If you put something in public domain, and it ends up say, as part of the coffee maker temperature controller on an airplane, malfunctions, and leads to loss of life.. you could be liable.
On what basis?
If you haven't received anything in return or given any sort of advice, commitment or guarantee, and someone chooses to use something that you happened to have released to the public domain in an inappropriate way, under what law(s) in what jurisdiction(s) can that affect you?
As an aside, in many places you can't effectively disclaim any liability you do have for things like causing loss of life anyway, and in some places you're likely to get a pretty rough time in court if you tried.
Please note that the "Steam's lock-in isn't that powerful" was one possibility, in contrast to the other possibility that the lock-in is powerful, not a statement of fact.
That said, gamers tend to upgrade their systems relatively quickly. In fact, with the average laptop today powerful enough to run e-mail, browsing or typical office software many times over, gamers are one of the few large markets with a real incentive to keep upgrading their hardware. Sooner or later -- probably sooner -- gamers will want to upgrade to something that supports DirectX 19,254 and 512x supercompositional linearised antialiasing. Of course a lot of them won't know what these buzzword technologies actually do for them, but the comparison screenshots will be pretty.:-)
That means all Microsoft has to do to push Windows 8 is to make new technologies that new games want to use into Windows 8 exclusives. They've done the same thing with Internet Explorer (can't get IE9 on Windows XP) and past versions of DirectX, after all. If they do that, and the new technologies are worth having, then large numbers of gamers will upgrade within a fairly small number of years to get the latest and greatest AAA titles, and after a while that old library won't look so appealing.
I'd be the first to agree that GFWL is not a strong offering right now. Then again, not so very long ago, the gaming community had a low opinion of a new service called Steam. The technology world moves fast, and while many markets can be fickle, gamers practically make a hobby out of changing their preferences every five minutes as fashions come and go.
Likewise, I agree that Microsoft aren't producing the top-level games on the Windows platform that perhaps they (and their subsidiary studios) did a few years ago. But they have the demonstrated capacity to do that, and I suspect they've still got the talent around if they choose to throw a huge wave of resources into it. MS have a long history of getting left behind in new or fast-moving markets but then shifting their entire culture to catch up if they decide it's the right way to go. Sometimes they succeed, sometimes they don't, but it's a brave man who would count them out before the starting bell.
The key point I'm really trying to make in this whole discussion is that Steam doesn't have some sort of highly defensible incumbent's position here. They are essentially a middleman, and middlemen introduce inefficiency that tends to get cut out eventually in competitive markets. I'm not saying their business is in any danger of failing tomorrow, but I think Gabe is right to be concerned (from his business's point of view) about the direction Microsoft is heading with Windows 8.
Because MS will kill Steam in order to boost its own platform with no games on it. That makes perfect sense.
You don't seem to understand the technical details here at all.
Games don't run on Steam. They run on Windows, or MacOS, or Linux, or whatever. Steam is just a distribution channel, and a bunch of related tools. All of the AAA Windows games that you can get on Steam today run on Microsoft's platform, every single one of them, and they could continue to do so even if Steam were completely removed from the picture.
Again, the reason you're talking shit is because you're making this all up. Windows 8's store doesn't have ANY advantage, none whatsoever. The ONLY thing they can do is lock down other players
You just contradicted your own statement.
and it is more likely to end up in a lawsuit
So a few people keep saying, but I don't see why. No-one's brought down Apple's store for having a lock on iOS apps. No-one's brought down Amazon for trying to establish a dominant end-to-end channel for ebooks. Where is this lawsuit you envisage coming from if Microsoft decide to lock down Windows 8, or any future versions of Windows, in the same way?
Steam will still start off with a better user base, an established brand, SteamWorks, all of VALVe's titles plus multiple AAA ones, and again, they'll have a Windows client, a Mac OS X client, and soon, a Linux client.
OK, let's be really absurd and actually look at the facts here.
A better user base: Everyone who is running Steam on Windows is already using Microsoft's platform.
An established brand: Microsoft have their own games console. They also have a long list of the most successful games in PC history under their belt, not least because these days they are the parent company behind several big name studios. By several sensible metrics, they are probably the most successful computer game company in the world today.
SteamWorks: Do you really think Microsoft couldn't build something equivalent? They already have a bunch of related technologies.
All of Valve's titles plus multiple AAA ones: Well, if the platform gets locked down, then Valve will either release its titles via whatever other channel is available or not release them at all, and so will everyone else.
Windows/OS X/Linux clients: Sure, but that doesn't magically mean everyone's games will run on all those different platforms. Hardly any AAA titles are available on Mac Steam right now, and Linux Steam is about on par with Linux gaming generally. Once again, games don't run on Steam, they run on the underlying OS. Steam is just a tool, and an entirely expendable one if anyone else produces a distribution platform with similar capabilities.
VALVe already has to deal with competition (Origin, GFWL), and Windows 8's upcoming app store will be no different.
Of course it could be different. It's operated by the company who own the underlying platform. They therefore have the power to do more or less anything, up to and including locking down competing products that would run on their platform or, perhaps more likely, imposing heavy fees as other channel-controlling vendors like Apple do.
To be clear, I'm not saying they are going to do this immediately with Windows 8, though they are certainly taking big steps in that direction already. But if you were running a company with a service like Steam, you'd have to seriously consider how future-proof your business model was at this point.
Anonymous Brave Guy is basically talking shit. There are way more reasons to use Steam as a distribution platform than the distribution channel
Such as what, exactly?
SteamWorks, a native Windows, Mac and Linux client, brand recognition in the gaming industry, there are many reasons to use Steam over anything else.
And exactly how many AAA titles are available via Steam on Apple, excluding Valve's own? (That's an easy question. You can almost count them on the fingers of one hand.)
