I worked for Healtheon back in the proverbial day. They had all of the software that the UK wanted to deploy up and working back before 2000. Lameness!
Oh well. $24B in venture funding could have produced one hell of a great electronic health care system...
I guess I'm missing something. Things being "enterpise" and "mission critical" has nothing to do with the medium, rather the software. Clearly Veritas Netbackup solves your problems on a software level. As far as off-site goes, I suppose a $20K/month Cogent Gigabit internet connection (or cheaper private gigabit connection) can take care of that.
I'm not interested in arguing the merits of backup software or replication/redundancy strategy. I'm complaining about Tape as a medium and tape drives as a transport.
650Mb/sec is pretty crappy. 2 streams "flooding" a server at 1 Gigabit is pretty crappy too. I'm sure Veritas can use a $200K 120TB rack of 40 boxes in parallel and saturate them at 4Gb/sec or (with gig cards) at 40Gb/sec. (making two copies of the data)
As far as costs being relative, I agree whole heartedly. The ability for an administrator to click on a button that does a full or incremental backup of hundreds of servers really has nothing to do with what the ultimate medium is that the data ends up on.
Cause tape doesn't work, simple as that. It's a crappy, slow and expensive medium. Why anyone at all, home users or enterprises still use it is beyond me.
The recent slashdot article about Capricorn selling you a 120TB rack of spinning disks with aggregate throughput of 40Gbps for $200K should put the final nail in tapes coffin.
Let's say you buy 2 for redundancy, show me a tape backup system that runs at 80Gbps and stores 240TB using only 2 racks worth of space and zero human physical intervention.
-B
I use these every day at the Internet Archive
on
3 Terabytes, 80 Watts
·
· Score: 2, Interesting
They are very low power, reasonably easy to work on and not hot-swap. They are not going to win any speed contests, but they store data cheaply and make it accessible at reasonable rates.
The VIA based systems are PATA, so they will not be RAID5 friendly because RAID5 on master/slave is simply stupid.
They are reasonable fast at delivering the data. Having only a 100Mb connection means that it takes a really really long time to fill it. At the Internet Archive we use the nodes in JBOD and do the redundancy at the application layer.
If I were doing it at home I would probably try out ATA over Ethernet and make all of these hosts/drives targets. Mirroring is always another option.
I had a TI99/4A given to me by my parents for Christmas. I was 9 and in 3rd grade. I got a subscription to HCM (Home Computer Magazine) and initially had to get my mom to type in the code at the back since she could type so much faster. I would modify the programs, save them to tape, etc.
In school, we had an Apple2 lab with Logo. People played Oregon Trail, had Math drills and such, but Logo was the shit.
By the time my friends and I hit Junior high, Quick BASIC was no big deal. In high school, we did Pascal, C and such.
Given that all of what I'm talking about above can be done on one of those $100 laptops, I don't see why people wouldn't do it.
You will get responses from people with good and bad experiences, but they are all jaded by their small particular case. After seeing what can happen with dozens of machines (8 drive and 4 drive) running Linux software RAID5, here is some concrete advice.
First, ensure that all of the drives are IDE masters. Don't double up slaves and masters.
Secondly, DON'T create gigantic partitions on each oft he 250's and then RAID them together, you will get bitten, and bitten hard.
Here's the skinny...
1) Ensure that your motherboard/IDE controllers will return SMART status information. Make sure you install the smartmon tools, configure them to run weekly self tests, and ensure you have smartd running so that you get alerted to potentially failing drives ahead of time.
2) Partition your 250GB drives into 40 GB partitions. Then use RAID5 to pull together the partitions across the drives. If you want a giant volume, create a Linear RAID group of all of the RAID5 groups you created and create the filesystem on top of that.
Here's why, this is the juice.
To keep it simple, let's say there are 20 secotrs per drive. When a drive gets an uncorrectable error on a sector, it will be kicked out of the array. By partitioning the drive into 5 or 6 partitions, let's say hd(a,c,e,g,i,k,l)1 are in one of the RAID5 groups, which contain sectors 1-4 (out of the fake 20 we made up earlier)
If sector 2 goes bad on/dev/hda1, Linux software RAID5 will kick/dev/hda1 out of the array. Now, it's likely that sector 11 might be bad on/dev/hdc. If you hadn't divided up the partitions, you would lose a second disk out of the array during a rebuild.
