Kernel space is only dangerous in the case of badly written graphics drivers. Windows NT has had kernel-level graphics hardware since its inception, and its only been a cause of problems with extremely buggy graphics drivers. And guess what - buggy graphics drivers are gonna cause problems whether they're in the kernel or not. Might as well put them there for the speed advantages.
I disagree. A properly separated model like the DRI/DRM has high speed userspace drivers and doesn't cause problems when there is a bug with the driver. The model requires a tiny kernel module called the DRM. It manages the hardware resources (eg, DMA) and queues the clients so they don't stomp on each other. The majority of the driver is written in userspace and links directly into the client application (via libGL).
Putting an entire video driver in the kernel isn't sensible. There is too much complexity and there is no actual benefit. It's actually faster with modern cards to link the driver directly to the client. The reason being that the client can fill the command buffer without context switches. If the entire video driver was in the kernel you would need two context switches per queue flush.
Network transparency's of only marginal value, particularly considering the cost (non-kernel graphics). Anyway, there's plenty of other methods of doing network logins without needing it built into the core graphics API.
The only cases when the network transparency causes a measurable impact is when a lot of data is being pushed from the client to the hardware. For those situations we have direct rendering in the client. For all other situations, the costs of network transparency are lost in the noise. I wouldn't be too concerned about it.
Anyways, the intention was more to point out the fact that CPUs are so fast that application load time is all but negligible, and certainly not so lengthy as to make me wish I could farm it all out to a central server.
That works fine for word processors. But there are several situations where client-server GUI is the preferred solution. For example, VPN clients are often implemented as Citrix over IPSec. In scientific and academic circles it's common for the applications to run on headless mainframes and/or supercomputers. In high security environments it's sometimes impossible to run a client locally; you must run it remotely and display it locally.
And when you start to consider other issues - how much does it cost to patch and maintain 3000 Windows desktops? - it quickly becomes obvious that per-user desktops aren't the be-all and end-all.
Load times, sure, I'll agree with you that's not a big deal.
Your point? When I double-click on the Word icon, it takes two seconds for the window to come up. Why should I care if the app is pre-loaded or not? If it's pre-loaded on everyone's system, why should we time it as if it weren't?
The problem is that there isn't enough RAM to preload all the applications. My PC during the day will run (and this is a typical work day) Word, Excel, Outlook, Visio, Project, Firefox, Internet Explorer, and an assortment of programs that don't concern me like virus scanners.
If all of these applications tried to preload themselves on startup then your swap would grind itself into dust and boottime would be in excess of 30 minutes.
It's false reasoning to say that Word takes only 2 seconds. It takes 2 seconds plus whatever time it added to the boot sequence. And if the first application you run isn't Word then there is a good chance that the preloaded Word will be swapped to disk anyway, making the next instance of Word take significantly longer than 2 seconds.
Take note that Mozilla also uses the preload trick. My work machine has consumed all 256MB of RAM and 450MB of swap after a fresh reboot and a login. That's 450MB of intensive swap activity that slowed down my boot sequence. If I just want to check my appointments in Outlook then why am I forced to wait for Word and Mozilla to fight over the swap? It's ludicrous.
A minute? What kind of PC do you have? My work PC barely takes a minute to completely load XP from the moment I hit the power button.
The work PCs all have 256MB RAM. Office pushed it into swap la-la land. It takes over 6 minutes from poweron until the thing stops trashing. Even though the desktop appears after a mere 2 minutes you can't do anything. Clicking the Word icon on the startbar just puts the cursor into busy mode.
Sigh, replying to my own posts, first sign of madness.
MIT-SHM extension of course does not go directly to the hardware. I reworded the second last sentence just before submitting and didn't recheck that the whole paragraph read properly.
In the common case, there is no client or server, just an app running on a PC. So don't build the assumption of networking into windowing.
OK. So let's run with that idea. We still have multiple clients and one set of hardware, so we need to arbitrate the access. We also need to have a common place where the clients can share information like window clip lists. Then there are issues like drag and drop, cut and paste, etc which also require inter-client communication. And how do you solve the issue of two clients seeing the mouse button being pressed, and both assuming that the click was for them?
At one stage you realise you need to have a program, somewhere, that coordinates all of the clients. Assuming this won't be the kernel, it must be another userspace program. We call this program "the X server". And because we have all these clients in userspace, and the X server is also in userspace, they need to use some form of inter-process communication. XFree86 and X.org already use UNIX sockets; one of the fastest IPC methods available. The only thing faster would be shared memory but that's been tried before and it's more hassle than it's worth.
