1) PC's are cheaper 2) Hardware vendors target PC for performance upgrades 3) PC's are much cheaper to upgrade 4) Nearly all games target Windows
By having Apple port games, you face an uphill battle on point 4, and do nothing about 1-3. Sorry, but what's my incentive to switch?
Why on earth would someone switch to Mac for gaming, when it will take longer to get their games, they'll have to dump more into the system, then dump more into the upgrades (which will also come after they do on PC, and many won't come at all). All so they get to say "Hey look at me! I'm a sucker!"
I work in the gaming business (albiet in a fringe corner of it). We have access to over 2 million "hardcore" gamers - those that like to play competitively. I have a hard time believing you'll convert any with this idea - it's simply not practical.
If you want to convert people, get some great Mac-only games. There's a long way to go before anyone serious will do that. If Valve launched CS:S and HL2 on Mac exclusively, they've already got 1 foot in the grave.
PC makers are pushing hardware limits. How will Mac play games? Let's see a port of HL2 to Mac, and have the hardware guys go nuts. If a Mac can consistently beat a top-end PC in the typical game benchmarks, then it will get some buyers.
While I hope Apple becomes a gaming system some day... I'm quite skeptical that it will ever happen.
I don't use MySQL for valuable data. But I do consider myself a professional, and I do use MySQL for "convenience records" - data that is merely a cache, or set of aggregate data that can be regenerated easily, and has to be queried a LOT (thousands of times per second). MySQL fits the bill here. Postgres, the database we use for the rest of our needs, couldn't handle so many queries without having locking issues.
Sometimes the lack of caring about data integrity (and therefore using less locks on a high-UPDATE table) can be good. To each their own.
Re:PayPal and eBay constantly push the envelope
on
Paypal Grinds To A Halt
·
· Score: 0, Flamebait
While I'm at it, I'll compare this apple, and this orange...
You obviously don't understand the difference between PayPal and a bank, yet still feel compelled to preach about it - which is typical of the normal Slashdot visitor I suppose. I guess I shouldn't have commented on something that I actually have experience with.
When you have dealt with such a system, I'll give two and a half shits about your suggestion to "re-think my post".
Re:PayPal and eBay constantly push the envelope
on
Paypal Grinds To A Halt
·
· Score: 2, Insightful
The "user systems" and associated features of these bank sites have nowheres near the relational properties of something like PayPal - it's not even close. They can segregate data based on customer sets, where you know where users are coming from, and you can tie them to their own set of data - data that can be archived and summarized into non-realtime aggregates.
PayPal has the task of linking transactions between users, and allowing users to create modifications or additions to them in real time.
These are not the same, by any stretch. It's the final bit of difference that makes it so much more complex.
There are two primary issues that banks deal with:
1) a large amount of data 2) a high volume of transactions
PayPal and eBay have the pleasure of dealing with a third:
3) highly versatile users with demanding needs, to both points #1 and #2.
And being a system that requires transactional integrity, you can add a fourth:
4) ACID compliant, highly relational database with proper locking, even on SELECT. NOT MYSQL.
#1 or #2 = very easily managed
#1 + #2 = get a good DBA
#1 + #2 + #3 = good DBA, good infrastructure
#1 + #2 + #3 + #4 = fackin gridlock
This is something that _WAS_ very rare until recently, and is only now becoming more prevalent, with the adoption of more data-oriented websites, more overall users, and them being more demanding.
I have the pleasure of working in a similar (although significantly smaller, though still "big") environment that fits these conditions. It's incredibly difficult to make a very large transaction based, highly-relational data set accessible to many (>500K) users at a time. Anyone who says otherwise hasn't had to do it.
The slashdot effect? Not a drop in the bucket!
on
Paypal Grinds To A Halt
·
· Score: 2, Interesting
For those who say "if its already slow, dont slashdot it!", seems you over estimate the slashdot effect. Slashdot links may take a normal web host down, but anyone with an infrastructure built to serve a lot of traffic shouldn't even notice it.
A site I run has been linked by Slashdot many times - aside from the number of referrals in web logs and perhaps a few mbit spike in traffic, you wouldn't know it even occured.
PayPal is having much larger issues. I've experienced the same types of problems when doing larger upgrades - it's usually a single piece of code that went unchecked, and the odd user(s) that actually get caught in an untested (or improperly tested) segment of code are throwing the system for a loop.
