It's not an argument. It is. The game is a multiplayer game. Just because that's a stupid idea doesn't mean it isn't the one they went with.
I'm sorry that your point of view is just wrong. But it is. The whole game was balanced around you being able to buy and sell from the auction house. That was a deliberate choice on blizzards part, and without the AH the game becomes prohibitively hard because you just can't get the right itemized gear and you need an astronomical amount of farming to get through the content. Again, I'm not saying that's a *good* design, but that is the design. If anything the game suffers because you almost never loot anything you actually want, I think I looted one inferno difficulty item I actually used, all of the rest I had to buy.
They certainly could have designed the itemization differently or had a full on single player mode with different itemization. But they didn't.
The 'core activity' of diablo is 'click'. I'll grant you that activity is mostly unchanged form previous versions. But most games are more than just one core activity.
they wouldnt even work with me on a refund, when I had issues 3 weeks after launch because I pre ordered it, and therefore it was more than 30 days out of date, eventhough i only had the game for aweek less than 30 days.
yes well, that's a whole other topic. But once they have your money they don't want to give it back.
That's actually pretty common when people do get hacked. If you have gold they immediately mail it off and sell it, and then try and bot farm whatever the best gold/hour is. That might be tradeskilling, that might be cash runs through bosses, sort of depended.
My lingering suspicions is that WoW was vulnerable to a session spoof attack at some point, or the usual exploit of a flash vulnerability to get your password, but their systems became overall pretty robust with authenticators added in.
In your case I'd guess a flash vulnerability, possibly a 0 day one, those are much less of a problem today than they were 2 or 3 years ago when browsers weren't well sandboxed etc. etc. But those sorts of things always got a few people.
Of the 56 unique players in my guild when we quit, only 2 had ever been hacked. We've certainly had people who were hacked off and on over time, (and most of them left the guild) but once they brought in authenticators it was pretty rare for people to get hacked. Even before that, you usually had to do something stupid to get your account hacked.
The most common culprits for it were from re-using passwords (especially on WoW fansites, because duh...) and people buying gold. Then there was the usual keyloggers and so on.
Diablo 3 is a multiplayer game with a where you can choose to not directly interact with other players, but without the auction house the whole itemization would need to be completely different.
That was one of the things they realized with D2, the reason it stuck around was the multiplayer, they just got the idea that the whole thing should be multiplayer. starcraft has less of an excuse because there's no meta economy in starcraft.
making you the target market for games, and modern MMO's. Especially so if you're male. Because you know, the people who actually work at blizzard want to play their own game, and they're mostly over 25 and have jobs. So if you're one of the 40 million or so people who ever created a battle.net account for starcraft or diablo or WoW then yes, this effects you. Because what was your security question, have you ever reused it, and was it publicly available information?
They said they're working on a change to the security question.
But yes, in general this is bad. Although that's sort of the idea behind salting and hashing passwords, that even if someone gets the passwords they still can't recover them.
I take exception to sensationalist reporting that blames the vehicle rather than the driver when the driver is clearly at fault
I think part of the problem is that like any software that fails, sometimes it's hard to tell if it's the user doing something stupid, or a flawed design or implementation, and since the media enjoys fear mongering anything that suggests the car you are driving might be unsafe is far more interesting than a person accidentally hitting the accelerator.
Then you get into broader cultural and societal issues. That we blinding accept speeding, or that the US accepts pervasive private gun ownership as a good idea all contribute to the statistical nature of problems. It's not so much that the car was designed to cause an accident, but someone made a deliberate choice that it only needed to pass whatever inspection regime, only needed to withstand whatever impact, that so many deaths due to accidental discharges, that so many deaths due to crazy people getting guns etc. etc. etc. are all acceptable. Societies as a whole (through governments) make deliberate choices that cars should only be sold if they meet particular criteria for example, there is, in that processes, the implicit acceptance that a certain number of people will die due to car failures that weren't the fault of the end user. Choosing that tolerance is an often buried part of governance.
Oh and people like to sue in the US, so if you can blame the car company and get millions it's like winning the lottery without the gambling.
That's the designer/manufacturer/operator's fault. I.e. people.
And? Once the device is built whatever caused the state it's in, if it fails in a way the end user in incapable of having prevented with archaic expertise it's the fault of the device, and the production chain that led to the device, or the business climate that let it get created (see support for IE6), which ceases being the fault of the individual user, and the fault of many people involved in the chain, from regulators, to physical assembly and distribution.
