Hmm... to me, the best thing about CrashPlan was the client. It would let me backup machine-to-mahcine, to a local drive, to a network drive, or to Crashplan Cloud -- all seamlessly from the same interface. I understand that the only thing they're eliminating is the cloud for home users product... but it seem from their site that they're also eliminating access to the client unless you have a Small Business account with a login to download the new ones.
Really, I bought their cloud product mostly because their client was good and their price was reasonable. Anyone know of a good and well-supported cross-platform client that lets you do machine-to-machine, NAS and cloud backups that maybe uses something like S3 or Glacier?
Corporations are taxed on profits, not on revenue. So anyone who will be taxed by definition has money, because they made a profit. Remember, they can write off just about anything that is involved with the operation of the corporation. So if they're losing money, they don't pay taxes.
Contrast that against income taxes, which are mostly (with the exception of a couple of write-offs like mortgage interest) on revenue.
An interesting thought experiment... what if income tax was treated the same way as corporate tax... you can write off ANYTHING that has to do with operations (living your life). That would be quite a stimulus, because it would encourage people to spend on all sorts of things, rather than just on real estate.
Your needs for 1000+ uniques are minimal. If I were to do it, I'd get a shared hosting account someplace and move on. Shared hosting can handle *way* more than that.
But if high availability (limited downtime) is part of the requirements, I'd say go out and buy an F5 BigIP. You plug your internet in the front, your machines in the built-in switch, configure your domain names in it using the web interface, and you're done. Set it to do service-checks, and it'll automatically pull out of the pool any machine that fails or that you take down for maintenance. So you get full up-time so long as your power and network don't fail.
Yes, you can get the same functionality using Linux HAProxy. But you sort of need to understand what you're doing. Reading the way your question is asked, I suspect you're learning this, and do you really want to make the mistakes on a real live project? Just go with the appliance until you have a solid understanding of what you're doing. Shoot -- I have a good solid understanding from years of experience, and I still use the BigIP when I have a budget (and HAProxy when I don't). It's just easier, and I can move on to more interesting problems with my time.
Once you've got this setup, set up a cron job to rdist the site to all the machines so that all your data is always on each machine. If you've got a database, you have some choices. For completely static data, I like to have it replicated to each machine, and have each web server just query localhost. If it's dynamic, have a replicated pair. At your levels, that can exist on the web servers.
I really dislike the cross-mounted disk architecture of traditional cluster solutions, because there are too many shared components. Each of those multiplies your possible points of failure for your whole setup. Better to keep everything completely separated, so if one component fails, that whole machine just drops out and the site keeps working because of the load balancer and because each machine can operate by itself.
When I changed to the Silverlight client, the quality improved, and I got to run it on my Mac too. What's not to like?
I love open source too... but you do yourselves a disservice when you fail to see the real reasons. They've got to stick to a DRM solution in order to get the film distributors to let them do rentals this way. It's how the distribution houses know who to pay royalties to. Without DRM, the major distribution houses would just say no. It's not Microsoft or Netflix forcing DRM on us -- it's the studios. And for a rental product (as opposed to a purchased one), it sort of makes sense.
Netflix planned this change for a year or more in order to deliver to the Macintosh market. They talked about it in their blogs and such -- they were just waiting for the Mac version of Silverlight to make it happen. I was sort of annoyed that it took so much longer than originally projected.
And for me the result has been significantly better quality with almost no re-buffering ever.
It not only ignores what they're doing, it ignores what they've done.
They've executed on their business plan. They're delivering the cars they promised at the price they promised them. Their current financial problem is purely due to the credit market drying up. If it weren't for that, they'd already have the financing they need in order to move from high-priced small production cars to medium-priced high-production cars.
If anyone deserves government backed loans, it's someone who's actually got a track record of delivering on their business plans and working toward the products of tomorrow.
Seems to me that this is the right thing to do 60-days before he actually gets into office -- gather information.
He didn't say he was going to cut anything, he asked for a cost-benefit analysis on various scenarios. If NASA can't deliver that, they don't deserve to keep operating. But I suspect they will give that, and it'll be fuel for the Obama administration to make (hopefully good) decisions.
