or somewhere else where there is a 'critical mass' of geeks. after that, it should be easy to show up for LUGs and go from there.
As for meeting women, I personally recommend Mensa. Worked for me. I actually met a nice embedded systems programmer on a mensa mailing list (rationalist-M-discussion, if you want to hunt down the archives) who happened to live a few miles from where I live. I have no idea why she chose me, but eh, we've been together for almost two years now, so I'm not complaining. I theorize that you just have to be clear about who you are, and eventually you will stumble upon someone looking for that.
But the density of nerds on the sf bay area makes everything much easier.
but it doesn't report correctable ECC errors if it doesn't know the chipset or if it doesn't have reliable support for the chipset (that's what it means if ecc is set to 'no' in memtest on a board with ecc ram)
that said, you are right, some motherboards don't support ECC. I'm just saying that's not what the 'ecc' 'on/off' field in memtest86 means.
people your own age with around 20 years experience. Obviously, those people will be more desirable than you will. (education is usually comparable to experience, year for year.)
I haven't seen much obvious ageism, but then I'm 28, (been doing this since 15... I'll have 20 years by the time I'm 35) I've worked with a whole lot of really awesome old folks. one guy had 40 years of experience, and it showed. he was really good. You will, however, have a hard time getting a toehold right now, just 'cause the economy is shit and you have no XP. You need to go grind XP. I would advise starting right now. relevant work experience and college beats college hands down.
my advice to you is to make sure that every job from now on is IT related, even if you have to take one that pays crap. There are always companies around (see mine, though I'm not hiring at the moment) that will hire you as a jr sysadmin if you are willing to work for slightly above retail wages. I know I put in my time at that rate. HE.net around here hires rack monkeys for $15/hr. that's around what I end up usually paying for people as well.
If you really can't get a related job (and that's possible, the economy really is shit right now. everyone I know who is even a little marginal is having a hard time of it.) start your own company. I'm serious. even if you have no money, go buy a $5 VPS (hey, how about you buy it from me? http://prgmr.com/xen/ - but seriously, I have lots of competition. servers are cheap now.) write a webapp, let people use it for free. write a blog about things you are figuring out while you write the webapp. You can then put that on your resume as industry experience. Maybe it will get big? who knows.
But yeah, until the economy lets up, you are going to have a hard time of it. good luck.
First, I assume you are working with other people on this. My current project, http://nostarch.com/xen.htm has two authors, along with a tech editor, a regular editor, and all the other people the publisher handles.
First, do not use a non-text format to store your book while you are still working on it. Sure, all modern GUIs have merge facilities and change tracking, but the tools are extremely clumsy compared to even the most basic text revision control system. Do not underestimate the power of diff.
Second, when dealing in text, write your rough drafts in mediawiki markup (unless you are super-familiar with Latex or the like) - it is simple, and it gives you a nice output format for dealing with rough drafts. Heck, it means that if you are working with editors that need a gui, they have one.
The idea is that everyone can use text (or wikis.) once you have the book done, you can get copyedit to put it in whatever format they like (or alternately, you can write a sed script to convert the basic mediawiki to basic latex or whatever) the basic idea is to separate the 'write the book' task from the 'format the book' task.
my experience has been that sales isn't so hard in this arena. (but then, I usually don't charge much over my dayjob hourly rate, while most contractors try to double that.) - Often you can greatly reduce sales effort by simply lowering your rate. For me, it's a great deal, 'cause I don't enjoy sales and I usually enjoy the work. I'd say that sales ate maybe 5-10% of my time when I was doing this (but then nearly all my work was repeat customers... I'd get a new customer once every few months.)
it is a *whole lot* cheaper for companies to only pay you when they call you. I mean, if you only get paid when you get called, naturally you might be unavailable sometimes when they call (I mean, I can't expect you to stand around waiting for my call unless I'm paying you to stand around waiting for my call, which can work out to a pretty nice retainer that works for both parties.)
The way to do this is as a contractor. when I did this, I brought my own laptop, my own health insurance, and usually I worked from my own office, so the companies real cost for hiring me was my rate + whatever it cost them to find me.
get around on craigslist. there are plenty of small corps that are happy to have someone every now and again. when I was doing it, I had one client who gave me a 40 hour retainer (I gave them a deep discount on the retainer, but they paid me for 40 hours a month even if they used less, and in exchange, they got a deep discount and prioritization if there was contention for my time. I don't think they ever wanted me less than 40 hours a month.) but that alone would cover rent and food, and I had a few other clients that would use me every now and again.