It doesn't seem to have changed their peak concurrent users any. They might have lost marginal customers, but they seem to have retained their core base.
At least not when you're measuring user numbers in the ~4-5 million range. Whether it cost 100k users here or there is hard to tell.
There's a rating system and apps are sorted into categories.
yep, and that's problematic, because like the iOS and Google play store, if you aren't on the front page you may as well not exist. Which is the whole problem with the app store model.
Which doesn't even work well for those of us in the games business. Gaming is actually a series of specialized niches, telling everyone they can go buy angry birds is fine, but if you want to sell prototype 2, or legends of pegasus you're never going to match angry birds on the chart, even if you could have a very successful game. Which is where steam comes in. If you want games for a 9 year old, steam isn't your place. If you want shitty mobile games, steam isn't the place to be. So right off the bat steam is a selection pressure on the market, you have to be good enough to warrant listing on steam in the first place, and you're not burying 60 million dollar games with 600 dollar games (production costs).
Adding in two dozen categories isn't going to help either. Steam and its ilk are useful because they roll through releases reasonably quickly and reasonably well, but you can still sensibly organize stuff. Although even that is becoming a problem with 2000 games on steam as we get into having games that are from 2002-2003 still prominently listed even though they don't really belong there.
That's why I say it's going to suck at being an app store: because they all suck at it. Even, as I say, something like steam, where you start adding in DLC and more than a couple of years of history and it becomes very cluttered and very hard to find something decent that was released 2 years ago unless you know exactly what you're looking for.
and the apps themselves won't connect you to other users
This one I definitely wasn't clear on. Steam connects you to your gamer friends, regardless of what game you're playing. XBL and PSN do the same thing. GFW actually is XBL, but the question becomes on of whether or not apps in any sort of sensible way will bundle in functionality, or if they'll all try and talk to each other reasonably well. With the windows app store you *might* be able to get into a contact list, you might not, you might use XBL/GFW, they might let you install steam, it might store lists on its own and now all the matchmaking services (for gaming) are going to try and bundle. It's going to go from 'use steam like everyone else' to an ungodly mess of dozens of companies all trying to be the matchmaking service. Which in and of itself isn't the end of the world, it's just going to be confusing compared to steam.
I'd like to point out where you are clearly talking out of your ass:
I'd like to point out that I'm in the process of trying to re-package a (non steamworks) game for windows 8, and confronting all of these problems, and more. And we're working on a release coming up that will have to support windows 8 somehow. Unfortunately we're a niche title (because gaming is a collection of niches), where 50-60K copies would be very successful, and it's not that we see windows 8 as a problem, we'll just use steam and gamersgate and impulse and.... and skip over the windows 8 store, because it's not adding value to us, and if anything it's just making things harder for our users. Not that I'm in charge of business decisions so the boss guy could decide he wants to go all in on windows 8 or steam or whatever.
if only. Then we'd never have recalls on cars, never have to patch bugs in IDE's and never have never have guns that accidentally discharge.
2012 jeep review in sweden the vehicle is consistently blowing tyres and nearly rolled over once. Most of the video is trying to figure out why it nearly rolled over once but not on subsequent attempts. But the tyres blowing at 70-80Km/h is uh... bad. Really really bad.
Bugs fixed in Visual studio 2012 some of this stuff goes back years and has to be compared against the c++ STL precisely because the way it does work, and the way it should work are not the same.
Don't get me started on years of various nVIDIA and AMD tools not playing nice with OpenGl or Directx.
Wiki on accidental discharge lists two scenarios, where a weapon is dropped, or when a weapon overheats that it can accidentally fire if it was improperly designed (e.g. a poor choice of materials).
So yes, cars themselves can cause accidents, guns do kill people, and IDE's can cause bugs in your code.
That's my point. They seem to be spending a huge pile of money, and exerting a huge effort in trying to hedge against windows 8, though I'm not convinced they're in any real risk.
Then again, the reason you invest in a hedge like that is because if you guessed wrong and didn't have a plan B you'd be out of business.
The difference is that the average user can't see what previously bundled products did badly, and developers kept supporting it.
Windows 8 is different. From the moment you boot it up on a desktop you can see what's wrong with it. You try and use anything, and you can see what is wrong with it. With the app store it will probably suck at actually being an app store, you won't be able to find good apps easily (unless you know the name of what you want), and the apps themselves won't connect you to other users, which is critical to a multiplayer environment.
