No, I don't think that has anything to do with it. Did you ever have a chance to see the Hotmail and Yahoo Mail front pages (just after you log in) before GMail launched? There were easily half a dozen ads covering more than half the screen. GMail had, at launch, a single column of text-only ads that were not animated, not flashing, and actually had something to do with what was on the page.
Did Hotmail and Yahoo Mail tone down their ads because of GMail? I don't know. But that isn't my point; my point is that in a world of cramming ads down our throats, Google has proven they can make money on unobtrusive ads in even unlikely places.
Rewriting OpenOffice to use AJAX is a lot more involved than a simple rebranding. It involves taking a native GUI and representing it inside a web browser using javascript and DHTML. No small feat.
I personally have no problem with a computer program scanning my document to generate advertising subjects. They are already STORING those personal documents, if a computer program is to scan it and generate advertising subjects, I have no objections. What harm does this cause to me?
Most likely. Google makes their money by providing stuff for free and making money off advertising. They make a lot of money doing it.
Back in the days of Yahoo! Mail and Hotmail, this practice was extremely annoying; you got half the screen filled with colour animated generic ads. Google proved that if you used targetted ads you could replace half a screen worth of ads with just one single group of text advertisements. I suspect they'll do something similar for an office suite, perhaps with the ads targetted to the content of your document.
Google proved that clustering could be fault tolerant, while costing less than true fault-tolerant hardware.
Google built massive clusters of thousands of machines out of very cheap unreliable hardware. They have tons of hardware failure due to the extremely cheap components (and sheer number of machines), but everything is redundant (And fully fault tolerant).
Any visual difference you see in a game between an AMD64 processor and a 32-bit processor is a scam; people should boycott any game that includes content that loads exclusively on 64-bit processors.
The reality is that 64-bit processors simply run games slightly faster than the same processor in 32-bit mode. This minor speed boost DOES NOT JUSTIFY higher detail via more doodads, higher poly models, higher res textures, more shaders, or anything else.
This is just an extremely shady promotion, where most gamers don't get the quality they should because someone arbitrarily decided to lock out content if it detects their processor.
Don't get me wrong, I am all for porting games to 64-bit architechtures. What I am NOT for is fake improvements by disabling content on 32-bit procs.
You've used bad examples. It doesn't cost the contractor anything when you sell your house. It does't cost Ford anything when you sell your mustang. It doesn't cost WoC anything when you sell your old cards (Indeed whoever buys them probably will buy more). It doesn't cost the publisher anything when you sell your books used.
It does cost developers something when you sell your old videogames.
Here's an analogy: Imagine if Barnes & Noble, Chapters, or tried to sell every customer a used book before they sold them a new book. They would make more money, but the lost sales to publishers would sink many of them.
Do you have a right to sell your old books? Yes. Should megastores be pushing to sell used copies to people before retail because they make more money? No.
I don't know about you, but when buying a PC game I'd rather have my money go to the developer so they can make more games than to a greedy store that won't sell their used stock at a price that is expected for used goods.
Phone support. Email support. Website support. Bandwidth for master servers, patch servers, websites, forums, other servers, etc.
This adds up to a tangible amount per person. If 25% of users are playing resold copies, that IS a large extra expense that wouldn't normally be there.
Think about this; the stores pay 10% for a used game, and resell it for 90% of retail cost. They're pocketing 80% of the retail value of the game, way more than they do with a retail game. Because the stores make so much more money off these used copies, they promote selling them BEFORE retail copies. This COSTS SALES of new games. They could give game companies 10 or 20 percent and still be making WAY more than they do at retail.
Keep in mind nobody is talking about charging Billy when he sells HyperDeathMatch to Jimmy at school. This is about stores like EB reselling games for close to retail cost. Not every company is an Epic or a Valve and can afford to pay higher costs to support each game, and suffer lower sales because people are buying used rather than new. The stores are getting an unfair percentage of the profit considering THEY DID NOT MAKE THE GAME. The developers put their sweat into that game, shouldn't they get something for it?
The problem is that you can replace Microsoft with Mozilla, and your statement changes to:
Patent Infringement Charges against Mozilla=Superawesome!
You see, Mozilla's browsers infringe on this patent in exactly the same way as Internet Explorer. Microsoft just got sued first, and while they claim they won't sue Mozilla, it is just that, a claim.
Why would a company run a bleeding-edge distro and X11 on a production machine?
If you're running a mission critical box, you pick something similar to Redhat Enterprise Linux 4, with a support contract from RedHat. You then put it on known supported hardware (RedHat will tell you) to ensure compatibility. You don't run useless extra software such as X11 that will increase the chance of failure.
