with a console, you buy a game, you go home, you stick it in the drive and you play the game.
Yes, that is their biggest strength. The biggest weakness is that you're on your own if that ever fails to work. I don't like the idea of spending $50 on a game -- most games I buy for the PC are half that, by the way -- and having that all hinge on the health of one DVD or Blu-Ray disc.
You have, however, missed perhaps the greatest strength of consoles: You can rent a console game. You can even rent a whole console.
By contrast, two of the last 4 PC games I bought (Spore and Far Cry 2) have required me to faff around with drivers before they would run.
You're right -- that does suck. A quick story:
When I bought this computer, it came with Ubuntu -- but it was a 32-bit Ubuntu, and a 64-bit machine. So I formatted it. The nvidia drivers were included, and worked out of the box. When I need to upgrade them, I tell my package manager to upgrade anything on my system, and it's handled.
On Windows was another story. Had it been a desktop, I could have simply gone to nvidia.com, and downloaded a driver -- heh, "simply", but that's not too bad. Instead, by some twisted licensing deal, nvidia does not have the rights to distribute drivers for this card -- only the manufacturer (Dell) did. Since they sold it with Vista, getting XP drivers was not easy -- I had to actually get on a chat with support, and have them feed me links.
And that was just the video. Everything else was similar -- ten or twenty different drivers to install on Windows, plus updates, versus exactly one tweak on Linux. And now I'll have to check back periodically for updates to those ten or twenty drivers (individually!), versus exactly one place to update everything on Linux.
This is not a pro-Linux rant. As much as I wish developers would see how much better of a gaming platform Linux could be, I doubt this is going to happen. No, the point here is that there's two very easy things Microsoft could do to make Windows approach Linux in terms of user-friendliness in that area:
1: Open up Windows Update -- or Microsoft Update -- into a single Update service, which third parties can plug into. 2: Include, as part of the WHQL requirements, a requirement that the driver be made available through that Update service. No excuses.
Then, when you install those new games, if they find you don't have new enough drivers, they could simply tell you to push your one big "Update" button, and all would be well.
There's also the price issue. A console will set you back a few hundred dollars, but you then don't need to replace it for 4-5 years. A gaming PC will set you back at least twice as much (and frequently more) and will generally be obsolete within two and a half years, unless you're willing to sink a lot of money into interim upgrades.
PC games are often cheaper, and much more often are available via digital distribution. Steam seems to be bigger than Xbox Live Arcade, last I checked.
And the price you've quoted isn't exactly accurate -- compare a PS3 for $400, maybe $500, to a decent gaming PC for $550. Hardly twice as much, and it means you've also got a desktop PC -- can you buy a PC that does everything else you want except gaming for $150?
Obsolete faster? Hardly -- games can be scaled back for older computers. Just because you have the option of buying a newer, faster computer every two years doesn't mean you have to -- any more than you have to buy a new PS3 because the new version has a bigger hard drive. (What's more, guess what you can buy for a computer? That's right, a bigger hard drive! Much cheaper than the bigger-hard-drive PS3, too...)
On top of which, in order to play all console games, you would have to own all three consoles, which adds up. Granted, that's going to be an issue for PCs too, but at least here, there's more competition on the hardware. Or, put another way: Cha
A console gets installed like a DVD player. A laptop will "wander" around the house. So it's disconnect/reconnect.
My laptop has HDMI out. So it's one connection, maybe two -- not a big deal.
And the "wandering" is a bonus -- I now have the same games when I go somewhere else. People keep talking about how economical it is to have split-screen multiplayer, rather than everyone having their own console -- but in my experience, people end up buying their own anyway, and now it's irritating to have to carry them around. A laptop, I can just drop in a backpack and walk over to a LAN party.
Finding a good spot to put the laptop, and on top ugly cables in front of your TV instead of nicely bundled at the back.
The majority of TVs I've seen have one of two setups: Either there is a large cabinet where everything goes, with cables neatly tucked behind, or there's just a small stand for the TV and everything's on the floor.