I hope he's doing the same with every government agency -- identifying their top line-items and looking at whether or not those items are really best done by continuing on the current paths.
Being that the email record of the development of these features is pretty widely distributed, they'd have a tough time defending that patent if anyone makes them try. Here's the original proposal by Marc Andressen:
In proposing the IMG tag, he explicitly says that it can be embedded in an anchor, and he describes its action. I have my doubts that these guys have prior art on web pages dating back to before 1993.
They aren't going to spend any more tax dollars. They're just going to make a law requiring that businesses do the work for them, thus forcing more overhead for all businesses.
Don't fight it. Inform them of exactly which tasks you won't be able to do without privileges, and then do the things you can (getting knowledge onto paper and into hands of your co-workers is the primary thing).
Their policy is their problem, and you're no longer responsible for the company's long-term welfare. But if you're as friendly and helpful as they allow you to be over the next few weeks, you maintain relationships that may be helpful in future years.
In the end, most tech companies I've been with have accepted my 2 weeks notice, spent one day with me handing off information, then asked me not to come back (but they pay out those weeks anyway). It's a policy designed to make sure the employee doesn't get bored enough to cause any mischief.
But if they want you to come in anyway once everything you have left to do is off limits, come in to be available for your co-workers questions. A coupe of weeks of letting other people's hands do the work while you just are available for questions is actually a very valuable training period for handing off your gig, and probably should be in your transition plan even if they'd left your access on. And spend the bored time prepping for your new job.
Hmm... a relationship consisting of short, efficient and flexible commands that can be combined in interesting ways. Sounds pretty good to me. But I've only seen it work in the BDSM world.
Especially the 'self-taught' ones. That's the point of self-taught computer people... they started it as a hobby, and eventually realized they know enough to get a job at it.
How many women have computer programming as a hobby? I know several... and they all became kick-ass computer engineers. But they're absolutely the minority.
I think the answer doesn't have much to do with the IT profession, but more to do with the roles our society teaches kids. Girls as teens pick up different hobbies from boys. We push girls at MyLittlePony and Barbie, and guys at erector sets and GoBots, it's no wondeer the majority of them go that way later on with their hobbies.
The stylings were done by the people who did the stylings fo the Lotus Elise. And Tesla hired the Lotus factory to build the Teslas. But it's its own car....and it's a legit question. Tesla is starting with a low production $100k car because that's all they can afford to do at first. If they had the ability to ramp up production like GM does, they could jump right to the $30k mark IMHO.
Sounds to me like the managers suck. They should be seeing these things happening, and creating better coding guidelines and working with the developers to improve their practices. No guidance from management means no improvement ever.
When I was at UCLA in the '80s, they had already long since stopped using the SSN as any sort of student ID number. It was already understood that exposing the SSN had the potential for fraud.
I suspect this database was a finantial one of some sort... one where they actually needed the SSN for its real purpose -- reporting earnings and such to the IRS and the Social Security Administration.
Now why they still retain that information for people who've been out of the system for years is beyond me. That'll probably get them in trouble.
But I've still got a beef with the credit agencies... they should not be using such a freely available number (historically) as the primary password for using someone's credit. 2-factor authentication (something remembered plus something physical) has been possible for far too long for us to ignore it. If they just shifted to SSN plus fingerprint it'd be enough. Or go a step further... SSN plus fingerprint plus photo. Make it illegal to retain the fingerprint beyond initial identification to the credit checking institution, and provide fingerprint readers that protect it just that way.
No, Peak Oil has never predicted that we'd "run out".
They predict that the demand will outstrip the supply of cheap oil, forcing us to shift to more expensive supplies and creating shortages that drive the price beyond a reasonable means. They draw a standard set of supply and demand curves, and show where they cross. What's most interesting to me is that it's not the supply curve that's the issue -- it's the demand curve.
And they're less worried about cars than they are about what that steep rise in prices will do to all manufacturing and industry in the west.