Given that it takes dozens or even hundreds of people multiple years and multi-million budgets to develop a modern AAA game, I find it amazing how many people posting in all the discussions like this one today think those games are magically going to be ported to run on other platforms just because of an off-hand comment by Gabe. They don't get ported right now, even to Mac, where Steam is already available. They won't just magically work on Linux either. I'd be the first one to cheer if all those titles did suddenly become available on Linux, preferably via some reasonable and not offensively DRM'd channel, but achieving that would be a vast undertaking, and it's not clear that even Valve/Steam have the power to make it happen.
So I contend that the advantages of your "reasons to use Steam over anything else" are often illusory/wishful thinking.
Well, Windows 8 isn't out yet, so everything any of us says here is a moderately informed guess at best.
With that said, MS are already locking down Windows 8 on ARM. They are already locking down Metro. They are already restricting the use of their inexpensive/hobbyist developer tools to target the grown up UI. That's going to affect a lot of Windows 8 applications, potentially including a lot of games, which Steam won't be able to support as things stand.
But the thing that gets me is that this is a pretty clear change of direction from Microsoft, a statement of intent. If they want to continue that trend and lock down the grown up Windows UI as well, which is hardly implausible given the above and how extraordinarily successful Apple have been with that approach, there is absolutely nothing Valve can do to stop them. Like the disk compression and anti-virus vendors of yesteryear, their market is at risk of being obliterated as the platform itself starts to offer equivalent functionality and with a home ground advantage.
In short, Steam's catalogue disappearing overnight isn't the issue. Whether Valve even have access to install games from that catalogue on their customers' new computers is. Windows 8 is already taking steps -- and they are big steps -- in the direction of locking down the platform. Even if you can still develop and side-load today's games into the main UI under Windows 8, it would be a brave business that bet its future on that continuing in future Windows versions. And if it turns out that a lot of people who have spent a lot of money buying games on Steam don't have as much portability as they thought, because MS pulled the rug out from under Valve, then a lot of the goodwill Steam has accumulated recently is going to evaporate in a heartbeat, just like every other subscription/library/DRM-locked multimedia service that looked great right up until its operators pulled the plug.
The problem for Valve is that Windows 8 is going down the app store route, and the main point of Steam is really to be an easy download and auto-update platform for games. Sure, Steam does other things too, but if it weren't for the distribution channel (which is the only distribution channel for Valve's own big name games) I don't really believe anyone would stick with it just for the minor perks. This leaves only two possibilities:
1. Steam has a powerful lock-in. In this case, a lot of people who have spent a lot of money with them based basically on trust are about to have their faith questioned. Since Steam's standards terms and conditions are a joke as far as guaranteeing anything to anyone but themselves, this leaves two variations:
1a. They will do right by their customers at almost any price, assuming this is even possible with whatever technical and commercial infrastructure MS adopts to go with Windows 8. This might save their reputation and business model, but would surely hurt Valve's bottom line significantly.
1b. They can't or won't pay that price and customers who move to Windows 8 will suffer a worse user experience, limited ability to buy new games, or in the worst case lose access to the existing library they've already paid for. In any case, Steam will take a huge PR hit that will at best severely damage Valve's credibility.
2. Steam's lock-in isn't that powerful. In this case, Microsoft can beat them at their own game (no pun intended) and outright steal their business.
There are exactly zero outcomes in there that are positive for Steam, and some represent an existential threat.
Though they don't appear to be too bothered further down on Mill Road, where cyclists ride at night without lights with impunity, run red lights, and travel at high speed along the inside of cars ready to smack into the side of cars turning out of junctions and turning left. Jumping off pavements without looking etc...
For what little it's worth, it's not with impunity. Mill Road has one of the worst accident records of any road in Cambridge, and many of the people who get hurt or worse were on bikes.
The idiots who overtake cyclists over the blind summit on the bridge do occasionally run into (very literally) someone similarly idiotic coming the other way, too.
Basically, what Cederic said. A NAS is a very specific piece of equipment, designed and engineered for a specific task, that of storing data/programs in a location reachable on a network.
Who said anything about building their own? Though building even quite substantial servers is easy these days if you're used to DIY PCs, it's not for everyone and there are also plenty of places who will build you a custom server box with the right kind of hardware. And of course you can buy a server of just about any scale off-the-shelf from a vendor like HP or Dell if you're willing to spend the money on it or if you need to place really large orders. Since you work in a bank, you might actually be in that position, though of course most businesses won't be.
Also, don't your servers also need things like resilience, fail-over, throughput, and low power consumption? And even if someone did a custom build themselves or bought one in, why would you assume it would somehow be inferior to a pre-built NAS box in features and/or reliability?
In a few minutes, I'm going back to work on a web-based project that uses several parts of both HTML5 and CSS3, and which runs just fine on IE9.
Perhaps you mean some subset of HTML5 and CSS3, possibly one that also isn't actually standardised yet, and that isn't supported reliably and portably by other browsers either, and where as a result few real-world projects are actually using the features yet anyway?
But even if you only want lots of storage, you can buy anything up to and including rack-mount servers with lots of hot-swap drive bays and redundant power/networking easily enough, and then set them up however you want using standard tools.
That's basically the conclusion we came to as well, though any backup plan is always a matter of risk and how much you're prepared to spend to mitigate how much potential danger, and you have to draw the line at some point.
Still, if a backup service isn't even claiming to offer sensible guarantees that they will actually provide the service advertised, that doesn't seem like a great starting point. We decided that in the current economic climate, there is a significant risk of any business failing or any on-line service being closed down as the result of some commercial transaction involving the company that operates it.
FWIW, in the end, we wound up using a combination of:
(a) some basic off-line storage procedures involving encrypted removable media, which we update every few days (or sometimes more frequently) based on how much has happened since the last time, and
(b) some near-real-time on-line storage using a small local hosting firm whose people we have personally met.