Obviously the reason this company is having trouble is because they ignored common sense (what I listed above). They probably picked random hardware, a random distro, loaded up a default install set with useless apps, and then expected everything to work in a production environment.
If you use non-faulty compatible hardware and put a stable distribution on it (Such as RHEL4), the system will NOT crash every few weeks.
Assuming you already have a cheap PC, throw a SATA replication card in (Cheapest way to get RAID-5) and you have yourself an array.
Using cheap drives (WD 7200RPM 250GB has the lowest cost-per-gig here) you can get 1TB of total storage for for $488 USD. Of course to get 1TB in RAID5 that means an extra drive, so the cost for 1.25TB of total storage is $610 USD.
Of course if you don't have a PC to use, a NAS server ($130 USD), two USB enclosures ($25 each) and of the 250GB drives and you have half a terrabytes in unreliable storage on the cheap.
Simply use GMail Delete Button, which adds a delete button to messages. Really, of all the things to treat as a showstopper, this is a fairly minor thing. Especially since it can be so easily solved.
Archiving reports vary by manufacture, but 70 years would be low for a quality CD-R with the norm being 100 years. Some manufacturers of Gold CD-R's claim 100-200 years!
Obviously 40 years won't be a problem if the proper CD-Rs are chosen. Obviously there are more permanent digital storage mediums if you need extremely long term storage, but I'm uncertain as to why anyone would want to ensure that a piece of data could exist for more than 200 years. Over that sort of a time frame, any important piece of data is unlikely to stay on one piece of media. Want to store family photographs? Put them on a CD-R. If your kids care enough to pass them on to their kids, they can copy the image files to whatever the media-du-jour is.
I was referring to North American voltage, IE that in Canada.
AVR works based on percentages. If your voltage drops more than x percent below optimal, it boosts by x percent. It does not boost back up to optimal, it boosts by x percent, no matter how low it goes.
On my APC unit, AVR is totally useless; the UPS unit is configured to switch to battery power before AVR could kick in. TO be honest I sort of feel like APC is misrepresenting the capabilities of AVR, and what AVR does. When you actually email APC to ask them exactly how AVR works, the response they give you shows how AVR really isn't nearly as useful or effective as they make it out to be.
Why do I sort of feel cheated? I purchased the unit because I have dirty power and I wanted to have it cleaned. Turns out even though APC seems to claim as much in their product documentation, line-interactive UPS do nothing of the sort. You need super expensive online models to do that, or some overpriced MonsterPower component.
Personally I'd be happy with 128-bit AES, as it is still way more secure than protocols such as the one that MSN Messenger uses.
I've personally been using SimpLite, a free tool that can seamlessly encrypt MSN messenger traffic (with versions for YIM, ICQ, and AIM) by acting as a local SOCKS proxy that understands the protocol. It uses 2048-bit RSA keys with AES 128-bit encryption.
Line-interactive UPS units like the 3000NET DO NOT filter power. They have AVR, which only filters huge changes in voltage. It will not kick in for even extremely dirty power. It is simply meant to boost up the voltage if your power drops down to, say, 90 volts, or goes up to 150 volts.
So keep in mind that really all the UPS is going to give you that a good power supply can't is battery backup and surge protection.
Look up SessionSaver for firefox; it is what I have been using. Does what you describe, loading up all pages, cookies/session states, and most text forms. I understand there were some restrictions on which text forms, but I haven't been following the changelog so that might have changed.
However I have experienced memory leaks, with lots of tabs with images open. Each time I switched tabs, Firefox sucked up another meg or so of memory. By the end it was sucking up something like 400MB of real memory and even more than that in virtual memory.
Such things are bugs, though, which get fixed. Looking at the long-term, such a problem isn't a reason to prefer one browser over another. If I wanted I could give Deerpark a whirl, assuming such problems are fixed there.
Anyone who was on the verge of switching before now have virtually no reason not to.
And anyone who wasn't on the verge of switching has virtually no reason to do so. I mean, this is all well and good, but Firefox is working rather nicely, why should I switch to Opera? How is Opera going to make my browsing experience better in a way that cannot be replicated via Firefox extensions? And how will Opera provide to me the functionality that I have via Firefox extensions that isn't part of Opera?
You misunderstand latency correction. It doesn't affect player movement, it affects hit detection. In a nutshell the server keeps a history of player locations so that when your client tells it you fired, the server sees where your target was when you fired, as opposed to when the server received your packet. That's it, that's all, movement isn't affected.
The idea behind this is that you don't have to lead your shots no matter how much latency there is.
