Yeah, you wonder why HI hasn't invested more heavily in solar and wind. There are spots on the islands where the wind is just about always going strong. No shortage of sun on the leeward side of any Hawaiian island. And oil makes electricity there very expensive. So if solar and wind are even vaguely close to cost effective, why hasn't Hawaii invested heavily?
For sites I trust I want a different application. A true remote display language without all the godawful workarounds it takes to do that in html.
I'm the kind of guy who surfs with noscript turned on and usually leaves a website that would require me to enable javascript to use it. For all the untrusted web sites I want a *simple* document markup language. Heck, I was happy with html3. And I don't feel any compelling need to access untrusted sites that require more functionality on my computer.
Can I please have a button that limits my browser to html 4.01? HTML 5 has thrown in everything including the kitchen sink. I don't want that functionality available to the web hack at http://random.web.site/. It's too much. I only want it on a web site I've decided I trust and I'd rather have it in a completely different program.
...leave it. Phones. Printers. Networking gear. If it doesn't have a hard drive, leave it behind.
Take the servers. Take the dozen workstations and laptops most likely to have stored mission-critical work on local drives. You already know who the knuckleheads are. And then go.
Let insurance take care of the rest. If you've time, take some pictures so you can prove the state of the office. And if you're self-insured, well, now you know better for next time.
First thing I want is something I can auto-log in which will place a set of browser windows in specific spots on the screen and then kill and replace them when I feel like it, also at the chosen spots on the screen. For my unattended network status display. When I can do this, I'll *consider* using one of them for my actual desktop. Until then, they lack utility.
If you want a spot on a professional basketball team, you should already know what a layup is and be able to do one. It's not trivia. It's "basics."
Here's a clue: a highly motivated, smart person doesn't show up without already knowing the basics. So when someone does show up ignorant of the basics, that's a really strong clue that they're far from highly motivated.
In my particular backwater of this fortune 500 company, 3 people on 3 different teams would hamstring the project and kill it with competing interests, none of them the customer's. If I can find a jack of all trades moderately clueful on security, I'll take him.
Yeah, he eventually decided it must be a hash. A hash is a good guess but the wrong answer. You're matching against multiple patterns with different don't-care bits. Every bit is significant in a hash lookup; they don't handle floating don't-care conditions.
If you're doing it purely in software, you're using something like a radix tree (alternately called a Trie). Otherwise you're sorting patterns by preference, stuffing them into a TCAM (a tri-state SRAM-like device) and burning lots of electricity and heat to get your answer in a single clock.
The guy I'm hiring for now doesn't have to know this. I'm not looking for a developer to build me software, I'm looking for a guy who can make damn sure operability is adequately addressed in the architecture, to make sure the logs spit out usable knowledge, to make sure the plans don't call for breaking traffic the devs didn't realize was ordinarily on a production network and then to run both that and other equipment.
When I ask, "If I block all ICMP packets passing my firewall, what breaks in TCP?" he should know enough about how TCP works to at least puzzle out the answer. Otherwise he just won't be able to spot the errors until he tries to take the product from lab to production and it falls flat. I can get an install monkey cheap. I need someone who can actually find the problems *before* the install.
The guy I need is a generalist. A little bit sysadmin. A little bit architect. A little bit tester. A little bit developer. And a whole lot of IP networking knowledge.
Am I any good at recognizing talent? Yeah, actually, I am. The much more challenging question is evaluating whether the individual fits. I won't claim any great skill there.
I don't want a good networking technician. I don't have enough work for a good networking technician. I want a good generalist with a background in networking and network security. Him I have work for.
As for programmers, I have plenty of good ones and I haven't been tasked with finding more. I'm looking for a generalist, not a heads-down programmer.
Scripting passes my test for a phone interview. If you can tell me why you're the right guy for the job on the phone, you get an in-person interview regardless of any checkboxes on my list. I'll ask you some questions that require a deep knowledge of TCP/IP. Neither programming nor operations per se, but about how the protocol itself works when you poke at it in various ways. If your path has led you to the answers, I won't be particular about how you got there.
