Ehm... "again ancient aliens" and "good show" in the same sentence... Don't get me started on all the factual errors in that show.... But it's always fun to watch for a BIG laugh..
Yep... and suddenly you will issue a patch that will transform their $100 bills be a $50. If you they dont want that they can always skip the patch but then they can only use it as 5x$10 bills.
And in large parts of the world so called "shrink-wrap licenses" don't mean shit... Here in Sweden all contracts or obligations needs to be presented before the actual purchase or the licenses/contracts are not valid.
Even more that that...
"All documents related to *any* service that you have provider to the owner of the "geohot" account identified in Request No. 1 above at any time"
I read that as.. Google, send all information stored about this use. Web-searches, gmail-account(?) with all mails, any google documents stored... logs of all stuff he might have viewed on youtube.. Logs of all comments, private messages etc.
Option 1:
Computer with HDMI
Enough cpu-power to play games
Good enough GFX-card
Small case that will fit at the TV
Good wireless controllers that the games has been designed for... Extremely hard.
And all parts should be quiet..... almost impossible.. = around ~700EUR is in the lower end.. add a windows-license +50eur....
Option 2:
Get a PS3 - around 250EUR + a bit bigger harddrive and some extra controllers... so lets say 350EUR.
Pros/Cons of option 1:
+ big bonus + You have full control over everything... Do whatever you want...
- Games are not designed for specific controllers
- Games are not designed to be run on a TV.
- Secondhand-games are hard to find, uncertain if it's possible to reactivate the game.
- Expensive
- Big case at the TV, hard to find something that can fit at the tv.
- Windows GUI not designed to be used on a TV. Can be hard to read...
If running XBMC or similar this goes away but then you have one gui for games (windows desktop) and one gui for media-playback (xbmc)..
- Inability to easily control windows or other desktop-os via a standardized controller.
Pros/Cons of option 2:
+ Small and quiet box
+ Cheap
+ Lots of secondhand-games
+ Games are designed to be controlled with the supplied controllers
+ Games are designed to be run on a larger screen.
+ Now with homebrew (linux, emulators, mediacenters etc)
- No possibility to upgrade..
- New games costs alot..
And for the geeks:)
+ Possibility to play around with a Cell processor...
- Few Perl Developers Disagree. Lots of people are using it to do small fixes or administrative stuff like parsing logs etc.
- Difficult (or impossible) to maintain NO. It's the same as any other language, it's only hard to maintain if the code is written badly and without comments.
- There are better alternatives Like what? In what other language can you write a simple log-parser that creates some graph's over usages in less than 5 minutes and only requires minimal system-resources?
- Easy to write badly difficult to write well (e.g. Language doesn't lend its self to good practices) It's easy to write well but it requires someone with some degree of development-skills to do so.
Perl is easy to learn, easy to use. BUT it's too easy to get started with and that causes lots of new developers (or sysadmins with shell-scripting experience) to try it out and that can only result in lots and lots of ugly code.
But hey, those 3% should not be too hard to fix with trees and other growing plants.
And btw, concrete is releases around 8% of the human caused CO2 production... Lets go back to wood, both helping out with reducing the co2 amount AND supplying building-material..
Well, with the above it could be made available... 'if X people sign up for the series we will produce another show' and since the production company would never see the money before the produce the show you dont have to worry about those grab'n'run scams..
But also, if all shows would be airing at 'primetime' it would greatly increase the number of viewings... And you would also get rid of the war between the channels as they fight for viewers to watch their channel.. This is mainly why i setup my HTPC here and record everything before i view it... then all things i want to view is available at the click of a button instead of when the broadcaster wants me to view the show..
