It works out to ~$2800USD according to one of the currency converters. For wow, let's see...a single volume spanning 4 disks...amazing, now where have I heard that idea before? Hmm, RAID sets, volume managers...so maybe there's something else there worth something? Uhh...great, a few Maxtor SATA drives. Yep, just what *I* want to put important info on...and wait, it gets better...USB!
Ok, so it's a consumer device obviously...I'll try to ignore my personal (including working with RAID and other enterprise storage) observations on failure rates of ATA and SATA (and Maxtor..) drives versus FC(yeah, $$$) and SCSI disks.
So now we're left with what- I can't read Japanese, but considering the size it's not doing redundant RAID of any kind, but now you have 4x the chance of losing a single disk that's going to wipe out your pr0n. True, what a giant collection of pr0n that would be, but having important data on any IDE/ATA/SATA disk without at least doing mirroring or RAID5, is just insane, even for 'a 'consumer.' I might be interested if you could buy two of these along with a RAID card, however...for $1k or so:-)
Believe me, I understand the sentiments myself. I worked for half of my normal salary for a year and a half, where I wound up leading a group of people that had been at that particular company for years that couldn't do the job I did, yet were paid double what I was. This isn't a 'pity party,' it's what happens occasionally, starting with an 'S';-) I went from _knowing_ I was working with the best people for the job, regardless of nationality, to a company who was a poster child for hiring incompetents as long as they were a minority. If they're useless, promote them while cutting back contract salaries to insanity.
Having said that, I have also worked at Cisco after the 'special job.' The difference was immeasurable; you have no idea. Going from working overtime so I could eat...to unpaid overtime but doing it because I enjoyed the job and the people I worked with. I left of my own accord to pursue something else, but _very_ tough choice- I really think the group I worked with (not 'standard Cisco' however) was possibly my most enjoyed job in a long time. I met some _excellent_ people at Cisco, of all nationalities, and more 'long-timers' (6-10+ years) than I've seen at most companies at any time...but they _could_ have left with their skills, they just liked the work and the company that much. That says something about a company, regardless of them being hit like all the other IT companies did and having to cut costs.
As far as the cost of Cisco hardware....no, it's not inexpensive by any means, and it wasn't before working there that I got the beginning of an idea of just how much work goes into their gear, and into IOS. In one way or another, almost _everything_ Cisco does, including people, including their ($$$) purchased dev hardware and servers, revolves around IOS and their hardware, period. Quite a large amount of $ to say the least. That really opened my eyes as to _why_ a Cisco 6500 or any other piece of their gear cost 'so much.' The cost to manufacture it really is insignificant, but their R&D costs are enormous in people and in support and dev hardware.
I won't argue that they could certainly 'spend more wisely' like almost any other largish company. They have projects fail or lose funding like everyone else, and yeah, inevitably 'unwise decisions' at times as well. I'll say two things here and wrap this up, but:
1. If you think a 'Linux box with a few cards' is truly going to stack up against a _real_ enterprise switch with multiple gig-ethernet 48 or 96 port blades, and the addition of ipchains/iptables and a few other pieces of software are going to stack up......then please go back to school, or something. I love Linux, I love alternatives as much as the other, and more so in many cases, but there is a _lot_ of work, and not only in software, before something like XORP comes close. I'm not discouraging it, but set some realistic expectations. 5 years after I'd started using Linux (0.8 kernel or so, 1993), I was still being asked- 'what's that system you're setting up? Linux, what's that?' by _technical_ people. Part of that is of course mindshare, media (or lack of then), etc...but by the same token there have also been some tasks that Linux was NOT the answer for (much more so then as it evolved). Anything like this is going to take a significant amount of time, PLUS (not OR) a significant amount of money as well. Building competition for NetGear? Ok, how's a few months? (Nothing against them, but mostly unmanaged simplish hardware). Replace Cisco switches and routers that are handling 400+ systems each, with the same functionality (or even close) to IOS...5 years plus 'a good media wind'..maybe.