I'm sure microsoft will *try* to put other people out of business, but if the windows app stuck is as bad as windows 8 or GFW live steam isn't in any real danger. XBL competes directly with valve in style, but it's also too expensive for most indie devs, which is the point of app stores.
Microsoft going to have distribution deals with every developer
having had the unfortunate 'pleasure' of working with a team of people doing business with valve, valve very much builds deals with individual developers. It's basically the same deal for everyone, 30% steam cut, but they do talk to you individually, sometimes not very nicely (and sometimes deservedly so). I suspect MS will have to automate everything and not really build individual relationships. This gives them an advantage in volume, and ease from a developer perspective, but it's worse from an end user perspective.
And ya, steam has everything except blizzard and EA's top tier games, and a few MMO's (because if you're an MMO maker steam is really not something you want to be on if you can avoid it). I see them as continuing to be the premium game and online service provider. I wouldn't be surprised if they lose out on their indie business largely - but even there they can pick up the premium versions.
It will be interesting to see if Microsoft bans bundling steam into games on the windows store. Gamersgate did that for a while, but had to give up on that plan.
I've had windows live bundled with various games for a while, but yes, MS has managed to seriously screw up pretty much every attempt at an online store for consumers so far.
I think he means with the windows 8 store. Valve is desperately (not necessarily correctly) trying to find something to keep them alive when the windows 8 app store rolls around. They are thinking (possibly completely wrongly) that people won't want to use steam when there's an official MS store.
I suspect they're wrong though, I suspect that the windows store will end up full of crap, including apps for webpages and that nonsense, and valve can own a chunk of the premium store market for windows (and linux).
People actually use readyboost? For what? I guess if you have a gig of ram you might, but otherwise readboost won't ever do anything for you. That's kind of the thrust of some of the SSD tools is to use them as a fast cache (like readyboost but at the HDD rather than RAM level), and that's not a bad idea. I believe windows 8 was going to support that natively.
But the windows caching algorithm seems to just be that it holds the last X GB of data in ram and doesn't ever actually guess what you're going to use next. it might just be that mine can't guess accurately because I use a lot of different stuff, but the algorithm seems pretty simplistic.
A raid 0 HDD is, at best going to be like half an SSD in performance, but it's a lot less money.
And yes, I write games for a living, and games *do* load faster from an SSD. But it depends on the game, and whether or not windows has cached it. My perennial favourites (that I didn't work on) the hearts of iron/victoria/europa universalis games all benefit a lot from an SSD on startup time, but they're all pretty small (under a gig each without mods) so they tend to stay in RAM cache, so the 'load time' benefit is only seen once after a reboot.
Something like WoW or SWTOR, well, it depends. My SSD is certainly faster than my HDD. The question is, when it's about 2x the throughput is it more than 2x as fast due to latency reduction? And the answer is..... no. Those two are mostly throughput limited. So yes, a SSD at 400MB/s will completely blow away even a 5 disk RAID 5 setup, but it's not the latency making the difference.
Again, depends what you do. If you play both WoW and SWTOR, and my aforementioned paradox titles then sure, you could daily see a benefit from the SSD latency difference, but that's still mostly overshadowed by it having easily double the throughput of a cheap raid.
Why do people buy better CPU's and more RAM? Most do it to cut down on the time they have to sit and wait on the computer. SSD's can cut that a lot. Really a lot.
yes, I'm not saying SSD's aren't better than traditional HDD's. They're even better than most raid setups. It's that they aren't enough better for every problem to justify the cost and the hassle of managing one small drive and one big drive. Depends on what you do.
It depends very much on your usage scenario. I said myself I have a 240 gig SSD. I just don't get a lot out of it because most of what I do is large files with programs that I opened a week ago. Speeding up how fast windows does things that don't impact the user experience doesn't matter to me, but it really depends on what you're doing what things will matter to the user experience.
Ya, patch tuesday for windows machines and not even all of those.
It drives me crazy that flash insists on checking for updates on startup, but not the rest of the time, opera only notifies me of updates when I close it and reopen, and a few other things like that. I'm sure a lot of it is configurable better, but because it happens so rarely I usually forget, and then get annoyed when I found out there was an update a week ago that I didn't get a notification for.
