore the player.
You could also introduce new portals into levels (that appeared as teleport pads) to create links to other levels/servers. All configured via a text file.
I was planning on joining all of the Quake2 single player maps together into a huge chain, and setting up a huge internet team game. ie Blue team starts at the beginning map, Red team starts at the end map, and then it becomes a territory fight for everything in between. When there were only players from one team on a map (ie they got there first, or killed off the other team players) then they would "own" that map and get points for it.
I was going to call it "Quake World War", and have the whole campaign spread out across dozens of servers on the net (1 map per server, for load reasons). There could have been potentially hundreds of players involved, spread across all the servers. It could have lead to some interesting strategies/tactics. Since some players would have better ping times on some servers than others, you'd want to deploy your troops to the right places to take advantage of that.
Obviously, I never finished it. Just another project that went on the pile.:)
In reference to patents, this is my idea: Term of protection = Term of creation
ie if it took you 10 years to "invent" the process, you get 10 years of protection.
Of course, the development time must be properly documented. Anyone who is seriously developing something that is truly worthy of a patent should have no problem with that. And the "development time" ends at the moment of patent application or time of first sale, whichever is first. And it must be active development. No fair just "sitting" on the invention for a few years to extend the protection period.
An interesting side effect of this: software patents are ok, because software development goes SO FAST that the protection period will be almost nothing. Case in point: the 1-click patent. Go ahead, patent it. Your protection period will be all of 5 minutes.:)
I ran into this same problem with our web caching software when I ported it to CE (iPaq). Fortunetly, a little web searching turned up a lifesaver:
http://www.stlport.org/
I classify them as "realistic setting" as opposed to "realism". And some realism factors come into play (fall too far, you die) but as you say, TRUE realism is sacrificed for fun, which is good.
As for DoD, I'd suggest you give it another try, but avoid that beach map. I feel it is HOPELESSLY unbalanced. Maybe I've just never played on a server that had a GOOD Allies team, but the Axis kick their ass every time. There are only two ways up from the beach. Put two MG guys at each point and they can hold back an infinite number of attackers. Maybe they designed the map this way (true to what actually happened at Omaha) but I don't find it fun. Even playing as Axis, where half your team can just screw around doing nothing and you'll still have no trouble holding back the Allies. Not fun.
However, some of the other maps ARE really fun. The inner-city maps (I think Cannes is my favorite) are really fantastic. Part of the fun might be that these maps are very complicated, and it can take a long time to learn all the little secrets. Unlike most CS maps where you can learn the patterns and strategic points pretty quickly. (Note: I was, and still am, a CS addict.) But they have put a HELL of a lot of work in DoD, and I'm impressed. The atmosphere of the game (background sounds, explosions, bullets whizzing throught the air, etc) is about the best I've ever seen.
I've been tossing this idea around for a while, and I think it works.
Solution: The length of patent protection is equal to the amount of time it took to develop.
So, if you spend 10 years perfecting some technology, you get a 10 year patent. If it took you 5 minutes (ie 1-click shopping), you get a 5 minute patent.
Of course, you have to PROVE how long it took you to develop it. Some type of verifiable documentation should do it. Anyone seriously interested in getting a patent for something should have no problem keeping accurate documentation. The patent period starts from the date of first sale. If someone else sells the same technology before you do, no patent (prior art). This will prevent people from artifically extending the development period to get a longer patent.
Note all the benefits. Software patents will pretty much disappear, since it doesn't take THAT long to create it. And it seems fair. If you spend a year developing a piece of software, odds are after about it year it will be pretty much obselete anyways.
Will they ever implement this policy? Of course not. It goes in the bucket with all the other reasonable patent-reform ideas. *sigh*
Actually, with proper deployment of web proxy caches, the current system would be just fine. And by "proper deployment" I mean a cache at the end of every pipe bigger than a T1. In an ideal setup, the origin web server would have to serve up each page *once* (assuming static content), across its pipe to its upstream, where it would be cached, and further distributed to the rest of the net. Ideally I'd like to see cache functionality built into every router (except maybe the core routers, thats an aweful lot of data to cache, the harddrive speed drags you down).
The holy grail, I think would be a system that still allowed interactive/dynamic content. Imagine a distributed/.:P
Done. It even uses existing technology almost entirely. Hopefully we'll be able to announce something about it in the near future.
I'm not sure why everyone (inc. government) assumes they are using crypto. Try this for foolproof, nigh-untraceable, safe communication.
1. Get a piece of paper and pen
2. Write "Abdul, please perform xxx terrorist act on xxx date. Signed, Osama." on the paper with the pen.
3. Fold, insert in envelope.
4. Put stamp on envelope, address to Abdul.
5. Drop in mailbox.
Now, explain to me how carnivore, echelon, or backdoors in crypto will ever catch this.