2. As far as the supposed 'extreme H1 hiring.' Cisco lost a _lot_ of people during the bust...some areas were flat out gloomy because there are entire sections of buildings just _empty_. A large number of companies slit their throats for the future due to focusing on short term cash-flow (can you say 3 out of 4 employees
Like some others here, I was involved in MUDs for some time, before the 'MMORPG outbreak.' Several of which were extremely successful by comparison to the tons of 'yay, yet another boring uninteresting game online' types...and enjoyed success for several years, with many of the starting players still there . To put it shortly, I've got a pretty good idea of what it takes to make a game work for more than the initial checking it out phase...and it's nowhere near as easy as many people (and _companies_) think.
When the first graphical MM engines were available (Quake of course!:-), a few of us had already been thinking about building a MMORPG, and were waiting for the right technology...which never really did materialize- most that came out drew people in initially by the graphics + multi-player RPG aspect, but they all lacked depth and weren't much more than MM FPSes in the end. Some had 'the usual quests' and puzzles, but all of them fell short IMHO 'in the long run.' They had the same fundamental issues as their MUD predecessors about Player Killing, Balance, finding and dealing with people cheating (or selling items on eBay;-)...but sooner or later, most of the players moved on to 'the next best thing,' which is a _critical_ issue.
When a game team manages to finally realize that they need to focus on _long term interest_ as one of their major goals, and achieves it, I'll be amazed. They are getting better, but not there yet..
What that would take in my mind, for starters: 1. Keeping in site where the MMORPGs all came from, and what kept players interests there- text MUDs/MUSHes/MOOs etc...the successful ones told a story, or several stories, and (generally) propogated much of that throughout their entire 'universe.'
2. Np pain, no gain. Sorry, but the description of 'you die and lose nothing' makes me think the game's going to have players bailing once they explore the world (pretty pictures) enough, do a few interesting quests and puzzles...and that's about it- unless there's something truly novel to keep people paying a monthly fee for the 'extraordinary social environment' that keeps some people around..(but I expect to see more of the same instead). Skills, guilds, trades, are great...but if you level your character without worry of losing _anything_ significant...yawn, it's just a matter of time. If your char dies, big deal, go back try again. Nothing lost = gains mean less.
3. NPC interactivity. This is getting better (much)....things like NPC warrning factions/countries, natural disasters, economies...going on and being reacted to by the NPCs _in the long term_- in other words, the first time you see an NPC doing 'it's thing,' for example, crying because her husband went off to war and she suspects he's dead...have you any word of him? May be cute/interesting/whatever the first time or two you run into her...but then what? She disappears to never be seen again, or she becomes a player quest...then 'next week' (or whatever timeframe...it's the same thing again, or perhaps exactly the same story form a different NPC. There needs to be more variety in NPC actions, more logic (or 'emotions') that aren't scripted, elements of (changing) randomness thrown in that affect the world...so that when someone wants a break from 'hack n slash,' it's actually _worth_ it to 'people watch' the NPCs as much six months later as the first month when it's all new. THAT goes a long way towards longetivity...
4. Something I can't define yet, but a unique enough form of player to player interaction (besides the 'who needs a free +20 Battle Broom'(or whatever...or for sale etc)...to keep people coming back for that almost by itself.
Some of the above may be a given, and some games have certainly gone down those paths...BUT think of your favorite games, whether FPS or RPG...and the elements of each that you love, versus 'all the other games you _used_ to think were great'...until you got bored. With MMORPGs, I hesitate to
The subject says it all. While it would be certainly _nice_ to simply download the ISOs for free, think about this: 1. Sun's come a pretty far way towards at least being more 'open source friendly,' and making free downloads available for products that would normally go for (IMHO exaggerated) prices otherwise. Java, SunONE (used to be Forte), etc. Contrast this to Microsoft if you'd like- has anyone seen MS make a version of Windows free, even opened the source to (agh!) Win3.1, or given a free version of MSVC/C++? I think not. Redhat and others sell their free versions of Linux on CDs and have increased prices over the years. Companies DO need to make some money, and no matter how much we wish _everything_ were free (I do as well), I highly doubt that at $20/download they will ever come close to even recovering 10% of their investment in Solaris 9/x86.