In the environments I run, it's entirely due to the release cycle. Once I've installed RHEL 4, I'm guerenteed that they'll keep doing security and bug-fix patches for 5-7 years. And those will be REAL patches, not forced upgrades (RedHat backports fixes to the old versions rather than forcing packages to new versions.)
This means that my company doesn't have to constantly pay to re-test and re-verify the software we write every time someone feels like adding a nifty new feature, yet I can still keep the systems patched and secure. Re-doing QA is really expensive.
I run fedora on my desktop at home. But every time I install a new version, I end up with things breaking that I've got to fix. Imagining doing that with all of the hundreds of machines i run, and then having to explain to management why so many of our applications had downtime down due to a little OS upgrade... that gives me nightmares. I'll stick to enterprise-level stuff on the job.
$6 Million rocket. $800,000 payload. The cost of the payload is pretty small, all things considered. It's worth the risk to go ahead and fly the payload the first time. Saves you $5 million if it works, and cost you less than $1 million if it fails....and when you add in that everything's going to be insured, it makes finantial sense.
Not only that, but it only runs 1920x1024 in analog mode. It won't even run my 23" display!...and the processor in it is slower than the slowest of the coreDuo laptops. They could have at least matched the laptop specs.
sort of opposite concepts. In one, people are choosing to work together on something. in the other there's a central power who's trying to get a group to buy something.
The biggest failure of "online communities" in the 'net days is that most of them are corporate sponsored marketing schemes rather than actual communities.
Yeah -- not that useful for a data center because space is no longer the big issue for a lot of people.
But for musicians, having a quiet machine that sits in a rack that isn't too deep means I can put a system in the same rack as the rest of my gear. Can you say portable pro-tools? and for my home setup, I've got 3 or 4 systems. For years I've wanted to get them all rack-mount so that they can be in a neat stack rather than sitting on the floor under desks... but then i stop when I realize how loud rackmount systems are.
The only way Douglas Adams got his script done was to die... then the arguements him and the studios about what constitutes a good script ended, and the movie got made. I suspect the same will be true of Enders... we'll see it in the theaters about five years after Card dies.
First of all, once you've had your first job, no one really gives a crap where you went to school. They care about what quality of work you did at your last job. They care that you *did* go to school. That's about it.
Second... someone talked about the 'reputation' of your school. No one cares. Tulane is a name-recognition school, and in most parts of the country no one will even know that it no longer has a CS program. They'll either recognize the school name or not. Did you know that UC Irvine has a great CS program? Or DeAnza College? I thought not. No one knows (or really cares mostly).
What you should be worried about is what kind of education you're going to get in the next two years... because that's what's going to determine how you do at those first few jobs.
Talk to your faculty advisors. Talk to the faculty. Find out which of them are staying and which of them are going to jump ship. It's reasonably likely that all the good CS staff will jump ship, since there's no longer going to be a career path at the school for them. No department = no research = no publishing = no career. That's what you should worry about -- losing all the good teachers.
If the teachers are going, you should go. If they're sticking around until you're gonna graduate, stick around with them (assuming you like the place and the program).
Yeah -- but this one is going to be a battle of definitions I believe. I was around during the beginnings of XML and working with several of the companies that created XML (on other projeects)... SoftQuad and several other SGML product makers. They originally envisioned XML as a "simpler" version of SGML... and a way to capitalize on the web market that they were rapidly losing to companies that made products that were simpler in concept....and this all happened right around 1997. And I have some vague memory of Paul Odom, currently of Scientigo was involved. If he was the only one to file a patent, then people are screwed on this.
But I think it'll boil down to whether or not XML is actually different from SGML, or just a re-definition or derivation. IMHO it's a simpler re-definition/derivation. But others would differ on that.
Slashdot Mirror
Hmm... to me, the best thing about CrashPlan was the client. It would let me backup machine-to-mahcine, to a local drive, to a network drive, or to Crashplan Cloud -- all seamlessly from the same interface. I understand that the only thing they're eliminating is the cloud for home users product... but it seem from their site that they're also eliminating access to the client unless you have a Small Business account with a login to download the new ones.