You note wrongly. If I could expect to get someone through the clearance process in two months, the requirement would read "clearable" instead of "cleared."
I'm told it currently takes about 12 months to go from zero to TS for someone with a clean background. Money I have. 12 months I don't.
I'm willing to pay EXACTLY what I think a candidate is worth. That's usually in the high five figures to low six figures. My HR department even has a group that figures out comparable salaries that other folks in the area are paid. They look at several similar job titles and help me narrow in on what lines up most closely with the guy I'm looking at. And if I really want the person, I have the latitude to push that up a little. No point spending the effort interviewing folks only to the lose the one I wanted because I made a low ball offer.
Why aren't the other people on the other projects capable of handling this?
We have a couple who are. They're fully subscribed on other projects and the time stolen to keep this project moving is missed. For the next year or two, this project needs someone who can give 50%-100% instead of 10%.
You expect someone who is a master to be able to explain the complex concepts to a novice in such a way that the novice can turn those complex concepts into code.
I'm looking for a senior level operations guy who can (A) explain stuff to senior level programmers without operations experience and (B) run some nifty BGP-using networks with interesting architectures.
That's your preference I suppose. The on call person tends to get hit with an after-hours event about once every two to three weeks. With three folks in the rotation, that's around 6 times a year each. Usually it's ack the alert and then go back to sleep because the redundancy in the system allows it to wait until morning without any customer impact. Once in a while it's more serious.
So a very real possibility of a very limited career there?
Very few folks here work just one project. Those who do, do so by choice.
If the people writing the code for the firewall need someone else to tell them what the packets look like then there is a problem. And that kind of education is a couple weeks at maximum.
Big difference between what stuff looks like in the lab and what it looks like on a production network. I need someone who knows the qualitative difference. That takes years of operations experience.
Too much work. Too few people. Couple weak skill areas. Hiring the right person for this position would greatly reduce the whole team's stress level, would make the project run more smoothly and since we're paid by the hour would make us more money.
No kidding. I interviewed someone with alleged experience writing high data rate deep packet inspection software. I asked him what packet characteristics he matched on. He explained that he received the packet profiles from someone else's software. I asked him what sort of data structure he used to manage matching packets against the profiles. He explained that he used a vendor library that was "really fast."
My dream candidate has a CS degree, a year or two developing network security software and 6+ years running routers and Linux servers on a US DoD network. Present that resume and the perfunctory interview will just be about making sure you're not lying.
Not shift work. Some flex to the hours. A little bit of travel but you'll like the destination. Share the on call in a rotation with two other folks and expect to have the authority to make changes which reduce the occurrence of 2 am emergencies. I'm one of the two and I'll insist on it.:)
What is this "minor software development"? Scripting? C or C++ or Java?
Anything at all gets you to a phone interview. The position is for the operations guy on a product team building a specialty firewall product. He's the guy who has to keep the devs grounded in both what can actually be maintained in the field and what the packets on deployed networks actually look like.
How intense is the security clearance you're looking for? Credit check? 10 years previous legal problems? Top Secret?
I want TS. SCI would be a bonus. I'll consider Secret for someone with an otherwise terrific skill set with the assumption I'll have to find side roles for them until they can get up to TS+SCI.
hire someone with 1 of the 3, and then let him or her learn the other two as they go
I've already done that. He's doing great work at a junior level and I expect fantastic things from him as he grows. In the mean time, I need a senior level person for a product that's due next year.
Slashdot Mirror
Yeah, you wonder why HI hasn't invested more heavily in solar and wind. There are spots on the islands where the wind is just about always going strong. No shortage of sun on the leeward side of any Hawaiian island. And oil makes electricity there very expensive. So if solar and wind are even vaguely close to cost effective, why hasn't Hawaii invested heavily?
For sites I trust I want a different application. A true remote display language without all the godawful workarounds it takes to do that in html.