We're happy that the odds of losing the on-line storage permanently are fairly slim. If that does happen, or if there's a major connectivity problem, then there's always a limit to how much data we'll lose access to, because we have the physical media as well. The whole set-up is fully encrypted on our side, and the total cost including overheads seems to be about the same as renting a bunch of space on something like Amazon's web services, and significantly less than subscribing to a lot of the dedicated on-line backup services.
My view of on-line backup services has wound up similar to my view of NAS devices: they provide a valuable option for people who don't have the technical expertise to go with something better, but if you do have that expertise, I just don't see what advantages the dedicated backup services really offer in return for the lock-in, limited flexibility, and typically higher costs. Again, maybe there are certain sizes/types of business where the cost/benefit trade-off looks different, but I haven't personally encountered one.
Just in case anyone else reading this thread is curious, that's a "let me Google that for you" link for the phrase "enterprise class nas devices".
I don't normally feed trolls, but I did take a look at a few results from the first page of search results, just in case I really had completely missed something. The top result goes to a page where the animated "New!" GIFs flash red and yellow. For those of you young enough that you only remember slapping "Beta" graphics on every web page you ever made, flashing graphics are what we used to mock before that, along with the blink tag and scrolling marquees.
But let's be fair and focus on the content. I followed the link to the first product described on that first page, which would surely put my recent experience in perspective. And so it did. The first device there offers a magnificent 2TB of storage in only 4U of rack space, based on a combination of 8x250GB drives. It also comes with such advanced features as an expandable RAID array, in case you don't want the full 2TB initially I assume, and automatically synchronising data with other devices of the same type. It even works on heterogeneous networks with PCs, Macs and Linux boxes! Truly, this is in a different class to the hardware I'm working with today. In fact, even our entry-level workstations can match that kind of spec, never mind the servers.
Just to be really fair, I did also check out some of the serious hardware that appears further down the page. Some of them make compelling sales pitches. If you buy a Thecus N16000, then "Whether rain, shine, or Armageddon, your data will always be there" apparently.
When all is said and done, I'm still waiting to find a counter-example to my position that these are essentially just inflexible servers that happen to be built in cases with lots of hot-swap disk enclosures, though. There is nothing in even the highest-spec devices I found on the first results page that you couldn't also achieve with a general server. There appear to be similar concerns about, for example, integrating with only a very limited number of vendor-chosen back-up services as standard, just like we had with the Cisco box we bought. And as it turns out, despite being only a "small business" product, the Cisco NAS was actually pretty close to the feature spec of a lot of these so-called enterprise-class devices, it just has a smaller number of drive bays and not as much redundancy in the power and networking.
So I stand by my earlier questions: who is the market for these high-end NAS devices, and what advantages does this kind of device offer given that anyone buying one of these boxes presumably has the expertise in-house to configure a real server anyway and will surely need to do a lot of manual configuration either way to integrate with the rest of the network?
Of course, for the OP, who was asking about storage solutions for a small business with around 10-20 staff, this is all rather academic, because they'd be looking at an entry-level NAS or an equivalent server anyway. So I stand by my original advice there as well: as long as you have either moderate Linux sysadmin skills in-house or the budget to bring in suitable help occasionally, there is very little advantage to choosing a NAS over a real server with a similar spec, at least among the various options we've evaluated over the past couple of years.
So you needed a real server, and were surprised when a NAS didn't meet your requirements?
No, at the time we set the network up, we just needed some reliable mass storage. We assumed, foolishly as it turned out, that getting a NAS would be the quickest and easiest way to achieve that.
We soon discovered that setting up the NAS was no easier than setting up a real server. In fact, it was more awkward, in the sense that it's probably still using much the same Linux tools as a real server under the hood, but you can't use all the familiar sysadmin knowledge your team has because you have to configure the device using a web interface with pretty bar charts and non-standard "friendly" terminology and the occasional undocumented corner case.
And as requirements involve -- and requirements always evolve for IT in small businesses -- you can adapt or completely repurpose a general server, while a NAS is forever a NAS.
Perhaps you're right, and instead of a NAS, we should have bought a real server all along. But in that case, my question to you is: who should buy a NAS in the first place? The only answer I've got is "someone who doesn't know enough to set up/administer a real server and can't bring in help who does know enough", which might apply to a lot of home users but probably doesn't apply to anyone responsible for setting up IT facilities for a business.
Your snarky reply does not change the fact that a real server could do everything the NAS could do and many other things as well. A NAS is just a server with a lot of hard disks, a lot of limitations, and hopefully at least some decent preinstalled networking, access control and storage management tools so it has some vague semblance of value.
If you disagree then perhaps, instead of making vague allusions to some mysterious high-end kind of NAS, you would do us all the courtesy of stating specific models and specific features they have that a general server box of a similar specification couldn't match, so we can all have an informed discussion.
I thought that when you said "running a real business" you actually meant a real business. Who the fuck installs email, calendaring or DHCP servers on their NAS device?
It seems likely that quite a lot of people do, given that basically every NAS on the market comes with options to do at least some of those things, and the vendors' support forums are full of people trying to figure out how to root them to install the missing ones.
I notice that your selective quoting missed out the "scheduling regular, secure off-site back-ups" problem I also mentioned. Maybe real businesses like banks are above such things. Do you by any chance work for RBS?
But if cloud-based backups (especially automated, encrypted cloud-based backups) let us mitigate our disaster risk and cut out the oops-forgot-to-change-the-tape factor, they're the lesser of two evils.
Since you're a lawyer, I'll just strongly suggest here that you read the terms of any on-line backup service you're considering using with the same care that you would review a document for a client. We did, and despite not being lawyers, we decided pretty quickly that we wouldn't use any of the ones we were considering.