By partitioning the disks you localize the failures a little, thus creating a more likely recovery scenario.
You wind up with a few RAID5 sets that are more resilient to multiple drive failures.
If you are using a hot spare, your rebuild time will also be less, at least for the RAID5 set that failed.
I hope this makes sense.
My advice to you is to bite the bullet and simply mirror the disks. That way, no matter how badly they fail you'll have some chance of getting some of the data off.
You cracker. Sitting with my mom typing in code from from the back pages of HCM magazine for my TI-99/4A when I was 9 are very happy memories indeed. Going to my dad's work and printing out text art of the Enterprise from Star Trek on fanfold paper from files on the mainframe when I was 8 was the bomb!
Does anyone remember this? It was so kick ass for the day... It was the late 90's wasn't it?
I remember downloading a tarball, throwing it onto an apache server and viola! I pointed a browser at it and up came a nice MS Word clone, all in Java. It was beatiful. It came, it went, and poof, 5 years later people are looking for it.
I'm sick and tired of all these weenies, be it from the enterprise or from their living room whining and complaining about storage and RAID and how do I do this, how do I do that.
Here's the scoop, or poop, or whatever.
Buy a case, buy a systemboard with X number of connectors. Connect all the drives, format them as one big partition each.
Buy a little via dinky dude to be a netboot server, do root NFS or whatever (that way no OS partitions no your storage box.)
Now, buy a second one storage node.
You have sitting in front of you, 2 boxes, each with 4, 6 8 or whatever 400GB Hitachi drives.
DON'T! do raid! You're talking commodity IDE, chances are that one little bad block around GB 3 on/dev/hda and another little bad block around GB 310 on/dev/hdc will mean your RAID5 array is screwed.
Mirror each disk onto another disk on the second machine.
Even if you are carrying your case to a friends house and drop it, chances are, if the drives aren't raided, you'll at least be able to get some/most of the data off.
Use a filesystem with distributed metadata (reiser and XFS if I'm not mistaken)
I guarantee you will have a catastrophic failure with RAID when using cheap IDE disks.
SO I'm sure it's worth millions, but wouldn't it be cool if you were able to have a website that was robust enough to survive if it was mentioned in a slashdot article... an article with only 80 comments, no less.
So I would like, get some bandwidth, and like, get some servers that didn't suck, and maybe like hire this dude to write an app that didn't suck. So like, if you do that, then there's like this weird chance that when you get bombed with traffic people will still get a page back...
I dunno though, I think it'll cost like $20 million. Oh my gawd, so hardware alone would cost like $5K, and bandwidth would be aboue $1K/month... (cogent) oh, and then there's the dude, another $100K/year.... so for $20M you should be able to keep this running for about 178 years. Wheee!!!
Any one of these can swing fuel consumption. Driving my Jeep like a granny, shifting once I hit 1,500 rpm (not a problem since it has so much torque) gets me close to 18 or 20 mpg, even with big tires. If I drive around like "normal" people I'd be getting closer to 11 or 12 mpg in the city.
However, having lived in a suburb where I would have to drive 45 miles a day just for work, to now driving less than 5 miles/day if at all has meant that by moving to "the city" I have reduced my personal emissions/day tremendously.
One interesting to note about tire pressure. I left San Francisco (65 degrees) and pumped the tires to 33psi cold. By the time I got down to Fresno it was getting a bit squirrely. When I checked the pressure it had hit 40psi! I let it back down to 32psi and continued onward. Coming down the 10 heading into Phoenix it felt weird again. Lo and behold, the 115 degree ambient temp had pushed the pressure back up to 38.
The point being that you need to keep your tires inflated to the pressure that gives you the lowest rolling resistance. I'll be changing out the 33x12.5 to 33x9.5 tires.
Ok, since this is your personal fileserver, I assume you are talking about storing music, movies, files, whatever. I also take the word "personal" to mean that if you are in the middle of watching 2001 and an error on your drive causes an interruption in the movie you are not going to be overly upset. You may be sad about having wasted certain 2001 enhacing substances, but anyway.