Now admittedly there are some situations where the clients simply need to talk directly to the hardware. For example the client needs to upload a 3D texture or render an MPEG-2 frame. For those situations it makes no sense to send that data to the X server first. So for those situations we do have solutions that bypass the X server and go directly to the hardware. These include the DRI extension, the MIT-SHM extension and the DGA extension.
It takes about 2 seconds for MS Word to come up on my laptop when running on batteries. When plugged in, that would presumably be a tad faster.
More precisely, it takes 2 seconds for the Word process that is already running in the background to render a new window. Word is preloaded on startup and it adds at least a minute to the boot time on my work PC.
I can't count how many times I hear on/. someone saying that X is too bulky, etc, etc. And here's Gosling saying (2 years ago) that X is headed in the direction of slim and lightweight.
Am I misreading what he's saying?
No. You've read him correctly. What Gosling is saying is a simplified version of the X.org roadmap.
For example, X11 contains a font renderer. The design is really ancient. No anti-aliasing. Poor kerning. Clients couldn't access the glyphs very easily, which made it impossible to do arbitrary things like strokepaths or proper printing. It kind of sucked. A number of font extensions were considered for XFree86. Any one of them would have addressed all of the existing issues but they were heavyweight solutions.
So in the end Keith Packard wrote a better solution. He implemented the XRender extension. This extension simply knows how to draw rows of glyphs. It also knows about alpha masks (Porter Duff compositing). The client now turns the font (typically TrueType) into alpha-masked glyphs and sends the glyphs to the X server. If you're using a GNOME or KDE desktop with antialiased fonts then you're using Keith's XRender extension and client-side font rendering instead of the X11 font renderer. This is only practical because the client-side libraries (eg, libxft2) are shared.
Another interesting example of "slimming down" the X server is the Composite extension. Rather than implement a heavy compositing engine in the X server, Keith designed this extension so it simply renders the window into offscreen memory. Another extension, XDamage, tells a special client called the "compositor" when any region of the window changes. The compositor then uses the XRender extension to render the damaged region with appropriate drop shadows and/or alpha masks. Notice how the rendering is still done by the X server so it can be hardware accelerated.
For the future of X.org there is more of this "slimming down" being planned. Jim Gettys and Keith Packard gave a presentation in July 2004 where they suggest the future of X is as an OpenGL client. They are both keen on a new design where the X server stops being the arbitrator of video hardware. Instead it becomes an OpenGL client with direct access to the video hardware through the DRM, just like every other DRI client. There is a simpler version of that paper in the short slideshow Life in X Land.
The whole situation in this country is just getting rediculous. Is it possible for people to believe that George W. Bush is a terrible politician, but a decent guy who just has a difference of opinion with you?
Hey, I'm not American and I've no attachment to either party in your two-party system, but it must be said that GWB is a fricking idiot. He's dumber than a box of hammers. He's a goose. A galah. A drongo. He couldn't argue his way out of a wet paper bag. When he's "lost in thought" you can almost hear his remaining pair of brain cells clicking together.
Evil? Nah, I don't reckon he's evil. I'm a generally positive person and I believe there are very few truly evil people. But there are stupid people. And there are ignorant people. And there are righteous people and they are the scariest of all. GWB manages to represent himself well in all three categories. How he managed to find himself in such a position of such power is beyond me. I thought your democracy was supposed to elect the best and brightest? You've managed to elect the dimmest bulb in the lighting store.
Ever seen the movie "Being There"? I thought it was a comedy. Now I wonder.
No as for the desirability of open-source firmware... If you plan on telling me that you intend to take this open-source firmware and modify it so your card can do different things with its radio, pay me no mind while I laugh in your face.
I think you're unfairly trivialising the need for FLOSS friendly firmware.
The actual problem is that the licensing on the firmware often prevents redistribution. This means you can't just install a distribution like Debian and start using your wireless card. There is an additional user-hostile step to extract the binary firmware from the Vendor Supplied Driver Disc to use with the otherwise FLOSS driver. I'm using a Prism54 chipset that has binary firmware so I speak from my own frustrated experience.
