DOn't be silly. Of course it's supposed to work in Sally's Dounut Shop, and airports and hotels and visitor centers at other companies. What's the point of a mobile worker if not being mobile?
Clearly the lawyers disagree with you (IP rights and patents) and lawyers are important in corporations.
Don't get me wrong, I would love all open source drivers. But it's just not feasible at the moment. We might get there some day.
Read the/. article here for reasons why nvidia has not open sourced their driver. And that is just one example, you would get the same mantra from most providers of binary only drivers
Imagine you're in a large corporation. You have windows, linux and unix boxes, with multi terrabyte storage systems. You use one of the few available backup/storage solutions that is delivered from one of the large players in enterprise level computing, IBM. With your Tivoli management solution you have Tivoli Storage Manager (TSM).
To those TSM servers in your storage hierarchy that are attache to tape libraries and tape drives you need special drivers that are supplied with TSM. These drivers are kernel modules and they are compiled against specific kernel levels of Red Hat Enterprise Linux and SUSE Enterprise Linx. The drivers are fully supported by IBM, but they require you to run them on the exact kernel version supplied by Red Hat or SUSE. And they are binary only.
So what do you do in your multi-million dollar SAN/storage solution? Port to amanda? Use cpio and library scripts in perl? Remember, you need full support for your backup solution, and it must work on Linux, Windows, AIX, HP-UX, Solaris, Mac and Netware servers.
Know if the drivers are secure? You don't care. They are supported, that's what count. (besides, your SAN/backup network should be closed and secured anyway:)
It's relatively easy to configure linux on a specific laptop with a specific wifi card to work on a specific wifi network. Having it work seamlessly with *all* wifis out there is much much harder.
We're talking about standard drivers and something like Windows XP SP2's "Choose a Wifi" application. Search, point, enter WPA passphrase and off you go. Corporate enterprise installations are not entirely static and predictable.
Corporate users expects the WiFi to just work, whatever the hotspot the user is in.
But that's hard. Cell phones, PDAs, home printers and equipment already installed, hotel equipment like broadband and printer services... it's easy to define a standard desktop and laptop, make a preloaded image that works and install that. But esp. for laptop users, when they are out there and have to interface with the real world, then you get problems.
Like on a hotel room I was last week with a special combo USB and Ethernet connector... only worked in Windows. How do you explain that to your CFO that will have a presentation tomorrow for a large financing group? Ask the receptionist? Dial up using his GSM cell phone and download this driver and compile it?
It got to be plug-and-play -- ideally just play -- the "plug" part can be hard too sometimes;)
Yes, and an abstraction level like this is what is in use in many cases. You get a binary only driver and a small software stub you can compile against your driver to integrate it.
If the kernel provided such a layer by default, that would surely solve the problem -- because then you would have a standard API layer you make your binary only driver for. There is work going on for this with big-name IT companies behind it, so let's cross our fingers:)
Yes, because those last 10% is what gives you problems. If you just go to your local electronic store and buy a Wifi PC Card (both for the Radius servers at work and with WPA for the users home nets, and open or WEP or WPA encrypted customer/coffee shot nets), you buy a MP3 player where you want do up/download music and use it as a portable storage device, you buy a label printer and a scanner for desktop use. Will it all "just work"? Nope.
Sure, you can find stuff that will work in Linux, but some requires 3rd party drivers (madwifi? how can you support that in a corporation) or binary only drivers (video cards, custom high end storage devices) or you have to use "vi" to configure it.
It has to be easily installed even by Joe Sixpak, else your support costs will skyrocket. IMO, this is the largest stumbling block for Linux Desktops.
A stable driver API is one of the things that is much needed. This is even a problem for server environments. In a perfect world, all drivers would be open source and easy to include, but that is just a pipe dream at the moment. There is a need for binary only drivers for several reasons, where a) support and b) it includes patented/licensed code are two of the biggest.
As it is now, Linux on the Desktop is only feasible for very specific desktop environments. And on laptops? Power management and wireless networking are not automatic, and with several different hardware versions and with users that roam the world... it's a pain.
Linux is getting there though, but slowly. The support cost for linux on desktops and laptops in corporations today would be too high I fear.
Re:NOONE COUNTED THE COLD CATHODES!!!
on
AMD 90nm Evaluated
·
· Score: 1
The baby I'm working on now got dual 1400 Watt PSUs:)
Because AIX in some repspects is better than Linux. Many reasons for that; more stable, better performance and enterprise level features and management tools that just isn't there on Linux (yet).