I would say rsync and scrub the disks.
Every night, simply rsync one disk onto the other one. You have two copies, you have no hardware or software raid configurations to deal with. It's cake to recover and to resync once you havea chance to run down to the store to pick up a drive.
Weekly, or while you are at work, run something like:
find . -type f | xargs md5sum >/filemanifest
You'll have a history of your disk and can track bitrot or whatever. You'll also be touching every single bit on the disk. No better way to detect errors, might want to run badblocks.
You are adding complexity and other people software with other peoples bugs trying to use weirdo ide hardware raid, etc. Linux sw raid1 is great because the disks aren't special if you split them, they are just plain ext2, or whatever you put on there.
Quick history lesson first. Ok, it's a bad one since I can't remeber the exact year, but it's right after Windows 95 came out with NetMeeting. I was using dial-up on the east coast, while my friend was dialed up in the midwest. We were playing Yahoo Chess on-line and had a netmeeting connection opened.
Conversations were more than acceptable and we would sit there for an hour or more, not loose a call and converse about all sorts of things while playing chess. On Dial-up!! I realize this was a point-to-point connection, but so what, it was 8 years ago with dial-up!!
I've used vonage in the Bay Area and it does work great over fast connections (dsl and up.)
A co-worker took his vonage phone to France and plugged it into his hotels broadband connection. The delay was over 1.5 seconds, probably 3 seconds. After less than a minute of conversation we simply started an interleaved conversation that took the gap into account (cheesy, I know, but afterwards I thought it was a really cool and efficient way to maximize productivity over lag (kind of like IRC, except with voice I guess.)
Regarding the troubles with Wifi, I have a feeling it has more to do with the queueing and QoS on the Wifi network than anything else. If it's really as bad as people described, then it should be totally impossible to play games over Wifi, etc...
There are cell phones coming out which will switch to VoIP when near an accessp point, and I'm sure they've done some testing on these things.
Please update the front page... I know that slashdot is a timesink, and severely decreases productivity. Yes, I realize it also has a healing effect on the damage psyches of those addicted to reading it every so often. However, this story reaches a new low.
Some dude, who's failed to spend a couple of days figuring out what's really going on posts an inflammatory and stress inducing story on the front page of this green rag.
From posted comments (no I never read slashdot comments) there appear to be binary drivers and XF86 drivers available.
An update on the front page would probably save abou 358,654 hours of wasted time by slashdot readers.
http://www.schneider-digital.de/html/download_at i. html
$150 for a wireless bridge that will work at several miles with a $50 antenna and you'll get about 3 Mb/sec.
Lets see, $200/year vs $50/month for DSL?
This is such a nobrainer its not even funny. I could put an Omni on my roof and 20 households could share the connection by putting a bridge in their front window.
Although W2K performance is a little sketchy, Linux seems to do quite well. I have the Japanese version., I don't know whether it is out here yet. It has the Transmeta Crusoe 667Mhz chip, 256Mb RAM, and 30 Gig hard drive. Linux runs great, including pcmcia and usb. The camera is supposedly supported under 2.4.7, I know, i'm lame for not trying yet. I have the quad capacity battery, which still leaves the damn little thing so damn portable. Coupled with a pcmcia wireless web device *used to be ricochet, we'll see what I will use next.
Anyway, the battery allowed me over 10 hours of use. I played around a bit in the airport in tokyo, closed the lid, hopped on the plane... opened it as soon as I could. Watched a bunch of TV episodes on the plane, listened to a bunch of music (it was on during the entire portion of the flight that laptops were allowed.)
After getting home I used it for a few more hours.
As a portable computing device, I really love it. It's not the fastest, it's not the smallest... actually, maybe it is... but its the first laptop that I could type on that I carry around in my non laptop bag along with the rest of the equipment required for a stroll in Golden Gate Park.
People keep talking abut their mother. "Let me tell you about my mother." -Leon, Blade Runner.
Windows sucks, as does MacOS. GNOME is ok, except for that damn panel. I don't think that modelling a desktop environment around a users expected behavior is what is needed. It's been tried before. The Mac took the symbols approach, and Microsoft soon followed. The problem with either of these two systems is that they were written in a time when we were trying to make difficult machines talk to simple people who really didn't know what to expect.