There is also the issue of bugs in firmware. Consider the recent bugs that several vendors had in their wireless encryption implementations. I might not have the skills to fix those bugs but I know some people do. Hell, some people patched the binary firmware even without access to source code, though their task was obviously harder. I don't want to be stuck in a situation where I'm hoping the vendor cares enough to fix the bugs for me. Especially when the vendor might have a conflict of interest; they want you to purchase their new equipment, not waste their time fixing your existing equipment.
There is an additional problem for long-term maintenance. What happens when the vendor goes bankrupt (as often happens), their website stops carrying downloadable drivers, and I can't find my Vendor Supplied Driver Disc to extract the firmware. I now have a useless doorstop instead of otherwise useful hardware. Consider this situation that happened to me recently. A decade old system. Poorly maintained. Original software long since lost. The vendor of the SCSI board had gone bankrupt. The system disk had just blown and the backups were useless. The client was in a state of shock; they had 200GB of data (on other disks) that they couldn't access. They couldn't find drivers anywhere. Linux saved their bacon because Linux had FLOSS drivers for a decade old 16-bit ISA SCSI card. In a world with non-redistributable binary firmware, they really would have been screwed.
So you trivialise the problem as being about wanting to hack on the driver. That's not what I dislike about binary firmwares. I'm considering ease of installation, long-term maintenance, the ability to control the software baseline, and I don't want to be dependent on the future of some two-bit company that might drop support for the product. Intel is unlikely to go bankrupt, but I predict they'll remove the firmware from their website as soon as the current product becomes unprofitable or a successor product is released.
There are options that the vendors could give me that would make me happy. I don't necessarily want the source code for the firmware. I wouldn't be able to do anything with it anyway. But I do want:
Licensing which permits unlimited copying of the firmware, including the ability to bundle it with distributions, no restrictions on charging money for the privilege, and so on.
Cleanly documented interface to the firmware so I can use FLOSS drivers running on the host CPU. I don't really care how the firmware works but I do want to know how to use it.
A guarantee from the vendor that if they stop supporting the firmware (eg, fixing bugs) then they will release the source code under a FLOSS license so people who are interested can maintain the firmware.
I don't think any of those wishlist items are unreasonable.
There is also one from the Line By Line analisis done by the Law Student.
Except he isn't a law student. He posted a reply in the Groklaw comments section pointing out that he's an IT student who is considering changing his degree to law in 12 months time.
Once again, the Slashdot writ-up and the actual story do not quite jive together. The story itself is NOT about the fact that the music industry is foisting off crappy CDs that they need to get rid of, as a part of the settlement. What the story ***IS*** about is rap music. The Kansas AG does not like the politics of the music, saying (rightly so) that it is overly sexual and violent.
Oh yes, Lou Reed being that famous rap artist known for his violent and sexual lyrics.
RTFA!!! And Slashdot editors: Do you even RTFA???? Clearly not.
You demonstrate the difference between reading and understanding.
I can't believe that this crowd thinks Outkast is crap music. Outkast has many excellent songs, some of which are very political and some of which are about other complex themes.
Yeah, like that one that reminds everybody that shit smells like poo-poo./rolleyes
I'm a Linux user but Mac laptops are lovely, with excellent battery life. Too bad Airport Express (and power management?) isn't supported on Linux PPC.
Airport Extreme? Yeah, that's a bummer.
Power Management is partially supported. You can read the PMU (AC status, battery status) and you can turn the computer off. You can patch your kernel with PMDISK and get suspend-to-disk. But sadly no support for sleep.
My biggest annoyance is the nvidia chipset. The FLOSS drivers don't support 3D. The binary drivers which do support 3D are for x86 only. The unsupported sleep is due to the nvidia chipset too; LinuxPPC can put the computer to sleep but it can't wake the nvidia chipset after returning from sleep. I really dislike nvidia's attitude towards Linux. I'd prefer they were icy cold and provided nothing, rather than this half-hearted sorry excuse for support they do provide.
Now THAT is intuitive, and unbelivably user friendly. Wow.
If you want to run in expert mode then you need to be an expert to enable it.
Not that it's exactly hard to find that option. Gconf-editor is in the main menu (Applications -> System Tools -> Configuration Editor). All of the "registry keys" are described. That particular key had this description.
If set to true, then all Nautilus windows will be browser windows. This is how Nautilus used to behave before version 2.6, and some people prefer this behavior.
If you couldn't find that then you're a newbie and you don't want to be running in complicated browser mode. Stick with the easier nautilus mode until you gain some more skills.