Why Linux has not adopted the nice AIX command structure, where commands are ch*, ls*, rm* and mk* for change, list, remove and make stuff. I.e chlv, lslv, rmlv and mklv for dealing with logical volumes in the LVM.
Oh, they are code compatible (though may need different optimizations). POWER5 are dual-core chips and with wastly larger caches than the PPC970 chips. For large workloads they are quite faster than POWER4 and PPC970.
In my country (Norway) we have quite good Fair Use laws (y'all will probably remember the "DVD-Jon" case and its positive outcome).
We can: 1. Copy any media (CD, DVD, LP, MC, VHS, whatever) for our own use as much as we want to. 2. Share with family (parents, siblings... maybe 1. cousins but that's it) 3. Share with close friends, and this is interpreted in a strick sense. Your best friend that you grew up with? Sure! Someone from class or a cow-orker? Nope, not close enough.
This complies with my sense of justice pretty well. After all, is it fair use to share your new CD with music with people you meet on the bus?
As a compensation for this, artists get paid from a fund. It's the same fund that was started when MC copying started and is/was funded by sale of empty music cassettes.
I bet that most audio copying today does not go straight to P2P networks. How many of you rip your CDs to a) play them on your DAP (mp3/ogg) player or b) have a copy in your car, but do not put the ripped files on P2P? I bet there are a lot of you out there. Maybe this can be used to make statistics to counter the RIAA drivel.
Any time I rip a CD with CDex, it does a lookup to freecddb.org. There are other services for this, like gracenote and others. We could assume that the total hits to these pages, minus lets say 10%, are legitimate rips of CDs. Then we would have an estimate of the amount of legal (legal as in fair use) ripping out there.
In CGA adapters you had low-res and hi-res modes for graphics:
1. 320x200 pixels with four colors (from two palettes: black, white, cyan, magenta or black, yellow, green and red) 2. 640x200 pixes for black and white
Then you had text mode, with 80x25 characters, each in a 8x8 pixel box - giving you the same 640x200 pixels resolution. But you had in text mode 16 colors: black, blue, red, cyan, magenta, green, yellow and white, with high intensity versions OR blinking instead. Later on, the low intensity yellow became brown and low intensity white became gray. Oh yes, you could also have a 40x25 character text mode... with 320x200 pixels resolution.
Around the same period you had MDA and the famous Hercules MDA cards, providing monochrome (green or in some cases yellow or amber), but much better looking characters since they used 9x12 font size.
Then later came EGA with 640x400 as baseline resolution and later VGA with 640x480 and the rest is probably familiar territory for most.
Good! This is quite a nifty little SQL engine with a lot of features. Will be exciting to follow the progress of Derby, could provide competition for mysql and postgresql.
Windows virtual memory handling is b0rked from my point of view.
1. It pages out minimised applications. Why? 2. It is much to aggressive with filesystem caching.
So when I minimise the app, it will page it out to cache in a file that I will read just once, then when I restore the window... paging hell. Because now it has to page out another app, free some filesystem cache, then page in the newly restore app.
If I could lower focus on fs cache and keep minimised apps in memory just as foreground ones (with a LRU algorithm to page out if needed) my PC would be much faster.
Re:That's exactly the quote I remembered
on
Ethernet at 10 Gbps
·
· Score: 3, Interesting
I haven't said that, but a columnist in Byte Magazine in the mid-80s had a rant about this.
He programmed on a Mac, and the compilation took typically 5 to 10 minutes. Enough to get a cup of coffee, check the newspaper and have a quick chat with a cow-orker. Then he got a new Mac, and it compiled the program in a minute or so. No time for coffee, no time for news, no time for smalltalk.
So the new, faster computer was too fast... he had to wait at his desk more with the new computer.
Slashdot Mirror
You have been X'ed!
DOn't be silly. Of course it's supposed to work in Sally's Dounut Shop, and airports and hotels and visitor centers at other companies. What's the point of a mobile worker if not being mobile?
Clearly the lawyers disagree with you (IP rights and patents) and lawyers are important in corporations.
Don't get me wrong, I would love all open source drivers. But it's just not feasible at the moment. We might get there some day.
Read the /. article here for reasons why nvidia has not open sourced their driver. And that is just one example, you would get the same mantra from most providers of binary only drivers
OK, real world example:
:)
Imagine you're in a large corporation. You have windows, linux and unix boxes, with multi terrabyte storage systems. You use one of the few available backup/storage solutions that is delivered from one of the large players in enterprise level computing, IBM. With your Tivoli management solution you have Tivoli Storage Manager (TSM).