I scored the anchor client because I interviewed with the place, and they liked me, but I didn't like them enough to abandon my own business and be a W2 for them, and while they liked me, they would have liked someone cheaper. Most of my other clients I picked up through craigslist or word of mouth.
It is not steady though, not even a little bit, so you have to be prepared for that. On the other hand, at what even a mediocre linux guy charges, there's really no reason to work full time, and the tax structure in this country seems to be setup to encourage you to keep your income below $50,000 or entirely in capital gains - so part time is much better than a few years on, a few years off, well, unless you can turn your earnings into capital gains.
That's another way to structure it... work for a year at a time, then take a year off. But structure it such that you are working 6 months out of every tax year so you pay taxes like you aren't making much money.
Many hosting providers give you ridiculously high bandwidth limits, counting on the fact that it's usually quite difficult to use much bandwidth without violating the AUP. So I'm guessing what happened was that these guys got popular and started costing the hosting company money (but were still within the stated limit) and that's the real reason they got disconnected.
It's a lot like the 'unlimited dialup' popular in the mid 90s... it was unlimited, unless you used it a whole lot, in which case they billed you or disconnected you.
If you look up my site on archive.org, you will see that my limits used to be much higher, comparable to my competitors. Being as my AUP is much less restrictive in the non-spam areas, it was a very expensive lesson for me to learn.
Is that If there is much IPv6 hype in the future, or if IPv6 wins, you can say "I have x years experience with IPv6" (well, and more to the point, you have had x years of experience with IPv6. you have made your newbie mistakes that break things while IPv6 still doesn't really matter.)
what about http://www.sixxs.net/ ? they support AYIYA tunnels which should work through nat, and they have European POPs, so it sounds like they might work much better for you.
really? define 'sucks terribly' (as in ping times)
I guess they figure that if you are advanced enough to want IPv6 you can setup a PC router. Hm. I wonder what the operational costs of running a free V6-UDP-V4 tunnel broker would be.
"a lot" ? examples? everything I use on a daily basis in the server space is v6 compatible and has been for a few years now. Clearly, servers will need to remain dual-stack until the v4-only client-side software cycles through, but old stuff that hasn't been updated in a while belongs behind a firewall/gateway no matter what transport the public internet uses.
the problem is that undoing that waste is very difficult. do you really expect corporations to go through the expense of renumbering inefficient assignments, and then to *give away* a scarce resource? a scarce resource that may become quite valuable in 3 years?
accessible via IPv6 (and the number of clients able to reach IPv6 hosts) is probably more important than the amount of traffic. all of my customers are dual-stack, and while only a trickle of data flows over IPv6, any customer can seamlessly connect to an IPv6 server over IPv6.
the problem is that a large percentage of people on the net need to be dual-stack (or otherwise able to access IPv6) before anyone serious goes IPv6-only.
the thing is, IPv6 is a much happier world for us all (both admins and end users) than nat hell. Sure, Nat seems to work ok at one level, when you split a single IPv4 address with your entire home network... but what happens when the ISP does the NAT and you have no public IPs at all? this will cut into your bittorrent habit.
Personally, I suspect that we will see a huge amount of IPv6 hype as the v4 pool runs out... I'm not saying that IPv6 will win, just that there will be a lot of hype.
Either way, things will get interesting. NAT hell will finalize the 'two tier' internet that was begun when ISPs started handing out dynamic addresses to dialup/dsl customers. Internet access behind NAT is largely a client-only experience, and many existing apps don't work well behind NAT (like bittorrent... remember, you can forward ports now, but at v4 runout, ISPs will start handing out NATted addresses to end users. and they are unlikely to forward ports for you.)
IPv6 has its own challenges... the biggest problem with IPv6, though, is the chicken and the egg. Until more hosts support IPv6, there's not much point to using it.
position? Obviously, if you are in a position where you get paid the same no matter how much you work, you (the worker) are going to try to work as little as possible, while your employer is going to try to get you to work as much as possible.
