The difference is in market share. Of the percentage of users running Linux on a desktop, I'm sure the percentage that run on non-x86 is inconsequential.
Come on now, be reasonable. Any company releasing desktop software for linux is already catering to what, 1% of the desktop market? Less? Asking them to support 1% of that 1% is insane. There's no point expending the resources for something a handful of people will use.
The company that created Exeem is located in the carribean. The MPAA/RIAA can't touch them.
The site/servers seem to be hosted in the netherlands. The RIAA/MPAA might be able to get something done there, but that won't stop Swarm Systems from moving their servers to another country, even the carribean.
From within the program: "Future updates will include bug fixes, more customizable program, skinning program, linux console version, and additional functions to ease the use of the program! "
So there you have it, your headless linux version is on the way. And of course there is always WINE.
NAO Design has built a functional Sandcrawler, ala Star Wars Episode IV. Complete with blaster-deflecting sidewalls, full interior carpeting, seating for five and a 400 watt stereo system.
NAO design has built a vehicle, a "Land Cruiser/Tank", not a sandcrawler. The vehicle they built has absolutely nothing to do with Star Wars so far as I can tell, nor does it have anything to do with sandcrawlers. It doesn't even look remotely like one.
Just because you have a brown metal vehicle doesn't make it a sandcrawler. If anything it looks more like Jabba's sandspeeders from Episode VI.
This year we're expecting the max size on 7200RPM notebook (2.5") drives to jump from 60GB all the way to 100GB, a huge jump.
And I'd also expect to see a jump in 5400RPM storage capacity from the current 100GB.
My ideal notebook drive for 2005 would be a 100GB 7200RPM drive with a 16MB cache, SATA(2?), and NCQ. But who knows when that will happen. The best drive available today is a 60GB 7200RPM drive with 8MB of cache, though as I mentioned earlier that will jump to 100GB this year.
I don't think that will happen to Steam. There would be a huge uproar from mod developers if Valve started charging money for mods and didn't compensate the mod teams.
On the other hand, if they DID compensate the mod team, I'm all for it. Actual budgets means nothing but better mods. One of the reasons NS has been such a great mod is that it followed the commercial development style, with professional paid voice actors, and the main coder working full time. The only difference is Natural-Selection was paid for out of the creator's own pocket, and donations from the community.
Personally, I think mods going retail are a good thing. I have no problem paying for a mod if it's a good enough game (I paid for NS by donating, I'd have bought it in retail too had that happened). So long as the mod is taken commercial by a relatively small company (I consider Valve to be very small, compared to say, EA) such as Valve, or even better, a company formed just for that product.
Mods going retail almost always means higher quality. The primary reason is that there are more people working full-time instead of part-time. This doesn't usually translate to a faster development cycle, merely more work put into each release, which leads to a higher quality product.
And if you look at NS for an example, it just wouldn't have been the same without the professional voice actors that cost the owner a lot of his savings.
That said I'm not against free mods. I just like when the good ones go commercial.
It's a real relic. It's an Apple OneScanner, supporting 300dpi and 4-bit grayscale. Enormous and weighs a ton, it uses a Centronix SCSI cable (adapter cable) to plug into a 33mhz macintosh. The old Motorola MC68LC040 ones.
Nearly every scanner program I've ever used first does some sort of low-res preview pass, usually at something that looks to be decidedly under 72dpi (20 r 30dpi perhaps? I don't know). You then select the the section of the page to scan and do the high-res scan.
It looks like existing scanners already do what you want.
Ultra Wide-Band could offer much higher bandwidth than traditional RF devices could ever offer, at greater ranges with less power. Giving traditional RF devices precedence over it permanently is a mistake.
Fortunately the nature of Ultra Wide-Band means it can co-exist with traditional RF devices, allowing for a graceful transition.
While the FCC has limited UWB devices' power output to a thousandth of what manufacturers were hoping for, Pulse-Link is doing some interesting things. Many of you have probably read Cringley's column about them doing wired UWB, circumventing the FCC's limits (which doesn't apply to wired, only wireless). UWB co-exists on wires with regular RF just as well as it does in wireless.