In the first case, it's not hard to find a shelf where the laptop can be tucked. So, that's two cables to plug in -- power, HDMI -- and push it onto the shelf. When it's gone, the cables are still there on the shelf, not "ugly and in front".
In the second case, the cables are already there on the floor -- maybe technically "behind" things, but you could call them ugly anyway. In which case, it's even easier to just set the laptop down, plug it in, and use it.
It might be possible to build something like that for 500â (Still I doubt it), but in two years time he'll have exactly the same problem when GTA5 comes out. Not so with a console.
No, with a console, in two years time, GTA5 will either not have terribly improved graphics, or require a completely new console.
Conversely, with a PC, even if you ignore the possibility of a partial upgrade, even a brand-new PC now is still going to perform as well as or better than most consoles -- so, in 2-3 years time, you simply dial down the settings on the game, and it still works.
Now, I consider the existence of things like a keyboard, a mouse, and some of the better mods and free games to be worth the price of a PC. But I think there's still a fair case to be made that a PC can be as good a platform as a console.
And the point which I've been trying to hammer home lately:
An important lesson that Google should take away from this is the importance of developing cross platform software on all platforms at the same time. Assigning just one guy six months ago to the Linux port would have significantly improved the source on day 1.
And just one guy certainly wouldn't have been a huge setback to getting it working on Windows, either, which is the frequent excuse given in this thread for the lack of non-Windows support.
When I used Firefox and Adblock, I maintained my own blocklist, refusing to simply grab a prefabricated block-everything list. I did this because there are a lot of ads I don't actually mind seeing, and sometimes I do actually click through and buy things.
But, I can't stand moving ads, especially Flash ads. Moving ads distract me, and Flash ads eat CPU while doing that -- and, in Flash 9, actually made my browser unstable. And the IntelliTXT people just piss me off... So I started blocking specific networks.
Now, I've switched to Konqueror for most of my browsing, and blocking a given ad just isn't as easy anymore -- no simple right-lick, and no Firebug to fall back on. Sure, I can run Firefox, but it's too much hassle to do all that just to maintain my own list...
So I don't. I just let Konqueror pull down one of the common blacklists, and I don't worry about it.
Moral of the story: If you want us not to block ads, make them less annoying. Maybe even make ads we want to see -- ads we go out of our way to watch. But do it quickly, before everyone is using an Adblock of some sort.
I mean yeah it's supposed to stop people from stealing your much more valuable personal data but that should be password protected anyway with a directory hider/protector (not like a compressed archive file with a password cuz that's too slow) so why bother?
Your ignorance is showing...
Compressed archive files are plenty fast, depending on what you're trying to protect. The real problem is, what happens when you "open" them? Most of the time, it'll be unpacking them to a temporary directory, opening them with some random program on your (unencrypted) hard drive (likely without anything to prevent it from being swapped out, so now your stuff is on disk in the clear twice), saved back to the temporary folder (three times, if you're still counting), and put back into the archive.
Plus, there's now a mention in Recent Documents, and all kinds of other information letting people know, at the very least, that you have some encrypted files, and what their names are.
This applies to Truecrypt also, by the way, unless you're using it for fulldisk encryption.
And if you're encrypting the whole disk -- where will you keep the encryption keys? How will you boot? Doing it in hardware suddenly makes sense -- probably a slight performance boost, also.
And once you're doing that, having a way to remotely destroy the crypto keys also makes sense -- if you're paranoid enough to encrypt your whole hard drive, this is the next best thing to putting thermite in the case and triggering that remotely instead.
It's not a deterrent, it's a way to make the crypto much more secure.
Well, if an OS already does really well with a normal hard drive (say, because it caches things effectively, schedules I/O well, and so on), then switching to an SSD will produce a smaller speedup.
On the other hand, if you're doing that, you're probably also making assumptions about the fact that you're on a spinning disk. So, for example:
The fact that Win98 fared best probably has more to say about its lackluster I/O scheduling and caching than it does about its suitability to SSDs.