I'm the kind of guy who surfs with noscript turned on and usually leaves a website that would require me to enable javascript to use it. For all the untrusted web sites I want a *simple* document markup language. Heck, I was happy with html3. And I don't feel any compelling need to access untrusted sites that require more functionality on my computer.
Can I please have a button that limits my browser to html 4.01? HTML 5 has thrown in everything including the kitchen sink. I don't want that functionality available to the web hack at http://random.web.site/. It's too much. I only want it on a web site I've decided I trust and I'd rather have it in a completely different program.
In my case, crashed meant that a bunch of processes got stuck in an I/O wait ("ps" reported state "D")
...leave it. Phones. Printers. Networking gear. If it doesn't have a hard drive, leave it behind.
Take the servers. Take the dozen workstations and laptops most likely to have stored mission-critical work on local drives. You already know who the knuckleheads are. And then go.
Let insurance take care of the rest. If you've time, take some pictures so you can prove the state of the office. And if you're self-insured, well, now you know better for next time.
First thing I want is something I can auto-log in which will place a set of browser windows in specific spots on the screen and then kill and replace them when I feel like it, also at the chosen spots on the screen. For my unattended network status display. When I can do this, I'll *consider* using one of them for my actual desktop. Until then, they lack utility.
NC-20's position: http://www.nc-20.com/sealevelrise.htm
"It is NC 20’s position that any SLR projections should be based on science, not computer models based on human speculation."
Couldn't have said it better myself.
Insurance companies will use whatever sources they think are reasonable, so some of this to-be-planned development may be hard to insure.
Nice theory but private insurers don't offer flood insurance in coastal areas. That's all done through the Federal National Flood Insurance Program.
http://www.floodsmart.gov/floodsmart/
If you want a spot on a professional basketball team, you should already know what a layup is and be able to do one. It's not trivia. It's "basics."
Here's a clue: a highly motivated, smart person doesn't show up without already knowing the basics. So when someone does show up ignorant of the basics, that's a really strong clue that they're far from highly motivated.
In my particular backwater of this fortune 500 company, 3 people on 3 different teams would hamstring the project and kill it with competing interests, none of them the customer's. If I can find a jack of all trades moderately clueful on security, I'll take him.
Yeah, he eventually decided it must be a hash. A hash is a good guess but the wrong answer. You're matching against multiple patterns with different don't-care bits. Every bit is significant in a hash lookup; they don't handle floating don't-care conditions.
If you're doing it purely in software, you're using something like a radix tree (alternately called a Trie). Otherwise you're sorting patterns by preference, stuffing them into a TCAM (a tri-state SRAM-like device) and burning lots of electricity and heat to get your answer in a single clock.
The guy I'm hiring for now doesn't have to know this. I'm not looking for a developer to build me software, I'm looking for a guy who can make damn sure operability is adequately addressed in the architecture, to make sure the logs spit out usable knowledge, to make sure the plans don't call for breaking traffic the devs didn't realize was ordinarily on a production network and then to run both that and other equipment.
When I ask, "If I block all ICMP packets passing my firewall, what breaks in TCP?" he should know enough about how TCP works to at least puzzle out the answer. Otherwise he just won't be able to spot the errors until he tries to take the product from lab to production and it falls flat. I can get an install monkey cheap. I need someone who can actually find the problems *before* the install.
The guy I need is a generalist. A little bit sysadmin. A little bit architect. A little bit tester. A little bit developer. And a whole lot of IP networking knowledge.
Am I any good at recognizing talent? Yeah, actually, I am. The much more challenging question is evaluating whether the individual fits. I won't claim any great skill there.
I don't want a good networking technician. I don't have enough work for a good networking technician. I want a good generalist with a background in networking and network security. Him I have work for.
As for programmers, I have plenty of good ones and I haven't been tasked with finding more. I'm looking for a generalist, not a heads-down programmer.