Really, I bought their cloud product mostly because their client was good and their price was reasonable. Anyone know of a good and well-supported cross-platform client that lets you do machine-to-machine, NAS and cloud backups that maybe uses something like S3 or Glacier?
Which can afford it? All of them!
Corporations are taxed on profits, not on revenue. So anyone who will be taxed by definition has money, because they made a profit. Remember, they can write off just about anything that is involved with the operation of the corporation. So if they're losing money, they don't pay taxes.
Contrast that against income taxes, which are mostly (with the exception of a couple of write-offs like mortgage interest) on revenue.
An interesting thought experiment... what if income tax was treated the same way as corporate tax... you can write off ANYTHING that has to do with operations (living your life). That would be quite a stimulus, because it would encourage people to spend on all sorts of things, rather than just on real estate.
Your needs for 1000+ uniques are minimal. If I were to do it, I'd get a shared hosting account someplace and move on. Shared hosting can handle *way* more than that.
But if high availability (limited downtime) is part of the requirements, I'd say go out and buy an F5 BigIP. You plug your internet in the front, your machines in the built-in switch, configure your domain names in it using the web interface, and you're done. Set it to do service-checks, and it'll automatically pull out of the pool any machine that fails or that you take down for maintenance. So you get full up-time so long as your power and network don't fail.
Yes, you can get the same functionality using Linux HAProxy. But you sort of need to understand what you're doing. Reading the way your question is asked, I suspect you're learning this, and do you really want to make the mistakes on a real live project? Just go with the appliance until you have a solid understanding of what you're doing. Shoot -- I have a good solid understanding from years of experience, and I still use the BigIP when I have a budget (and HAProxy when I don't). It's just easier, and I can move on to more interesting problems with my time.
Once you've got this setup, set up a cron job to rdist the site to all the machines so that all your data is always on each machine. If you've got a database, you have some choices. For completely static data, I like to have it replicated to each machine, and have each web server just query localhost. If it's dynamic, have a replicated pair. At your levels, that can exist on the web servers.
I really dislike the cross-mounted disk architecture of traditional cluster solutions, because there are too many shared components. Each of those multiplies your possible points of failure for your whole setup. Better to keep everything completely separated, so if one component fails, that whole machine just drops out and the site keeps working because of the load balancer and because each machine can operate by itself.
When I changed to the Silverlight client, the quality improved, and I got to run it on my Mac too. What's not to like?
I love open source too... but you do yourselves a disservice when you fail to see the real reasons. They've got to stick to a DRM solution in order to get the film distributors to let them do rentals this way. It's how the distribution houses know who to pay royalties to. Without DRM, the major distribution houses would just say no. It's not Microsoft or Netflix forcing DRM on us -- it's the studios. And for a rental product (as opposed to a purchased one), it sort of makes sense.
Netflix planned this change for a year or more in order to deliver to the Macintosh market. They talked about it in their blogs and such -- they were just waiting for the Mac version of Silverlight to make it happen. I was sort of annoyed that it took so much longer than originally projected.
And for me the result has been significantly better quality with almost no re-buffering ever.
It not only ignores what they're doing, it ignores what they've done.
They've executed on their business plan. They're delivering the cars they promised at the price they promised them. Their current financial problem is purely due to the credit market drying up. If it weren't for that, they'd already have the financing they need in order to move from high-priced small production cars to medium-priced high-production cars.
If anyone deserves government backed loans, it's someone who's actually got a track record of delivering on their business plans and working toward the products of tomorrow.
Seems to me that this is the right thing to do 60-days before he actually gets into office -- gather information.
He didn't say he was going to cut anything, he asked for a cost-benefit analysis on various scenarios. If NASA can't deliver that, they don't deserve to keep operating. But I suspect they will give that, and it'll be fuel for the Obama administration to make (hopefully good) decisions.
I hope he's doing the same with every government agency -- identifying their top line-items and looking at whether or not those items are really best done by continuing on the current paths.