One of the hacks I've heard recently -- but haven't tried yet -- is actually disabling the default intelligent IO scheduler and using a dumber one. Since there's no seek time, it doesn't particularly matter what order you read from the device -- so any time spent doing things like ordering and prioritizing reads/writes is wasted.
I'm not saying that you shouldn't implement Windows. I'm saying that if you do POSIX first, you've still got most other OSes, which means lots of practice porting and making it portable.
Which means a cleaner architecture, and probably an easier port to Windows.
At least, I know from personal experience that it's easier porting web-standard sites to IE than it is porting IE-only sites to standards.
Chrome is open source. Just as we now have portable Firefox, people with an interest in running Chrome in hostile environments like yours are likely to develop portable Chrome, with or without Google's consent.
It looks as though you'll have to invest in some actual security.
But I find it hard to believe that one can't understand hitting 90% of the market as quickly as possible and them filling in as much of the remaining 10%.
If you ever intend to hit that 10%, you're better off starting there.
Keep in mind, Windows/IE is always the odd one out. Stick to POSIX, and it pretty much works everywhere except Windows. Stick to standard html/css/javascript, and it pretty much works everywhere except IE.
We keep an ie.css and an ie.js file, for that reason. We develop in Firefox, and only fire up IE to verify that it still works. If it doesn't, we then add some hack to support IE -- much easier than trying to take a completely hacked-up version developed for IE, and adding the cross-browser compatibility after the fact.
Bad analogy. This is a totally different issue. Memory fragmentation means you leave little gaps in the allocated blocks which are too small to reuse, effectively wasting memory.
Given a sufficient number of small files, this can happen with filesystems, too. Of course, it's usually not as pronounced, as most filesystems won't pack more than one file into a block...
Keep in mind, the block leveling algorithm will be abstracting the actual disk organization; defragmenting wouldn't accomplish anything on an SSD
It would, at the very least, consolidate extents.
I would also say, it's sad that SSDs have block leveling in hardware -- I mostly blame Windows for that.
And how long does it take you to build that proper defense? If you aren't building as fast as you can click, are you actually prepared when they swarm in?
No, Zerg Rush isn't the only strategy. However, raw speed of micromanaging is a necessary component -- until you have that, it's very difficult to win on strategy alone.
This attitude really comes through about two seconds in:
But going back to Dune II today is an eye-opening experience, as it becomes clear how very little the genre has moved forwards in the last 16 years â" like finding out Halo had really been released in 1982.
It was released in 1993 -- as Doom. And Dune II wasn't 1982, it was 1992, only one year before Doom.
Granted, there have been significant improvements in gameplay and visuals since then -- there's no equivalent to Halo's melee weapons, or melee use of a weapon -- but the fundamentals are the same.
That's also true of Dune II, if I remember -- Starcraft is years ahead of it, if only for the ability to select multiple units at once and direct them at a target -- better AI helps, too.
I don't think you can really say, though, that Starcraft 2 is that much farther ahead of Dune II than Halo 3 is ahead of Doom.
I think that one thing we can forget about with SSD is file fragmentation. The contiguity of data will be far less of a problem with the negligible latency times of these new disks.
Well, no, actually -- remember how Firefox used to leak memory? A lot of that was actually memory fragmentation. And RAM has exactly as much seek time as an SSD -- that is, none at all.
It's less of an issue, certainly. You probably don't want to run a defragger nearly as much, maybe not at all. But there's still an advantage -- especially if the block layer is doing wear-leveling for you -- to storing large files in contiguous extents, and packing small files together into a single block.
Well, the article is crap in many ways -- doesn't cite what benchmarks they actually used, just throws around meaningless percentages and outright lies (Linux and Win2k absolutely do run things in the background).
One of the other little details they left out is which filesystem they used, on any platform. My guess is, they just used the default -- I wonder if they were even aware of alternatives (you can install Win2k on FAT32, if you really want).
So, to answer your comment, Linux has at least two filesystems that are designed to work directly on solid-state media. Unfortunately, most SSDs pretend to be ATA hard drives, but the point still stands -- Linux has many filesystems. I wonder which one actually performs best on that ATA drive?