This has lead us to our current situation. We have people that have "learned" Windows, or "learned" the MacOS. They have learned it simply to use their applications and write e-mail to their grandchildren.
I've had the fortune of getting a couple of people behind a computer running Linux (enlightenment once and fvwm another time) I've found that the gnome panel is nothing more than a crutch for them to get into the swing of things. And that's only if they come from a Mac or Windows world. In both cases, once they saw my desktop, no panel, epplets and the pager and icon box, they wanted that... why? because it looked cool. As far as the user interface is concerned, I think that on the application level, GNOME has it right.
Common themes, etc. But in terms of what a UI really needs, it's this: A meaningful gateway into the applications a processes running on top of your operating system. The enlightenment menuing is brilliant. Why should anyone have to go to some corner foot/start button for anything?
Between that and a simple lesson on gtop (for misbehaving applications) I believe the UI is perfect. A user should be asked to do nothing more than know how to launch applications and how to get rid of them when they are failing.
In Windows and MacOS this isn't an option, since you usually have to reboot in order to clear up any "weirdness" you may be experiencing. This doesn't leave the posibility open for a system that will be up for a year that you simply log in and out of (if that) or simply run your applications on.
People talk about how isolated the user is in MacOS and Windows, but really, the user is never more isolated from an operating system than the stability of the OS allows. Six desktops with all of your application usually open and at the ready is far more productive and user friendly than something where you feel like you need to start over every day or two.
The idea of a system that is so customizeable that it can be rendered almost inusable by someone other than "you" is perfect. Let the UI be an extension of the person and how they wish to work. This is in sharp contrast to the "we know what's best for you and this is how you are going to work" mantra being chanted by the other side.
Back to my mother. There are two directions to come at this. In several cases where the user has already known windows, I've found that they suddenly have this revelation that the UI is nothing more than another application (as it should be.) The next revelation is usually how much easier X/enlightenment or even fvwm are, in the sense that they are not in your face.
Anyone that has used linux as a desktop OS for more than a few days (i.e. feels comfortable, can kill a java dead netscape, etc) feels like Windows/Mac is lame. They can still use it, obviously, but they feel betrayed at the lack of power that has been handed to them in these sugar coated excersizes of futility.
If my mother has never used a computer before, it is even easier to sit her down in front Linux. You can safely say that she should not be afraid. She cannot possibly break the system. Removing that fear from a new user is one of the greatest enabling factors in learning any kind of technology.
I believe that customizability and stability should be the foremost thought going into any UI. A close second, I would argue, is transparency. I fear that discussions about what a UI needs to have will inevitably lead down a road that creates YADDUI (Yet Another Dumbed Down User Interface) to which some subset of the masses must conform. It doesn't get any simpler than clicking "somewhere" on your desktop to bring up a menu...
I don't like the tone of Mr. Katz when he suggests that these authors didn't anticipate the Ueber Corporation. In a lot of SF, even in something like Clans of the Alphane Moon (Phillip K. Dick) the insane asylum is the result of an ueber corporation. I don't think there is anything anyone missed.... on the contrary, I think it is a running theme that results in the weaponry and technology that is present in these novels. The end of government....
Believe it or not, 24/96 I believe is worth it. 16/44 is just plain lame. While I don't claim to be able to hear the samples, anyone that "listens" to music feels the inherent supoeriority of vinyl. Now, the problem is that the qualities of vinyl that people like (analog, fullness, more natural stereo seperation) are mitigated by the imperfections of atoms. Bits are the answer, but CD's weren't it.
24/96 is ~ digital vinyl, or digital analog... I would argue that anything beyond that is in fact superfluous, but CD's are very lacking when it comes to accurately reproducing the original.
Why are we willing to put up with MP3's? Because they fit onto devices we can afford, and are transferrable at bandwidth's that we can afford... in the long run, neither of these are an issue.
Slashdot Mirror
if they care enough for you to have an extraordinary experience when you have to use the restroom, that means they really care.
Too bad all these companies are making millions off the forethought of Brewster. They should each donate 1 shipping container to the Internet Archive.
I worked for Healtheon back in the proverbial day. They had all of the software that the UK wanted to deploy up and working back before 2000. Lameness!