I'm well aware of seek times. And honestly, they don't matter all that much. For a very small file they take you from 1-2 seconds to effectively instant yes, but for a significant file you're throughput limiting yourself anyway.
I guess it depends on the types of files you access. If you're regularly hitting small files, seek times matter, especially large collections of smallf iles, if you're regularly hitting big files, not as much. I do a lot more with 'big' (70-100MB files) than I do with small ones.
A pair of drives in a RAID isn't as good as an SSD in performance. But for 60 bucks you still only have one blob of 1.5 TB, compared to 100 bucks for a lone SSD that would be lucky to be 120 GB. At some point you're trading your time to manage the data on your SSD versus the time to access the data on the SSD, and how much money that's worth to you.
I've got a pair of windows 7 machines, a windows 8 machine and a pair of ubuntu machines between me and the GF. When you factor in the 10 or 15 seconds it can take for the bios to do its thing, and then you need to login, 35 seconds from cold isn't really very long. On my x58 based system the best I can do is 19 seconds from cold to windows 7 fully started, including typing my password. Of that about 10 is purely bios stuff. I haven't timed out ubuntu, but it's not much different with SSD's from windows 7. Windows 8 is a tiny bit faster, but on a machine with a bios it's not much different.
If you want to load a desktop full of icons, and power on all of the various USB and networked devices I have (and discover all of those), and then have any sort of graphical UI, language packs etc on a high res display there isn't much you can do to make things blazing fast other than improve the HDD speed.
In the old days when you had lower res displays, less graphical clutter (sorry, themes), less stuff trying to be automatically discovered on both your computer and network it could be faster. But for the vast majority of people an extra second or two to automatically discover the printer, the media server, to scan all the USB devices and so on before giving them the chance to do stuff is better than having them try and load it all real time. We didn't start throwing in multiple gigabytes of memory to not use it.
You're missing out then. I picked up a 240 GB ocz vertex 3 for 180 bucks, and in the right MOBO (one that can actually do full speed) it makes a huge difference to the user experience.
I had a 120 GB SSD, and ya, that, between development tools, windows, office productivity and 1 game (whichever game I was playing the most) I was chronically pushing my luck on space. 240 is enough that it behaves pretty well. I don't use it for a 'data' drive, for that I use a RAID 1 traditional drive, which is a combination of capacity and convenience. It's not all that hard to replace anything on the drive if it fails (reinstall from disks, copy over a directory from another machine), and the boost in day to day performance on web browsers, office tools etc. makes a big difference. Not because waiting 10 seconds to load Word or Powerpoint or whatever should actually be a problem, but after about 3 seconds I've opened a web browser and started reading/. and reddit etc. and then commending on/. and then 10 minutes later I've wasted 9 minutes and 53 seconds because of an extra 7 second loading time.
That was true back when you were talking about MB numbers. It's definitely not true when talking in that rage of GB.
To use windows as an example, it will try and cache what it thinks you're going to load based on I think a fairly simple algorithm. Which means it's usually wrong. If you never access more than 32 GB worth of data from your HDD then sure, 32 GB of Ram will do the trick. But, for example, if you play WoW or SWTOR (both of which flutter around 20GB), any other game + web browser + windows you could fairly easily waltz past 32 GB of data, at which point you're into 'cache misses'. And yes, this is conceptually the exact same problem as cache hit ratios, just working at a different level (logical files or directories rather than lines of memory).
I had virtually no performance increase going from 12 to 24 GB of RAM on general disk use.
You can get a big boost from an SSD, and especially, getting something that will actually work a SATAIII connection at full speed. My x58 board is lucky to pull more than 200MB/s from even a very good SSD, whereas the same drive on a sandy bridge board will do 450-550 range.
Now keep in mind, a regular HDD is about 70MB/s for sustained data. Put that in a raid 1 (mobo hardware, or software) and you can see 120-130, so an SSD on a bad connection may not be that much better than much less expensive RAID.
You can get insurance for anything. Whether or not you are smart enough to get enough insurance is another matter entirely.
Even the whole Facebook NASDAQ thing and there's an insurance fund to cover those problems, which were technical but not HFT. Now the insurance fund isn't big enough for that sort of problem, and I'm not sure how exactly it's supposed to work given that, but people who have small losses like that would have been dealt with already.