As for teleporting players, there are solutions to this. Half-Life (1 and 2) use interpolation. By default the buffer is 100ms. What it does is puts all data it receives into a 100ms buffer, and then interpolates that for display. While it does mean that what the player sees is delayed by 100ms, in practice this doesn't matter. Latency correction handles the delay for shots, and the interpolation itself ensures other players move smoothly.
In practice what it means is that you can lose or miss packets and the game can just interpolate around them.
Obviously if you're going to be mucking about with netcode to put Quake on a grid, adding such things as client-side interpolation and latency correction are certainly things that need to be considered.
IBM's Quake 2 port is a proof of concept, not a practical implementation.
This is what latency correction is designed to solve. Obviously such a system as this would almost require latency correction.
However with latency correction, the added 70ms are trivial, as in they wouldn't affect performance.
In my experience anything over 50ms is too much for me, unless there is latency correction. Then I'm comfortable up to about 200 or 250ms. Not that I couldn't play with more, it's just that when you're on a DSL connection and pinging more than 250ms to a server, chances are you've got high packetloss and other problems.
They could have saved some space by compressing assets, however uncompressed files lead to faster loading and lower CPU usage. On a PC game, this is an acceptable tradeoff.
Textures and sound account for about 2GB out of that 3GB, so if we assume that texture and audio compression would have netted a 50% size reduction, the game would indeed cut down to 1.3GB, which would fit on a 2GB SD card. However, it is important to keep in mind that the GP2X has a slow CPU (200MHz), and does not have any hardware 3D acceleration. All 3D rendering must be done in software, which is why gbax reccomends the Quake 1 engine for all 3D development; it is about as advanced as you can get with the available power.
I would be much more excited about the XGP, depending on what degree of homebrew support they have. The GXP has a higher resolution screen, and a 400mhz processor; it looks to be similar to the PSP in capabilities, though there isn't any info about it's 3D support.
Anyhow, with Quake 1 level graphics 2GB is way more than we'll ever need. Quake 1 took up what, 40MB?
Slashdot Mirror
No, I don't think that has anything to do with it. Did you ever have a chance to see the Hotmail and Yahoo Mail front pages (just after you log in) before GMail launched? There were easily half a dozen ads covering more than half the screen. GMail had, at launch, a single column of text-only ads that were not animated, not flashing, and actually had something to do with what was on the page.
Did Hotmail and Yahoo Mail tone down their ads because of GMail? I don't know. But that isn't my point; my point is that in a world of cramming ads down our throats, Google has proven they can make money on unobtrusive ads in even unlikely places.
Rewriting OpenOffice to use AJAX is a lot more involved than a simple rebranding. It involves taking a native GUI and representing it inside a web browser using javascript and DHTML. No small feat.
I personally have no problem with a computer program scanning my document to generate advertising subjects. They are already STORING those personal documents, if a computer program is to scan it and generate advertising subjects, I have no objections. What harm does this cause to me?
Most likely. Google makes their money by providing stuff for free and making money off advertising. They make a lot of money doing it.
Back in the days of Yahoo! Mail and Hotmail, this practice was extremely annoying; you got half the screen filled with colour animated generic ads. Google proved that if you used targetted ads you could replace half a screen worth of ads with just one single group of text advertisements. I suspect they'll do something similar for an office suite, perhaps with the ads targetted to the content of your document.
Google proved that clustering could be fault tolerant, while costing less than true fault-tolerant hardware.
Google built massive clusters of thousands of machines out of very cheap unreliable hardware. They have tons of hardware failure due to the extremely cheap components (and sheer number of machines), but everything is redundant (And fully fault tolerant).
They did this, again, using dirt cheap hardware.
Any visual difference you see in a game between an AMD64 processor and a 32-bit processor is a scam; people should boycott any game that includes content that loads exclusively on 64-bit processors.
The reality is that 64-bit processors simply run games slightly faster than the same processor in 32-bit mode. This minor speed boost DOES NOT JUSTIFY higher detail via more doodads, higher poly models, higher res textures, more shaders, or anything else.
This is just an extremely shady promotion, where most gamers don't get the quality they should because someone arbitrarily decided to lock out content if it detects their processor.
Don't get me wrong, I am all for porting games to 64-bit architechtures. What I am NOT for is fake improvements by disabling content on 32-bit procs.
You've used bad examples. It doesn't cost the contractor anything when you sell your house. It does't cost Ford anything when you sell your mustang. It doesn't cost WoC anything when you sell your old cards (Indeed whoever buys them probably will buy more). It doesn't cost the publisher anything when you sell your books used.