And even more promising, I was having a short email conversation with Cringley, who told me that Pulse-Link has managed to pull off Wi-Fi type range while staying in the FCC's power limits. Apparently they're already deploying units for the TSA, and Cringley expects consumer devices on the market in 2006. The first-gen Pulse-Link UWB devices are 1.3gigabit, and as I understand it, multiple UWB networks can co-exist without sharing bandwidth.
When the news post said this: " It is not known by Charlie "Flayra" Cleveland and the rest of the NS Team whether NS:S will be a direct port or whether NS2 will use the Source Engine, UT2004 Engine, etc."
To clarify that, NS:Source and NS2 are two different products. NS:Source would be a free Source mod, and NS2 will be a commercial product, who's engine hasn't yet been decided (though Source and Unreal Engine 3 are two engines that Flayra mentioned as possibilities).
According to Flayra (Charlie Cleveland), NS2 will be a ~$50 game (In other words mainstream), and will require a multimillion dollar budget to pull off, in addition to requiring several years of development.
Normally for a first time developer to get such a huge budget would be rare, but Charlie has proven with the original NS that he can put out a cutting-edge genre-defining game that pushes the rendering engine to the limit. That will count for something.
The performance hit might not be as big as you think.
The biggest problem with Natural-Selection is it's heavy use of entities. The Half-Life engine wasn't designed to handle maps with hundreds of entities, which caused enormous load on the server (And much more on the client than was normal).
Source, on the other hand, can efficiently handle a great number more entities than Half-Life could. There WILL be a performance hit going to source, but part of it will be offset by substantially better entity handling. Whereas NS is by far the most taxing Half-Life mod, it would very likely be on-par with existing Source mods/modes.
Also, keep in mind that if NS:Source were to enter development, the actual release would be something like 6 months or a year away. In that time average computer power will increase a fair bit, and while HL2's performance level today is acceptable (Enough people are playing it and it's mods to cast doubt on your "barely runs on most people's PC"), in 6 to 12 months one to two videocard refreshs (And possible one generation jump) will have occured, not to mention significant changes in the CPU business. In other words in 6 to 12 months the average PC will be a bit stronger.
They have already. Introduced better compression formats that is.
The next stage in the evolution of image compression is wavelet compression, such as the sucessor to JPEG, which is called JPEG2000. As you can tell by the name, it hasn't caught on yet. Which is unfortunately, because it's a vastly superior format (Though there are better wavelet still image codecs out there).
This new stuffit format will suffer the same problem as JPEG2000: adoption. Any new image format will require client support, and since no browsers support it, nobody uses it. And because nobody uses it, no browsers support it. And so on.
While JPEG-2000 only provides a ~20% increase in compression over JPEG, it provides a host of other architechtural benefits over JPEG, not to mention that there are many wavelet codecs out there that are better than JPEG-2000 because they're not bogged down by the standardization process.
Contact a neighbour or two (or three, or four) about setting up a WDS (mesh) wireless network. In this mode multiple routers communicate with eachother as part of the same network.
The security risks involved with sharing the same network with multiple other people (who would each have their own router, their own default gateway) can be easily solved with a firewall blocking access to your net connection (or other machines) by all but the authorized IPs.
Any linksys WRT54G wireless router should be able to do this with something like the Sveasoft firmware.
However it also seems to lack Exeem's kazaa-like simplicity.
Each has it's pros and cons. Exeem will very likely be a much easier and more streamlined solution, but BT/K, being opensourced, opens the possibility of support in other opensourced clients such as Azureus (Which seems to be the most popular client based on what I see in swarms).
If you compare Canadian and US broadband, the US gets half the speed for twice the price, and it's even worse when you compare the US to Korea, Japan, or even India.
One could argue that price is irrelevant, but the US is far behind on average connection speed, which does matter.
Indeed, the company (Petroglyph) is made up mainly of ex-Westwood employees, who were responsible for every Command & Conquer game up until (Not including) Generals.
But even bigger news is that the music will be provided by the award winning Frank Klepacki, who did the music for every single C&C game up until (not including) Generals.
I'm looking at this project as a new Command & Conquer game in a different universe. It will be interesting to see how much of C&C's gameplay mechanics they keep. Certainly EA threw out all of C&C's trademark gameplay mechanics when they made Generals.
Regardless, Former-westwood and Frank Klepacki working on a game, you know it's going to be good fun.
I seem to recall there being computer-controlled players, at least in the vesion I played. However this was all ten years ago so my memory could be faulty.