Slashdot Mirror
It's not an argument. It is. The game is a multiplayer game. Just because that's a stupid idea doesn't mean it isn't the one they went with.
I'm sorry that your point of view is just wrong. But it is. The whole game was balanced around you being able to buy and sell from the auction house. That was a deliberate choice on blizzards part, and without the AH the game becomes prohibitively hard because you just can't get the right itemized gear and you need an astronomical amount of farming to get through the content. Again, I'm not saying that's a *good* design, but that is the design. If anything the game suffers because you almost never loot anything you actually want, I think I looted one inferno difficulty item I actually used, all of the rest I had to buy.
They certainly could have designed the itemization differently or had a full on single player mode with different itemization. But they didn't.
The 'core activity' of diablo is 'click'. I'll grant you that activity is mostly unchanged form previous versions. But most games are more than just one core activity.
they wouldnt even work with me on a refund, when I had issues 3 weeks after launch because I pre ordered it, and therefore it was more than 30 days out of date, eventhough i only had the game for aweek less than 30 days.
yes well, that's a whole other topic. But once they have your money they don't want to give it back.
That's actually pretty common when people do get hacked. If you have gold they immediately mail it off and sell it, and then try and bot farm whatever the best gold/hour is. That might be tradeskilling, that might be cash runs through bosses, sort of depended.
My lingering suspicions is that WoW was vulnerable to a session spoof attack at some point, or the usual exploit of a flash vulnerability to get your password, but their systems became overall pretty robust with authenticators added in.
In your case I'd guess a flash vulnerability, possibly a 0 day one, those are much less of a problem today than they were 2 or 3 years ago when browsers weren't well sandboxed etc. etc. But those sorts of things always got a few people.
Of the 56 unique players in my guild when we quit, only 2 had ever been hacked. We've certainly had people who were hacked off and on over time, (and most of them left the guild) but once they brought in authenticators it was pretty rare for people to get hacked. Even before that, you usually had to do something stupid to get your account hacked.
The most common culprits for it were from re-using passwords (especially on WoW fansites, because duh...) and people buying gold. Then there was the usual keyloggers and so on.
Diablo 3 is a multiplayer game with a where you can choose to not directly interact with other players, but without the auction house the whole itemization would need to be completely different.
That was one of the things they realized with D2, the reason it stuck around was the multiplayer, they just got the idea that the whole thing should be multiplayer. starcraft has less of an excuse because there's no meta economy in starcraft.
Since I''m over 25 and work for a living
making you the target market for games, and modern MMO's. Especially so if you're male. Because you know, the people who actually work at blizzard want to play their own game, and they're mostly over 25 and have jobs. So if you're one of the 40 million or so people who ever created a battle.net account for starcraft or diablo or WoW then yes, this effects you. Because what was your security question, have you ever reused it, and was it publicly available information?
They said they're working on a change to the security question.
But yes, in general this is bad. Although that's sort of the idea behind salting and hashing passwords, that even if someone gets the passwords they still can't recover them.
I take exception to sensationalist reporting that blames the vehicle rather than the driver when the driver is clearly at fault
I think part of the problem is that like any software that fails, sometimes it's hard to tell if it's the user doing something stupid, or a flawed design or implementation, and since the media enjoys fear mongering anything that suggests the car you are driving might be unsafe is far more interesting than a person accidentally hitting the accelerator.
Then you get into broader cultural and societal issues. That we blinding accept speeding, or that the US accepts pervasive private gun ownership as a good idea all contribute to the statistical nature of problems. It's not so much that the car was designed to cause an accident, but someone made a deliberate choice that it only needed to pass whatever inspection regime, only needed to withstand whatever impact, that so many deaths due to accidental discharges, that so many deaths due to crazy people getting guns etc. etc. etc. are all acceptable. Societies as a whole (through governments) make deliberate choices that cars should only be sold if they meet particular criteria for example, there is, in that processes, the implicit acceptance that a certain number of people will die due to car failures that weren't the fault of the end user. Choosing that tolerance is an often buried part of governance.
Oh and people like to sue in the US, so if you can blame the car company and get millions it's like winning the lottery without the gambling.
That's the designer/manufacturer/operator's fault. I.e. people.
And? Once the device is built whatever caused the state it's in, if it fails in a way the end user in incapable of having prevented with archaic expertise it's the fault of the device, and the production chain that led to the device, or the business climate that let it get created (see support for IE6), which ceases being the fault of the individual user, and the fault of many people involved in the chain, from regulators, to physical assembly and distribution.