The clue was in the way they could typically shut down their respective services at about five minutes' notice with little if any guarantee that we would be able to retrieve backed-up data in the event of a disaster affecting us around the same time. Typically, all the useful mass-restore options like getting everything sent in the mail on DVDs/hard drives seem to be cut off the moment they announce that their service is closing down. Your only remaining option is usually to download every byte over an Internet connection within perhaps 72 hours of the notice being given, at a time when obviously every customer they have will be trying to do the same thing. That wouldn't even get you the full amount of data that the entry-level plans for these services claim to support, on a saturated fast leased line, with no contention for access to the servers.
I'd say a minimum acceptable requirement for an off-site backup service is enough notice of any change/closure that we could make other arrangements with continuous cover from one service or the other, and a meaningful and technically credible promise that we could get every byte of our data back before they closed down the service. If they aren't willing to give a contractual guarantee that they can do that, and demonstrate that they are maintaining the necessary resources to honour that commitment immediately if anything happens to their business, then how can they possibly be a trustworthy partner for large amounts of sensitive data? It's like an insurance policy that ends with the words "Policy may be arbitrarily and unpredictably void in the event of a claim being required". Sure, it's not likely that they my business will suffer a catastrophe within a couple of weeks of a backup service closing down, but then it's not likely that we'll suffer a catastrophe at all, is it? If we were going to pay a silly amount of money for robust off-site backups, we'd want some sort of legally and technically credible promises in return.
Can you give any concrete examples where IE9 doesn't work properly? Not new features in HTML5 and CSS3 that it doesn't support, but things where it's supposed to support them but is actually broken?
We also went through this a while ago, but the other way around. After kitting out a small office network, the one purchase we really regretted was the NAS (a Cisco-branded device, which in fact is a rebadged QNAP).
The hardware has not failed and supports hot-swapping drives if necessary, but those are about the only good things I have to say about this unit. It is in all other respects just a very limited and relatively expensive Linux server, where essential operations like scheduling regular, secure off-site back-ups are absurdly difficult, and where you can't easily install other server software (e-mail, calendars, DHCP, RADIUS, whatever) unless whoever supplied your NAS happens to make some sort of plug-in available for their particular style of firmware. Even Cisco gave up trying to provide any meaningful support in this area within a few months of the device launching, eventually just providing a mechanism for people to upgrade their firmware to QNAP's own.
When we were investigating options for a new device earlier this year, it looked like more recent NAS devices from other suppliers were little better, maybe differing in some of the details but essentially still the same old story.
My conclusion: NAS devices are for non-technical home users who want to plug in and go. If you're running a real business with serious requirements, and you have moderate Linux skills and/or a modest budget to bring in someone who does when you need them, then buy a real server with a specification suitable for your requirements. There is absolutely no advantage to buying a NAS for someone in that position, IME.
1. The cloud services you mentioned aren't even close to secure enough for legally sensitive documents.
2. Judges are unlikely to accept "my Internet connection was down" as a valid reason for not filing documentation properly.
3. Legal documents are written using serious software, not trivial web apps. They have numerous technical requirements and typographical conventions that must be strictly adhered to, in some cases to the point where courts will specify the precise font you must use for all submissions, for example. You don't write this sort of thing in Google Docs, where the concept of a cross-reference has yet to appear and the numbering styles available are one small step past "numbered" and "not numbered".
If you're genuinely interested, you shouldn't trust some random post on Slashdot anyway, you should go experiment for yourself and see how easy it is to break.
Here are a couple of starting points to try out. Both of these areas get broken all the time from direct personal experience and have easily observable problems with the current Chrome release:
Set up a site with an HTML5 video. Try serving the video in different encoding/container formats. Try using different combinations of attributes on the video element itself. There have been so many bugs in this area that you're bound to hit one if you're even a little creative. If you want concrete citations, just try searching for "chrome html5 video bug youtube".
Set up a site using a plug-in like Flash or Java. Have some JavaScript (which has a single-threaded execution model) do something obvious like showing a "please wait" message with some sort of animated GIF spinner graphic, and then immediately call into your Flash or applet to do something time consuming like, say, fetching a file from the server. Observe that the whole browser locks up while this is going on. Conclude that Chrome appears to be either executing external code directly in its main UI thread or at least failing to synchronize things properly. Never mind the reliability issues, this is an obvious security vulnerability, and ever-more-obnoxious blocking of plug-in-based content is not an acceptable "fix". It's also a newbie error that most desktop app developers learned not to make 10 or 20 years ago, so it's very surprising that an application with the support of Chrome would have such a poor architecture, but there you go.
Businesses are shunning it in droves. There is a striking difference today between browser usage patterns during office hours and during evenings/weekends.
The problem with updates for home users is that when one of the web apps I work on breaks because of a bad Chrome update, I can promise you it's not Google who get paged at 4am to deal with it. And the end user doesn't know where the problem is, because they don't even realise Chrome has been updating. All they know is that yesterday they could log in and see their account details and today they can't, and so they call the support number or file a bug report on the support page, and my life for the next few days gets more painful. How much do you think this makes me, as a developer, want to use any new browser feature that isn't likely to be 100% robust and future-proof?
Unfortunately, even the LTS releases aren't much use for a big business that runs numerous internal web apps.
For one thing, the period of support is still way too short: the first one, based on Firefox 3.6, already ran out some time ago. Moving new feature development to another branch after a few months is fine, but security updates have to last for the useful lifetime of the platform if you want to be taken seriously.
For another thing, the overlap between support for one LTS version and the next is only a few weeks, which is an absurdly short time to upgrade an entire organisation on the scale I'm talking about there. They do need to do extensive testing during that upgrade process, because for example the last Firefox LTS release seriously broke plug-ins like Flash and Java applets, which of course many of those in-house web apps rely on.
In short, while most of the small businesses I work with could cope fine with an annual upgrade cycle, the large businesses simply can't and won't. For them, having a stable, secure platform for the long term is very much more important than having the latest shiny HTML5 gimmicks.
You obviously feel passionately about the way Microsoft behaved in the past and the IE6 legacy.