Oh well. $24B in venture funding could have produced one hell of a great electronic health care system...
Have clue, will travel.
I guess I'm missing something. Things being "enterpise" and "mission critical" has nothing to do with the medium, rather the software. Clearly Veritas Netbackup solves your problems on a software level. As far as off-site goes, I suppose a $20K/month Cogent Gigabit internet connection (or cheaper private gigabit connection) can take care of that.
I'm not interested in arguing the merits of backup software or replication/redundancy strategy. I'm complaining about Tape as a medium and tape drives as a transport.
650Mb/sec is pretty crappy. 2 streams "flooding" a server at 1 Gigabit is pretty crappy too. I'm sure Veritas can use a $200K 120TB rack of 40 boxes in parallel and saturate them at 4Gb/sec or (with gig cards) at 40Gb/sec. (making two copies of the data)
As far as costs being relative, I agree whole heartedly. The ability for an administrator to click on a button that does a full or incremental backup of hundreds of servers really has nothing to do with what the ultimate medium is that the data ends up on.
Cause tape doesn't work, simple as that. It's a crappy, slow and expensive medium. Why anyone at all, home users or enterprises still use it is beyond me.
The recent slashdot article about Capricorn selling you a 120TB rack of spinning disks with aggregate throughput of 40Gbps for $200K should put the final nail in tapes coffin.
Let's say you buy 2 for redundancy, show me a tape backup system that runs at 80Gbps and stores 240TB using only 2 racks worth of space and zero human physical intervention.
-B
They are very low power, reasonably easy to work on and not hot-swap. They are not going to win any speed contests, but they store data cheaply and make it accessible at reasonable rates.
The VIA based systems are PATA, so they will not be RAID5 friendly because RAID5 on master/slave is simply stupid.
They are reasonable fast at delivering the data. Having only a 100Mb connection means that it takes a really really long time to fill it. At the Internet Archive we use the nodes in JBOD and do the redundancy at the application layer.
If I were doing it at home I would probably try out ATA over Ethernet and make all of these hosts/drives targets. Mirroring is always another option.
The year, 1981.
The place, suburbs of Minneapolis, MN.
I had a TI99/4A given to me by my parents for Christmas. I was 9 and in 3rd grade. I got a subscription to HCM (Home Computer Magazine) and initially had to get my mom to type in the code at the back since she could type so much faster.
I would modify the programs, save them to tape, etc.
In school, we had an Apple2 lab with Logo. People played Oregon Trail, had Math drills and such, but Logo was the shit.
By the time my friends and I hit Junior high, Quick BASIC was no big deal. In high school, we did Pascal, C and such.
Given that all of what I'm talking about above can be done on one of those $100 laptops, I don't see why people wouldn't do it.
You will get responses from people with good and bad experiences, but they are all jaded by their small particular case. After seeing what can happen with dozens of machines (8 drive and 4 drive) running Linux software RAID5, here is some concrete advice.
/dev/hda1, Linux software RAID5 will kick /dev/hda1 out of the array. Now, it's likely that sector 11 might be bad on /dev/hdc. If you hadn't divided up the partitions, you would lose a second disk out of the array during a rebuild.
First, ensure that all of the drives are IDE masters. Don't double up slaves and masters.
Secondly, DON'T create gigantic partitions on each oft he 250's and then RAID them together, you will get bitten, and bitten hard.
Here's the skinny...
1) Ensure that your motherboard/IDE controllers will return SMART status information. Make sure you install the smartmon tools, configure them to run weekly self tests, and ensure you have smartd running so that you get alerted to potentially failing drives ahead of time.
2) Partition your 250GB drives into 40 GB partitions. Then use RAID5 to pull together the partitions across the drives. If you want a giant volume, create a Linear RAID group of all of the RAID5 groups you created and create the filesystem on top of that.
Here's why, this is the juice.
To keep it simple, let's say there are 20 secotrs per drive. When a drive gets an uncorrectable error on a sector, it will be kicked out of the array. By partitioning the drive into 5 or 6 partitions, let's say hd(a,c,e,g,i,k,l)1 are in one of the RAID5 groups, which contain sectors 1-4 (out of the fake 20 we made up earlier)
If sector 2 goes bad on
By partitioning the disks you localize the failures a little, thus creating a more likely recovery scenario.