It doesn't seem to have changed their peak concurrent users any. They might have lost marginal customers, but they seem to have retained their core base.
At least not when you're measuring user numbers in the ~4-5 million range. Whether it cost 100k users here or there is hard to tell.
There's a rating system and apps are sorted into categories.
yep, and that's problematic, because like the iOS and Google play store, if you aren't on the front page you may as well not exist. Which is the whole problem with the app store model.
Which doesn't even work well for those of us in the games business. Gaming is actually a series of specialized niches, telling everyone they can go buy angry birds is fine, but if you want to sell prototype 2, or legends of pegasus you're never going to match angry birds on the chart, even if you could have a very successful game. Which is where steam comes in. If you want games for a 9 year old, steam isn't your place. If you want shitty mobile games, steam isn't the place to be. So right off the bat steam is a selection pressure on the market, you have to be good enough to warrant listing on steam in the first place, and you're not burying 60 million dollar games with 600 dollar games (production costs).
Adding in two dozen categories isn't going to help either. Steam and its ilk are useful because they roll through releases reasonably quickly and reasonably well, but you can still sensibly organize stuff. Although even that is becoming a problem with 2000 games on steam as we get into having games that are from 2002-2003 still prominently listed even though they don't really belong there.
That's why I say it's going to suck at being an app store: because they all suck at it. Even, as I say, something like steam, where you start adding in DLC and more than a couple of years of history and it becomes very cluttered and very hard to find something decent that was released 2 years ago unless you know exactly what you're looking for.
and the apps themselves won't connect you to other users
This one I definitely wasn't clear on. Steam connects you to your gamer friends, regardless of what game you're playing. XBL and PSN do the same thing. GFW actually is XBL, but the question becomes on of whether or not apps in any sort of sensible way will bundle in functionality, or if they'll all try and talk to each other reasonably well. With the windows app store you *might* be able to get into a contact list, you might not, you might use XBL/GFW, they might let you install steam, it might store lists on its own and now all the matchmaking services (for gaming) are going to try and bundle. It's going to go from 'use steam like everyone else' to an ungodly mess of dozens of companies all trying to be the matchmaking service. Which in and of itself isn't the end of the world, it's just going to be confusing compared to steam.
I'd like to point out where you are clearly talking out of your ass:
I'd like to point out that I'm in the process of trying to re-package a (non steamworks) game for windows 8, and confronting all of these problems, and more. And we're working on a release coming up that will have to support windows 8 somehow. Unfortunately we're a niche title (because gaming is a collection of niches), where 50-60K copies would be very successful, and it's not that we see windows 8 as a problem, we'll just use steam and gamersgate and impulse and .... and skip over the windows 8 store, because it's not adding value to us, and if anything it's just making things harder for our users. Not that I'm in charge of business decisions so the boss guy could decide he wants to go all in on windows 8 or steam or whatever.
Cars don't cause accidents. Guns don't kill people. IDEs don't write bugs.
if only. Then we'd never have recalls on cars, never have to patch bugs in IDE's and never have never have guns that accidentally discharge.
2012 jeep review in sweden the vehicle is consistently blowing tyres and nearly rolled over once. Most of the video is trying to figure out why it nearly rolled over once but not on subsequent attempts. But the tyres blowing at 70-80Km/h is uh... bad. Really really bad.
Bugs fixed in Visual studio 2012 some of this stuff goes back years and has to be compared against the c++ STL precisely because the way it does work, and the way it should work are not the same.
Don't get me started on years of various nVIDIA and AMD tools not playing nice with OpenGl or Directx.
Wiki on accidental discharge lists two scenarios, where a weapon is dropped, or when a weapon overheats that it can accidentally fire if it was improperly designed (e.g. a poor choice of materials).
So yes, cars themselves can cause accidents, guns do kill people, and IDE's can cause bugs in your code.
That's my point. They seem to be spending a huge pile of money, and exerting a huge effort in trying to hedge against windows 8, though I'm not convinced they're in any real risk.
Then again, the reason you invest in a hedge like that is because if you guessed wrong and didn't have a plan B you'd be out of business.
The difference is that the average user can't see what previously bundled products did badly, and developers kept supporting it.