To those TSM servers in your storage hierarchy that are attache to tape libraries and tape drives you need special drivers that are supplied with TSM. These drivers are kernel modules and they are compiled against specific kernel levels of Red Hat Enterprise Linux and SUSE Enterprise Linx. The drivers are fully supported by IBM, but they require you to run them on the exact kernel version supplied by Red Hat or SUSE. And they are binary only.
So what do you do in your multi-million dollar SAN/storage solution? Port to amanda? Use cpio and library scripts in perl? Remember, you need full support for your backup solution, and it must work on Linux, Windows, AIX, HP-UX, Solaris, Mac and Netware servers.
Know if the drivers are secure? You don't care. They are supported, that's what count. (besides, your SAN/backup network should be closed and secured anyway
It's relatively easy to configure linux on a specific laptop with a specific wifi card to work on a specific wifi network. Having it work seamlessly with *all* wifis out there is much much harder.
We're talking about standard drivers and something like Windows XP SP2's "Choose a Wifi" application. Search, point, enter WPA passphrase and off you go. Corporate enterprise installations are not entirely static and predictable.
Corporate users expects the WiFi to just work, whatever the hotspot the user is in.
But that's hard. Cell phones, PDAs, home printers and equipment already installed, hotel equipment like broadband and printer services... it's easy to define a standard desktop and laptop, make a preloaded image that works and install that. But esp. for laptop users, when they are out there and have to interface with the real world, then you get problems.
;)
Like on a hotel room I was last week with a special combo USB and Ethernet connector... only worked in Windows. How do you explain that to your CFO that will have a presentation tomorrow for a large financing group? Ask the receptionist? Dial up using his GSM cell phone and download this driver and compile it?
It got to be plug-and-play -- ideally just play -- the "plug" part can be hard too sometimes
Yes, and an abstraction level like this is what is in use in many cases. You get a binary only driver and a small software stub you can compile against your driver to integrate it.
:)
If the kernel provided such a layer by default, that would surely solve the problem -- because then you would have a standard API layer you make your binary only driver for. There is work going on for this with big-name IT companies behind it, so let's cross our fingers
Yes, because those last 10% is what gives you problems. If you just go to your local electronic store and buy a Wifi PC Card (both for the Radius servers at work and with WPA for the users home nets, and open or WEP or WPA encrypted customer/coffee shot nets), you buy a MP3 player where you want do up/download music and use it as a portable storage device, you buy a label printer and a scanner for desktop use. Will it all "just work"? Nope.
Sure, you can find stuff that will work in Linux, but some requires 3rd party drivers (madwifi? how can you support that in a corporation) or binary only drivers (video cards, custom high end storage devices) or you have to use "vi" to configure it.
It has to be easily installed even by Joe Sixpak, else your support costs will skyrocket. IMO, this is the largest stumbling block for Linux Desktops.
A stable driver API is one of the things that is much needed. This is even a problem for server environments. In a perfect world, all drivers would be open source and easy to include, but that is just a pipe dream at the moment. There is a need for binary only drivers for several reasons, where a) support and b) it includes patented/licensed code are two of the biggest.
As it is now, Linux on the Desktop is only feasible for very specific desktop environments. And on laptops? Power management and wireless networking are not automatic, and with several different hardware versions and with users that roam the world... it's a pain.
Linux is getting there though, but slowly. The support cost for linux on desktops and laptops in corporations today would be too high I fear.
The baby I'm working on now got dual 1400 Watt PSUs :)
And yes, it's a big server.
Because AIX in some repspects is better than Linux. Many reasons for that; more stable, better performance and enterprise level features and management tools that just isn't there on Linux (yet).
Why Linux has not adopted the nice AIX command structure, where commands are ch*, ls*, rm* and mk* for change, list, remove and make stuff. I.e chlv, lslv, rmlv and mklv for dealing with logical volumes in the LVM.
Oh, they are code compatible (though may need different optimizations). POWER5 are dual-core chips and with wastly larger caches than the PPC970 chips. For large workloads they are quite faster than POWER4 and PPC970.
POWER5 > PPC970/G5
First, IANAL.
In my country (Norway) we have quite good Fair Use laws (y'all will probably remember the "DVD-Jon" case and its positive outcome).