Personally, I like hourly contracting gigs more. If I come in late and leave early, I get paid less. If I show up early and stay late, I get paid more. this way management doesn't have an incentive to overwork me, and I don't have an incentive to slack off.
Red Hat has a lot of things going for them... but 'cheap' isn't one of them.
In a grid computing situation like EC2, 99% of the work is going to be dealing with bad hardware- most of the time, that means having a guy with a screwdriver and a shopping cart full of hard drives on-site. Most of the more skilled work is going to involve dealing with your provisioning system, your networking system, and making sure your hardware matches up with your operating systems. Still, they are largely operational problems, and RedHat is a development company.
every large corp I've worked for paid big bucks for Linux support. some of those big corps have gotten a lot of value from that support, while others have gotten not much value at all, but nearly all feel the need to pay for it.
Usually it is the smaller shops (who have less in-house skill) who choose CentOS over RHEL.
I'm not qualified to say why, but like everyone else, I'll hazard a guess- I think this has more to do with risk tolerance than with skill.
I think having all your servers with one provider (even if you are that provider) is usually a bad idea. People make mistakes. hardware dies. natural disasters happen.
abandoned after ec2 came out. EC2 does most of what it sounds like you want... except for the redundancy in providers.
Personally, I'm working on building a better provisioning system for my own VPS services at http://prgmr.com/xen but the idea is that it's not that hard, even with the way ec2 is now, to take your ec2 image and run it on another xen host, or take a xen image and run it on ec2. (now getting a 'public' image off amazon ec2 and downloading it, that's hard. but if it's your image, downloading it is trivial. and amazon has all the tools you need to turn it into a plain file or tarball that any Xen provider can use)
nameserver RHN package, caching-nameserver.arch to serve authoritative zones. it's a caching nameserver, it's not supposed to serve authoritative zones! if you are using the regular nameserver package, bind.arch, it breaks nothing. it keeps the old config and copies the new config to.rpmnew.
why not run an inward facing IDS- something like snort. It's easy enough to setup a script that automatically terminates accounts of people sending abuse, and to do it on the first instance of that abuse.
Slashdot Mirror
or somewhere else where there is a 'critical mass' of geeks. after that, it should be easy to show up for LUGs and go from there. As for meeting women, I personally recommend Mensa. Worked for me. I actually met a nice embedded systems programmer on a mensa mailing list (rationalist-M-discussion, if you want to hunt down the archives) who happened to live a few miles from where I live. I have no idea why she chose me, but eh, we've been together for almost two years now, so I'm not complaining. I theorize that you just have to be clear about who you are, and eventually you will stumble upon someone looking for that. But the density of nerds on the sf bay area makes everything much easier.
but it doesn't report correctable ECC errors if it doesn't know the chipset or if it doesn't have reliable support for the chipset (that's what it means if ecc is set to 'no' in memtest on a board with ecc ram)
that said, you are right, some motherboards don't support ECC. I'm just saying that's not what the 'ecc' 'on/off' field in memtest86 means.
really, it's not that much more expensive. Search newegg for unbuffered ecc, if you are using a desktop class system that can't handle registered ram.
You wouldn't put data you care about on a hard drive without raid, would you?
people your own age with around 20 years experience. Obviously, those people will be more desirable than you will. (education is usually comparable to experience, year for year.)
I haven't seen much obvious ageism, but then I'm 28, (been doing this since 15... I'll have 20 years by the time I'm 35) I've worked with a whole lot of really awesome old folks. one guy had 40 years of experience, and it showed. he was really good. You will, however, have a hard time getting a toehold right now, just 'cause the economy is shit and you have no XP. You need to go grind XP. I would advise starting right now. relevant work experience and college beats college hands down.
my advice to you is to make sure that every job from now on is IT related, even if you have to take one that pays crap. There are always companies around (see mine, though I'm not hiring at the moment) that will hire you as a jr sysadmin if you are willing to work for slightly above retail wages. I know I put in my time at that rate. HE.net around here hires rack monkeys for $15/hr. that's around what I end up usually paying for people as well.