But, harder fights won't accomplish that goal of making them rare. Players will just get better.
The drop rate is just as crude, but at least there, you've got the possibility of cheating and completely arbitrating it -- of setting a drop rate by some unit of time, say, or as a percentage of the player base, rather than simply a percent chance on every attempt.
So, the choice is anal rape, with or without lube?
To say that with somewhat less flamebait: If you really want to make it about choice, try at least offering languages from more than one vendor, preferably more than one paradigm. My choices would be, in no particular order:
I know I'm late to the thread, but I've used Try Ruby with some success on that age group.
I'm suggesting the tutorial more than the language -- it's easier to pick up a new language than it is to start programming in the first place.
There are reasons I'd suggest Ruby, of course -- if they find they want to write performance-critical apps, it's Algol-like enough that C won't be too hard to learn. But if they do stick with Ruby, unlike JavaScript, there's a gradual, natural progression to more and more advanced things. JavaScript mutates as you learn more about it.
I didn't go to class and they gave me an honours degree. A good one too.
See, that worked somewhat for the comp sci stuff -- I wouldn't go to class, but I would do the weekly assignments and show up for tests. Roughly two hours a week of work, and I got a B.
That didn't work for the other classes.
Any specific career related knowledge gained is and should be purely coincidental. It's a fucking university, not a tech college.
I think you're confusing "university" with "high school".
In high school, it really doesn't matter what you're taught, because most of it is simply to give you very basic skills (like "how to type"), and to get you in the habit of writing papers, doing homework, and all the other sort of meta-work-related stuff that is actually helpful. But the knowledge itself can be completely thrown away.
College is meant to actually teach you things -- the idea being to give you a well-rounded education, and to teach you about things you're specifically interested in. There are fundamental skills I did not learn in college, but did learn very quickly at my first development job.
Yes, I might throw out the specific language they used to teach me to program. But I would expect to at least learn some programming, and learn about unit tests, version control, communicating with a team... Basic, basic stuff, which applies no matter what else I do in software.
Maybe they saved that for later years -- which still seems idiotic. I'd argue, version control should be in comp sci 101.
Which is, by the way, why starting with Java is a bad idea. Every language benefits from automated testing. Not every language has a concept of public methods, static methods, return types, or even classes, all of which I was expected to understand in the very first lecture. (I was fine; I'd played with Java before...)
I guess some would say that it's impossible to code PHP well.
I thought so, too. Then I discovered Drupal.
I am now convinced that it is possible to code any language well, but you'll spend much of your time re-implementing things that you get for free in other languages. Drupal was written before PHP started to (sort of, half-assedly) make itself object-oriented, so they built an object-like plugin system on top of it.
I'd much rather work in a language which goes out of its way to make it easy to code well.
How do you make a price that is competitive with "free"?
By also providing enough more value -- that people actually want -- which isn't available for free. And, in some cases, by actually making it free.
It is possible, in absolute terms, to provide a service which saves enough of your customers' time (versus the free alternative) for it to be worth the price. Even if this is not the case, it may be worth the money for the convenience.
And that assumes that the products are identical. There are ways, even with music, to make the legitimate version of a product more desirable than a pirated version.
And yes, it is possible to make money by giving the product away entirely. How do you think Pandora makes money?
Let me rephrase it: Microsoft and Apple both seem to have found a price that is competitive with "free". I've paid for Windows to dual-boot on this machine, because I want to play games which are not available for Linux, and don't necessarily work under Wine.
Slashdot Mirror
You can, however, make life very difficult for a good one by giving them lousy tools, or by not teaching them how to use their tools.
with a console, you buy a game, you go home, you stick it in the drive and you play the game.
Yes, that is their biggest strength. The biggest weakness is that you're on your own if that ever fails to work. I don't like the idea of spending $50 on a game -- most games I buy for the PC are half that, by the way -- and having that all hinge on the health of one DVD or Blu-Ray disc.