Re:Completely silent-Dead giveaway.
on
Sun Rays For Linux
·
· Score: 2, Informative
Not sure what the pricing is like now, but I looked into buying 50 of them at one point. In the UK it was going to cost me around £500 per station, plus of course the extra beef the servers needed (our network was ok so at least I didn't need to upgrade that). I couldn't justify the cost in comparison to PCs, which we were buying at arouns £1k at the time; for £500 extra the PC could be repurposed as a build machine, a test server, or whatever a project needed; they also came with 40Gb of disk, which would have cost us $$$ on the server. We could also save on PCs by not buying a new monitor for each PC replaced.
You only compared hardware prices. Throw in the cost of PC software (including all the support software like Ghost, Antivirus, etc) plus the extra maintenance of having to package software for deployment, and a SunRay solution suddenly is surprisingly cheap.
User's PCs weren't backed up, everyone had space on the servers which was raided and backed up; the cost of providing that much disk space, and backing it up, with the Sunray solution was prohibitive, and it would have been a single point of failure.
You don't have an entire PC image per user on the server, silly person. User directories run 1-2GB each. The system software is shared across all users. You can get away with 200GB RAID-1 for 50 users without any trouble. Get a V880 with 12x FCAL slots and you have in-box growth up to 1TB easily.
OSRM is the company PJ (you know, of Groklaw [groklaw.com]) joined a few months back to provide indemnification for Linux users. This organization isn't the enemy, folks.
Why? Because PJ works there? Be serious. The company might be a scam for all we know. Just because PJ and Bruce work there doesn't mean a damn thing. We had FLOSS friends working at Caldera/SCO too.
This is the exact concern that several of us raised months ago: OSRM hiring PJ and Bruce looks exactly like an attempt to buy credibility. And apparently it has already worked.
Slashdot Mirror
I disagree. A properly separated model like the DRI/DRM has high speed userspace drivers and doesn't cause problems when there is a bug with the driver. The model requires a tiny kernel module called the DRM. It manages the hardware resources (eg, DMA) and queues the clients so they don't stomp on each other. The majority of the driver is written in userspace and links directly into the client application (via libGL).
Putting an entire video driver in the kernel isn't sensible. There is too much complexity and there is no actual benefit. It's actually faster with modern cards to link the driver directly to the client. The reason being that the client can fill the command buffer without context switches. If the entire video driver was in the kernel you would need two context switches per queue flush.
The only cases when the network transparency causes a measurable impact is when a lot of data is being pushed from the client to the hardware. For those situations we have direct rendering in the client. For all other situations, the costs of network transparency are lost in the noise. I wouldn't be too concerned about it.
That works fine for word processors. But there are several situations where client-server GUI is the preferred solution. For example, VPN clients are often implemented as Citrix over IPSec. In scientific and academic circles it's common for the applications to run on headless mainframes and/or supercomputers. In high security environments it's sometimes impossible to run a client locally; you must run it remotely and display it locally.
And when you start to consider other issues - how much does it cost to patch and maintain 3000 Windows desktops? - it quickly becomes obvious that per-user desktops aren't the be-all and end-all.
Load times, sure, I'll agree with you that's not a big deal.
The problem is that there isn't enough RAM to preload all the applications. My PC during the day will run (and this is a typical work day) Word, Excel, Outlook, Visio, Project, Firefox, Internet Explorer, and an assortment of programs that don't concern me like virus scanners.
If all of these applications tried to preload themselves on startup then your swap would grind itself into dust and boottime would be in excess of 30 minutes.
It's false reasoning to say that Word takes only 2 seconds. It takes 2 seconds plus whatever time it added to the boot sequence. And if the first application you run isn't Word then there is a good chance that the preloaded Word will be swapped to disk anyway, making the next instance of Word take significantly longer than 2 seconds.
Take note that Mozilla also uses the preload trick. My work machine has consumed all 256MB of RAM and 450MB of swap after a fresh reboot and a login. That's 450MB of intensive swap activity that slowed down my boot sequence. If I just want to check my appointments in Outlook then why am I forced to wait for Word and Mozilla to fight over the swap? It's ludicrous.
The work PCs all have 256MB RAM. Office pushed it into swap la-la land. It takes over 6 minutes from poweron until the thing stops trashing. Even though the desktop appears after a mere 2 minutes you can't do anything. Clicking the Word icon on the startbar just puts the cursor into busy mode.