Or, just bad performance planning. I'm not surprised - these things have to happen every now and then, they're quite unavoidable when you're dealing with often changes to an infrastructure with that many _DYNAMIC_ users, and so much _DYNAMIC_ data.
Regardless, it's a growing pain. I've been through a few of these on a 150 GB database with 2 MM users and 15 servers. PayPal is a lot larger, and they surely have much more difficult problems. Hell, if they have to do something such as change the way a single piece of data is stored, that means a lot of downtime (or slowtime) processing!
I'm sure they'll sort it out, unfortunately lots of people are losing lots of business in the process.
Of course they will support songs from other stores. That way, you get used to the new features in WMP10, play all your songs there, and you're in musical nirvana.
How many employees does SCO have working on this? I generated six figures in the last few by months selling mouse pads to users coming from ads on a single website... If investors consider the ability to generate revenue a test of your worth, SCO should be considered quite worthless.
Also, I ONLY use it plugged into the wall (by requirement, not choice, heh). Additionally, my next machine (very soon now) will have at least 5 hours during desktop usage - travel has forced me into it.
My Sony VAIO PCG-FXA36. AMD Athlon 1 GHz, 256MB, 20GB disk, DVD drive, 15" screen that doesn't adjust itself down when you go to battery power.
You can let this notebook sit idle w/ the screen on (and and tell XP not to turn off the display or spin down the disk) and it will be off in about 30 minutes.
It's a great machine otherwise, but the battery is just horrible...
Over the course of the past 5 years, for mostly business reasons (but often personal as well), I've used my share of hard drives - and had to deal with the support for them as well. I'm talking about around 300-500 drives, mostly Seagate, IBM, and a few Maxtors (until I learned to avoid them). All but a handful of them ATA-66/100, 7200 RPM.
I can say without a doubt that I've seen the lowest failure rate from Seagate drives. Out of all the drives we shipped, over half were Seagate, and we've had around 5 failures. The best news? They were pretty good failures, i.e. the drive may not spin up every time - but when it does the data is fine (good for recovery).
Maxtor gets the highest failure rate, period. Of the 25 or so drives I've used, nearly all of them have failed - many within 6 months. Additionally, once they fail, it's nearly impossible to get them spinning/reading again. The only thing good that I can say about Maxtor drives is that their 5400RPM's are more reliable than the 7200RPM's.
There was a whole series of IBM drives that we shipped, which had quite consistent long-term (3+ year) failures. I believe it was the 75GXP line. IBM never admitted any problem, though it was widely reported (even here on Slashdot) that it was a very common issue. Otherwise, I've been quite happy with the performance and reliability of other lines of IBM drives.
I've heard good and bad about Western Digital, but haven't used but a handful of WD drives (we used all these drives in hot-swap RAID, and WD likes to make their back-plugs incompatible with the "standard", so WD is a non-option).
At the end of the day, Seagate is the IDE drive I would recommend, and I think that with this much experience dealing with the day-to-day support of these types of disks, that recommendation is quite justified.
I'm glad they've increased the warranty length. Perhaps it will make Maxtor do the same, and in the process make them turn out higher quality disks.
So now, in addition to dealing with the kids using all the bandwidth to download mp3's, you now also have to deal with the nerds who are using all of the CPU time calculating prime numbers.
The only way this will work is if you can "kill -9" other user's processes. Make sure it comes with boxing gloves.
When you need to verify the absolute identity of someone you're dealing with, these companies, with their background checks (which aren't strong by any stretch of the imagination), certs by Verisign/etc may make sense.
However, the most common usage of SSL cert's is simply to enable encryption between two points. For this, there's nothing wrong with even a home-brew cert - validation of the cert via it matching the domain should be sufficient. A SSL cert generated by a 3rd party adds absolutely nothing to security, and it shouldn't do anything to reassure the customer/client that they're dealing with a legitimate operation.
Encryption between a website and their visitor shouldn't cost anything, and having companies issuing these certificates will (er, has) lent itself to corruption.
If we're going to have 3rd party CA's, it should be done by the government. These companies (i.e. Verisign) shouldn't have the right to monopolize this technology, which has become an absolute necessity for online commerce.