You have, however, missed perhaps the greatest strength of consoles: You can rent a console game. You can even rent a whole console.
By contrast, two of the last 4 PC games I bought (Spore and Far Cry 2) have required me to faff around with drivers before they would run.
You're right -- that does suck. A quick story:
When I bought this computer, it came with Ubuntu -- but it was a 32-bit Ubuntu, and a 64-bit machine. So I formatted it. The nvidia drivers were included, and worked out of the box. When I need to upgrade them, I tell my package manager to upgrade anything on my system, and it's handled.
On Windows was another story. Had it been a desktop, I could have simply gone to nvidia.com, and downloaded a driver -- heh, "simply", but that's not too bad. Instead, by some twisted licensing deal, nvidia does not have the rights to distribute drivers for this card -- only the manufacturer (Dell) did. Since they sold it with Vista, getting XP drivers was not easy -- I had to actually get on a chat with support, and have them feed me links.
And that was just the video. Everything else was similar -- ten or twenty different drivers to install on Windows, plus updates, versus exactly one tweak on Linux. And now I'll have to check back periodically for updates to those ten or twenty drivers (individually!), versus exactly one place to update everything on Linux.
This is not a pro-Linux rant. As much as I wish developers would see how much better of a gaming platform Linux could be, I doubt this is going to happen. No, the point here is that there's two very easy things Microsoft could do to make Windows approach Linux in terms of user-friendliness in that area:
1: Open up Windows Update -- or Microsoft Update -- into a single Update service, which third parties can plug into.
2: Include, as part of the WHQL requirements, a requirement that the driver be made available through that Update service. No excuses.
Then, when you install those new games, if they find you don't have new enough drivers, they could simply tell you to push your one big "Update" button, and all would be well.
There's also the price issue. A console will set you back a few hundred dollars, but you then don't need to replace it for 4-5 years. A gaming PC will set you back at least twice as much (and frequently more) and will generally be obsolete within two and a half years, unless you're willing to sink a lot of money into interim upgrades.
PC games are often cheaper, and much more often are available via digital distribution. Steam seems to be bigger than Xbox Live Arcade, last I checked.
And the price you've quoted isn't exactly accurate -- compare a PS3 for $400, maybe $500, to a decent gaming PC for $550. Hardly twice as much, and it means you've also got a desktop PC -- can you buy a PC that does everything else you want except gaming for $150?
Obsolete faster? Hardly -- games can be scaled back for older computers. Just because you have the option of buying a newer, faster computer every two years doesn't mean you have to -- any more than you have to buy a new PS3 because the new version has a bigger hard drive. (What's more, guess what you can buy for a computer? That's right, a bigger hard drive! Much cheaper than the bigger-hard-drive PS3, too...)
On top of which, in order to play all console games, you would have to own all three consoles, which adds up. Granted, that's going to be an issue for PCs too, but at least here, there's more competition on the hardware. Or, put another way: Cha
A console gets installed like a DVD player. A laptop will "wander" around the house. So it's disconnect/reconnect.
My laptop has HDMI out. So it's one connection, maybe two -- not a big deal.
And the "wandering" is a bonus -- I now have the same games when I go somewhere else. People keep talking about how economical it is to have split-screen multiplayer, rather than everyone having their own console -- but in my experience, people end up buying their own anyway, and now it's irritating to have to carry them around. A laptop, I can just drop in a backpack and walk over to a LAN party.
Finding a good spot to put the laptop, and on top ugly cables in front of your TV instead of nicely bundled at the back.
The majority of TVs I've seen have one of two setups: Either there is a large cabinet where everything goes, with cables neatly tucked behind, or there's just a small stand for the TV and everything's on the floor.
In the first case, it's not hard to find a shelf where the laptop can be tucked. So, that's two cables to plug in -- power, HDMI -- and push it onto the shelf. When it's gone, the cables are still there on the shelf, not "ugly and in front".
In the second case, the cables are already there on the floor -- maybe technically "behind" things, but you could call them ugly anyway. In which case, it's even easier to just set the laptop down, plug it in, and use it.