If you really can't get a related job (and that's possible, the economy really is shit right now. everyone I know who is even a little marginal is having a hard time of it.) start your own company. I'm serious. even if you have no money, go buy a $5 VPS (hey, how about you buy it from me? http://prgmr.com/xen/ - but seriously, I have lots of competition. servers are cheap now.) write a webapp, let people use it for free. write a blog about things you are figuring out while you write the webapp. You can then put that on your resume as industry experience. Maybe it will get big? who knows.
But yeah, until the economy lets up, you are going to have a hard time of it. good luck.
about what I have learned so far.
First, I assume you are working with other people on this. My current project, http://nostarch.com/xen.htm has two authors, along with a tech editor, a regular editor, and all the other people the publisher handles.
First, do not use a non-text format to store your book while you are still working on it. Sure, all modern GUIs have merge facilities and change tracking, but the tools are extremely clumsy compared to even the most basic text revision control system. Do not underestimate the power of diff.
Second, when dealing in text, write your rough drafts in mediawiki markup (unless you are super-familiar with Latex or the like) - it is simple, and it gives you a nice output format for dealing with rough drafts. Heck, it means that if you are working with editors that need a gui, they have one.
The idea is that everyone can use text (or wikis.) once you have the book done, you can get copyedit to put it in whatever format they like (or alternately, you can write a sed script to convert the basic mediawiki to basic latex or whatever) the basic idea is to separate the 'write the book' task from the 'format the book' task.
my experience has been that sales isn't so hard in this arena. (but then, I usually don't charge much over my dayjob hourly rate, while most contractors try to double that.) - Often you can greatly reduce sales effort by simply lowering your rate. For me, it's a great deal, 'cause I don't enjoy sales and I usually enjoy the work. I'd say that sales ate maybe 5-10% of my time when I was doing this (but then nearly all my work was repeat customers... I'd get a new customer once every few months.)
it is a *whole lot* cheaper for companies to only pay you when they call you. I mean, if you only get paid when you get called, naturally you might be unavailable sometimes when they call (I mean, I can't expect you to stand around waiting for my call unless I'm paying you to stand around waiting for my call, which can work out to a pretty nice retainer that works for both parties.)
The way to do this is as a contractor. when I did this, I brought my own laptop, my own health insurance, and usually I worked from my own office, so the companies real cost for hiring me was my rate + whatever it cost them to find me.
I scored the anchor client because I interviewed with the place, and they liked me, but I didn't like them enough to abandon my own business and be a W2 for them, and while they liked me, they would have liked someone cheaper. Most of my other clients I picked up through craigslist or word of mouth.
It is not steady though, not even a little bit, so you have to be prepared for that. On the other hand, at what even a mediocre linux guy charges, there's really no reason to work full time, and the tax structure in this country seems to be setup to encourage you to keep your income below $50,000 or entirely in capital gains - so part time is much better than a few years on, a few years off, well, unless you can turn your earnings into capital gains.
That's another way to structure it... work for a year at a time, then take a year off. But structure it such that you are working 6 months out of every tax year so you pay taxes like you aren't making much money.
Many hosting providers give you ridiculously high bandwidth limits, counting on the fact that it's usually quite difficult to use much bandwidth without violating the AUP. So I'm guessing what happened was that these guys got popular and started costing the hosting company money (but were still within the stated limit) and that's the real reason they got disconnected.
It's a lot like the 'unlimited dialup' popular in the mid 90s... it was unlimited, unless you used it a whole lot, in which case they billed you or disconnected you.
If you look up my site on archive.org, you will see that my limits used to be much higher, comparable to my competitors. Being as my AUP is much less restrictive in the non-spam areas, it was a very expensive lesson for me to learn.
Is that If there is much IPv6 hype in the future, or if IPv6 wins, you can say "I have x years experience with IPv6" (well, and more to the point, you have had x years of experience with IPv6. you have made your newbie mistakes that break things while IPv6 still doesn't really matter.)
what about http://www.sixxs.net/ ? they support AYIYA tunnels which should work through nat, and they have European POPs, so it sounds like they might work much better for you.
who has jurisdiction over the internet? many of the biggest wastes were allocated before ICANN existed.
before you start calling for government intervention, remember that it isn't just your government that is involved here.
I guess they figure that if you are advanced enough to want IPv6 you can setup a PC router. Hm. I wonder what the operational costs of running a free V6-UDP-V4 tunnel broker would be.