You wind up with a few RAID5 sets that are more resilient to multiple drive failures.
If you are using a hot spare, your rebuild time will also be less, at least for the RAID5 set that failed.
I hope this makes sense.
My advice to you is to bite the bullet and simply mirror the disks. That way, no matter how badly they fail you'll have some chance of getting some of the data off.
You cracker. Sitting with my mom typing in code from from the back pages of HCM magazine for my TI-99/4A when I was 9 are very happy memories indeed. Going to my dad's work and printing out text art of the Enterprise from Star Trek on fanfold paper from files on the mainframe when I was 8 was the bomb!
Here is a google groups link to a discussion about it. It was kick ass.
http://groups.google.com/groups?q=corel+office+
Does anyone remember this? It was so kick ass for the day... It was the late 90's wasn't it?
I remember downloading a tarball, throwing it onto an apache server and viola! I pointed a browser at it and up came a nice MS Word clone, all in Java. It was beatiful. It came, it went, and poof, 5 years later people are looking for it.
I'm sick and tired of all these weenies, be it from the enterprise or from their living room whining and complaining about storage and RAID and how do I do this, how do I do that.
/dev/hda and another little bad block around GB 310 on /dev/hdc will mean your RAID5 array is screwed.
Here's the scoop, or poop, or whatever.
Buy a case, buy a systemboard with X number of connectors. Connect all the drives, format them as one big partition each.
Buy a little via dinky dude to be a netboot server, do root NFS or whatever (that way no OS partitions no your storage box.)
Now, buy a second one storage node.
You have sitting in front of you, 2 boxes, each with 4, 6 8 or whatever 400GB Hitachi drives.
DON'T! do raid! You're talking commodity IDE, chances are that one little bad block around GB 3 on
Mirror each disk onto another disk on the second machine.
Even if you are carrying your case to a friends house and drop it, chances are, if the drives aren't raided, you'll at least be able to get some/most of the data off.
Use a filesystem with distributed metadata (reiser and XFS if I'm not mistaken)
I guarantee you will have a catastrophic failure with RAID when using cheap IDE disks.
SO I'm sure it's worth millions, but wouldn't it be cool if you were able to have a website that was robust enough to survive if it was mentioned in a slashdot article... an article with only 80 comments, no less.
So I would like, get some bandwidth, and like, get some servers that didn't suck, and maybe like hire this dude to write an app that didn't suck. So like, if you do that, then there's like this weird chance that when you get bombed with traffic people will still get a page back...
I dunno though, I think it'll cost like $20 million. Oh my gawd, so hardware alone would cost like $5K, and bandwidth would be aboue $1K/month... (cogent) oh, and then there's the dude, another $100K/year.... so for $20M you should be able to keep this running for about 178 years. Wheee!!!
Any one of these can swing fuel consumption. Driving my Jeep like a granny, shifting once I hit 1,500 rpm (not a problem since it has so much torque) gets me close to 18 or 20 mpg, even with big tires. If I drive around like "normal" people I'd be getting closer to 11 or 12 mpg in the city.
However, having lived in a suburb where I would have to drive 45 miles a day just for work, to now driving less than 5 miles/day if at all has meant that by moving to "the city" I have reduced my personal emissions/day tremendously.
One interesting to note about tire pressure. I left San Francisco (65 degrees) and pumped the tires to 33psi cold. By the time I got down to Fresno it was getting a bit squirrely. When I checked the pressure it had hit 40psi! I let it back down to 32psi and continued onward. Coming down the 10 heading into Phoenix it felt weird again. Lo and behold, the 115 degree ambient temp had pushed the pressure back up to 38.
The point being that you need to keep your tires inflated to the pressure that gives you the lowest rolling resistance. I'll be changing out the 33x12.5 to 33x9.5 tires.
Keeping the car well maintained is also a bonus.
Ok, since this is your personal fileserver, I assume you are talking about storing music, movies, files, whatever. I also take the word "personal" to mean that if you are in the middle of watching 2001 and an error on your drive causes an interruption in the movie you are not going to be overly upset. You may be sad about having wasted certain 2001 enhacing substances, but anyway.