Being that the email record of the development of these features is pretty widely distributed, they'd have a tough time defending that patent if anyone makes them try. Here's the original proposal by Marc Andressen:
http://1997.webhistory.org/www.lists/www-talk.1993q1/0182.html
In proposing the IMG tag, he explicitly says that it can be embedded in an anchor, and he describes its action. I have my doubts that these guys have prior art on web pages dating back to before 1993.
They aren't going to spend any more tax dollars. They're just going to make a law requiring that businesses do the work for them, thus forcing more overhead for all businesses.
Don't fight it. Inform them of exactly which tasks you won't be able to do without privileges, and then do the things you can (getting knowledge onto paper and into hands of your co-workers is the primary thing).
Their policy is their problem, and you're no longer responsible for the company's long-term welfare. But if you're as friendly and helpful as they allow you to be over the next few weeks, you maintain relationships that may be helpful in future years.
In the end, most tech companies I've been with have accepted my 2 weeks notice, spent one day with me handing off information, then asked me not to come back (but they pay out those weeks anyway). It's a policy designed to make sure the employee doesn't get bored enough to cause any mischief.
But if they want you to come in anyway once everything you have left to do is off limits, come in to be available for your co-workers questions. A coupe of weeks of letting other people's hands do the work while you just are available for questions is actually a very valuable training period for handing off your gig, and probably should be in your transition plan even if they'd left your access on. And spend the bored time prepping for your new job.
Seems like he's going to have a lot of time to write a lot of code now.
Hmm... a relationship consisting of short, efficient and flexible commands that can be combined in interesting ways. Sounds pretty good to me. But I've only seen it work in the BDSM world.
Especially the 'self-taught' ones. That's the point of self-taught computer people... they started it as a hobby, and eventually realized they know enough to get a job at it.
How many women have computer programming as a hobby? I know several... and they all became kick-ass computer engineers. But they're absolutely the minority.
I think the answer doesn't have much to do with the IT profession, but more to do with the roles our society teaches kids. Girls as teens pick up different hobbies from boys. We push girls at MyLittlePony and Barbie, and guys at erector sets and GoBots, it's no wondeer the majority of them go that way later on with their hobbies.
No, the Tesla is built from the ground up.
...and it's a legit question. Tesla is starting with a low production $100k car because that's all they can afford to do at first. If they had the ability to ramp up production like GM does, they could jump right to the $30k mark IMHO.
The stylings were done by the people who did the stylings fo the Lotus Elise. And Tesla hired the Lotus factory to build the Teslas. But it's its own car.
Sounds to me like the managers suck. They should be seeing these things happening, and creating better coding guidelines and working with the developers to improve their practices. No guidance from management means no improvement ever.
When I was at UCLA in the '80s, they had already long since stopped using the SSN as any sort of student ID number. It was already understood that exposing the SSN had the potential for fraud.
I suspect this database was a finantial one of some sort... one where they actually needed the SSN for its real purpose -- reporting earnings and such to the IRS and the Social Security Administration.
Now why they still retain that information for people who've been out of the system for years is beyond me. That'll probably get them in trouble.
But I've still got a beef with the credit agencies... they should not be using such a freely available number (historically) as the primary password for using someone's credit. 2-factor authentication (something remembered plus something physical) has been possible for far too long for us to ignore it. If they just shifted to SSN plus fingerprint it'd be enough. Or go a step further... SSN plus fingerprint plus photo. Make it illegal to retain the fingerprint beyond initial identification to the credit checking institution, and provide fingerprint readers that protect it just that way.
No, Peak Oil has never predicted that we'd "run out".
They predict that the demand will outstrip the supply of cheap oil, forcing us to shift to more expensive supplies and creating shortages that drive the price beyond a reasonable means. They draw a standard set of supply and demand curves, and show where they cross. What's most interesting to me is that it's not the supply curve that's the issue -- it's the demand curve.
And they're less worried about cars than they are about what that steep rise in prices will do to all manufacturing and industry in the west.