Sigh, replying to my own posts, first sign of madness. MIT-SHM extension of course does not go directly to the hardware. I reworded the second last sentence just before submitting and didn't recheck that the whole paragraph read properly.
OK. So let's run with that idea. We still have multiple clients and one set of hardware, so we need to arbitrate the access. We also need to have a common place where the clients can share information like window clip lists. Then there are issues like drag and drop, cut and paste, etc which also require inter-client communication. And how do you solve the issue of two clients seeing the mouse button being pressed, and both assuming that the click was for them?
At one stage you realise you need to have a program, somewhere, that coordinates all of the clients. Assuming this won't be the kernel, it must be another userspace program. We call this program "the X server". And because we have all these clients in userspace, and the X server is also in userspace, they need to use some form of inter-process communication. XFree86 and X.org already use UNIX sockets; one of the fastest IPC methods available. The only thing faster would be shared memory but that's been tried before and it's more hassle than it's worth.
Now admittedly there are some situations where the clients simply need to talk directly to the hardware. For example the client needs to upload a 3D texture or render an MPEG-2 frame. For those situations it makes no sense to send that data to the X server first. So for those situations we do have solutions that bypass the X server and go directly to the hardware. These include the DRI extension, the MIT-SHM extension and the DGA extension.
More precisely, it takes 2 seconds for the Word process that is already running in the background to render a new window. Word is preloaded on startup and it adds at least a minute to the boot time on my work PC.
No. You've read him correctly. What Gosling is saying is a simplified version of the X.org roadmap.
For example, X11 contains a font renderer. The design is really ancient. No anti-aliasing. Poor kerning. Clients couldn't access the glyphs very easily, which made it impossible to do arbitrary things like strokepaths or proper printing. It kind of sucked. A number of font extensions were considered for XFree86. Any one of them would have addressed all of the existing issues but they were heavyweight solutions.
So in the end Keith Packard wrote a better solution. He implemented the XRender extension. This extension simply knows how to draw rows of glyphs. It also knows about alpha masks (Porter Duff compositing). The client now turns the font (typically TrueType) into alpha-masked glyphs and sends the glyphs to the X server. If you're using a GNOME or KDE desktop with antialiased fonts then you're using Keith's XRender extension and client-side font rendering instead of the X11 font renderer. This is only practical because the client-side libraries (eg, libxft2) are shared.
Another interesting example of "slimming down" the X server is the Composite extension. Rather than implement a heavy compositing engine in the X server, Keith designed this extension so it simply renders the window into offscreen memory. Another extension, XDamage, tells a special client called the "compositor" when any region of the window changes. The compositor then uses the XRender extension to render the damaged region with appropriate drop shadows and/or alpha masks. Notice how the rendering is still done by the X server so it can be hardware accelerated.
For the future of X.org there is more of this "slimming down" being planned. Jim Gettys and Keith Packard gave a presentation in July 2004 where they suggest the future of X is as an OpenGL client. They are both keen on a new design where the X server stops being the arbitrator of video hardware. Instead it becomes an OpenGL client with direct access to the video hardware through the DRM, just like every other DRI client. There is a simpler version of that paper in the short slideshow Life in X Land.
Hey, I'm not American and I've no attachment to either party in your two-party system, but it must be said that GWB is a fricking idiot. He's dumber than a box of hammers. He's a goose. A galah. A drongo. He couldn't argue his way out of a wet paper bag. When he's "lost in thought" you can almost hear his remaining pair of brain cells clicking together.
Evil? Nah, I don't reckon he's evil. I'm a generally positive person and I believe there are very few truly evil people. But there are stupid people. And there are ignorant people. And there are righteous people and they are the scariest of all. GWB manages to represent himself well in all three categories. How he managed to find himself in such a position of such power is beyond me. I thought your democracy was supposed to elect the best and brightest? You've managed to elect the dimmest bulb in the lighting store.
Ever seen the movie "Being There"? I thought it was a comedy. Now I wonder.
That's not the only thing Bush has trouble with.
"Tribal sovereignty means that. It's sovereign. You're a ... you're a ... you've been given sovereignty and you're viewed as a sovereign entity."
I sense a great disturbance in the force. As if millions of hands raised in unison.
He has a cold.
I think you're unfairly trivialising the need for FLOSS friendly firmware.