Hehe, well i said that they where heading in the right direction with online-distribution and 'buy the whole season' options... but i agree that $2 per episode is a bit steep for 17 minutes.. Have written about that on their blog but never got any response from them except for other viewers agreeing with that the minute-price of the show was too high.. I did read that they where looking over the pricing of the show but i never heard anything more about that, maybe that will come for the next season.. The only problem i have with these type of series is that you never know when they will get terminated, they really need to make a commitment that 'we will air atleast 4 seasons' and maybe even allow people to buy the 4 seasons even before they are produced as a way of securing capital for the production.. Ie the users decide if they want a show to continue.
But i think Sanctuary is doing a good job to test the market and hopefully after the first season they will get closer to what people are willing to pay...
Well, to distribute 50000 copies of a show via some bittorrent type of distribution is not hard and don't cost much..
But lets talk some numbers... Lets say one series has 10,000,000 online-viewers over the world, and that is a quite low figure, and put the price at 2 bucks per viewing. With lets say 24 episodes per year that would then get them around $480,000,000 per year or $20,000,000 per episode.
You always have to think world-wide about this. The problem is currently that they air something in the states and if we are lucky here in Europe we might get it 6-12 months later and sometimes they even skip a season or broadcast them out of order.
What i would love to see is some type of product with the following functionality. - Downloads the show as soon as it's released. - Compatible with a multitude of providers to avoid have to get one box per provider. (some standardized protocol for getting it or one provider that acts as a proxy between the users and the studios, like the current networks are doing) - If P2P distribution then let the users gain credits by seeding or allow them to turn of seeding if they don't want to. - Very large archive of old movies and shows!!! (preferably all movies/shows ever produced) - Fixed Monthly cost for as much shows i want to see. (maybe 2 separate accounts, one for shows and one for movies) - No quality-degradation just because it's via the net. - Must be a couple of different boxes to allow people to hook it up their preferred way (DVI/Scart/SVideo etc) and should at least have support for 5.1 surround. And the limitations i could agree too: - 2 price models maybe where one that costs $0.5 per view and one that costs $2 but that allows me to burn it to DVD/CD. - Hardware lockdown (video-decryption in the GPU maybe?), just as long as the hardware allows me to receive content from multiple provides. - Access-card requirement per provider, but then there should be multiple slots or allow to connect extra card-readers (USB?) to the box, but it must support multiple boxes so i can use the same subscription from multiple boxes. (just have one 'master' that the slave-boxes connect to, and this would reduce the cost for the small boxes to) - Free shows with advertisements, but never advertisements in shows you have to pay for.
With something like this we could get around the idiotic people at the networks that cancels good shows just because a show gets bad ratings at some off-hour... And to our and their delight every show would get prime-time airing since people can view it when they prefer and not when the networks has some slot to fill.
The only way to take down Microsoft, or make them improve their ways is through serious competition. And, I means s e r i o u s.
Not allow companies to forcefully bundle MS with their computers? (dell/hp and a few other have been taking a few steps in the right direction)
Mainstream a Linux desktop, and by mainstreaming, I mean make it commercial. Make it so Joe Notageek, and his grandmother, can install it with less clicks than it takes to install Windows. Provide apps for it.
It is really easy to install, if not even easier than windows. It just depends on what distribution you choose.. The problem is that people don't want to be classified as 'stupid' so the few of those that actually want to use it choose one of the hardest just to prove that they are 'smart'... This is a big problem because then they will tell everyone they know that it's really hard.
Mainstream a Linux server. Yes, I know there a lots out there, but again, only a few companies are really commercial. This is probably where Linux is most strong.
Redhat, Suse, Debian, Oracle.. Just to name a few large ones with their own linux-distributions.. Microsoft is just one company, but with massive amounts of advertisements.
Not a Linux problem, but a parallel issue: Mainstream Linux apps. The killers are office and games, then accounting, then graphics. Open Office is quaint, but users still want MS. If the new commercial Linux Desktop seriously competes with MS, MS will start an Office Linux version. AND, game developers will create games for it that don't suck. Creating an auto-WINE that will allow a user to load existing Windows apps in Linux would help. Getting the industry to create a logo for Windows apps that are compatible under a WINE or other emu system would be great.