It might be possible to build something like that for 500â (Still I doubt it), but in two years time he'll have exactly the same problem when GTA5 comes out. Not so with a console.
No, with a console, in two years time, GTA5 will either not have terribly improved graphics, or require a completely new console.
Conversely, with a PC, even if you ignore the possibility of a partial upgrade, even a brand-new PC now is still going to perform as well as or better than most consoles -- so, in 2-3 years time, you simply dial down the settings on the game, and it still works.
Now, I consider the existence of things like a keyboard, a mouse, and some of the better mods and free games to be worth the price of a PC. But I think there's still a fair case to be made that a PC can be as good a platform as a console.
Why would I want more Monday? If I've learned anything from Garfield, it's that Monday should be outlawed!
Now, if they can tell me how to get more Friday or Saturday, I'm in!
And the point which I've been trying to hammer home lately:
An important lesson that Google should take away from this is the importance of developing cross platform software on all platforms at the same time. Assigning just one guy six months ago to the Linux port would have significantly improved the source on day 1.
And just one guy certainly wouldn't have been a huge setback to getting it working on Windows, either, which is the frequent excuse given in this thread for the lack of non-Windows support.
When I used Firefox and Adblock, I maintained my own blocklist, refusing to simply grab a prefabricated block-everything list. I did this because there are a lot of ads I don't actually mind seeing, and sometimes I do actually click through and buy things.
But, I can't stand moving ads, especially Flash ads. Moving ads distract me, and Flash ads eat CPU while doing that -- and, in Flash 9, actually made my browser unstable. And the IntelliTXT people just piss me off... So I started blocking specific networks.
Now, I've switched to Konqueror for most of my browsing, and blocking a given ad just isn't as easy anymore -- no simple right-lick, and no Firebug to fall back on. Sure, I can run Firefox, but it's too much hassle to do all that just to maintain my own list...
So I don't. I just let Konqueror pull down one of the common blacklists, and I don't worry about it.
Moral of the story: If you want us not to block ads, make them less annoying. Maybe even make ads we want to see -- ads we go out of our way to watch. But do it quickly, before everyone is using an Adblock of some sort.
I mean yeah it's supposed to stop people from stealing your much more valuable personal data but that should be password protected anyway with a directory hider/protector (not like a compressed archive file with a password cuz that's too slow) so why bother?
Your ignorance is showing...
Compressed archive files are plenty fast, depending on what you're trying to protect. The real problem is, what happens when you "open" them? Most of the time, it'll be unpacking them to a temporary directory, opening them with some random program on your (unencrypted) hard drive (likely without anything to prevent it from being swapped out, so now your stuff is on disk in the clear twice), saved back to the temporary folder (three times, if you're still counting), and put back into the archive.
Plus, there's now a mention in Recent Documents, and all kinds of other information letting people know, at the very least, that you have some encrypted files, and what their names are.
This applies to Truecrypt also, by the way, unless you're using it for fulldisk encryption.
And if you're encrypting the whole disk -- where will you keep the encryption keys? How will you boot? Doing it in hardware suddenly makes sense -- probably a slight performance boost, also.
And once you're doing that, having a way to remotely destroy the crypto keys also makes sense -- if you're paranoid enough to encrypt your whole hard drive, this is the next best thing to putting thermite in the case and triggering that remotely instead.
It's not a deterrent, it's a way to make the crypto much more secure.
Well, if an OS already does really well with a normal hard drive (say, because it caches things effectively, schedules I/O well, and so on), then switching to an SSD will produce a smaller speedup.
On the other hand, if you're doing that, you're probably also making assumptions about the fact that you're on a spinning disk. So, for example:
The fact that Win98 fared best probably has more to say about its lackluster I/O scheduling and caching than it does about its suitability to SSDs.
One of the hacks I've heard recently -- but haven't tried yet -- is actually disabling the default intelligent IO scheduler and using a dumber one. Since there's no seek time, it doesn't particularly matter what order you read from the device -- so any time spent doing things like ordering and prioritizing reads/writes is wasted.