2. It isn't a bad OS. The x86 versions of Solaris have some definite differences from the Sparc/UltaSparc version as far as development goes (some library differences), but it's a pretty stable, decent OS, and most Solaris open source software can be made to build on it fairly easily. You'll need to go to GCC/G++ for development obviously. It's quite stable, even if earlier versions (I've run x86 Solaris 2.6, 7, and 8 previously) don't have near the HCL that Linux does.
3. It's another step in the right direction for Sun. Bearing in mind that they won't make any $ at all off of the x86 line, all hey're really doing with it is trying to gain a few more supporters in the 'new to Unix' camp, which may help in Solaris/Sparc sales down the line, and get some good will in the open source/tech community. Not a bad deal all around.
Ok, ranting off. Lest I be 'attacked' for any sort of anti open-source, anti-Linux, or anything else, I've been pushing Linux at every company I've worked for as developer and admin, as well as open-source options where they're available. I've replaced many a Windows server in my time...but do realize that companies do ultimately need to regain _something_ on their investment...so those of you that still buy RedHat or Suse on CDs in order to 'give something back' to their respective companies...$20 isn't a bad deal at all.
I'll agree with others this is a scary idea...however, it seems to me the fact that Intel & company are _trying_ to do this in the first place is scarier by far than their specs on this thing- what's to stop everyone from running a modified version of VMWare for example, one that simply always 'validates' any key requests? As it's already doing actual HD emulation already it's pretty doubtful this would prove to be difficult.. Or someone writes a new device driver for NT/Windows-something that does the same thing?
Granted, I haven't looked at the specs themselves closely(are they available now?), but this seems like it will be more an enormous waste of time and money for anyone involved in it, followed by a short time of 'chaos' before the solution(s) come out- ways around this useless waste of an idea..
Most of you seem to know this, but others are acting like this is the end of the world or something. While I'm not overly thrilled about some of RedHats decisions, the network update is NOT a big deal-
Do you REALLY think that ONLY Network subscribers will be able to update their installs? Look at the past and get a grip. Years ago, you could go and purchase a RH CD, OR download it for free. TODAY, you can purchase RH 6.2 or download it...for free. This isn't likely to change.
A subscription service is a GOOD idea for some people- dedicated bandwidth for subscribers. Does that mean the 'free' update servers will get less monitoring and bandwidth? Maybe, but I live right down the road from RH and their servers are STILL slow when major updates or releases come out..no big deal. If I could receive a quarterly update CD, and CD or guaranteed bandwidth for getting new releases at a reasonable cost, I might think about buying it..note the _reasonable cost_ part, that's the only potential problem here. I'd give RH $100/year as an individual for quarterly CDs and guaranteed updates and releases, but more than that I'll contuinue to just get it for free.
If they price this as mentioned(according to usage, size of company etc), this is a GOOD THING for some people-the only 'problem' is whether or not they will offer 'personal' subscriptions that aren't insanely priced. We'll see.
Ok, not every NEW Linux or Unix user knows to disable some inetd services. The same can be said for file and print sharing on Windows. Or worse, does that wonderful 'share your printers with your friends over the Net' thing still exist on the MS downloads page?
Back on topic- you're right, not everyone knows this. At this stage of Linux distros, you need to do a bit of research for a 'relatively secure' system. It would be a HUGE plus if the distros offered a 'secure install' that simply disabled EVERYTHING you didn't specifically ask for, or perhaps a 'disable all lesser used daemons' option.
Let's face it- most distros installs seem to try to make it 'super simple and easy' for new users. Just hit enter a few times and it's basically installed. RHs 'groups' you can select leave a bit to be desired- with so many packages it gets monotonous to install via selecting every individual package, but there's no 'networked server with secure settings' option at this point, so regardless of which install mode you choose, you usually wind up removing or at least disabling things.
We're making progress, but it would be a big benefit if the installs had 'safer' default configurations. (eg don't start apache just because it's installed, don't run daytime, finger, ftpd, etc etc without the user specifically enabling them.)