I cannot help but notice that today's situation where Mozilla and Google are throwing new extensions around every few weeks looks a lot like the IE vs. Netscape problem repeating itself, except that at least IE6 won so all the users had the same client with the same proprietary extensions for those of us writing web pages to target.
Today, we don't even have that, and all this "living standard" madness seems like a guaranteed way to make sure we will never again have a stable, reliable platform to code against.
Slashdot Mirror
That's mostly just repeating what the other guy already said. My question is under which specific laws that liability would arise.
If you put something in public domain, and it ends up say, as part of the coffee maker temperature controller on an airplane, malfunctions, and leads to loss of life.. you could be liable.
On what basis?
If you haven't received anything in return or given any sort of advice, commitment or guarantee, and someone chooses to use something that you happened to have released to the public domain in an inappropriate way, under what law(s) in what jurisdiction(s) can that affect you?
As an aside, in many places you can't effectively disclaim any liability you do have for things like causing loss of life anyway, and in some places you're likely to get a pretty rough time in court if you tried.
Please note that the "Steam's lock-in isn't that powerful" was one possibility, in contrast to the other possibility that the lock-in is powerful, not a statement of fact.
That said, gamers tend to upgrade their systems relatively quickly. In fact, with the average laptop today powerful enough to run e-mail, browsing or typical office software many times over, gamers are one of the few large markets with a real incentive to keep upgrading their hardware. Sooner or later -- probably sooner -- gamers will want to upgrade to something that supports DirectX 19,254 and 512x supercompositional linearised antialiasing. Of course a lot of them won't know what these buzzword technologies actually do for them, but the comparison screenshots will be pretty. :-)
That means all Microsoft has to do to push Windows 8 is to make new technologies that new games want to use into Windows 8 exclusives. They've done the same thing with Internet Explorer (can't get IE9 on Windows XP) and past versions of DirectX, after all. If they do that, and the new technologies are worth having, then large numbers of gamers will upgrade within a fairly small number of years to get the latest and greatest AAA titles, and after a while that old library won't look so appealing.
I'd be the first to agree that GFWL is not a strong offering right now. Then again, not so very long ago, the gaming community had a low opinion of a new service called Steam. The technology world moves fast, and while many markets can be fickle, gamers practically make a hobby out of changing their preferences every five minutes as fashions come and go.
Likewise, I agree that Microsoft aren't producing the top-level games on the Windows platform that perhaps they (and their subsidiary studios) did a few years ago. But they have the demonstrated capacity to do that, and I suspect they've still got the talent around if they choose to throw a huge wave of resources into it. MS have a long history of getting left behind in new or fast-moving markets but then shifting their entire culture to catch up if they decide it's the right way to go. Sometimes they succeed, sometimes they don't, but it's a brave man who would count them out before the starting bell.
The key point I'm really trying to make in this whole discussion is that Steam doesn't have some sort of highly defensible incumbent's position here. They are essentially a middleman, and middlemen introduce inefficiency that tends to get cut out eventually in competitive markets. I'm not saying their business is in any danger of failing tomorrow, but I think Gabe is right to be concerned (from his business's point of view) about the direction Microsoft is heading with Windows 8.
Because MS will kill Steam in order to boost its own platform with no games on it. That makes perfect sense.
You don't seem to understand the technical details here at all.
Games don't run on Steam. They run on Windows, or MacOS, or Linux, or whatever. Steam is just a distribution channel, and a bunch of related tools. All of the AAA Windows games that you can get on Steam today run on Microsoft's platform, every single one of them, and they could continue to do so even if Steam were completely removed from the picture.
Again, the reason you're talking shit is because you're making this all up. Windows 8's store doesn't have ANY advantage, none whatsoever. The ONLY thing they can do is lock down other players
You just contradicted your own statement.
and it is more likely to end up in a lawsuit
So a few people keep saying, but I don't see why. No-one's brought down Apple's store for having a lock on iOS apps. No-one's brought down Amazon for trying to establish a dominant end-to-end channel for ebooks. Where is this lawsuit you envisage coming from if Microsoft decide to lock down Windows 8, or any future versions of Windows, in the same way?
Steam will still start off with a better user base, an established brand, SteamWorks, all of VALVe's titles plus multiple AAA ones, and again, they'll have a Windows client, a Mac OS X client, and soon, a Linux client.
OK, let's be really absurd and actually look at the facts here.
A better user base: Everyone who is running Steam on Windows is already using Microsoft's platform.
An established brand: Microsoft have their own games console. They also have a long list of the most successful games in PC history under their belt, not least because these days they are the parent company behind several big name studios. By several sensible metrics, they are probably the most successful computer game company in the world today.
SteamWorks: Do you really think Microsoft couldn't build something equivalent? They already have a bunch of related technologies.
All of Valve's titles plus multiple AAA ones: Well, if the platform gets locked down, then Valve will either release its titles via whatever other channel is available or not release them at all, and so will everyone else.
Windows/OS X/Linux clients: Sure, but that doesn't magically mean everyone's games will run on all those different platforms. Hardly any AAA titles are available on Mac Steam right now, and Linux Steam is about on par with Linux gaming generally. Once again, games don't run on Steam, they run on the underlying OS. Steam is just a tool, and an entirely expendable one if anyone else produces a distribution platform with similar capabilities.
VALVe already has to deal with competition (Origin, GFWL), and Windows 8's upcoming app store will be no different.
Of course it could be different. It's operated by the company who own the underlying platform. They therefore have the power to do more or less anything, up to and including locking down competing products that would run on their platform or, perhaps more likely, imposing heavy fees as other channel-controlling vendors like Apple do.
To be clear, I'm not saying they are going to do this immediately with Windows 8, though they are certainly taking big steps in that direction already. But if you were running a company with a service like Steam, you'd have to seriously consider how future-proof your business model was at this point.