Odds of message arriving: nearly 100% (sure, mail sometimes get lost, but its REALLY rare.)
Odds of being intercepted: nearly 0%.
Use a P.o. box and you don't even have to put a name on it (concievably the post office *might* be able to somehow watch for mail addresses to suspected terrorists).
Or, possibly even better, send the resupply ships BEFORE the manned one, programmed to "stop" and await the manned flight at specified intervals. I, for one, would feel better about flying off to Mars if I knew there was food and water WAITING for me ahead, rather than hoping it will catch up.:)
The whole might work if you if you just redefine it a bit. Instead of "end-to-end" try "edge-to-edge". The article is right, you want to keep the core of the network simple. BUT, the simplicity doesn't have to extend to the END of the connection...just the edge. Caching is a great example. Caches in the core of the network just doesn't work, there is just too much damn load. Howerver, a cache at the EDGE (say, the ISP level) makes alot of sense. Same for VoIP. Have gateways at the edges, which the end's use to establish communication.
Slashdot Mirror
Works perfectly!
Top 3 is our 3 favorite local shops, Joey's, Caz's and Stoyle's. Not in our preferred order of course (Stoyle's is best), but Google can't be expected to know that (yet).
Now I just need a "Download location to in-car navigation system" button.
BSA.
NEVER worry about licensing issues again. Go all open source and when the BSA comes to your door you can tell them to go take a flying leap.
Back in university we used to play Wolf3D (and later Doom) in small groups, one guy playing and the rest watching. We had one guy who used to give out advice like that; we took to calling him Coach. His most sage piece of advice: "Don't get hurt!"
Interesting to compare these guys to Truckstop.net. Currently have rolled out 420 or so hotspots in 5 months, with plans to get around 3000 locations. Of course, their typical customers are truck drivers (hence the name), but anyone with with wireless and proximity to a location can sign up. Also, they charge the user directly, instead of charging the location. Still, once you sign up you can use any Truckstop hotspot at all, which can be pretty useful if you are travelling.
You can get to a nifty map on their page that shows all current and planned locations.
Paranoia was awesome. Some of the weapons were hysterical. There was one in particular I remember, the Nuclear Hand Grenade (I can't recall if this was official or one of my friends made it up). Had specs something like this:
Max throwing distance: 10m
Blast radius: 50m
I can't comment on XMLTV's quality vs Tivo, but I can tell you that it is better (more accurate, better data) than the Guide+ feature built into my RCA TV (automatically aquires guide data from the cable signal somehow). I used Guide+ for quite some time after I bought my TV, and it was really great, particularly because I wasn't even aware of the feature when I bought the TV, I only discovered it after I got home. But since then I have built myself a MythTV box, so now I'm using XMLTV's guide data (which gets its data from clickTV since I'm in Canada).
Why one is better I cannot say, since you'd figure all guide data originally stems from a single source, but I guess this is not the case.
There is a Yahoo group setup for the game. Come and help out!
http://groups.yahoo.com/group/metacortechs
I had to deal with this for the web GUI for one of our products. The solution has gone through a few revisions. The graphs consisted of two pie charts and a bar chart.
:)
Version 1: images generated on the server (perl script using GD). This worked well, but I came to dislike how it had to download a new image every time you refreshed the page. Some of our customers were on pretty slow links. Also, we wanted the next version to work on Windows (v1 was on linux), and the perl/GD thing was going to be a pain. Our product is a web cache, so obviously I'm a bandwidth conservationist, and the bandwidth needed to download the images over and over offended me.
Version 2: images generated on the client using a java applet (purchased...it was only $50 or so). Nice, because the images are generated client-side, but the applet took awhile to download on slow links, and the startup time is BRUTAL. I came to hate this solution.
Version 2.1: pie charts done with Flash, bar charts done with DHTML/javascript. This has been working great! Flash starts up instantly, and almost everyone already has it. The numbers for the graphs are just included in the HTML code, and Flash/javascript uses those numbers to generate graphs. There are 2 numbers for each pie chart, and 60 numbers for the bar graph. Being text, they take up very little bandwidth.
But Joe Programmer works for The Company. Therefore, acting as a representative of The Company, he can quite legally release the stuff under GPL.
Now, whether or not he was supposed to do that or not is an internal matter at The Company. If he was not supposed to, Joe will probably find himself looking for a new job really fast. But I believe that doesn't give the company the ability to revoke the GPL release. Its released, its under GPL, sorry about your luck. Next time hire employees that are more trustworthy.