Scott
Slashdot Mirror
It works out to ~$2800USD according to one of the currency converters. For wow, let's see...a single volume spanning 4 disks...amazing, now where have I heard that idea before? Hmm, RAID sets, volume managers...so maybe there's something else there worth something? Uhh...great, a few Maxtor SATA drives. Yep, just what *I* want to put important info on...and wait, it gets better...USB!
:-)
Ok, so it's a consumer device obviously...I'll try to ignore my personal (including working with RAID and other enterprise storage) observations on failure rates of ATA and SATA (and Maxtor..) drives versus FC(yeah, $$$) and SCSI disks.
So now we're left with what- I can't read Japanese, but considering the size it's not doing redundant RAID of any kind, but now you have 4x the chance of losing a single disk that's going to wipe out your pr0n. True, what a giant collection of pr0n that would be, but having important data on any IDE/ATA/SATA disk without at least doing mirroring or RAID5, is just insane, even for 'a 'consumer.' I might be interested if you could buy two of these along with a RAID card, however...for $1k or so
Believe me, I understand the sentiments myself. I worked for half of my normal salary for a year and a half, where I wound up leading a group of people that had been at that particular company for years that couldn't do the job I did, yet were paid double what I was. This isn't a 'pity party,' it's what happens occasionally, starting with an 'S' ;-) I went from _knowing_ I was working with the best people for the job, regardless of nationality, to a company who was a poster child for hiring incompetents as long as they were a minority. If they're useless, promote them while cutting back contract salaries to insanity.
Having said that, I have also worked at Cisco after the 'special job.' The difference was immeasurable; you have no idea. Going from working overtime so I could eat...to unpaid overtime but doing it because I enjoyed the job and the people I worked with. I left of my own accord to pursue something else, but _very_ tough choice- I really think the group I worked with (not 'standard Cisco' however) was possibly my most enjoyed job in a long time. I met some _excellent_ people at Cisco, of all nationalities, and more 'long-timers' (6-10+ years) than I've seen at most companies at any time...but they _could_ have left with their skills, they just liked the work and the company that much. That says something about a company, regardless of them being hit like all the other IT companies did and having to cut costs.
As far as the cost of Cisco hardware....no, it's not inexpensive by any means, and it wasn't before working there that I got the beginning of an idea of just how much work goes into their gear, and into IOS. In one way or another, almost _everything_ Cisco does, including people, including their ($$$) purchased dev hardware and servers, revolves around IOS and their hardware, period. Quite a large amount of $ to say the least. That really opened my eyes as to _why_ a Cisco 6500 or any other piece of their gear cost 'so much.' The cost to manufacture it really is insignificant, but their R&D costs are enormous in people and in support and dev hardware.
I won't argue that they could certainly 'spend more wisely' like almost any other largish company. They have projects fail or lose funding like everyone else, and yeah, inevitably 'unwise decisions' at times as well. I'll say two things here and wrap this up, but:
1. If you think a 'Linux box with a few cards' is truly going to stack up against a _real_ enterprise switch with multiple gig-ethernet 48 or 96 port blades, and the addition of ipchains/iptables and a few other pieces of software are going to stack up......then please go back to school, or something. I love Linux, I love alternatives as much as the other, and more so in many cases, but there is a _lot_ of work, and not only in software, before something like XORP comes close. I'm not discouraging it, but set some realistic expectations. 5 years after I'd started using Linux (0.8 kernel or so, 1993), I was still being asked- 'what's that system you're setting up? Linux, what's that?' by _technical_ people. Part of that is of course mindshare, media (or lack of then), etc...but by the same token there have also been some tasks that Linux was NOT the answer for (much more so then as it evolved). Anything like this is going to take a significant amount of time, PLUS (not OR) a significant amount of money as well. Building competition for NetGear? Ok, how's a few months? (Nothing against them, but mostly unmanaged simplish hardware). Replace Cisco switches and routers that are handling 400+ systems each, with the same functionality (or even close) to IOS...5 years plus 'a good media wind'..maybe.