Anonymous Brave Guy is basically talking shit. There are way more reasons to use Steam as a distribution platform than the distribution channel
Such as what, exactly?
SteamWorks, a native Windows, Mac and Linux client, brand recognition in the gaming industry, there are many reasons to use Steam over anything else.
And exactly how many AAA titles are available via Steam on Apple, excluding Valve's own? (That's an easy question. You can almost count them on the fingers of one hand.)
Given that it takes dozens or even hundreds of people multiple years and multi-million budgets to develop a modern AAA game, I find it amazing how many people posting in all the discussions like this one today think those games are magically going to be ported to run on other platforms just because of an off-hand comment by Gabe. They don't get ported right now, even to Mac, where Steam is already available. They won't just magically work on Linux either. I'd be the first one to cheer if all those titles did suddenly become available on Linux, preferably via some reasonable and not offensively DRM'd channel, but achieving that would be a vast undertaking, and it's not clear that even Valve/Steam have the power to make it happen.
So I contend that the advantages of your "reasons to use Steam over anything else" are often illusory/wishful thinking.
Well, Windows 8 isn't out yet, so everything any of us says here is a moderately informed guess at best.
With that said, MS are already locking down Windows 8 on ARM. They are already locking down Metro. They are already restricting the use of their inexpensive/hobbyist developer tools to target the grown up UI. That's going to affect a lot of Windows 8 applications, potentially including a lot of games, which Steam won't be able to support as things stand.
But the thing that gets me is that this is a pretty clear change of direction from Microsoft, a statement of intent. If they want to continue that trend and lock down the grown up Windows UI as well, which is hardly implausible given the above and how extraordinarily successful Apple have been with that approach, there is absolutely nothing Valve can do to stop them. Like the disk compression and anti-virus vendors of yesteryear, their market is at risk of being obliterated as the platform itself starts to offer equivalent functionality and with a home ground advantage.
In short, Steam's catalogue disappearing overnight isn't the issue. Whether Valve even have access to install games from that catalogue on their customers' new computers is. Windows 8 is already taking steps -- and they are big steps -- in the direction of locking down the platform. Even if you can still develop and side-load today's games into the main UI under Windows 8, it would be a brave business that bet its future on that continuing in future Windows versions. And if it turns out that a lot of people who have spent a lot of money buying games on Steam don't have as much portability as they thought, because MS pulled the rug out from under Valve, then a lot of the goodwill Steam has accumulated recently is going to evaporate in a heartbeat, just like every other subscription/library/DRM-locked multimedia service that looked great right up until its operators pulled the plug.
The problem for Valve is that Windows 8 is going down the app store route, and the main point of Steam is really to be an easy download and auto-update platform for games. Sure, Steam does other things too, but if it weren't for the distribution channel (which is the only distribution channel for Valve's own big name games) I don't really believe anyone would stick with it just for the minor perks. This leaves only two possibilities:
1. Steam has a powerful lock-in. In this case, a lot of people who have spent a lot of money with them based basically on trust are about to have their faith questioned. Since Steam's standards terms and conditions are a joke as far as guaranteeing anything to anyone but themselves, this leaves two variations:
1a. They will do right by their customers at almost any price, assuming this is even possible with whatever technical and commercial infrastructure MS adopts to go with Windows 8. This might save their reputation and business model, but would surely hurt Valve's bottom line significantly.
1b. They can't or won't pay that price and customers who move to Windows 8 will suffer a worse user experience, limited ability to buy new games, or in the worst case lose access to the existing library they've already paid for. In any case, Steam will take a huge PR hit that will at best severely damage Valve's credibility.
2. Steam's lock-in isn't that powerful. In this case, Microsoft can beat them at their own game (no pun intended) and outright steal their business.
There are exactly zero outcomes in there that are positive for Steam, and some represent an existential threat.
Though they don't appear to be too bothered further down on Mill Road, where cyclists ride at night without lights with impunity, run red lights, and travel at high speed along the inside of cars ready to smack into the side of cars turning out of junctions and turning left. Jumping off pavements without looking etc...
For what little it's worth, it's not with impunity. Mill Road has one of the worst accident records of any road in Cambridge, and many of the people who get hurt or worse were on bikes.
The idiots who overtake cyclists over the blind summit on the bridge do occasionally run into (very literally) someone similarly idiotic coming the other way, too.
Basically, what Cederic said. A NAS is a very specific piece of equipment, designed and engineered for a specific task, that of storing data/programs in a location reachable on a network.
In what ways, exactly?
Who said anything about building their own? Though building even quite substantial servers is easy these days if you're used to DIY PCs, it's not for everyone and there are also plenty of places who will build you a custom server box with the right kind of hardware. And of course you can buy a server of just about any scale off-the-shelf from a vendor like HP or Dell if you're willing to spend the money on it or if you need to place really large orders. Since you work in a bank, you might actually be in that position, though of course most businesses won't be.
Also, don't your servers also need things like resilience, fail-over, throughput, and low power consumption? And even if someone did a custom build themselves or bought one in, why would you assume it would somehow be inferior to a pre-built NAS box in features and/or reliability?
In a few minutes, I'm going back to work on a web-based project that uses several parts of both HTML5 and CSS3, and which runs just fine on IE9.
Perhaps you mean some subset of HTML5 and CSS3, possibly one that also isn't actually standardised yet, and that isn't supported reliably and portably by other browsers either, and where as a result few real-world projects are actually using the features yet anyway?
But even if you only want lots of storage, you can buy anything up to and including rack-mount servers with lots of hot-swap drive bays and redundant power/networking easily enough, and then set them up however you want using standard tools.
That's basically the conclusion we came to as well, though any backup plan is always a matter of risk and how much you're prepared to spend to mitigate how much potential danger, and you have to draw the line at some point.