I'm not saying that you shouldn't implement Windows. I'm saying that if you do POSIX first, you've still got most other OSes, which means lots of practice porting and making it portable.
Which means a cleaner architecture, and probably an easier port to Windows.
At least, I know from personal experience that it's easier porting web-standard sites to IE than it is porting IE-only sites to standards.
Chrome is open source. Just as we now have portable Firefox, people with an interest in running Chrome in hostile environments like yours are likely to develop portable Chrome, with or without Google's consent.
It looks as though you'll have to invest in some actual security.
But I find it hard to believe that one can't understand hitting 90% of the market as quickly as possible and them filling in as much of the remaining 10%.
If you ever intend to hit that 10%, you're better off starting there.
Keep in mind, Windows/IE is always the odd one out. Stick to POSIX, and it pretty much works everywhere except Windows. Stick to standard html/css/javascript, and it pretty much works everywhere except IE.
We keep an ie.css and an ie.js file, for that reason. We develop in Firefox, and only fire up IE to verify that it still works. If it doesn't, we then add some hack to support IE -- much easier than trying to take a completely hacked-up version developed for IE, and adding the cross-browser compatibility after the fact.
What I find interesting is that even someone who isn't terribly impressed is calling it "her".
That's not particularly new -- some languages assign gender to everything. Even in English, we call ships female, and we name storms after people.
And, yes, we have had obviously-fake devices, robotic or not, which have had names and identities. Think children with stuffed animals.
Bad analogy. This is a totally different issue. Memory fragmentation means you leave little gaps in the allocated blocks which are too small to reuse, effectively wasting memory.
Given a sufficient number of small files, this can happen with filesystems, too. Of course, it's usually not as pronounced, as most filesystems won't pack more than one file into a block...
Keep in mind, the block leveling algorithm will be abstracting the actual disk organization; defragmenting wouldn't accomplish anything on an SSD
It would, at the very least, consolidate extents.
I would also say, it's sad that SSDs have block leveling in hardware -- I mostly blame Windows for that.
Slashdot is not a single user with a single mind, despite what you'd like to think.
But when you think about it, if copyright didn't exist, there'd be far less need for something like the GPL.
And how long does it take you to build that proper defense? If you aren't building as fast as you can click, are you actually prepared when they swarm in?
No, Zerg Rush isn't the only strategy. However, raw speed of micromanaging is a necessary component -- until you have that, it's very difficult to win on strategy alone.
This attitude really comes through about two seconds in:
But going back to Dune II today is an eye-opening experience, as it becomes clear how very little the genre has moved forwards in the last 16 years â" like finding out Halo had really been released in 1982.
It was released in 1993 -- as Doom. And Dune II wasn't 1982, it was 1992, only one year before Doom.
Granted, there have been significant improvements in gameplay and visuals since then -- there's no equivalent to Halo's melee weapons, or melee use of a weapon -- but the fundamentals are the same.
That's also true of Dune II, if I remember -- Starcraft is years ahead of it, if only for the ability to select multiple units at once and direct them at a target -- better AI helps, too.
I don't think you can really say, though, that Starcraft 2 is that much farther ahead of Dune II than Halo 3 is ahead of Doom.
I think that one thing we can forget about with SSD is file fragmentation. The contiguity of data will be far less of a problem with the negligible latency times of these new disks.
Well, no, actually -- remember how Firefox used to leak memory? A lot of that was actually memory fragmentation. And RAM has exactly as much seek time as an SSD -- that is, none at all.
It's less of an issue, certainly. You probably don't want to run a defragger nearly as much, maybe not at all. But there's still an advantage -- especially if the block layer is doing wear-leveling for you -- to storing large files in contiguous extents, and packing small files together into a single block.
Well, the article is crap in many ways -- doesn't cite what benchmarks they actually used, just throws around meaningless percentages and outright lies (Linux and Win2k absolutely do run things in the background).
One of the other little details they left out is which filesystem they used, on any platform. My guess is, they just used the default -- I wonder if they were even aware of alternatives (you can install Win2k on FAT32, if you really want).