Windows 8 is different. From the moment you boot it up on a desktop you can see what's wrong with it. You try and use anything, and you can see what is wrong with it. With the app store it will probably suck at actually being an app store, you won't be able to find good apps easily (unless you know the name of what you want), and the apps themselves won't connect you to other users, which is critical to a multiplayer environment.
I'm sure microsoft will *try* to put other people out of business, but if the windows app stuck is as bad as windows 8 or GFW live steam isn't in any real danger. XBL competes directly with valve in style, but it's also too expensive for most indie devs, which is the point of app stores.
Microsoft going to have distribution deals with every developer
having had the unfortunate 'pleasure' of working with a team of people doing business with valve, valve very much builds deals with individual developers. It's basically the same deal for everyone, 30% steam cut, but they do talk to you individually, sometimes not very nicely (and sometimes deservedly so). I suspect MS will have to automate everything and not really build individual relationships. This gives them an advantage in volume, and ease from a developer perspective, but it's worse from an end user perspective.
And ya, steam has everything except blizzard and EA's top tier games, and a few MMO's (because if you're an MMO maker steam is really not something you want to be on if you can avoid it). I see them as continuing to be the premium game and online service provider. I wouldn't be surprised if they lose out on their indie business largely - but even there they can pick up the premium versions.
It will be interesting to see if Microsoft bans bundling steam into games on the windows store. Gamersgate did that for a while, but had to give up on that plan.
I've had windows live bundled with various games for a while, but yes, MS has managed to seriously screw up pretty much every attempt at an online store for consumers so far.
I think he means with the windows 8 store. Valve is desperately (not necessarily correctly) trying to find something to keep them alive when the windows 8 app store rolls around. They are thinking (possibly completely wrongly) that people won't want to use steam when there's an official MS store.
I suspect they're wrong though, I suspect that the windows store will end up full of crap, including apps for webpages and that nonsense, and valve can own a chunk of the premium store market for windows (and linux).
People actually use readyboost? For what? I guess if you have a gig of ram you might, but otherwise readboost won't ever do anything for you. That's kind of the thrust of some of the SSD tools is to use them as a fast cache (like readyboost but at the HDD rather than RAM level), and that's not a bad idea. I believe windows 8 was going to support that natively.
But the windows caching algorithm seems to just be that it holds the last X GB of data in ram and doesn't ever actually guess what you're going to use next. it might just be that mine can't guess accurately because I use a lot of different stuff, but the algorithm seems pretty simplistic.
A raid 0 HDD is, at best going to be like half an SSD in performance, but it's a lot less money.
And yes, I write games for a living, and games *do* load faster from an SSD. But it depends on the game, and whether or not windows has cached it. My perennial favourites (that I didn't work on) the hearts of iron/victoria/europa universalis games all benefit a lot from an SSD on startup time, but they're all pretty small (under a gig each without mods) so they tend to stay in RAM cache, so the 'load time' benefit is only seen once after a reboot.
Something like WoW or SWTOR, well, it depends. My SSD is certainly faster than my HDD. The question is, when it's about 2x the throughput is it more than 2x as fast due to latency reduction? And the answer is..... no. Those two are mostly throughput limited. So yes, a SSD at 400MB/s will completely blow away even a 5 disk RAID 5 setup, but it's not the latency making the difference.
Again, depends what you do. If you play both WoW and SWTOR, and my aforementioned paradox titles then sure, you could daily see a benefit from the SSD latency difference, but that's still mostly overshadowed by it having easily double the throughput of a cheap raid.
Why do people buy better CPU's and more RAM? Most do it to cut down on the time they have to sit and wait on the computer. SSD's can cut that a lot. Really a lot.
yes, I'm not saying SSD's aren't better than traditional HDD's. They're even better than most raid setups. It's that they aren't enough better for every problem to justify the cost and the hassle of managing one small drive and one big drive. Depends on what you do.
It depends very much on your usage scenario. I said myself I have a 240 gig SSD. I just don't get a lot out of it because most of what I do is large files with programs that I opened a week ago. Speeding up how fast windows does things that don't impact the user experience doesn't matter to me, but it really depends on what you're doing what things will matter to the user experience.
Ya, patch tuesday for windows machines and not even all of those.
It drives me crazy that flash insists on checking for updates on startup, but not the rest of the time, opera only notifies me of updates when I close it and reopen, and a few other things like that. I'm sure a lot of it is configurable better, but because it happens so rarely I usually forget, and then get annoyed when I found out there was an update a week ago that I didn't get a notification for.