It does cost developers something when you sell your old videogames.
Here's an analogy: Imagine if Barnes & Noble, Chapters, or tried to sell every customer a used book before they sold them a new book. They would make more money, but the lost sales to publishers would sink many of them.
Do you have a right to sell your old books? Yes. Should megastores be pushing to sell used copies to people before retail because they make more money? No.
I don't know about you, but when buying a PC game I'd rather have my money go to the developer so they can make more games than to a greedy store that won't sell their used stock at a price that is expected for used goods.
Phone support. Email support. Website support. Bandwidth for master servers, patch servers, websites, forums, other servers, etc.
This adds up to a tangible amount per person. If 25% of users are playing resold copies, that IS a large extra expense that wouldn't normally be there.
Think about this; the stores pay 10% for a used game, and resell it for 90% of retail cost. They're pocketing 80% of the retail value of the game, way more than they do with a retail game. Because the stores make so much more money off these used copies, they promote selling them BEFORE retail copies. This COSTS SALES of new games. They could give game companies 10 or 20 percent and still be making WAY more than they do at retail.
Keep in mind nobody is talking about charging Billy when he sells HyperDeathMatch to Jimmy at school. This is about stores like EB reselling games for close to retail cost. Not every company is an Epic or a Valve and can afford to pay higher costs to support each game, and suffer lower sales because people are buying used rather than new. The stores are getting an unfair percentage of the profit considering THEY DID NOT MAKE THE GAME. The developers put their sweat into that game, shouldn't they get something for it?
The problem is that you can replace Microsoft with Mozilla, and your statement changes to:
Patent Infringement Charges against Mozilla=Superawesome!
You see, Mozilla's browsers infringe on this patent in exactly the same way as Internet Explorer. Microsoft just got sued first, and while they claim they won't sue Mozilla, it is just that, a claim.
Why would a company run a bleeding-edge distro and X11 on a production machine?
If you're running a mission critical box, you pick something similar to Redhat Enterprise Linux 4, with a support contract from RedHat. You then put it on known supported hardware (RedHat will tell you) to ensure compatibility. You don't run useless extra software such as X11 that will increase the chance of failure.
Obviously the reason this company is having trouble is because they ignored common sense (what I listed above). They probably picked random hardware, a random distro, loaded up a default install set with useless apps, and then expected everything to work in a production environment.
If you use non-faulty compatible hardware and put a stable distribution on it (Such as RHEL4), the system will NOT crash every few weeks.
Cheap PC and RAID-5.
Assuming you already have a cheap PC, throw a SATA replication card in (Cheapest way to get RAID-5) and you have yourself an array.
Using cheap drives (WD 7200RPM 250GB has the lowest cost-per-gig here) you can get 1TB of total storage for for $488 USD. Of course to get 1TB in RAID5 that means an extra drive, so the cost for 1.25TB of total storage is $610 USD.
Of course if you don't have a PC to use, a NAS server ($130 USD), two USB enclosures ($25 each) and of the 250GB drives and you have half a terrabytes in unreliable storage on the cheap.
This book contains all the info you need for this project.
Simply use GMail Delete Button, which adds a delete button to messages. Really, of all the things to treat as a showstopper, this is a fairly minor thing. Especially since it can be so easily solved.
Archiving reports vary by manufacture, but 70 years would be low for a quality CD-R with the norm being 100 years. Some manufacturers of Gold CD-R's claim 100-200 years!
Obviously 40 years won't be a problem if the proper CD-Rs are chosen. Obviously there are more permanent digital storage mediums if you need extremely long term storage, but I'm uncertain as to why anyone would want to ensure that a piece of data could exist for more than 200 years. Over that sort of a time frame, any important piece of data is unlikely to stay on one piece of media. Want to store family photographs? Put them on a CD-R. If your kids care enough to pass them on to their kids, they can copy the image files to whatever the media-du-jour is.
I was referring to North American voltage, IE that in Canada.
AVR works based on percentages. If your voltage drops more than x percent below optimal, it boosts by x percent. It does not boost back up to optimal, it boosts by x percent, no matter how low it goes.
On my APC unit, AVR is totally useless; the UPS unit is configured to switch to battery power before AVR could kick in. TO be honest I sort of feel like APC is misrepresenting the capabilities of AVR, and what AVR does. When you actually email APC to ask them exactly how AVR works, the response they give you shows how AVR really isn't nearly as useful or effective as they make it out to be.