So, to answer your comment, Linux has at least two filesystems that are designed to work directly on solid-state media. Unfortunately, most SSDs pretend to be ATA hard drives, but the point still stands -- Linux has many filesystems. I wonder which one actually performs best on that ATA drive?
But, harder fights won't accomplish that goal of making them rare. Players will just get better.
The drop rate is just as crude, but at least there, you've got the possibility of cheating and completely arbitrating it -- of setting a drop rate by some unit of time, say, or as a percentage of the player base, rather than simply a percent chance on every attempt.
So, the choice is anal rape, with or without lube?
To say that with somewhat less flamebait: If you really want to make it about choice, try at least offering languages from more than one vendor, preferably more than one paradigm. My choices would be, in no particular order:
- ASM
- Ruby
- Erlang
- Common LISP (or Scheme)
- Haskell
Each of those is sufficiently different as to force you to change the way you think, going from one to the other.
I know I'm late to the thread, but I've used Try Ruby with some success on that age group.
I'm suggesting the tutorial more than the language -- it's easier to pick up a new language than it is to start programming in the first place.
There are reasons I'd suggest Ruby, of course -- if they find they want to write performance-critical apps, it's Algol-like enough that C won't be too hard to learn. But if they do stick with Ruby, unlike JavaScript, there's a gradual, natural progression to more and more advanced things. JavaScript mutates as you learn more about it.
Considering how many of them will be playing WoW, I think Lua would probably be the most immediately useful, yes.
I didn't go to class and they gave me an honours degree. A good one too.
See, that worked somewhat for the comp sci stuff -- I wouldn't go to class, but I would do the weekly assignments and show up for tests. Roughly two hours a week of work, and I got a B.
That didn't work for the other classes.
Any specific career related knowledge gained is and should be purely coincidental. It's a fucking university, not a tech college.
I think you're confusing "university" with "high school".
In high school, it really doesn't matter what you're taught, because most of it is simply to give you very basic skills (like "how to type"), and to get you in the habit of writing papers, doing homework, and all the other sort of meta-work-related stuff that is actually helpful. But the knowledge itself can be completely thrown away.
College is meant to actually teach you things -- the idea being to give you a well-rounded education, and to teach you about things you're specifically interested in. There are fundamental skills I did not learn in college, but did learn very quickly at my first development job.
Yes, I might throw out the specific language they used to teach me to program. But I would expect to at least learn some programming, and learn about unit tests, version control, communicating with a team... Basic, basic stuff, which applies no matter what else I do in software.
Maybe they saved that for later years -- which still seems idiotic. I'd argue, version control should be in comp sci 101.
Which is, by the way, why starting with Java is a bad idea. Every language benefits from automated testing. Not every language has a concept of public methods, static methods, return types, or even classes, all of which I was expected to understand in the very first lecture. (I was fine; I'd played with Java before...)
I guess some would say that it's impossible to code PHP well.
I thought so, too. Then I discovered Drupal.
I am now convinced that it is possible to code any language well, but you'll spend much of your time re-implementing things that you get for free in other languages. Drupal was written before PHP started to (sort of, half-assedly) make itself object-oriented, so they built an object-like plugin system on top of it.
I'd much rather work in a language which goes out of its way to make it easy to code well.
How do you make a price that is competitive with "free"?
By also providing enough more value -- that people actually want -- which isn't available for free. And, in some cases, by actually making it free.
It is possible, in absolute terms, to provide a service which saves enough of your customers' time (versus the free alternative) for it to be worth the price. Even if this is not the case, it may be worth the money for the convenience.
And that assumes that the products are identical. There are ways, even with music, to make the legitimate version of a product more desirable than a pirated version.
And yes, it is possible to make money by giving the product away entirely. How do you think Pandora makes money?
Let me rephrase it: Microsoft and Apple both seem to have found a price that is competitive with "free". I've paid for Windows to dual-boot on this machine, because I want to play games which are not available for Linux, and don't necessarily work under Wine.