This is probably the first valid point you make, but with a twist.. People don't want MS Office, but it's all they know.. I have switched quite a few people over to OpenOffice and most of them has asked how it can be free..
The games are a big problem... gnu/linux will never be adopted by the masses before games are available, and the game-developers will hesitate to develop games for such a small market, so what comes first.. Chicken or the egg...
And to the gems you wrote... I have never heard any of those comments, except when used as jokes...
The real problem is that people don't know what 'linux' is, and they think it's some obscure unusable thing that only 'nerds' use. Just take my dad, he is a complete computer illiterate and thought it was going to be something completely new to learn and hard to use, but after installing it on his pc he has become a linux-advertiser...
My view on the way i would think it will be adopted is:
1. GNU/Linux will probably first be adopted by companies that don't need/want games on the company computes and to reduce the cost of HW.
2. When people have become familiar with it at their workplace they will start to think about installing it at home opening up a market for support of home-users.
3. When there is a critical mass of home-users more companies will start shipping gnu/linux with their pc's if the user want to..
4. When we have another critical mass of computers out there companies will start to think about creating games for them, or at least provide some way (Wine maybe) to play them. One way could be to include a stable wine environment for the specific game on the install CD..
But it will never happen overnight.
And i know, windows are good on some things, different linux-distributions are good in another way.. But to sum it up, if linux would have got the same amount of advertisement as Windows have got then linux would probably be the leading OS today. In the 'OSS-world' more users = more apps/games = potential profits = more companies that supports it and helps out with the development.
Just something to think about.
And i post this as a anonymous coward since i'm probably gonna get flamebaited for it.
$5000 is nothing if you really think about it, and it wont get you any good rack-servers... Those are usually around $2-3k at least.
$5000 system.
Good chassis with only passive cooling, around $1000 CPU around $700 AMD Athlon64 FX RAM around $800 4Gb of some good low-latency memory or maybe 4x2Gb DDR2's MB around $200-$300 Soundcard $300-$400 HD's $800 for a few 400Gb SATA drives in a raid0+1. Or maybe a bit smaller and faster SCSI drives, but then you would need a scsi card too. I-RAM drive $400 Gigabytes i-ram disk so you can have a disk with real low latency, 150 for the i-ram card and 250 for the DDR memory. Screen 24" $1200 Gfx-Card $2400 ATI FireGL T2 V8650 Dual-DVI 2GB... And Nvidia Quadro goes up to around 8000 for the-top-of-the-line cards.
= around $8000-$9000
So $5000 for a maxed out machine is not really that much...
The prices are converted from swedish SEK without knowing the current rate, but it should give you a rough estimate...
Is not Virtual Server a software you run on top of windows? So then you still have to buy a windows license for that, and if you are going to be running virtual machines you probably want a few more cpu's in the box.. Don't that cost extra too?
VMWare ESX is a standalone OS that gives you almost the full performance of the actual hardware without having a underlying OS doing all sorts of stuff that will slow down the virtual-machines..
And IMHO all disk-management in windows is crap... If you want to take that route... Facts count, opinions don't.:D
Another few things that you can do in GNU/Linux... - On the fly addition of disks into a raid5. - On the fly file-system expansion. - Raid5 on partitions, create a 10Gb partition on each disk and create a raid5 on those and use the rest of the diskspace as unsecure storage for tmp files or whatever. - Free and good iscsi targets/initiators that you even can use to boot the servers via. Ie no per-machine disk-config.. If you want a bit more security then have 2 iscsi servers that you mirror the data between. Possible in windows, but then you need to buy those extras again.. - And it's all 'free' and included in the basic OS-support if you buy support from somewhere.
Redhat, as an example, supports all applications they deliver on the install cd's... With MS you have to buy almost everything extra, or at least buy the support extra.