/filemanifest
I would say rsync and scrub the disks.
Every night, simply rsync one disk onto the other one. You have two copies, you have no hardware or software raid configurations to deal with. It's cake to recover and to resync once you havea chance to run down to the store to pick up a drive.
Weekly, or while you are at work, run something like:
find . -type f | xargs md5sum >
You'll have a history of your disk and can track bitrot or whatever. You'll also be touching every single bit on the disk. No better way to detect errors, might want to run badblocks.
You are adding complexity and other people software with other peoples bugs trying to use weirdo ide hardware raid, etc. Linux sw raid1 is great because the disks aren't special if you split them, they are just plain ext2, or whatever you put on there.
Quick history lesson first. Ok, it's a bad one since I can't remeber the exact year, but it's right after Windows 95 came out with NetMeeting. I was using dial-up on the east coast, while my friend was dialed up in the midwest. We were playing Yahoo Chess on-line and had a netmeeting connection opened.
Conversations were more than acceptable and we would sit there for an hour or more, not loose a call and converse about all sorts of things while playing chess. On Dial-up!! I realize this was a point-to-point connection, but so what, it was 8 years ago with dial-up!!
I've used vonage in the Bay Area and it does work great over fast connections (dsl and up.)
A co-worker took his vonage phone to France and plugged it into his hotels broadband connection. The delay was over 1.5 seconds, probably 3 seconds. After less than a minute of conversation we simply started an interleaved conversation that took the gap into account (cheesy, I know, but afterwards I thought it was a really cool and efficient way to maximize productivity over lag (kind of like IRC, except with voice I guess.)
Regarding the troubles with Wifi, I have a feeling it has more to do with the queueing and QoS on the Wifi network than anything else. If it's really as bad as people described, then it should be totally impossible to play games over Wifi, etc...
There are cell phones coming out which will switch to VoIP when near an accessp point, and I'm sure they've done some testing on these things.
Please update the front page... I know that slashdot is a timesink, and severely decreases productivity. Yes, I realize it also has a healing effect on the damage psyches of those addicted to reading it every so often. However, this story reaches a new low.
Some dude, who's failed to spend a couple of days figuring out what's really going on posts an inflammatory and stress inducing story on the front page of this green rag.
From posted comments (no I never read slashdot comments) there appear to be binary drivers and XF86 drivers available.
An update on the front page would probably save abou 358,654 hours of wasted time by slashdot readers.
http://www.schneider-digital.de/html/download_a
What the heck are you talking about?
$150 for a wireless bridge that will work at several miles with a $50 antenna and you'll get about 3 Mb/sec.
Lets see, $200/year vs $50/month for DSL?
This is such a nobrainer its not even funny. I could put an Omni on my roof and 20 households could share the connection by putting a bridge in their front window.
Although W2K performance is a little sketchy, Linux seems to do quite well. I have the Japanese version., I don't know whether it is out here yet. It has the Transmeta Crusoe 667Mhz chip, 256Mb RAM, and 30 Gig hard drive. Linux runs great, including pcmcia and usb. The camera is supposedly supported under 2.4.7, I know, i'm lame for not trying yet. I have the quad capacity battery, which still leaves the damn little thing so damn portable. Coupled with a pcmcia wireless web device *used to be ricochet, we'll see what I will use next. Anyway, the battery allowed me over 10 hours of use. I played around a bit in the airport in tokyo, closed the lid, hopped on the plane... opened it as soon as I could. Watched a bunch of TV episodes on the plane, listened to a bunch of music (it was on during the entire portion of the flight that laptops were allowed.) After getting home I used it for a few more hours. As a portable computing device, I really love it. It's not the fastest, it's not the smallest... actually, maybe it is... but its the first laptop that I could type on that I carry around in my non laptop bag along with the rest of the equipment required for a stroll in Golden Gate Park.
Agreed. My download whizzed to package 57, and the last 5 have taken 20 minutes now, or more.... those bastards. Wait... I am one.
People keep talking abut their mother. "Let me tell you about my mother." -Leon, Blade Runner.