And using a PC in the living room, connected to your TV wasn't innovative? With the Xbox, Microsoft took the first big step in bridging home entertainment, computing, and the Internet.
The Xbox was very Innovative, and it's still the best console out there (despite the PS2's stronger library of games)
Seriously, why would you leave data on a free hosting service's servers? You can't count on them. If I use a Hotmail or Yahoo email account, I have to understand it could drop off the planet tomorrow.
It takes big ones to complain about a free service.
I would love to stop using Windows. Problem is, I can't - on my desktop at least (15/17 servers run Open/FreeBSD).
There are some simple reasons why the desktop switch won't work for me:
1) Application Support!
The work I do fits into 2 categories, artistic/creative and technical - mainly for the web, homebrew intranet apps, and the oddball video production.
I need Photoshop (Gimp, while mature, is not a good replacement). I need Premiere. I need IE (for testing purposes, I swear!).
I need to be able to encode to Windows Media A/V formats (the best in streaming for 90% of any web author's target audience - Quicktime doesn't have the install base, and Real is... well Real is Real!)
2) Game Support
While I don't play games much for Leisure, I do need them for work (www.gotfrag.com).
If they would all run under Wine easily, legally, and the first time without and screwing around, I'd be game in this dept - but they dont, and therefore I'm not. There's been a lot of progress here, but there are those of us who can't spend hours to get a game running.
3) Desktop Support
No matter how much I try, I still can't get used to KDE/GNOME. It's not that I'm adverse to using something without a start button (haha.. well, nevermind that in this case) - I love OS X, but the feel that KDE and Gnome exhibit is, well, a bit rough around the edges. Not to mention the problem of having to choose one and live with all of the repercussions of not being in the other.
In my opinion (as the average user), here's what Linux/BSD needs to be king of the desktop:
1) A standardized UI/API that the developers can get behind. Sorry, but someone has to champion this thing. Microsoft is GREAT at getting developers behind their UI design choices, KDE/GNOME haven't done so well. Apps need to feel right to all users regardless of settings, etc.
2) Commercial software developers have to have reasons to port their software. I don't have the answers here, but 9/10 software companies won't devote the engineering resources to port software unless they see the money in it. I think that one real shot here may be to work through distributors/VAR's to put the pressure on here, and show the sales potential (hopefully it exists).
3) DirectX. Native. OpenGL (and other fringe, unrelated libraries) are no longer useful. DirectX is the platform, and rightly so - it's the best out there. Linux needs it in the worst way, and having it would make porting games incredibly easy. Not to mention that many multimedia related desktop apps are using DX components too!
4) Developer Environment and tools support. Linux/BSD are doing well here. Eclipse is where it's at, everyone should rally around it with the proper plugins to make a fully universal IDE. It works on Windows, perfectly. It will allow more Windows developers to work at porting their software to other systems, because they can jump right in without re-learning the tools and techniques.
That's about all I have, but there's a long way to go. We're making good progress though.
One important note, Linux doesn't have to have a 70% desktop share to win, not even close. What does need to happen, is for MS share to drop significantly. If MS were to drop to around 50% of the market (with Apple, Linux, BSD, WHATEVER!! eating up the rest), it will force developers to port software, OR it will force developers to standardize their users on a single platform. While the 2nd will be messy, it will make them consider what platform to standardize on. Linux does have a lower TCO in most situations, hopefully by that point the masses will be more educated about it's requirements, and the do's and dont's.
I'm sure that meant 1xAGP as "one AGP slot". Not having AGP 8X would be suicide. It should even have PCI Express if it's going to take a few months to be released.
It's been reported in the OEM market that Iomega has not only replaced, but upgraded the famous Zip Click-Of-Death(TM?) for their new Son of Jaz model. It seems that when my SoJ disks start to fail, the device will begin playing soulful tunes from the always enjoyable John Coltrane.
Iomega may not understand market pricing, quality assurance or customer service. It's good to know that they have figured out something that their customers have known for a while now - when you lose data, soothing music helps ease the pain!
Slashdot Mirror
First, let's take some facts:
1) PC's are cheaper
2) Hardware vendors target PC for performance upgrades
3) PC's are much cheaper to upgrade
4) Nearly all games target Windows
By having Apple port games, you face an uphill battle on point 4, and do nothing about 1-3. Sorry, but what's my incentive to switch?