But there is a BIG downside with all of this too... Too much freedom is hell, at least for us that have to help customers get the things going, or implement the setup they wanted... Things can get COMPLEX when they start listing feature after feature that they just have read about...
If not implemented in hardware i do see some problems like how to get access to the page as a normal user.. In software there are easier ways...
1. IRQ's on the system since it started, for one or more more devices. 2. cpu-serial 3. check the number of cpu-cycles that passed since the seed-generator started (a bit like the chaos theory if you have many processes running) 4. uptime of the system 5. local time (do have have a little randomness due to variations op the local clock-circuit) 6. serial of the harddrives 7. do some quick benchmark on the RAM and use the latency and/or bandwidth 8. do some quick benchmark on the disk of some semi-random blocks and use the latency and/or bandwidth 9. do some benchmark on the gfx-card and use the latency or bandwidth 10. Free ram / used ram / current usage of the buffer-cache 11. Check free or used diskspace on the filesystems. 12. Got a analog tv/radio-card? Tune to a semi-random frequency read some noise.. 13. read some static noise from the mic-input (with or without having a mic plugged in, usually you get some noise if you have a 'bad' soundcard) 14. ping a few semi-random addresses on the network and use the latency.
Use one or more depending on how 'random' you wanna get and use something like this: seed(step 1 + rand()) seed(step 2 + rand()) seed(step 3 + rand()) seed(step 4 + rand()) X = rand() and then use X as you final seed value...
Or maybe just use the number of hits per second you get when being \.'ed:)
Slashdot Mirror
Use your imagination
Ehm... "again ancient aliens" and "good show" in the same sentence... Don't get me started on all the factual errors in that show.... But it's always fun to watch for a BIG laugh..
http://arstechnica.com/tech-policy/news/2010/09/claimed-hdcp-master-key-leak-could-be-fatal-to-drm-scheme.ars
the DRM schemes are not for protecting content, it's for making high-level management think they are protecting content.
Yep... and suddenly you will issue a patch that will transform their $100 bills be a $50. If you they dont want that they can always skip the patch but then they can only use it as 5x$10 bills.
And in large parts of the world so called "shrink-wrap licenses" don't mean shit... Here in Sweden all contracts or obligations needs to be presented before the actual purchase or the licenses/contracts are not valid.
Even more that that... "All documents related to *any* service that you have provider to the owner of the "geohot" account identified in Request No. 1 above at any time" I read that as.. Google, send all information stored about this use. Web-searches, gmail-account(?) with all mails, any google documents stored... logs of all stuff he might have viewed on youtube.. Logs of all comments, private messages etc.
Option 1:
Computer with HDMI
Enough cpu-power to play games
Good enough GFX-card
Small case that will fit at the TV
Good wireless controllers that the games has been designed for... Extremely hard.
And all parts should be quiet..... almost impossible..
= around ~700EUR is in the lower end.. add a windows-license +50eur....
Option 2:
Get a PS3 - around 250EUR + a bit bigger harddrive and some extra controllers... so lets say 350EUR.
Pros/Cons of option 1:
+ big bonus + You have full control over everything... Do whatever you want...
- Games are not designed for specific controllers
- Games are not designed to be run on a TV.
- Secondhand-games are hard to find, uncertain if it's possible to reactivate the game.
- Expensive
- Big case at the TV, hard to find something that can fit at the tv.
- Windows GUI not designed to be used on a TV. Can be hard to read...
If running XBMC or similar this goes away but then you have one gui for games (windows desktop) and one gui for media-playback (xbmc)..
- Inability to easily control windows or other desktop-os via a standardized controller.
Pros/Cons of option 2:
+ Small and quiet box
+ Cheap
+ Lots of secondhand-games
+ Games are designed to be controlled with the supplied controllers
+ Games are designed to be run on a larger screen.
+ Now with homebrew (linux, emulators, mediacenters etc)
- No possibility to upgrade..
- New games costs alot..