They don't port UnrealEd to the XBox or PlayStation 2 either, but both of those are supported platforms of Unreal Engine 2. Yes, the obvious reason is that they're consoles, but my point is that despite the fact that Unreal Engine 2 is cross-platform and runs on a variety of systems, its main platform is Windows. Editing tools don't need to be cross platform as much as the game itself does.
Who said the screenshot was of a real mailing list? More likely it was from a test. Would a good way to test such a piece of software not be to set up an email server that would accept any mail to any address, and then generate 450 thosuand random addresses at that domain?
If it's being filtered client-side, it's not so bad.
Mail is delivered to your server, assuming 20 emails per minute (That's about two hundred thousand mails a week), consuming roughly half a kilobyte per second of downstream. You could run your mail server on DIALUP and that would STILL be a small amount. These are bounce mails, they're all text and probably only about 2KB (I'm guessing, but they can't be that big)
And say you cleared out your local mail server to your client (downloaded your mail) once an hour, each hour you'd download over your LAN. That'd be about 2.4MB of email every hour, which would take less than one second to transfer over a LAN (Probably longer due to transfering each email seperately).
From there all the crap would be filtered, and as far as you're concerned, you are not getting any crap.
The volume is irrelevant. No matter where or how you deal with this, be it client side or server side, the server IS going to get a huge volume of mail. Even if you manage to blacklist it on the server it's still going to have to deal with each and every message.
Besides, a hundred thousand messages is still only about a hundred megabytes worth of messages, or two hundred megs, which is a drop in the bucket on a real server (Or even a budget one http://servermatrix.com). Heck, the "flood" wouldn't even stress the mail server I used to host on my DSL connection.
The post already says that you're using Thunderbird's built in bayesian filtering. So what's the problem here? Thunderbird should be (If you train it properly) filtering out all those nasty bounce emails into your spam folder.
Slashdot Mirror
The difference is in market share. Of the percentage of users running Linux on a desktop, I'm sure the percentage that run on non-x86 is inconsequential.
Come on now, be reasonable. Any company releasing desktop software for linux is already catering to what, 1% of the desktop market? Less? Asking them to support 1% of that 1% is insane. There's no point expending the resources for something a handful of people will use.
The company that created Exeem is located in the carribean. The MPAA/RIAA can't touch them.
The site/servers seem to be hosted in the netherlands. The RIAA/MPAA might be able to get something done there, but that won't stop Swarm Systems from moving their servers to another country, even the carribean.
Exeem has been mentioned on Slashdot before. It's the sucessor to the largest BitTorrent site on the internet. A followup is certainly in order.
From within the program: "Future updates will include bug fixes, more customizable program, skinning program, linux console version, and additional functions to ease the use of the program! "
So there you have it, your headless linux version is on the way. And of course there is always WINE.
NAO Design has built a functional Sandcrawler, ala Star Wars Episode IV. Complete with blaster-deflecting sidewalls, full interior carpeting, seating for five and a 400 watt stereo system.
NAO design has built a vehicle, a "Land Cruiser/Tank", not a sandcrawler. The vehicle they built has absolutely nothing to do with Star Wars so far as I can tell, nor does it have anything to do with sandcrawlers. It doesn't even look remotely like one.
Just because you have a brown metal vehicle doesn't make it a sandcrawler. If anything it looks more like Jabba's sandspeeders from Episode VI.
This year we're expecting the max size on 7200RPM notebook (2.5") drives to jump from 60GB all the way to 100GB, a huge jump.
And I'd also expect to see a jump in 5400RPM storage capacity from the current 100GB.
My ideal notebook drive for 2005 would be a 100GB 7200RPM drive with a 16MB cache, SATA(2?), and NCQ. But who knows when that will happen. The best drive available today is a 60GB 7200RPM drive with 8MB of cache, though as I mentioned earlier that will jump to 100GB this year.
I don't think that will happen to Steam. There would be a huge uproar from mod developers if Valve started charging money for mods and didn't compensate the mod teams.
On the other hand, if they DID compensate the mod team, I'm all for it. Actual budgets means nothing but better mods. One of the reasons NS has been such a great mod is that it followed the commercial development style, with professional paid voice actors, and the main coder working full time. The only difference is Natural-Selection was paid for out of the creator's own pocket, and donations from the community.