Windows sucks, as does MacOS. GNOME is ok, except for that damn panel. I don't think that modelling a desktop environment around a users expected behavior is what is needed. It's been tried before. The Mac took the symbols approach, and Microsoft soon followed. The problem with either of these two systems is that they were written in a time when we were trying to make difficult machines talk to simple people who really didn't know what to expect.
This has lead us to our current situation. We have people that have "learned" Windows, or "learned" the MacOS. They have learned it simply to use their applications and write e-mail to their grandchildren.
I've had the fortune of getting a couple of people behind a computer running Linux (enlightenment once and fvwm another time) I've found that the gnome panel is nothing more than a crutch for them to get into the swing of things. And that's only if they come from a Mac or Windows world. In both cases, once they saw my desktop, no panel, epplets and the pager and icon box, they wanted that... why? because it looked cool. As far as the user interface is concerned, I think that on the application level, GNOME has it right.
Common themes, etc. But in terms of what a UI really needs, it's this:
A meaningful gateway into the applications a processes running on top of your operating system. The enlightenment menuing is brilliant. Why should anyone have to go to some corner foot/start button for anything?
Between that and a simple lesson on gtop (for misbehaving applications) I believe the UI is perfect. A user should be asked to do nothing more than know how to launch applications and how to get rid of them when they are failing.
In Windows and MacOS this isn't an option, since you usually have to reboot in order to clear up any "weirdness" you may be experiencing. This doesn't leave the posibility open for a system that will be up for a year that you simply log in and out of (if that) or simply run your applications on.
People talk about how isolated the user is in MacOS and Windows, but really, the user is never more isolated from an operating system than the stability of the OS allows. Six desktops with all of your application usually open and at the ready is far more productive and user friendly than something where you feel like you need to start over every day or two.
The idea of a system that is so customizeable that it can be rendered almost inusable by someone other than "you" is perfect. Let the UI be an extension of the person and how they wish to work. This is in sharp contrast to the "we know what's best for you and this is how you are going to work" mantra being chanted by the other side.
Back to my mother. There are two directions to come at this. In several cases where the user has already known windows, I've found that they suddenly have this revelation that the UI is nothing more than another application (as it should be.) The next revelation is usually how much easier X/enlightenment or even fvwm are, in the sense that they are not in your face.
Anyone that has used linux as a desktop OS for more than a few days (i.e. feels comfortable, can kill a java dead netscape, etc) feels like Windows/Mac is lame. They can still use it, obviously, but they feel betrayed at the lack of power that has been handed to them in these sugar coated excersizes of futility.
If my mother has never used a computer before, it is even easier to sit her down in front Linux. You can safely say that she should not be afraid. She cannot possibly break the system. Removing that fear from a new user is one of the greatest enabling factors in learning any kind of technology.
I believe that customizability and stability should be the foremost thought going into any UI. A close second, I would argue, is transparency. I fear that discussions about what a UI needs to have will inevitably lead down a road that creates YADDUI (Yet Another Dumbed Down User Interface) to which some subset of the masses must conform. It doesn't get any simpler than clicking "somewhere" on your desktop to bring up a menu...
I don't like the tone of Mr. Katz when he suggests that these authors didn't anticipate the Ueber Corporation. In a lot of SF, even in something like Clans of the Alphane Moon (Phillip K. Dick) the insane asylum is the result of an ueber corporation. I don't think there is anything anyone missed.... on the contrary, I think it is a running theme that results in the weaponry and technology that is present in these novels. The end of government....
That might be even more frightening than some actually having picked up the domain name... I knew Bush was "right," but this is ridiculous...
Believe it or not, 24/96 I believe is worth it. 16/44 is just plain lame. While I don't claim to be able to hear the samples, anyone that "listens" to music feels the inherent supoeriority of vinyl. Now, the problem is that the qualities of vinyl that people like (analog, fullness, more natural stereo seperation) are mitigated by the imperfections of atoms. Bits are the answer, but CD's weren't it.
24/96 is ~ digital vinyl, or digital analog... I would argue that anything beyond that is in fact superfluous, but CD's are very lacking when it comes to accurately reproducing the original.
Why are we willing to put up with MP3's? Because they fit onto devices we can afford, and are transferrable at bandwidth's that we can afford... in the long run, neither of these are an issue.