Why do I sort of feel cheated? I purchased the unit because I have dirty power and I wanted to have it cleaned. Turns out even though APC seems to claim as much in their product documentation, line-interactive UPS do nothing of the sort. You need super expensive online models to do that, or some overpriced MonsterPower component.
Personally I'd be happy with 128-bit AES, as it is still way more secure than protocols such as the one that MSN Messenger uses.
I've personally been using SimpLite, a free tool that can seamlessly encrypt MSN messenger traffic (with versions for YIM, ICQ, and AIM) by acting as a local SOCKS proxy that understands the protocol. It uses 2048-bit RSA keys with AES 128-bit encryption.
Line-interactive UPS units like the 3000NET DO NOT filter power. They have AVR, which only filters huge changes in voltage. It will not kick in for even extremely dirty power. It is simply meant to boost up the voltage if your power drops down to, say, 90 volts, or goes up to 150 volts.
So keep in mind that really all the UPS is going to give you that a good power supply can't is battery backup and surge protection.
Look up SessionSaver for firefox; it is what I have been using. Does what you describe, loading up all pages, cookies/session states, and most text forms. I understand there were some restrictions on which text forms, but I haven't been following the changelog so that might have changed.
I can access it fine from Canada. Not redirecting me to google.ca or anything either.
40MB system and 30MB virtual for me right now.
However I have experienced memory leaks, with lots of tabs with images open. Each time I switched tabs, Firefox sucked up another meg or so of memory. By the end it was sucking up something like 400MB of real memory and even more than that in virtual memory.
Such things are bugs, though, which get fixed. Looking at the long-term, such a problem isn't a reason to prefer one browser over another. If I wanted I could give Deerpark a whirl, assuming such problems are fixed there.
Anyone who was on the verge of switching before now have virtually no reason not to.
And anyone who wasn't on the verge of switching has virtually no reason to do so. I mean, this is all well and good, but Firefox is working rather nicely, why should I switch to Opera? How is Opera going to make my browsing experience better in a way that cannot be replicated via Firefox extensions? And how will Opera provide to me the functionality that I have via Firefox extensions that isn't part of Opera?
Great, shine blinding lights at everybody wearing glasses. That's a wonderful idea.
Hopefully this anti-camera stuff never takes off, or camera manufacturers find a way to circumvent it.
You misunderstand latency correction. It doesn't affect player movement, it affects hit detection. In a nutshell the server keeps a history of player locations so that when your client tells it you fired, the server sees where your target was when you fired, as opposed to when the server received your packet. That's it, that's all, movement isn't affected.
The idea behind this is that you don't have to lead your shots no matter how much latency there is.
As for teleporting players, there are solutions to this. Half-Life (1 and 2) use interpolation. By default the buffer is 100ms. What it does is puts all data it receives into a 100ms buffer, and then interpolates that for display. While it does mean that what the player sees is delayed by 100ms, in practice this doesn't matter. Latency correction handles the delay for shots, and the interpolation itself ensures other players move smoothly.
In practice what it means is that you can lose or miss packets and the game can just interpolate around them.
Obviously if you're going to be mucking about with netcode to put Quake on a grid, adding such things as client-side interpolation and latency correction are certainly things that need to be considered.
IBM's Quake 2 port is a proof of concept, not a practical implementation.
This is what latency correction is designed to solve. Obviously such a system as this would almost require latency correction.
However with latency correction, the added 70ms are trivial, as in they wouldn't affect performance.
In my experience anything over 50ms is too much for me, unless there is latency correction. Then I'm comfortable up to about 200 or 250ms. Not that I couldn't play with more, it's just that when you're on a DSL connection and pinging more than 250ms to a server, chances are you've got high packetloss and other problems.
Check again. The GCF files for HL2 are 3.3GB.
They could have saved some space by compressing assets, however uncompressed files lead to faster loading and lower CPU usage. On a PC game, this is an acceptable tradeoff.
Textures and sound account for about 2GB out of that 3GB, so if we assume that texture and audio compression would have netted a 50% size reduction, the game would indeed cut down to 1.3GB, which would fit on a 2GB SD card. However, it is important to keep in mind that the GP2X has a slow CPU (200MHz), and does not have any hardware 3D acceleration. All 3D rendering must be done in software, which is why gbax reccomends the Quake 1 engine for all 3D development; it is about as advanced as you can get with the available power.
I would be much more excited about the XGP, depending on what degree of homebrew support they have. The GXP has a higher resolution screen, and a 400mhz processor; it looks to be similar to the PSP in capabilities, though there isn't any info about it's 3D support.
Anyhow, with Quake 1 level graphics 2GB is way more than we'll ever need. Quake 1 took up what, 40MB?