The actual problem is that the licensing on the firmware often prevents redistribution. This means you can't just install a distribution like Debian and start using your wireless card. There is an additional user-hostile step to extract the binary firmware from the Vendor Supplied Driver Disc to use with the otherwise FLOSS driver. I'm using a Prism54 chipset that has binary firmware so I speak from my own frustrated experience.
There is also the issue of bugs in firmware. Consider the recent bugs that several vendors had in their wireless encryption implementations. I might not have the skills to fix those bugs but I know some people do. Hell, some people patched the binary firmware even without access to source code, though their task was obviously harder. I don't want to be stuck in a situation where I'm hoping the vendor cares enough to fix the bugs for me. Especially when the vendor might have a conflict of interest; they want you to purchase their new equipment, not waste their time fixing your existing equipment.
There is an additional problem for long-term maintenance. What happens when the vendor goes bankrupt (as often happens), their website stops carrying downloadable drivers, and I can't find my Vendor Supplied Driver Disc to extract the firmware. I now have a useless doorstop instead of otherwise useful hardware. Consider this situation that happened to me recently. A decade old system. Poorly maintained. Original software long since lost. The vendor of the SCSI board had gone bankrupt. The system disk had just blown and the backups were useless. The client was in a state of shock; they had 200GB of data (on other disks) that they couldn't access. They couldn't find drivers anywhere. Linux saved their bacon because Linux had FLOSS drivers for a decade old 16-bit ISA SCSI card. In a world with non-redistributable binary firmware, they really would have been screwed.
So you trivialise the problem as being about wanting to hack on the driver. That's not what I dislike about binary firmwares. I'm considering ease of installation, long-term maintenance, the ability to control the software baseline, and I don't want to be dependent on the future of some two-bit company that might drop support for the product. Intel is unlikely to go bankrupt, but I predict they'll remove the firmware from their website as soon as the current product becomes unprofitable or a successor product is released.
There are options that the vendors could give me that would make me happy. I don't necessarily want the source code for the firmware. I wouldn't be able to do anything with it anyway. But I do want:
I don't think any of those wishlist items are unreasonable.
Except he isn't a law student. He posted a reply in the Groklaw comments section pointing out that he's an IT student who is considering changing his degree to law in 12 months time.
Yes.
Oh yes, Lou Reed being that famous rap artist known for his violent and sexual lyrics.
You demonstrate the difference between reading and understanding.
Yeah, like that one that reminds everybody that shit smells like poo-poo. /rolleyes
You can code free software apps for Windows too. What's your point?
Software freedom.
Shrug. I just did that second test and got 104wpm. I don't think my typing is that unusual.
Airport Extreme? Yeah, that's a bummer.
Power Management is partially supported. You can read the PMU (AC status, battery status) and you can turn the computer off. You can patch your kernel with PMDISK and get suspend-to-disk. But sadly no support for sleep.
My biggest annoyance is the nvidia chipset. The FLOSS drivers don't support 3D. The binary drivers which do support 3D are for x86 only. The unsupported sleep is due to the nvidia chipset too; LinuxPPC can put the computer to sleep but it can't wake the nvidia chipset after returning from sleep. I really dislike nvidia's attitude towards Linux. I'd prefer they were icy cold and provided nothing, rather than this half-hearted sorry excuse for support they do provide.
If you want to run in expert mode then you need to be an expert to enable it.
Not that it's exactly hard to find that option. Gconf-editor is in the main menu (Applications -> System Tools -> Configuration Editor). All of the "registry keys" are described. That particular key had this description.
If you couldn't find that then you're a newbie and you don't want to be running in complicated browser mode. Stick with the easier nautilus mode until you gain some more skills.
You only compared hardware prices. Throw in the cost of PC software (including all the support software like Ghost, Antivirus, etc) plus the extra maintenance of having to package software for deployment, and a SunRay solution suddenly is surprisingly cheap.
You don't have an entire PC image per user on the server, silly person. User directories run 1-2GB each. The system software is shared across all users. You can get away with 200GB RAID-1 for 50 users without any trouble. Get a V880 with 12x FCAL slots and you have in-box growth up to 1TB easily.
Hey, these jokes are funny... in Japan!
Why? Because PJ works there? Be serious. The company might be a scam for all we know. Just because PJ and Bruce work there doesn't mean a damn thing. We had FLOSS friends working at Caldera/SCO too.
This is the exact concern that several of us raised months ago: OSRM hiring PJ and Bruce looks exactly like an attempt to buy credibility. And apparently it has already worked.