Still, if a backup service isn't even claiming to offer sensible guarantees that they will actually provide the service advertised, that doesn't seem like a great starting point. We decided that in the current economic climate, there is a significant risk of any business failing or any on-line service being closed down as the result of some commercial transaction involving the company that operates it.
FWIW, in the end, we wound up using a combination of:
(a) some basic off-line storage procedures involving encrypted removable media, which we update every few days (or sometimes more frequently) based on how much has happened since the last time, and
(b) some near-real-time on-line storage using a small local hosting firm whose people we have personally met.
We're happy that the odds of losing the on-line storage permanently are fairly slim. If that does happen, or if there's a major connectivity problem, then there's always a limit to how much data we'll lose access to, because we have the physical media as well. The whole set-up is fully encrypted on our side, and the total cost including overheads seems to be about the same as renting a bunch of space on something like Amazon's web services, and significantly less than subscribing to a lot of the dedicated on-line backup services.
My view of on-line backup services has wound up similar to my view of NAS devices: they provide a valuable option for people who don't have the technical expertise to go with something better, but if you do have that expertise, I just don't see what advantages the dedicated backup services really offer in return for the lock-in, limited flexibility, and typically higher costs. Again, maybe there are certain sizes/types of business where the cost/benefit trade-off looks different, but I haven't personally encountered one.
Just in case anyone else reading this thread is curious, that's a "let me Google that for you" link for the phrase "enterprise class nas devices".
I don't normally feed trolls, but I did take a look at a few results from the first page of search results, just in case I really had completely missed something. The top result goes to a page where the animated "New!" GIFs flash red and yellow. For those of you young enough that you only remember slapping "Beta" graphics on every web page you ever made, flashing graphics are what we used to mock before that, along with the blink tag and scrolling marquees.
But let's be fair and focus on the content. I followed the link to the first product described on that first page, which would surely put my recent experience in perspective. And so it did. The first device there offers a magnificent 2TB of storage in only 4U of rack space, based on a combination of 8x250GB drives. It also comes with such advanced features as an expandable RAID array, in case you don't want the full 2TB initially I assume, and automatically synchronising data with other devices of the same type. It even works on heterogeneous networks with PCs, Macs and Linux boxes! Truly, this is in a different class to the hardware I'm working with today. In fact, even our entry-level workstations can match that kind of spec, never mind the servers.
Just to be really fair, I did also check out some of the serious hardware that appears further down the page. Some of them make compelling sales pitches. If you buy a Thecus N16000, then "Whether rain, shine, or Armageddon, your data will always be there" apparently.
When all is said and done, I'm still waiting to find a counter-example to my position that these are essentially just inflexible servers that happen to be built in cases with lots of hot-swap disk enclosures, though. There is nothing in even the highest-spec devices I found on the first results page that you couldn't also achieve with a general server. There appear to be similar concerns about, for example, integrating with only a very limited number of vendor-chosen back-up services as standard, just like we had with the Cisco box we bought. And as it turns out, despite being only a "small business" product, the Cisco NAS was actually pretty close to the feature spec of a lot of these so-called enterprise-class devices, it just has a smaller number of drive bays and not as much redundancy in the power and networking.
So I stand by my earlier questions: who is the market for these high-end NAS devices, and what advantages does this kind of device offer given that anyone buying one of these boxes presumably has the expertise in-house to configure a real server anyway and will surely need to do a lot of manual configuration either way to integrate with the rest of the network?
Of course, for the OP, who was asking about storage solutions for a small business with around 10-20 staff, this is all rather academic, because they'd be looking at an entry-level NAS or an equivalent server anyway. So I stand by my original advice there as well: as long as you have either moderate Linux sysadmin skills in-house or the budget to bring in suitable help occasionally, there is very little advantage to choosing a NAS over a real server with a similar spec, at least among the various options we've evaluated over the past couple of years.
So you needed a real server, and were surprised when a NAS didn't meet your requirements?
No, at the time we set the network up, we just needed some reliable mass storage. We assumed, foolishly as it turned out, that getting a NAS would be the quickest and easiest way to achieve that.
We soon discovered that setting up the NAS was no easier than setting up a real server. In fact, it was more awkward, in the sense that it's probably still using much the same Linux tools as a real server under the hood, but you can't use all the familiar sysadmin knowledge your team has because you have to configure the device using a web interface with pretty bar charts and non-standard "friendly" terminology and the occasional undocumented corner case.
And as requirements involve -- and requirements always evolve for IT in small businesses -- you can adapt or completely repurpose a general server, while a NAS is forever a NAS.
Perhaps you're right, and instead of a NAS, we should have bought a real server all along. But in that case, my question to you is: who should buy a NAS in the first place? The only answer I've got is "someone who doesn't know enough to set up/administer a real server and can't bring in help who does know enough", which might apply to a lot of home users but probably doesn't apply to anyone responsible for setting up IT facilities for a business.
Your snarky reply does not change the fact that a real server could do everything the NAS could do and many other things as well. A NAS is just a server with a lot of hard disks, a lot of limitations, and hopefully at least some decent preinstalled networking, access control and storage management tools so it has some vague semblance of value.
If you disagree then perhaps, instead of making vague allusions to some mysterious high-end kind of NAS, you would do us all the courtesy of stating specific models and specific features they have that a general server box of a similar specification couldn't match, so we can all have an informed discussion.
I thought that when you said "running a real business" you actually meant a real business. Who the fuck installs email, calendaring or DHCP servers on their NAS device?
It seems likely that quite a lot of people do, given that basically every NAS on the market comes with options to do at least some of those things, and the vendors' support forums are full of people trying to figure out how to root them to install the missing ones.
I notice that your selective quoting missed out the "scheduling regular, secure off-site back-ups" problem I also mentioned. Maybe real businesses like banks are above such things. Do you by any chance work for RBS?