Scripting passes my test for a phone interview. If you can tell me why you're the right guy for the job on the phone, you get an in-person interview regardless of any checkboxes on my list. I'll ask you some questions that require a deep knowledge of TCP/IP. Neither programming nor operations per se, but about how the protocol itself works when you poke at it in various ways. If your path has led you to the answers, I won't be particular about how you got there.
You note wrongly. If I could expect to get someone through the clearance process in two months, the requirement would read "clearable" instead of "cleared."
I'm told it currently takes about 12 months to go from zero to TS for someone with a clean background. Money I have. 12 months I don't.
I'm willing to pay EXACTLY what I think a candidate is worth. That's usually in the high five figures to low six figures. My HR department even has a group that figures out comparable salaries that other folks in the area are paid. They look at several similar job titles and help me narrow in on what lines up most closely with the guy I'm looking at. And if I really want the person, I have the latitude to push that up a little. No point spending the effort interviewing folks only to the lose the one I wanted because I made a low ball offer.
Why aren't the other people on the other projects capable of handling this?
We have a couple who are. They're fully subscribed on other projects and the time stolen to keep this project moving is missed. For the next year or two, this project needs someone who can give 50%-100% instead of 10%.
You expect someone who is a master to be able to explain the complex concepts to a novice in such a way that the novice can turn those complex concepts into code.
I'm looking for a senior level operations guy who can (A) explain stuff to senior level programmers without operations experience and (B) run some nifty BGP-using networks with interesting architectures.
That's your preference I suppose. The on call person tends to get hit with an after-hours event about once every two to three weeks. With three folks in the rotation, that's around 6 times a year each. Usually it's ack the alert and then go back to sleep because the redundancy in the system allows it to wait until morning without any customer impact. Once in a while it's more serious.
I mean the company is paid by the hour. The position is, of course, salaried with benefits.
So a very real possibility of a very limited career there?
Very few folks here work just one project. Those who do, do so by choice.
If the people writing the code for the firewall need someone else to tell them what the packets look like then there is a problem. And that kind of education is a couple weeks at maximum.
Big difference between what stuff looks like in the lab and what it looks like on a production network. I need someone who knows the qualitative difference. That takes years of operations experience.
Too much work. Too few people. Couple weak skill areas. Hiring the right person for this position would greatly reduce the whole team's stress level, would make the project run more smoothly and since we're paid by the hour would make us more money.
No kidding. I interviewed someone with alleged experience writing high data rate deep packet inspection software. I asked him what packet characteristics he matched on. He explained that he received the packet profiles from someone else's software. I asked him what sort of data structure he used to manage matching packets against the profiles. He explained that he used a vendor library that was "really fast."
My dream candidate has a CS degree, a year or two developing network security software and 6+ years running routers and Linux servers on a US DoD network. Present that resume and the perfunctory interview will just be about making sure you're not lying.
Not shift work. Some flex to the hours. A little bit of travel but you'll like the destination. Share the on call in a rotation with two other folks and expect to have the authority to make changes which reduce the occurrence of 2 am emergencies. I'm one of the two and I'll insist on it. :)
Accepting an interim clearance is the customer's choice. The customer refused.
The project isn't stalled for lack of this person but it would be doing a lot better with him.
Shouldn't you be [...] paying them the 6 figure salaries that they'd demand?
Something makes you think I'm not?
just tell us which state.
Dulles airport.
What is this "minor software development"?
Scripting? C or C++ or Java?
Anything at all gets you to a phone interview. The position is for the operations guy on a product team building a specialty firewall product. He's the guy who has to keep the devs grounded in both what can actually be maintained in the field and what the packets on deployed networks actually look like.
How intense is the security clearance you're looking for? Credit check? 10 years previous legal problems? Top Secret?
I want TS. SCI would be a bonus. I'll consider Secret for someone with an otherwise terrific skill set with the assumption I'll have to find side roles for them until they can get up to TS+SCI.
hire someone with 1 of the 3, and then let him or her learn the other two as they go
I've already done that. He's doing great work at a junior level and I expect fantastic things from him as he grows. In the mean time, I need a senior level person for a product that's due next year.