Why on earth would someone switch to Mac for gaming, when it will take longer to get their games, they'll have to dump more into the system, then dump more into the upgrades (which will also come after they do on PC, and many won't come at all). All so they get to say "Hey look at me! I'm a sucker!"
I work in the gaming business (albiet in a fringe corner of it). We have access to over 2 million "hardcore" gamers - those that like to play competitively. I have a hard time believing you'll convert any with this idea - it's simply not practical.
If you want to convert people, get some great Mac-only games. There's a long way to go before anyone serious will do that. If Valve launched CS:S and HL2 on Mac exclusively, they've already got 1 foot in the grave.
PC makers are pushing hardware limits. How will Mac play games? Let's see a port of HL2 to Mac, and have the hardware guys go nuts. If a Mac can consistently beat a top-end PC in the typical game benchmarks, then it will get some buyers.
While I hope Apple becomes a gaming system some day... I'm quite skeptical that it will ever happen.
I don't use MySQL for valuable data. But I do consider myself a professional, and I do use MySQL for "convenience records" - data that is merely a cache, or set of aggregate data that can be regenerated easily, and has to be queried a LOT (thousands of times per second). MySQL fits the bill here. Postgres, the database we use for the rest of our needs, couldn't handle so many queries without having locking issues.
Sometimes the lack of caring about data integrity (and therefore using less locks on a high-UPDATE table) can be good. To each their own.
While I'm at it, I'll compare this apple, and this orange...
You obviously don't understand the difference between PayPal and a bank, yet still feel compelled to preach about it - which is typical of the normal Slashdot visitor I suppose. I guess I shouldn't have commented on something that I actually have experience with.
When you have dealt with such a system, I'll give two and a half shits about your suggestion to "re-think my post".
The "user systems" and associated features of these bank sites have nowheres near the relational properties of something like PayPal - it's not even close. They can segregate data based on customer sets, where you know where users are coming from, and you can tie them to their own set of data - data that can be archived and summarized into non-realtime aggregates.
PayPal has the task of linking transactions between users, and allowing users to create modifications or additions to them in real time.
These are not the same, by any stretch. It's the final bit of difference that makes it so much more complex.
There are two primary issues that banks deal with:
1) a large amount of data
2) a high volume of transactions
PayPal and eBay have the pleasure of dealing with a third:
3) highly versatile users with demanding needs, to both points #1 and #2.
And being a system that requires transactional integrity, you can add a fourth:
4) ACID compliant, highly relational database with proper locking, even on SELECT. NOT MYSQL.
#1 or #2 = very easily managed
#1 + #2 = get a good DBA
#1 + #2 + #3 = good DBA, good infrastructure
#1 + #2 + #3 + #4 = fackin gridlock
This is something that _WAS_ very rare until recently, and is only now becoming more prevalent, with the adoption of more data-oriented websites, more overall users, and them being more demanding.
I have the pleasure of working in a similar (although significantly smaller, though still "big") environment that fits these conditions. It's incredibly difficult to make a very large transaction based, highly-relational data set accessible to many (>500K) users at a time. Anyone who says otherwise hasn't had to do it.
For those who say "if its already slow, dont slashdot it!", seems you over estimate the slashdot effect. Slashdot links may take a normal web host down, but anyone with an infrastructure built to serve a lot of traffic shouldn't even notice it.
A site I run has been linked by Slashdot many times - aside from the number of referrals in web logs and perhaps a few mbit spike in traffic, you wouldn't know it even occured.
PayPal is having much larger issues. I've experienced the same types of problems when doing larger upgrades - it's usually a single piece of code that went unchecked, and the odd user(s) that actually get caught in an untested (or improperly tested) segment of code are throwing the system for a loop.
Or, just bad performance planning. I'm not surprised - these things have to happen every now and then, they're quite unavoidable when you're dealing with often changes to an infrastructure with that many _DYNAMIC_ users, and so much _DYNAMIC_ data.
Regardless, it's a growing pain. I've been through a few of these on a 150 GB database with 2 MM users and 15 servers. PayPal is a lot larger, and they surely have much more difficult problems. Hell, if they have to do something such as change the way a single piece of data is stored, that means a lot of downtime (or slowtime) processing!
I'm sure they'll sort it out, unfortunately lots of people are losing lots of business in the process.