2. As far as the supposed 'extreme H1 hiring.' Cisco lost a _lot_ of people during the bust...some areas were flat out gloomy because there are entire sections of buildings just _empty_. A large number of companies slit their throats for the future due to focusing on short term cash-flow (can you say 3 out of 4 employees
Like some others here, I was involved in MUDs for some time, before the 'MMORPG outbreak.' Several of which were extremely successful by comparison to the tons of 'yay, yet another boring uninteresting game online' types...and enjoyed success for several years, with many of the starting players still there . To put it shortly, I've got a pretty good idea of what it takes to make a game work for more than the initial checking it out phase...and it's nowhere near as easy as many people (and _companies_) think.
:-), a few of us had already been thinking about building a MMORPG, and were waiting for the right technology...which never really did materialize- most that came out drew people in initially by the graphics + multi-player RPG aspect, but they all lacked depth and weren't much more than MM FPSes in the end. Some had 'the usual quests' and puzzles, but all of them fell short IMHO 'in the long run.' They had the same fundamental issues as their MUD predecessors about Player Killing, Balance, finding and dealing with people cheating (or selling items on eBay ;-)...but sooner or later, most of the players moved on to 'the next best thing,' which is a _critical_ issue.
When the first graphical MM engines were available (Quake of course!
When a game team manages to finally realize that they need to focus on _long term interest_ as one of their major goals, and achieves it, I'll be amazed. They are getting better, but not there yet..
What that would take in my mind, for starters:
1. Keeping in site where the MMORPGs all came from, and what kept players interests there- text MUDs/MUSHes/MOOs etc...the successful ones told a story, or several stories, and (generally) propogated much of that throughout their entire 'universe.'
2. Np pain, no gain. Sorry, but the description of 'you die and lose nothing' makes me think the game's going to have players bailing once they explore the world (pretty pictures) enough, do a few interesting quests and puzzles...and that's about it- unless there's something truly novel to keep people paying a monthly fee for the 'extraordinary social environment' that keeps some people around..(but I expect to see more of the same instead). Skills, guilds, trades, are great...but if you level your character without worry of losing _anything_ significant...yawn, it's just a matter of time. If your char dies, big deal, go back try again. Nothing lost = gains mean less.
3. NPC interactivity. This is getting better (much)....things like NPC warrning factions/countries, natural disasters, economies...going on and being reacted to by the NPCs _in the long term_- in other words, the first time you see an NPC doing 'it's thing,' for example, crying because her husband went off to war and she suspects he's dead...have you any word of him? May be cute/interesting/whatever the first time or two you run into her...but then what? She disappears to never be seen again, or she becomes a player quest...then 'next week' (or whatever timeframe...it's the same thing again, or perhaps exactly the same story form a different NPC. There needs to be more variety in NPC actions, more logic (or 'emotions') that aren't scripted, elements of (changing) randomness thrown in that affect the world...so that when someone wants a break from 'hack n slash,' it's actually _worth_ it to 'people watch' the NPCs as much six months later as the first month when it's all new. THAT goes a long way towards longetivity...
4. Something I can't define yet, but a unique enough form of player to player interaction (besides the 'who needs a free +20 Battle Broom'(or whatever...or for sale etc)...to keep people coming back for that almost by itself.
Some of the above may be a given, and some games have certainly gone down those paths...BUT think of your favorite games, whether FPS or RPG...and the elements of each that you love, versus 'all the other games you _used_ to think were great'...until you got bored. With MMORPGs, I hesitate to
The subject says it all. While it would be certainly _nice_ to simply download the ISOs for free, think about this:
1. Sun's come a pretty far way towards at least being more 'open source friendly,' and making free downloads available for products that would normally go for (IMHO exaggerated) prices otherwise. Java, SunONE (used to be Forte), etc. Contrast this to Microsoft if you'd like- has anyone seen MS make a version of Windows free, even opened the source to (agh!) Win3.1, or given a free version of MSVC/C++? I think not. Redhat and others sell their free versions of Linux on CDs and have increased prices over the years. Companies DO need to make some money, and no matter how much we wish _everything_ were free (I do as well), I highly doubt that at $20/download they will ever come close to even recovering 10% of their investment in Solaris 9/x86.