I learned this the other day from an article at Tom's Hardware. In retrospect it makes logical sense but I don't think it would have occured to me. We're sorta trained to think faster == more performance.
Anyways, what the article discovers is that you'll get BEST performance when memory speed == FSB speed. In benchmarks they find that a Athlon 3000+ (333Mhz FSB) with DDR333 is faster than the 3000+ with DDR400 (or DDR444). So, mental note, when shopping for a system, don't bother paying extra for that faster RAM, just get whatever matches your FSB.
"No one ever got fired for blaming Microsoft."
True, in some cases. Not, however, in my case. I would LOVE to buy a Tivo. I'd do it today. Tivo won't sell to me because I live in Canada. However, I do have an RCA TV, with Guide+ (which is great!). So this RCA PVR with Guide+ would be just the ticket for me. Sure, it doesn't have the Tivo suggestions, which I think would be cool, but its better than having nothing.
Thats slightly misleading. A well provisioned web cache will save approximately 25% of your bandwidth. You'll see a document hit rate of 40-50%, and a byte savings of about 20-30% (average 45% and 25% respectively). I'm a cache vendor, I see these stats all the time on customers servers. Its really interesting actually, the size doesn't seem to matter (T1, T3 or more) the numbers are almost always identical.
Some people don't seem to believe it, but I'll say it straight out: A web cache WILL reduce bandwidth usage on a pipe by 25% (well, 25% of web traffic, that is. For most ISPs and companies, thats the vast majority of traffic on the pipe).
I just went through this. My WinME system was starting to get goofy (don't laugh, I had really good stability with WinME, but lots of people seem to have trouble). Anyways, I decided to give XP a try. I've still got 40 days to activate, haven't decided wether to stay or not. My main concern was boot time. XP is *slightly* slower than WinME (I can live with it). Win2k was bad.
Anyways. For those that reinstall a lot: Do you tend to erase C:\windows and reinstall or just install on top of the existing install? It'd be nice to just re-install on top so you don't have to reinstall all your apps again, but I wonder if you still get that Just Installed(tm) speed feeling you get from a real virgin install. Experiences? I've never had a chance to experiment with it myself.
Not directly related, as there were no fires involved (that we know of), but a good story none the less.
Around 1990 or so (think "386's are brand new) I was working in a computer store. At this time motherboards didn't have I/O on them, it was all on a multi I/O ISA card, including the real-time clock w/battery. Well, the maker of the cards we used switched from a rechargable battery on this card to a long life lithium battery. However, they neglected to remove the recharging circuit from the design. Apparently it is not good to try to recharge lithium batteries, as we started getting frantic phone calls from our customers.
Me: "Hello, can I help you?"
Customer: My new computer just exploded!!
Me: You mean it crashed?
Customer: No, it EXPLODED! There was a loud BANG and sparks came out the front!!
Sure enough, they brought the systems in and we inspected them. The battery had disintegrated. In one case the floppy cable had been pretty much lying right on top of it. The cable was all black and scorched and had little bits of metal shrapnel embedded in it.
Fortunetly it didn't do much real damage. We just replaced the card, cleaned all the little metal bits out and sent them on their way.
Needless to say we did a quick recall on the others! (Only about 5 systems went out like that.)
Have you ever told a newbie to "not be afraid" and "you can't damage the computer no matter what you do"? Man, I felt sorry for these people!
I actually implemented this in Quake2. By hacking around with the game source, I turned the exit of each level into a "portal" to another server. I actually had it working. I could connect to Server1, run through e1m1 (or whatever the offical map name was, I forget) and hit the exit. Server1 would tell my client to connect to Server2, where I would spawn in at the beginning of e1m2, complete with all my stats/weapons/etc from when I exited the previous level.
:)
The warp was interesting. I repurposed the savegame/loadgame code to dump the players current status to disk, then I transmitted the file to other server via a TCP connection, and used the loadgame code on the other server to restore the player.
You could also introduce new portals into levels (that appeared as teleport pads) to create links to other levels/servers. All configured via a text file.
I was planning on joining all of the Quake2 single player maps together into a huge chain, and setting up a huge internet team game. ie Blue team starts at the beginning map, Red team starts at the end map, and then it becomes a territory fight for everything in between. When there were only players from one team on a map (ie they got there first, or killed off the other team players) then they would "own" that map and get points for it.
I was going to call it "Quake World War", and have the whole campaign spread out across dozens of servers on the net (1 map per server, for load reasons). There could have been potentially hundreds of players involved, spread across all the servers. It could have lead to some interesting strategies/tactics. Since some players would have better ping times on some servers than others, you'd want to deploy your troops to the right places to take advantage of that.