My bet is 100% Windows. It doesn't make sense to devote time for development to target anything else.
Of course they will support songs from other stores. That way, you get used to the new features in WMP10, play all your songs there, and you're in musical nirvana.
Where are you going to buy your next song from?
How many employees does SCO have working on this? I generated six figures in the last few by months selling mouse pads to users coming from ads on a single website... If investors consider the ability to generate revenue a test of your worth, SCO should be considered quite worthless.
Also, I ONLY use it plugged into the wall (by requirement, not choice, heh). Additionally, my next machine (very soon now) will have at least 5 hours during desktop usage - travel has forced me into it.
My Sony VAIO PCG-FXA36. AMD Athlon 1 GHz, 256MB, 20GB disk, DVD drive, 15" screen that doesn't adjust itself down when you go to battery power.
You can let this notebook sit idle w/ the screen on (and and tell XP not to turn off the display or spin down the disk) and it will be off in about 30 minutes.
It's a great machine otherwise, but the battery is just horrible...
Over the course of the past 5 years, for mostly business reasons (but often personal as well), I've used my share of hard drives - and had to deal with the support for them as well. I'm talking about around 300-500 drives, mostly Seagate, IBM, and a few Maxtors (until I learned to avoid them). All but a handful of them ATA-66/100, 7200 RPM.
I can say without a doubt that I've seen the lowest failure rate from Seagate drives. Out of all the drives we shipped, over half were Seagate, and we've had around 5 failures. The best news? They were pretty good failures, i.e. the drive may not spin up every time - but when it does the data is fine (good for recovery).
Maxtor gets the highest failure rate, period. Of the 25 or so drives I've used, nearly all of them have failed - many within 6 months. Additionally, once they fail, it's nearly impossible to get them spinning/reading again. The only thing good that I can say about Maxtor drives is that their 5400RPM's are more reliable than the 7200RPM's.
There was a whole series of IBM drives that we shipped, which had quite consistent long-term (3+ year) failures. I believe it was the 75GXP line. IBM never admitted any problem, though it was widely reported (even here on Slashdot) that it was a very common issue. Otherwise, I've been quite happy with the performance and reliability of other lines of IBM drives.
I've heard good and bad about Western Digital, but haven't used but a handful of WD drives (we used all these drives in hot-swap RAID, and WD likes to make their back-plugs incompatible with the "standard", so WD is a non-option).
At the end of the day, Seagate is the IDE drive I would recommend, and I think that with this much experience dealing with the day-to-day support of these types of disks, that recommendation is quite justified.
I'm glad they've increased the warranty length. Perhaps it will make Maxtor do the same, and in the process make them turn out higher quality disks.
So now, in addition to dealing with the kids using all the bandwidth to download mp3's, you now also have to deal with the nerds who are using all of the CPU time calculating prime numbers.
The only way this will work is if you can "kill -9" other user's processes. Make sure it comes with boxing gloves.
When you need to verify the absolute identity of someone you're dealing with, these companies, with their background checks (which aren't strong by any stretch of the imagination), certs by Verisign/etc may make sense.
However, the most common usage of SSL cert's is simply to enable encryption between two points. For this, there's nothing wrong with even a home-brew cert - validation of the cert via it matching the domain should be sufficient. A SSL cert generated by a 3rd party adds absolutely nothing to security, and it shouldn't do anything to reassure the customer/client that they're dealing with a legitimate operation.
Encryption between a website and their visitor shouldn't cost anything, and having companies issuing these certificates will (er, has) lent itself to corruption.
If we're going to have 3rd party CA's, it should be done by the government. These companies (i.e. Verisign) shouldn't have the right to monopolize this technology, which has become an absolute necessity for online commerce.
And using a PC in the living room, connected to your TV wasn't innovative? With the Xbox, Microsoft took the first big step in bridging home entertainment, computing, and the Internet.
The Xbox was very Innovative, and it's still the best console out there (despite the PS2's stronger library of games)
Wait 12 months, friend. I think it's enough to say that Half-Life 2 running on DirectX only is a statement that's hard to argue with.
... which in this case is... nothing :P
Seriously, why would you leave data on a free hosting service's servers? You can't count on them. If I use a Hotmail or Yahoo email account, I have to understand it could drop off the planet tomorrow.
It takes big ones to complain about a free service.