2. It isn't a bad OS. The x86 versions of Solaris have some definite differences from the Sparc/UltaSparc version as far as development goes (some library differences), but it's a pretty stable, decent OS, and most Solaris open source software can be made to build on it fairly easily. You'll need to go to GCC/G++ for development obviously. It's quite stable, even if earlier versions (I've run x86 Solaris 2.6, 7, and 8 previously) don't have near the HCL that Linux does.
3. It's another step in the right direction for Sun. Bearing in mind that they won't make any $ at all off of the x86 line, all hey're really doing with it is trying to gain a few more supporters in the 'new to Unix' camp, which may help in Solaris/Sparc sales down the line, and get some good will in the open source/tech community. Not a bad deal all around.
Ok, ranting off. Lest I be 'attacked' for any sort of anti open-source, anti-Linux, or anything else, I've been pushing Linux at every company I've worked for as developer and admin, as well as open-source options where they're available. I've replaced many a Windows server in my time...but do realize that companies do ultimately need to regain _something_ on their investment...so those of you that still buy RedHat or Suse on CDs in order to 'give something back' to their respective companies...$20 isn't a bad deal at all.
I'll agree with others this is a scary idea...however, it seems to me the fact that Intel & company are _trying_ to do this in the first place is scarier by far than their specs on this thing- what's to stop everyone from running a modified version of VMWare for example, one that simply always 'validates' any key requests? As it's already doing actual HD emulation already it's pretty doubtful this would prove to be difficult.. Or someone writes a new device driver for NT/Windows-something that does the same thing?
Granted, I haven't looked at the specs themselves closely(are they available now?), but this seems like it will be more an enormous waste of time and money for anyone involved in it, followed by a short time of 'chaos' before the solution(s) come out- ways around this useless waste of an idea..
Most of you seem to know this, but others are acting like this is the end of the world or something. While I'm not overly thrilled about some of RedHats decisions, the network update is NOT a big deal-
Do you REALLY think that ONLY Network subscribers will be able to update their installs? Look at the past and get a grip. Years ago, you could go and purchase a RH CD, OR download it for free. TODAY, you can purchase RH 6.2 or download it...for free. This isn't likely to change.
A subscription service is a GOOD idea for some people- dedicated bandwidth for subscribers. Does that mean the 'free' update servers will get less monitoring and bandwidth? Maybe, but I live right down the road from RH and their servers are STILL slow when major updates or releases come out..no big deal. If I could receive a quarterly update CD, and CD or guaranteed bandwidth for getting new releases at a reasonable cost, I might think about buying it..note the _reasonable cost_ part, that's the only potential problem here. I'd give RH $100/year as an individual for quarterly CDs and guaranteed updates and releases, but more than that I'll contuinue to just get it for free.
If they price this as mentioned(according to usage, size of company etc), this is a GOOD THING for some people-the only 'problem' is whether or not they will offer 'personal' subscriptions that aren't insanely priced. We'll see.
Ok, not every NEW Linux or Unix user knows to disable some inetd services. The same can be said for file and print sharing on Windows. Or worse, does that wonderful 'share your printers with your friends over the Net' thing still exist on the MS downloads page? Back on topic- you're right, not everyone knows this. At this stage of Linux distros, you need to do a bit of research for a 'relatively secure' system. It would be a HUGE plus if the distros offered a 'secure install' that simply disabled EVERYTHING you didn't specifically ask for, or perhaps a 'disable all lesser used daemons' option. Let's face it- most distros installs seem to try to make it 'super simple and easy' for new users. Just hit enter a few times and it's basically installed. RHs 'groups' you can select leave a bit to be desired- with so many packages it gets monotonous to install via selecting every individual package, but there's no 'networked server with secure settings' option at this point, so regardless of which install mode you choose, you usually wind up removing or at least disabling things. We're making progress, but it would be a big benefit if the installs had 'safer' default configurations. (eg don't start apache just because it's installed, don't run daytime, finger, ftpd, etc etc without the user specifically enabling them.) Scott