Personally, I think mods going retail are a good thing. I have no problem paying for a mod if it's a good enough game (I paid for NS by donating, I'd have bought it in retail too had that happened). So long as the mod is taken commercial by a relatively small company (I consider Valve to be very small, compared to say, EA) such as Valve, or even better, a company formed just for that product.
Mods going retail almost always means higher quality. The primary reason is that there are more people working full-time instead of part-time. This doesn't usually translate to a faster development cycle, merely more work put into each release, which leads to a higher quality product.
And if you look at NS for an example, it just wouldn't have been the same without the professional voice actors that cost the owner a lot of his savings.
That said I'm not against free mods. I just like when the good ones go commercial.
I do believe my scanner falls into that category.
It's a real relic. It's an Apple OneScanner, supporting 300dpi and 4-bit grayscale. Enormous and weighs a ton, it uses a Centronix SCSI cable (adapter cable) to plug into a 33mhz macintosh. The old Motorola MC68LC040 ones.
Nearly every scanner program I've ever used first does some sort of low-res preview pass, usually at something that looks to be decidedly under 72dpi (20 r 30dpi perhaps? I don't know). You then select the the section of the page to scan and do the high-res scan.
It looks like existing scanners already do what you want.
Ultra Wide-Band could offer much higher bandwidth than traditional RF devices could ever offer, at greater ranges with less power. Giving traditional RF devices precedence over it permanently is a mistake.
Fortunately the nature of Ultra Wide-Band means it can co-exist with traditional RF devices, allowing for a graceful transition.
While the FCC has limited UWB devices' power output to a thousandth of what manufacturers were hoping for, Pulse-Link is doing some interesting things. Many of you have probably read Cringley's column about them doing wired UWB, circumventing the FCC's limits (which doesn't apply to wired, only wireless). UWB co-exists on wires with regular RF just as well as it does in wireless.
And even more promising, I was having a short email conversation with Cringley, who told me that Pulse-Link has managed to pull off Wi-Fi type range while staying in the FCC's power limits. Apparently they're already deploying units for the TSA, and Cringley expects consumer devices on the market in 2006. The first-gen Pulse-Link UWB devices are 1.3gigabit, and as I understand it, multiple UWB networks can co-exist without sharing bandwidth.
When the news post said this: " It is not known by Charlie "Flayra" Cleveland and the rest of the NS Team whether NS:S will be a direct port or whether NS2 will use the Source Engine, UT2004 Engine, etc."
To clarify that, NS:Source and NS2 are two different products. NS:Source would be a free Source mod, and NS2 will be a commercial product, who's engine hasn't yet been decided (though Source and Unreal Engine 3 are two engines that Flayra mentioned as possibilities).
According to Flayra (Charlie Cleveland), NS2 will be a ~$50 game (In other words mainstream), and will require a multimillion dollar budget to pull off, in addition to requiring several years of development.
Normally for a first time developer to get such a huge budget would be rare, but Charlie has proven with the original NS that he can put out a cutting-edge genre-defining game that pushes the rendering engine to the limit. That will count for something.
The performance hit might not be as big as you think.
The biggest problem with Natural-Selection is it's heavy use of entities. The Half-Life engine wasn't designed to handle maps with hundreds of entities, which caused enormous load on the server (And much more on the client than was normal).
Source, on the other hand, can efficiently handle a great number more entities than Half-Life could. There WILL be a performance hit going to source, but part of it will be offset by substantially better entity handling. Whereas NS is by far the most taxing Half-Life mod, it would very likely be on-par with existing Source mods/modes.
Also, keep in mind that if NS:Source were to enter development, the actual release would be something like 6 months or a year away. In that time average computer power will increase a fair bit, and while HL2's performance level today is acceptable (Enough people are playing it and it's mods to cast doubt on your "barely runs on most people's PC"), in 6 to 12 months one to two videocard refreshs (And possible one generation jump) will have occured, not to mention significant changes in the CPU business. In other words in 6 to 12 months the average PC will be a bit stronger.
They have already. Introduced better compression formats that is.
The next stage in the evolution of image compression is wavelet compression, such as the sucessor to JPEG, which is called JPEG2000. As you can tell by the name, it hasn't caught on yet. Which is unfortunately, because it's a vastly superior format (Though there are better wavelet still image codecs out there).