Spammers would just get a copy of the do-not-spam list and start spamming it! There's absolutely nothing to stop them.
We need SMTP v2.0, and we need it soon.
AMD should be able to get a Sempron 2500+ (or whatever) to the $50-70 USD range - that goes a long way in the low-end OEM market.
I would love to stop using Windows. Problem is, I can't - on my desktop at least (15/17 servers run Open/FreeBSD).
There are some simple reasons why the desktop switch won't work for me:
1) Application Support!
The work I do fits into 2 categories, artistic/creative and technical - mainly for the web, homebrew intranet apps, and the oddball video production.
I need Photoshop (Gimp, while mature, is not a good replacement). I need Premiere. I need IE (for testing purposes, I swear!).
I need to be able to encode to Windows Media A/V formats (the best in streaming for 90% of any web author's target audience - Quicktime doesn't have the install base, and Real is... well Real is Real!)
2) Game Support
While I don't play games much for Leisure, I do need them for work (www.gotfrag.com).
If they would all run under Wine easily, legally, and the first time without and screwing around, I'd be game in this dept - but they dont, and therefore I'm not. There's been a lot of progress here, but there are those of us who can't spend hours to get a game running.
3) Desktop Support
No matter how much I try, I still can't get used to KDE/GNOME. It's not that I'm adverse to using something without a start button (haha.. well, nevermind that in this case) - I love OS X, but the feel that KDE and Gnome exhibit is, well, a bit rough around the edges. Not to mention the problem of having to choose one and live with all of the repercussions of not being in the other.
In my opinion (as the average user), here's what Linux/BSD needs to be king of the desktop:
1) A standardized UI/API that the developers can get behind. Sorry, but someone has to champion this thing. Microsoft is GREAT at getting developers behind their UI design choices, KDE/GNOME haven't done so well. Apps need to feel right to all users regardless of settings, etc.
2) Commercial software developers have to have reasons to port their software. I don't have the answers here, but 9/10 software companies won't devote the engineering resources to port software unless they see the money in it. I think that one real shot here may be to work through distributors/VAR's to put the pressure on here, and show the sales potential (hopefully it exists).
3) DirectX. Native. OpenGL (and other fringe, unrelated libraries) are no longer useful. DirectX is the platform, and rightly so - it's the best out there. Linux needs it in the worst way, and having it would make porting games incredibly easy. Not to mention that many multimedia related desktop apps are using DX components too!
4) Developer Environment and tools support. Linux/BSD are doing well here. Eclipse is where it's at, everyone should rally around it with the proper plugins to make a fully universal IDE. It works on Windows, perfectly. It will allow more Windows developers to work at porting their software to other systems, because they can jump right in without re-learning the tools and techniques.
That's about all I have, but there's a long way to go. We're making good progress though.
One important note, Linux doesn't have to have a 70% desktop share to win, not even close. What does need to happen, is for MS share to drop significantly. If MS were to drop to around 50% of the market (with Apple, Linux, BSD, WHATEVER!! eating up the rest), it will force developers to port software, OR it will force developers to standardize their users on a single platform. While the 2nd will be messy, it will make them consider what platform to standardize on. Linux does have a lower TCO in most situations, hopefully by that point the masses will be more educated about it's requirements, and the do's and dont's.
Anywho, I can't leave Windows yet. Soon maybe?
I'm sure that meant 1xAGP as "one AGP slot". Not having AGP 8X would be suicide. It should even have PCI Express if it's going to take a few months to be released.
Just export the CVS checkin log and get on with your day!
Amazon owns Alexa, which has a Toolbar that sends your browsing habits to Amazon for rankings and analysis.
a9 likely uses Alexa data to generate better search results, and the a9 toolbar likely sends data to Alexa and/or a9 for analysis.
Yep, I think that's right.
It's been reported in the OEM market that Iomega has not only replaced, but upgraded the famous Zip Click-Of-Death(TM?) for their new Son of Jaz model. It seems that when my SoJ disks start to fail, the device will begin playing soulful tunes from the always enjoyable John Coltrane.
Iomega may not understand market pricing, quality assurance or customer service. It's good to know that they have figured out something that their customers have known for a while now - when you lose data, soothing music helps ease the pain!
^H is geek for "I can't configut^Hre my terminal correctly"