And for the geeks :)
+ Possibility to play around with a Cell processor...
oops
- Few Perl Developers
Disagree. Lots of people are using it to do small fixes or administrative stuff like parsing logs etc.
- Difficult (or impossible) to maintain
NO. It's the same as any other language, it's only hard to maintain if the code is written badly and without comments.
- There are better alternatives
Like what? In what other language can you write a simple log-parser that creates some graph's over usages in less than 5 minutes and only requires minimal system-resources?
- Easy to write badly difficult to write well (e.g. Language doesn't lend its self to good practices)
It's easy to write well but it requires someone with some degree of development-skills to do so.
Perl is easy to learn, easy to use. BUT it's too easy to get started with and that causes lots of new developers (or sysadmins with shell-scripting experience) to try it out and that can only result in lots and lots of ugly code.
Well, did you think about what kind of crap music they succeed to sell?
97% of non-human or non-animal?
But hey, those 3% should not be too hard to fix with trees and other growing plants.
And btw, concrete is releases around 8% of the human caused CO2 production... Lets go back to wood, both helping out with reducing the co2 amount AND supplying building-material..
Well, with the above it could be made available... 'if X people sign up for the series we will produce another show' and since the production company would never see the money before the produce the show you dont have to worry about those grab'n'run scams..
But also, if all shows would be airing at 'primetime' it would greatly increase the number of viewings... And you would also get rid of the war between the channels as they fight for viewers to watch their channel..
This is mainly why i setup my HTPC here and record everything before i view it... then all things i want to view is available at the click of a button instead of when the broadcaster wants me to view the show..
I think it would be enough that they forced him write snail-mail to everyone he ever spammed..Handwritten ofcourse.. 14 hours a day..
Hehe, well i said that they where heading in the right direction with online-distribution and 'buy the whole season' options... but i agree that $2 per episode is a bit steep for 17 minutes.. Have written about that on their blog but never got any response from them except for other viewers agreeing with that the minute-price of the show was too high.. I did read that they where looking over the pricing of the show but i never heard anything more about that, maybe that will come for the next season..
The only problem i have with these type of series is that you never know when they will get terminated, they really need to make a commitment that 'we will air atleast 4 seasons' and maybe even allow people to buy the 4 seasons even before they are produced as a way of securing capital for the production.. Ie the users decide if they want a show to continue.
But i think Sanctuary is doing a good job to test the market and hopefully after the first season they will get closer to what people are willing to pay...
Oh, forgot to mention. This is how i view TV today, but using normal broadcasts for the distribution .. And from time to time i download something ;) ...
One show that has kicked off in the right direction is http://www.sanctuaryforall.com/ and it's only available online...
Well, to distribute 50000 copies of a show via some bittorrent type of distribution is not hard and don't cost much..
But lets talk some numbers...
Lets say one series has 10,000,000 online-viewers over the world, and that is a quite low figure, and put the price at 2 bucks per viewing. With lets say 24 episodes per year that would then get them around $480,000,000 per year or $20,000,000 per episode.
You always have to think world-wide about this. The problem is currently that they air something in the states and if we are lucky here in Europe we might get it 6-12 months later and sometimes they even skip a season or broadcast them out of order.
What i would love to see is some type of product with the following functionality.
- Downloads the show as soon as it's released.
- Compatible with a multitude of providers to avoid have to get one box per provider. (some standardized protocol for getting it or one provider that acts as a proxy between the users and the studios, like the current networks are doing)
- If P2P distribution then let the users gain credits by seeding or allow them to turn of seeding if they don't want to.
- Very large archive of old movies and shows!!! (preferably all movies/shows ever produced)
- Fixed Monthly cost for as much shows i want to see. (maybe 2 separate accounts, one for shows and one for movies)
- No quality-degradation just because it's via the net.
- Must be a couple of different boxes to allow people to hook it up their preferred way (DVI/Scart/SVideo etc) and should at least have support for 5.1 surround.