We can:
1. Copy any media (CD, DVD, LP, MC, VHS, whatever) for our own use as much as we want to.
2. Share with family (parents, siblings... maybe 1. cousins but that's it)
3. Share with close friends, and this is interpreted in a strick sense. Your best friend that you grew up with? Sure! Someone from class or a cow-orker? Nope, not close enough.
This complies with my sense of justice pretty well. After all, is it fair use to share your new CD with music with people you meet on the bus?
As a compensation for this, artists get paid from a fund. It's the same fund that was started when MC copying started and is/was funded by sale of empty music cassettes.
I bet that most audio copying today does not go straight to P2P networks. How many of you rip your CDs to a) play them on your DAP (mp3/ogg) player or b) have a copy in your car, but do not put the ripped files on P2P? I bet there are a lot of you out there. Maybe this can be used to make statistics to counter the RIAA drivel.
Any time I rip a CD with CDex, it does a lookup to freecddb.org. There are other services for this, like gracenote and others. We could assume that the total hits to these pages, minus lets say 10%, are legitimate rips of CDs. Then we would have an estimate of the amount of legal (legal as in fair use) ripping out there.
A TRUE story was the introduction of Honda Fitta in Norway (and the rest of Scandinavia).
Why was that bad? Check this link
You too? Didn't realise that y'all could get NRK2 where they aired that movie yesterday :)
Nope, not the IBM song, this is: Ever onwards
And that really swings :)
Was no 320x240 at all :) (Amiga 500 used 320x240)
In CGA adapters you had low-res and hi-res modes for graphics:
1. 320x200 pixels with four colors (from two palettes: black, white, cyan, magenta or black, yellow, green and red)
2. 640x200 pixes for black and white
Then you had text mode, with 80x25 characters, each in a 8x8 pixel box - giving you the same 640x200 pixels resolution. But you had in text mode 16 colors: black, blue, red, cyan, magenta, green, yellow and white, with high intensity versions OR blinking instead. Later on, the low intensity yellow became brown and low intensity white became gray. Oh yes, you could also have a 40x25 character text mode... with 320x200 pixels resolution.
Around the same period you had MDA and the famous Hercules MDA cards, providing monochrome (green or in some cases yellow or amber), but much better looking characters since they used 9x12 font size.
Then later came EGA with 640x400 as baseline resolution and later VGA with 640x480 and the rest is probably familiar territory for most.
*sigh* Yes :)
Oh well, maybe they just used some moderator points that smelled funny and was about to expire on the best-before date.
But in the future, I will take better care to direct moderators. Maybe include links to the parent post and detailed instructions.
HEHE! Thanks, you made my day brighter :)
:)
PS. Moderators: this is where you can use your funny points
Why all this fuzz about an old Token Ring PCI card driver?
Typically, 8086/88 had CGA or MDA, 80286 had EGA or BW Mono and 80386 had VGA or BW Mono. Not very likely to have a CGA in a 80386 machine... :)
BW Mono was considered state-of-the art for desktop publishing and text work, that started to enter the stage on 80286 class machines.
This is just meant as a general idea, I'm sure lots of you have examples of a 80386 PC with CGA in it -- but it wasn't the norm.
The Cloudscape homepage: Cloudscape
And more details with links to PDF documents: Features and Benefits
I would guess that mysql would be faster for simple stuff, but Cloudscape could give it a run for it's money with support for more complex SQL.
Wouldn't know how it compares agains postgresql...
Good! This is quite a nifty little SQL engine with a lot of features. Will be exciting to follow the progress of Derby, could provide competition for mysql and postgresql.
Windows virtual memory handling is b0rked from my point of view.
1. It pages out minimised applications. Why?
2. It is much to aggressive with filesystem caching.
So when I minimise the app, it will page it out to cache in a file that I will read just once, then when I restore the window... paging hell. Because now it has to page out another app, free some filesystem cache, then page in the newly restore app.
If I could lower focus on fs cache and keep minimised apps in memory just as foreground ones (with a LRU algorithm to page out if needed) my PC would be much faster.
I haven't said that, but a columnist in Byte Magazine in the mid-80s had a rant about this.
He programmed on a Mac, and the compilation took typically 5 to 10 minutes. Enough to get a cup of coffee, check the newspaper and have a quick chat with a cow-orker. Then he got a new Mac, and it compiled the program in a minute or so. No time for coffee, no time for news, no time for smalltalk.
So the new, faster computer was too fast... he had to wait at his desk more with the new computer.