Obviously, I never finished it. Just another project that went on the pile.
Great idea! Combine this with a geo-caching type idea (record the GPS coords of the antennae) to make it easy to find and aim your equipment.
In reference to patents, this is my idea:
:)
Term of protection = Term of creation
ie if it took you 10 years to "invent" the process, you get 10 years of protection.
Of course, the development time must be properly documented. Anyone who is seriously developing something that is truly worthy of a patent should have no problem with that. And the "development time" ends at the moment of patent application or time of first sale, whichever is first. And it must be active development. No fair just "sitting" on the invention for a few years to extend the protection period.
An interesting side effect of this: software patents are ok, because software development goes SO FAST that the protection period will be almost nothing. Case in point: the 1-click patent. Go ahead, patent it. Your protection period will be all of 5 minutes.
It compiles on many platforms.
I classify them as "realistic setting" as opposed to "realism". And some realism factors come into play (fall too far, you die) but as you say, TRUE realism is sacrificed for fun, which is good.
As for DoD, I'd suggest you give it another try, but avoid that beach map. I feel it is HOPELESSLY unbalanced. Maybe I've just never played on a server that had a GOOD Allies team, but the Axis kick their ass every time. There are only two ways up from the beach. Put two MG guys at each point and they can hold back an infinite number of attackers. Maybe they designed the map this way (true to what actually happened at Omaha) but I don't find it fun. Even playing as Axis, where half your team can just screw around doing nothing and you'll still have no trouble holding back the Allies. Not fun.
However, some of the other maps ARE really fun. The inner-city maps (I think Cannes is my favorite) are really fantastic. Part of the fun might be that these maps are very complicated, and it can take a long time to learn all the little secrets. Unlike most CS maps where you can learn the patterns and strategic points pretty quickly. (Note: I was, and still am, a CS addict.) But they have put a HELL of a lot of work in DoD, and I'm impressed. The atmosphere of the game (background sounds, explosions, bullets whizzing throught the air, etc) is about the best I've ever seen.
I've been tossing this idea around for a while, and I think it works.
Solution: The length of patent protection is equal to the amount of time it took to develop.
So, if you spend 10 years perfecting some technology, you get a 10 year patent. If it took you 5 minutes (ie 1-click shopping), you get a 5 minute patent.
Of course, you have to PROVE how long it took you to develop it. Some type of verifiable documentation should do it. Anyone seriously interested in getting a patent for something should have no problem keeping accurate documentation. The patent period starts from the date of first sale. If someone else sells the same technology before you do, no patent (prior art). This will prevent people from artifically extending the development period to get a longer patent.
Note all the benefits. Software patents will pretty much disappear, since it doesn't take THAT long to create it. And it seems fair. If you spend a year developing a piece of software, odds are after about it year it will be pretty much obselete anyways.
Will they ever implement this policy? Of course not. It goes in the bucket with all the other reasonable patent-reform ideas. *sigh*
The holy grail, I think would be a system that still allowed interactive/dynamic content. Imagine a distributed /. :P
Done. It even uses existing technology almost entirely. Hopefully we'll be able to announce something about it in the near future.
I'm not sure why everyone (inc. government) assumes they are using crypto. Try this for foolproof, nigh-untraceable, safe communication.
1. Get a piece of paper and pen
2. Write "Abdul, please perform xxx terrorist act on xxx date. Signed, Osama." on the paper with the pen.
3. Fold, insert in envelope.
4. Put stamp on envelope, address to Abdul.
5. Drop in mailbox.
Now, explain to me how carnivore, echelon, or backdoors in crypto will ever catch this.
Odds of message arriving: nearly 100% (sure, mail sometimes get lost, but its REALLY rare.)
Odds of being intercepted: nearly 0%.
Use a P.o. box and you don't even have to put a name on it (concievably the post office *might* be able to somehow watch for mail addresses to suspected terrorists).
Or, possibly even better, send the resupply ships BEFORE the manned one, programmed to "stop" and await the manned flight at specified intervals. I, for one, would feel better about flying off to Mars if I knew there was food and water WAITING for me ahead, rather than hoping it will catch up. :)
The whole might work if you if you just redefine it a bit. Instead of "end-to-end" try "edge-to-edge". The article is right, you want to keep the core of the network simple. BUT, the simplicity doesn't have to extend to the END of the connection...just the edge. Caching is a great example. Caches in the core of the network just doesn't work, there is just too much damn load. Howerver, a cache at the EDGE (say, the ISP level) makes alot of sense. Same for VoIP. Have gateways at the edges, which the end's use to establish communication.