And the limitations i could agree too:
- 2 price models maybe where one that costs $0.5 per view and one that costs $2 but that allows me to burn it to DVD/CD.
- Hardware lockdown (video-decryption in the GPU maybe?), just as long as the hardware allows me to receive content from multiple provides.
- Access-card requirement per provider, but then there should be multiple slots or allow to connect extra card-readers (USB?) to the box, but it must support multiple boxes so i can use the same subscription from multiple boxes. (just have one 'master' that the slave-boxes connect to, and this would reduce the cost for the small boxes to)
- Free shows with advertisements, but never advertisements in shows you have to pay for.
With something like this we could get around the idiotic people at the networks that cancels good shows just because a show gets bad ratings at some off-hour... And to our and their delight every show would get prime-time airing since people can view it when they prefer and not when the networks has some slot to fill.
Now i finally understand how they calculate the amount lost to piracy!
Nope... The slashdotters are the only ones that will survive cause of our tinfoil-hats that will block zombies from sensing our brains! :)
oops... i seem to be brave today ;)
Not allow companies to forcefully bundle MS with their computers? (dell/hp and a few other have been taking a few steps in the right direction)
Mainstream a Linux desktop, and by mainstreaming, I mean make it commercial. Make it so Joe Notageek, and his grandmother, can install it with less clicks than it takes to install Windows. Provide apps for it.
It is really easy to install, if not even easier than windows. It just depends on what distribution you choose.. The problem is that people don't want to be classified as 'stupid' so the few of those that actually want to use it choose one of the hardest just to prove that they are 'smart'... This is a big problem because then they will tell everyone they know that it's really hard.
Mainstream a Linux server. Yes, I know there a lots out there, but again, only a few companies are really commercial. This is probably where Linux is most strong.
Redhat, Suse, Debian, Oracle.. Just to name a few large ones with their own linux-distributions.. Microsoft is just one company, but with massive amounts of advertisements.
Not a Linux problem, but a parallel issue: Mainstream Linux apps. The killers are office and games, then accounting, then graphics. Open Office is quaint, but users still want MS. If the new commercial Linux Desktop seriously competes with MS, MS will start an Office Linux version. AND, game developers will create games for it that don't suck. Creating an auto-WINE that will allow a user to load existing Windows apps in Linux would help. Getting the industry to create a logo for Windows apps that are compatible under a WINE or other emu system would be great.
This is probably the first valid point you make, but with a twist.. People don't want MS Office, but it's all they know.. I have switched quite a few people over to OpenOffice and most of them has asked how it can be free.. The games are a big problem... gnu/linux will never be adopted by the masses before games are available, and the game-developers will hesitate to develop games for such a small market, so what comes first.. Chicken or the egg...
And to the gems you wrote... I have never heard any of those comments, except when used as jokes...
The real problem is that people don't know what 'linux' is, and they think it's some obscure unusable thing that only 'nerds' use. Just take my dad, he is a complete computer illiterate and thought it was going to be something completely new to learn and hard to use, but after installing it on his pc he has become a linux-advertiser...
My view on the way i would think it will be adopted is:
1. GNU/Linux will probably first be adopted by companies that don't need/want games on the company computes and to reduce the cost of HW.
2. When people have become familiar with it at their workplace they will start to think about installing it at home opening up a market for support of home-users.
3. When there is a critical mass of home-users more companies will start shipping gnu/linux with their pc's if the user want to..
4. When we have another critical mass of computers out there companies will start to think about creating games for them, or at least provide some way (Wine maybe) to play them. One way could be to include a stable wine environment for the specific game on the install CD..
But it will never happen overnight.
And i know, windows are good on some things, different linux-distributions are good in another way.. But to sum it up, if linux would have got the same amount of advertisement as Windows have got then linux would probably be the leading OS today. In the 'OSS-world' more users = more apps/games = potential profits = more companies that supports it and helps out with the development.
Just something to think about.