This new stuffit format will suffer the same problem as JPEG2000: adoption. Any new image format will require client support, and since no browsers support it, nobody uses it. And because nobody uses it, no browsers support it. And so on.
While JPEG-2000 only provides a ~20% increase in compression over JPEG, it provides a host of other architechtural benefits over JPEG, not to mention that there are many wavelet codecs out there that are better than JPEG-2000 because they're not bogged down by the standardization process.
Contact a neighbour or two (or three, or four) about setting up a WDS (mesh) wireless network. In this mode multiple routers communicate with eachother as part of the same network.
The security risks involved with sharing the same network with multiple other people (who would each have their own router, their own default gateway) can be easily solved with a firewall blocking access to your net connection (or other machines) by all but the authorized IPs.
Any linksys WRT54G wireless router should be able to do this with something like the Sveasoft firmware.
However it also seems to lack Exeem's kazaa-like simplicity.
Each has it's pros and cons. Exeem will very likely be a much easier and more streamlined solution, but BT/K, being opensourced, opens the possibility of support in other opensourced clients such as Azureus (Which seems to be the most popular client based on what I see in swarms).
If you compare Canadian and US broadband, the US gets half the speed for twice the price, and it's even worse when you compare the US to Korea, Japan, or even India.
One could argue that price is irrelevant, but the US is far behind on average connection speed, which does matter.
Indeed, the company (Petroglyph) is made up mainly of ex-Westwood employees, who were responsible for every Command & Conquer game up until (Not including) Generals.
But even bigger news is that the music will be provided by the award winning Frank Klepacki, who did the music for every single C&C game up until (not including) Generals.
I'm looking at this project as a new Command & Conquer game in a different universe. It will be interesting to see how much of C&C's gameplay mechanics they keep. Certainly EA threw out all of C&C's trademark gameplay mechanics when they made Generals.
Regardless, Former-westwood and Frank Klepacki working on a game, you know it's going to be good fun.
I seem to recall there being computer-controlled players, at least in the vesion I played. However this was all ten years ago so my memory could be faulty.
They don't port UnrealEd to the XBox or PlayStation 2 either, but both of those are supported platforms of Unreal Engine 2. Yes, the obvious reason is that they're consoles, but my point is that despite the fact that Unreal Engine 2 is cross-platform and runs on a variety of systems, its main platform is Windows. Editing tools don't need to be cross platform as much as the game itself does.
I used to play Bolo on my mac back in the mid 90s. Singleplayer though, I had no net connection.
Maybe it'll get hacked like that bike service in hongkong. Then people would be riding around in free cars.
Seriously though, this doesn't sound very secure/safe/good.
Who said the screenshot was of a real mailing list? More likely it was from a test. Would a good way to test such a piece of software not be to set up an email server that would accept any mail to any address, and then generate 450 thosuand random addresses at that domain?
On a home network it's no problem.
If it's being filtered client-side, it's not so bad.
Mail is delivered to your server, assuming 20 emails per minute (That's about two hundred thousand mails a week), consuming roughly half a kilobyte per second of downstream. You could run your mail server on DIALUP and that would STILL be a small amount. These are bounce mails, they're all text and probably only about 2KB (I'm guessing, but they can't be that big)
And say you cleared out your local mail server to your client (downloaded your mail) once an hour, each hour you'd download over your LAN. That'd be about 2.4MB of email every hour, which would take less than one second to transfer over a LAN (Probably longer due to transfering each email seperately).
From there all the crap would be filtered, and as far as you're concerned, you are not getting any crap.
The volume is irrelevant. No matter where or how you deal with this, be it client side or server side, the server IS going to get a huge volume of mail. Even if you manage to blacklist it on the server it's still going to have to deal with each and every message.
Besides, a hundred thousand messages is still only about a hundred megabytes worth of messages, or two hundred megs, which is a drop in the bucket on a real server (Or even a budget one http://servermatrix.com). Heck, the "flood" wouldn't even stress the mail server I used to host on my DSL connection.
The post already says that you're using Thunderbird's built in bayesian filtering. So what's the problem here? Thunderbird should be (If you train it properly) filtering out all those nasty bounce emails into your spam folder.
So, then, what's the problem?