"a lot" ? examples? everything I use on a daily basis in the server space is v6 compatible and has been for a few years now. Clearly, servers will need to remain dual-stack until the v4-only client-side software cycles through, but old stuff that hasn't been updated in a while belongs behind a firewall/gateway no matter what transport the public internet uses.
the problem is that undoing that waste is very difficult. do you really expect corporations to go through the expense of renumbering inefficient assignments, and then to *give away* a scarce resource? a scarce resource that may become quite valuable in 3 years?
as far as I can tell, most of the tunnels are pretty quick. And most reasonably decent ISPs provide local tunnels for their customers.
accessible via IPv6 (and the number of clients able to reach IPv6 hosts) is probably more important than the amount of traffic. all of my customers are dual-stack, and while only a trickle of data flows over IPv6, any customer can seamlessly connect to an IPv6 server over IPv6.
the problem is that a large percentage of people on the net need to be dual-stack (or otherwise able to access IPv6) before anyone serious goes IPv6-only.
the thing is, IPv6 is a much happier world for us all (both admins and end users) than nat hell. Sure, Nat seems to work ok at one level, when you split a single IPv4 address with your entire home network... but what happens when the ISP does the NAT and you have no public IPs at all? this will cut into your bittorrent habit.
Personally, I suspect that we will see a huge amount of IPv6 hype as the v4 pool runs out... I'm not saying that IPv6 will win, just that there will be a lot of hype.
Either way, things will get interesting. NAT hell will finalize the 'two tier' internet that was begun when ISPs started handing out dynamic addresses to dialup/dsl customers. Internet access behind NAT is largely a client-only experience, and many existing apps don't work well behind NAT (like bittorrent... remember, you can forward ports now, but at v4 runout, ISPs will start handing out NATted addresses to end users. and they are unlikely to forward ports for you.)
IPv6 has its own challenges... the biggest problem with IPv6, though, is the chicken and the egg. Until more hosts support IPv6, there's not much point to using it.
position? Obviously, if you are in a position where you get paid the same no matter how much you work, you (the worker) are going to try to work as little as possible, while your employer is going to try to get you to work as much as possible.
Personally, I like hourly contracting gigs more. If I come in late and leave early, I get paid less. If I show up early and stay late, I get paid more. this way management doesn't have an incentive to overwork me, and I don't have an incentive to slack off.
Red Hat has a lot of things going for them... but 'cheap' isn't one of them.
In a grid computing situation like EC2, 99% of the work is going to be dealing with bad hardware- most of the time, that means having a guy with a screwdriver and a shopping cart full of hard drives on-site. Most of the more skilled work is going to involve dealing with your provisioning system, your networking system, and making sure your hardware matches up with your operating systems. Still, they are largely operational problems, and RedHat is a development company.
every large corp I've worked for paid big bucks for Linux support. some of those big corps have gotten a lot of value from that support, while others have gotten not much value at all, but nearly all feel the need to pay for it.
Usually it is the smaller shops (who have less in-house skill) who choose CentOS over RHEL.
I'm not qualified to say why, but like everyone else, I'll hazard a guess- I think this has more to do with risk tolerance than with skill.
I think having all your servers with one provider (even if you are that provider) is usually a bad idea. People make mistakes. hardware dies. natural disasters happen.
abandoned after ec2 came out. EC2 does most of what it sounds like you want... except for the redundancy in providers. Personally, I'm working on building a better provisioning system for my own VPS services at http://prgmr.com/xen but the idea is that it's not that hard, even with the way ec2 is now, to take your ec2 image and run it on another xen host, or take a xen image and run it on ec2. (now getting a 'public' image off amazon ec2 and downloading it, that's hard. but if it's your image, downloading it is trivial. and amazon has all the tools you need to turn it into a plain file or tarball that any Xen provider can use)
nameserver RHN package, caching-nameserver.arch to serve authoritative zones. it's a caching nameserver, it's not supposed to serve authoritative zones! if you are using the regular nameserver package, bind.arch, it breaks nothing. it keeps the old config and copies the new config to .rpmnew.
why not run an inward facing IDS- something like snort. It's easy enough to setup a script that automatically terminates accounts of people sending abuse, and to do it on the first instance of that abuse.