And i post this as a anonymous coward since i'm probably gonna get flamebaited for it.
Oh? So you have sunlight all day/year around?
;)
You got to include the nights in the calculations too
$5000 is nothing if you really think about it, and it wont get you any good rack-servers... Those are usually around $2-3k at least.
$5000 system.
Good chassis with only passive cooling, around $1000
CPU around $700 AMD Athlon64 FX
RAM around $800 4Gb of some good low-latency memory or maybe 4x2Gb DDR2's
MB around $200-$300
Soundcard $300-$400
HD's $800 for a few 400Gb SATA drives in a raid0+1. Or maybe a bit smaller and faster SCSI drives, but then you would need a scsi card too.
I-RAM drive $400 Gigabytes i-ram disk so you can have a disk with real low latency, 150 for the i-ram card and 250 for the DDR memory.
Screen 24" $1200
Gfx-Card $2400 ATI FireGL T2 V8650 Dual-DVI 2GB... And Nvidia Quadro goes up to around 8000 for the-top-of-the-line cards.
= around $8000-$9000
So $5000 for a maxed out machine is not really that much...
The prices are converted from swedish SEK without knowing the current rate, but it should give you a rough estimate...
Would be cool to run the window-manager on the gfx-card instead of the cpu.. :)
Is not Virtual Server a software you run on top of windows? So then you still have to buy a windows license for that, and if you are going to be running virtual machines you probably want a few more cpu's in the box.. Don't that cost extra too?
:D
VMWare ESX is a standalone OS that gives you almost the full performance of the actual hardware without having a underlying OS doing all sorts of stuff that will slow down the virtual-machines..
And IMHO all disk-management in windows is crap... If you want to take that route... Facts count, opinions don't.
Another few things that you can do in GNU/Linux...
- On the fly addition of disks into a raid5.
- On the fly file-system expansion.
- Raid5 on partitions, create a 10Gb partition on each disk and create a raid5 on those and use the rest of the diskspace as unsecure storage for tmp files or whatever.
- Free and good iscsi targets/initiators that you even can use to boot the servers via. Ie no per-machine disk-config.. If you want a bit more security then have 2 iscsi servers that you mirror the data between. Possible in windows, but then you need to buy those extras again..
- And it's all 'free' and included in the basic OS-support if you buy support from somewhere.
Redhat, as an example, supports all applications they deliver on the install cd's... With MS you have to buy almost everything extra, or at least buy the support extra.
But there is a BIG downside with all of this too... Too much freedom is hell, at least for us that have to help customers get the things going, or implement the setup they wanted... Things can get COMPLEX when they start listing feature after feature that they just have read about...
If not implemented in hardware i do see some problems like how to get access to the page as a normal user.. In software there are easier ways...
:)
1. IRQ's on the system since it started, for one or more more devices.
2. cpu-serial
3. check the number of cpu-cycles that passed since the seed-generator started (a bit like the chaos theory if you have many processes running)
4. uptime of the system
5. local time (do have have a little randomness due to variations op the local clock-circuit)
6. serial of the harddrives
7. do some quick benchmark on the RAM and use the latency and/or bandwidth
8. do some quick benchmark on the disk of some semi-random blocks and use the latency and/or bandwidth
9. do some benchmark on the gfx-card and use the latency or bandwidth
10. Free ram / used ram / current usage of the buffer-cache
11. Check free or used diskspace on the filesystems.
12. Got a analog tv/radio-card? Tune to a semi-random frequency read some noise..
13. read some static noise from the mic-input (with or without having a mic plugged in, usually you get some noise if you have a 'bad' soundcard)
14. ping a few semi-random addresses on the network and use the latency.
Use one or more depending on how 'random' you wanna get and use something like this:
seed(step 1 + rand())
seed(step 2 + rand())
seed(step 3 + rand())
seed(step 4 + rand())
X = rand()
and then use X as you final seed value...
Or maybe just use the number of hits per second you get when being \.'ed