I'm well aware of seek times. And honestly, they don't matter all that much. For a very small file they take you from 1-2 seconds to effectively instant yes, but for a significant file you're throughput limiting yourself anyway.
I guess it depends on the types of files you access. If you're regularly hitting small files, seek times matter, especially large collections of smallf iles, if you're regularly hitting big files, not as much. I do a lot more with 'big' (70-100MB files) than I do with small ones.
A pair of drives in a RAID isn't as good as an SSD in performance. But for 60 bucks you still only have one blob of 1.5 TB, compared to 100 bucks for a lone SSD that would be lucky to be 120 GB. At some point you're trading your time to manage the data on your SSD versus the time to access the data on the SSD, and how much money that's worth to you.
I've got a pair of windows 7 machines, a windows 8 machine and a pair of ubuntu machines between me and the GF. When you factor in the 10 or 15 seconds it can take for the bios to do its thing, and then you need to login, 35 seconds from cold isn't really very long. On my x58 based system the best I can do is 19 seconds from cold to windows 7 fully started, including typing my password. Of that about 10 is purely bios stuff. I haven't timed out ubuntu, but it's not much different with SSD's from windows 7. Windows 8 is a tiny bit faster, but on a machine with a bios it's not much different.
If you want to load a desktop full of icons, and power on all of the various USB and networked devices I have (and discover all of those), and then have any sort of graphical UI, language packs etc on a high res display there isn't much you can do to make things blazing fast other than improve the HDD speed.
In the old days when you had lower res displays, less graphical clutter (sorry, themes), less stuff trying to be automatically discovered on both your computer and network it could be faster. But for the vast majority of people an extra second or two to automatically discover the printer, the media server, to scan all the USB devices and so on before giving them the chance to do stuff is better than having them try and load it all real time. We didn't start throwing in multiple gigabytes of memory to not use it.
You're missing out then. I picked up a 240 GB ocz vertex 3 for 180 bucks, and in the right MOBO (one that can actually do full speed) it makes a huge difference to the user experience.
I had a 120 GB SSD, and ya, that, between development tools, windows, office productivity and 1 game (whichever game I was playing the most) I was chronically pushing my luck on space. 240 is enough that it behaves pretty well. I don't use it for a 'data' drive, for that I use a RAID 1 traditional drive, which is a combination of capacity and convenience. It's not all that hard to replace anything on the drive if it fails (reinstall from disks, copy over a directory from another machine), and the boost in day to day performance on web browsers, office tools etc. makes a big difference. Not because waiting 10 seconds to load Word or Powerpoint or whatever should actually be a problem, but after about 3 seconds I've opened a web browser and started reading /. and reddit etc. and then commending on /. and then 10 minutes later I've wasted 9 minutes and 53 seconds because of an extra 7 second loading time.
That was true back when you were talking about MB numbers. It's definitely not true when talking in that rage of GB.
To use windows as an example, it will try and cache what it thinks you're going to load based on I think a fairly simple algorithm. Which means it's usually wrong. If you never access more than 32 GB worth of data from your HDD then sure, 32 GB of Ram will do the trick. But, for example, if you play WoW or SWTOR (both of which flutter around 20GB), any other game + web browser + windows you could fairly easily waltz past 32 GB of data, at which point you're into 'cache misses'. And yes, this is conceptually the exact same problem as cache hit ratios, just working at a different level (logical files or directories rather than lines of memory).
I had virtually no performance increase going from 12 to 24 GB of RAM on general disk use.
You can get a big boost from an SSD, and especially, getting something that will actually work a SATAIII connection at full speed. My x58 board is lucky to pull more than 200MB/s from even a very good SSD, whereas the same drive on a sandy bridge board will do 450-550 range.
Now keep in mind, a regular HDD is about 70MB/s for sustained data. Put that in a raid 1 (mobo hardware, or software) and you can see 120-130, so an SSD on a bad connection may not be that much better than much less expensive RAID.
You can get insurance for anything. Whether or not you are smart enough to get enough insurance is another matter entirely.
Even the whole Facebook NASDAQ thing and there's an insurance fund to cover those problems, which were technical but not HFT. Now the insurance fund isn't big enough for that sort of problem, and I'm not sure how exactly it's supposed to work given that, but people who have small losses like that would have been dealt with already.