But if cloud-based backups (especially automated, encrypted cloud-based backups) let us mitigate our disaster risk and cut out the oops-forgot-to-change-the-tape factor, they're the lesser of two evils.
Since you're a lawyer, I'll just strongly suggest here that you read the terms of any on-line backup service you're considering using with the same care that you would review a document for a client. We did, and despite not being lawyers, we decided pretty quickly that we wouldn't use any of the ones we were considering.
The clue was in the way they could typically shut down their respective services at about five minutes' notice with little if any guarantee that we would be able to retrieve backed-up data in the event of a disaster affecting us around the same time. Typically, all the useful mass-restore options like getting everything sent in the mail on DVDs/hard drives seem to be cut off the moment they announce that their service is closing down. Your only remaining option is usually to download every byte over an Internet connection within perhaps 72 hours of the notice being given, at a time when obviously every customer they have will be trying to do the same thing. That wouldn't even get you the full amount of data that the entry-level plans for these services claim to support, on a saturated fast leased line, with no contention for access to the servers.
I'd say a minimum acceptable requirement for an off-site backup service is enough notice of any change/closure that we could make other arrangements with continuous cover from one service or the other, and a meaningful and technically credible promise that we could get every byte of our data back before they closed down the service. If they aren't willing to give a contractual guarantee that they can do that, and demonstrate that they are maintaining the necessary resources to honour that commitment immediately if anything happens to their business, then how can they possibly be a trustworthy partner for large amounts of sensitive data? It's like an insurance policy that ends with the words "Policy may be arbitrarily and unpredictably void in the event of a claim being required". Sure, it's not likely that they my business will suffer a catastrophe within a couple of weeks of a backup service closing down, but then it's not likely that we'll suffer a catastrophe at all, is it? If we were going to pay a silly amount of money for robust off-site backups, we'd want some sort of legally and technically credible promises in return.
Can you give any concrete examples where IE9 doesn't work properly? Not new features in HTML5 and CSS3 that it doesn't support, but things where it's supposed to support them but is actually broken?
We also went through this a while ago, but the other way around. After kitting out a small office network, the one purchase we really regretted was the NAS (a Cisco-branded device, which in fact is a rebadged QNAP).
The hardware has not failed and supports hot-swapping drives if necessary, but those are about the only good things I have to say about this unit. It is in all other respects just a very limited and relatively expensive Linux server, where essential operations like scheduling regular, secure off-site back-ups are absurdly difficult, and where you can't easily install other server software (e-mail, calendars, DHCP, RADIUS, whatever) unless whoever supplied your NAS happens to make some sort of plug-in available for their particular style of firmware. Even Cisco gave up trying to provide any meaningful support in this area within a few months of the device launching, eventually just providing a mechanism for people to upgrade their firmware to QNAP's own.
When we were investigating options for a new device earlier this year, it looked like more recent NAS devices from other suppliers were little better, maybe differing in some of the details but essentially still the same old story.
My conclusion: NAS devices are for non-technical home users who want to plug in and go. If you're running a real business with serious requirements, and you have moderate Linux skills and/or a modest budget to bring in someone who does when you need them, then buy a real server with a specification suitable for your requirements. There is absolutely no advantage to buying a NAS for someone in that position, IME.
Among other obvious reasons:
1. The cloud services you mentioned aren't even close to secure enough for legally sensitive documents.
2. Judges are unlikely to accept "my Internet connection was down" as a valid reason for not filing documentation properly.
3. Legal documents are written using serious software, not trivial web apps. They have numerous technical requirements and typographical conventions that must be strictly adhered to, in some cases to the point where courts will specify the precise font you must use for all submissions, for example. You don't write this sort of thing in Google Docs, where the concept of a cross-reference has yet to appear and the numbering styles available are one small step past "numbered" and "not numbered".
If you're genuinely interested, you shouldn't trust some random post on Slashdot anyway, you should go experiment for yourself and see how easy it is to break.
Here are a couple of starting points to try out. Both of these areas get broken all the time from direct personal experience and have easily observable problems with the current Chrome release:
Businesses are shunning it in droves. There is a striking difference today between browser usage patterns during office hours and during evenings/weekends.
The problem with updates for home users is that when one of the web apps I work on breaks because of a bad Chrome update, I can promise you it's not Google who get paged at 4am to deal with it. And the end user doesn't know where the problem is, because they don't even realise Chrome has been updating. All they know is that yesterday they could log in and see their account details and today they can't, and so they call the support number or file a bug report on the support page, and my life for the next few days gets more painful. How much do you think this makes me, as a developer, want to use any new browser feature that isn't likely to be 100% robust and future-proof?
Unfortunately, even the LTS releases aren't much use for a big business that runs numerous internal web apps.
For one thing, the period of support is still way too short: the first one, based on Firefox 3.6, already ran out some time ago. Moving new feature development to another branch after a few months is fine, but security updates have to last for the useful lifetime of the platform if you want to be taken seriously.
For another thing, the overlap between support for one LTS version and the next is only a few weeks, which is an absurdly short time to upgrade an entire organisation on the scale I'm talking about there. They do need to do extensive testing during that upgrade process, because for example the last Firefox LTS release seriously broke plug-ins like Flash and Java applets, which of course many of those in-house web apps rely on.
In short, while most of the small businesses I work with could cope fine with an annual upgrade cycle, the large businesses simply can't and won't. For them, having a stable, secure platform for the long term is very much more important than having the latest shiny HTML5 gimmicks.
You obviously feel passionately about the way Microsoft behaved in the past and the IE6 legacy.
I cannot help but notice that today's situation where Mozilla and Google are throwing new extensions around every few weeks looks a lot like the IE vs. Netscape problem repeating itself, except that at least IE6 won so all the users had the same client with the same proprietary extensions for those of us writing web pages to target.
Today, we don't even have that, and all this "living standard" madness seems like a guaranteed way to make sure we will never again have a stable, reliable platform to code against.