In the environments I run, it's entirely due to the release cycle. Once I've installed RHEL 4, I'm guerenteed that they'll keep doing security and bug-fix patches for 5-7 years. And those will be REAL patches, not forced upgrades (RedHat backports fixes to the old versions rather than forcing packages to new versions.)
This means that my company doesn't have to constantly pay to re-test and re-verify the software we write every time someone feels like adding a nifty new feature, yet I can still keep the systems patched and secure. Re-doing QA is really expensive.
I run fedora on my desktop at home. But every time I install a new version, I end up with things breaking that I've got to fix. Imagining doing that with all of the hundreds of machines i run, and then having to explain to management why so many of our applications had downtime down due to a little OS upgrade... that gives me nightmares. I'll stick to enterprise-level stuff on the job.
$6 Million rocket. $800,000 payload. The cost of the payload is pretty small, all things considered. It's worth the risk to go ahead and fly the payload the first time. Saves you $5 million if it works, and cost you less than $1 million if it fails. ...and when you add in that everything's going to be insured, it makes finantial sense.
Not only that, but it only runs 1920x1024 in analog mode. It won't even run my 23" display! ...and the processor in it is slower than the slowest of the coreDuo laptops. They could have at least matched the laptop specs.
I find myself disappointed yet again.
"community" and "customers"
sort of opposite concepts. In one, people are choosing to work together on something. in the other there's a central power who's trying to get a group to buy something.
The biggest failure of "online communities" in the 'net days is that most of them are corporate sponsored marketing schemes rather than actual communities.
Yeah -- not that useful for a data center because space is no longer the big issue for a lot of people.
But for musicians, having a quiet machine that sits in a rack that isn't too deep means I can put a system in the same rack as the rest of my gear. Can you say portable pro-tools? and for my home setup, I've got 3 or 4 systems. For years I've wanted to get them all rack-mount so that they can be in a neat stack rather than sitting on the floor under desks... but then i stop when I realize how loud rackmount systems are.
So it's useful for some poeople I think.
The only way Douglas Adams got his script done was to die... then the arguements him and the studios about what constitutes a good script ended, and the movie got made. I suspect the same will be true of Enders... we'll see it in the theaters about five years after Card dies.
A couple of things.
First of all, once you've had your first job, no one really gives a crap where you went to school. They care about what quality of work you did at your last job. They care that you *did* go to school. That's about it.
Second... someone talked about the 'reputation' of your school. No one cares. Tulane is a name-recognition school, and in most parts of the country no one will even know that it no longer has a CS program. They'll either recognize the school name or not. Did you know that UC Irvine has a great CS program? Or DeAnza College? I thought not. No one knows (or really cares mostly).
What you should be worried about is what kind of education you're going to get in the next two years... because that's what's going to determine how you do at those first few jobs.
Talk to your faculty advisors. Talk to the faculty. Find out which of them are staying and which of them are going to jump ship. It's reasonably likely that all the good CS staff will jump ship, since there's no longer going to be a career path at the school for them. No department = no research = no publishing = no career. That's what you should worry about -- losing all the good teachers.
If the teachers are going, you should go. If they're sticking around until you're gonna graduate, stick around with them (assuming you like the place and the program).
Yeah -- but this one is going to be a battle of definitions I believe. I was around during the beginnings of XML and working with several of the companies that created XML (on other projeects)... SoftQuad and several other SGML product makers. They originally envisioned XML as a "simpler" version of SGML... and a way to capitalize on the web market that they were rapidly losing to companies that made products that were simpler in concept. ...and this all happened right around 1997. And I have some vague memory of Paul Odom, currently of Scientigo was involved. If he was the only one to file a patent, then people are screwed on this.
But I think it'll boil down to whether or not XML is actually different from SGML, or just a re-definition or derivation. IMHO it's a simpler re-definition/derivation. But others would differ on that.
If I recall correctly, at a demo he did just after the NAMM where he introduced them, I heard him pronounce them just that way -- MoagerFoager.
But he was a practical person. I suspect he didn't care how you pronounced them as long as you bought and used them.