I haven't had to interview too many people at my current job (boo, federal budget cuts), but when I did, on 80% or so of people, I asked the question:
"Star Wars or Star Trek?"
The thing is, I didn't really care which one you picked, so long as you could explain why. And if you picked something else (Firefly, Battlestar Glactica, Dr. Who, Red Dwarf, etc.) and could give a passionate answer, that's even better. The only wrong answer is the 'I'm not going to pick one or the other because I don't want to offend anyone' unless you could really impress me some creativity in the process.
And for anyone who complains that there might be people who haven't seen any of 'em (I still know some people who are almost 30 and qualify)... I work at a NASA center... if you haven't seen any of the TV shows I've listed, there's a *really* high probability that you wouldn't fit in.
... then why not have them be the keepers of the data?
Then they'd be able to accurately monitor how it's actually being used.
Having some third party manage it just seems like one of those 'well, technically we're not supposed to, so we found a loophole'... like how they're not allowed to operate spy satellites over the US, so they have to instead buy imagery from commercial businesses to get those images.
Back when I lived in DC (late 1990s) there were regular reports of exploding manholes... with the best guess of the cause being a combination of gas leaks and electrical shorts. Of couse, in the report on the problem blamed PEPCO (electrical) not Washington Gas.
About 10 years ago, they had a solution -- install manholes with vent holes in them, so the gas pressure can't build up as easily. Of course, you instead get extra water underground, which can lead to faster corrosion of pipes.
Last year, when the methane levels were first reported, the estimate was 38 exploding manholes per year... so I'm guessing the vented covers have been less than successful.
And then you sit there waiting 20+ minutes, and when it gets to 100%, it then goes back to 0%, because that was just the time to complete one file... but there's no indication of how many files need to be processed, so you're just left hanging.
I wonder why people even other trying to give estimates on time to completion.
I completely agree -- I've got one on the front porch, but nothing on the back deck. If I want to run a line without leaving the back door open, I end up stretching an extension cord from the detached garage behind the house.
Putting them on the sides of the house would be useful for hedge trimmers, the electric chainsaw for when I don't want to drag out the big one, paint sprayer, pressure washer, etc.
I've been places where there was conduit run so they popped up boxes about a foot high near the large trees. It was low enough so it wasn't a huge distraction, but gave the ability to run lights, sound and fans when having a party and decorations for the holidays.
You're making the mistake of assuming that means that you *must* have 'great content' to 'keep users'.
Which is either not true, or 'great content' has been defined as 'any content that keeps users', and then we might be able to say 'this review was great content' with a completely straight face.
It's been so long ago since I had a land-line account, I can't remember if I was paying $8/month for service plus $6/month to not be listed, or the other way around.
Either way, it was possible for pay the phone company so that I did *not* show up in the phone book. Which I happily paid, as I'm one of those people with a unique name, and previously had a stalker.
Butthead: Hey Beavis.
Beavis: What?
Butthead: I was just like...thinking and stuff. It was pretty cool.
Beavis: Yeah, I'm gonna try that.
Butthead: TV is cool.
Beavis: Yeah, yeah, TV rules. It rules! Yeah...
Butthead: Hey beavis...I heard that pretty soon, they're gonna have, like, 500
channels.
That's gonna be cool.
Beavis: Really? That would be cool.
Butthead: You know what would be really cool, though? If like, one of the
channels didn't suck.
Beavis: Yeah, but, like, if one of them didn't suck, why would you need the
other...um...
three hundred and twenty-seven?
Butthead: Because, you know all those tv shows that suck? It's like, you gotta
put them somewhere! You can't put 'em on the cool channel!
Beavis: Yeah, yeah! They should call it the cool channel!
Personally, I think it's spot-on, but I don't necessarily think that non-technical managers are a problem. If they know what their deficiencies are, and are willing to ask for help at the appropriate time, they might be just fine.
But I've also had 'technical' managers who were from different fields -- one only dealt with mainframes, and when our departments got merged, didn't understand that the unix team oversaw dozens of machines per person and didn't just have a single task; his boss wouldn't bother showing up to meetings and read the white boards afterwards and would report on that to upper management (with the boards sometimes being the 'no, that won't work, I'll show you why after the meeting' diagrams). He also didn't understand why a 35k user mail system change (spread across 15 hosts) wan't just insert a disk and double click. With the two of them together, they'd do things like plan a power outage to service the machine room UPS and not bother telling the sysadmins until the week before, so we had a mad scramble to coordinate between groups on what order the 200+ systems had to come down and back up.
I currently have more than one manager, because I'm a contractor... my boss (who assigns the tasks) has enough IT skills to be dangerous, but he delegates to the rest of us for the most part, so it's not a big deal; my manager deals with the contracting company's HR / corporate headaches. Both come from the sciences; neither one's business-school trained, but they're not IT, either. I actually think it's better than having washed up IT people in management jobs, who think they know what they're doing. (eg, insisting on specific hardware configurations, but not realizing that the process is single threaded so they would've been better off buying more 2-cpu boxes rather than the beefier ones they got some great deal on)
There are two types -- ones that get paid by the company doing the hiring, and those that you have to pay if you get the job.
In the first case, you'd stand a pretty good chance of getting the job -- they select a few few (maybe 3-4) people who they consider to be qualified for the job, and then pass it up to the company. Of course, they get paid 20-40% of the annual salary to do so, but it's basically the equivalent of the company outsourcing their HR department. (and they're not a company that's got a requirement that they post the jobs publicly, or the job's been open for too long with no good hits)
In the second case, they have absolutely no problem with wasting your time, because they're not wasting theirs, either... they can just throw a bunch of people at the company until one sticks, and then they get their money. All for the cost of a person making a few phone calls and maybe restructuring your resume. For a couple of days work (for all of the people they process), they get a nice fat check... but *you* have to pay them for it. (maybe a month or two of salary).
If you get a call from a headhunter, ask them what company the job is with... if they're in the second group, they'll never tell you, as you can make the end-run around them and not have to pay them. The first one is often the gatekeeper to the job, and so you can't go around them... but if you can, you might be able to talk the company into a signing bonus (that they'd have otherwise had to pay to the headhunter).
In the worst case I heard of, the headhunter started pressuring someone to take certification classes to make them more 'attractive' to employers. The way I heard the story, I wouldn't be surprised if they were getting kickbacks from the training company.
Year before last I had an hour long conversation with my grandmother, as she had called me up to ask me about the whole '2012' thing, and I would know if the sun was going to blow up or not.
It took a while, but I managed to get her to accept that the weathermen can't even tell us if it's going to rain next week, and we think there was some group hundreds of years ago that knew down to the day when something was going to happen? It'd be like claiming to know years in advance of when an earthquake, volcano eruption or tornado was going to happen.
Imagine being one of the people who read the e-mail address placed on error messages on the STEREO servers.
Then imagine that STEREO had a highly compressed near-real time data stream that was used for space weather forecasting. And that the files were replaced once the full-resolution data was downlinked and processed.
And, by working for the government, you have a duty to respond to requests for information from the public, even when they're being completely abusive in their messages.
And there's no reason to view some GIF that someone made when you can just view a slideshow of the images directly. It'll let you speed up / slow down the images (once you've downloaded them). After that, it should move into the field of view of COR2A
He took a list of 10 languages, added another list of languages, then looked for those...
Ie, there could well be other languages that he didn't look for that are more in demand than the one he looked for.
I don't know exactly how he handled only 'programmer' jobs for Dice.com... but they've got 31 jobs that match 'perl'. Add that to the 3 from the other site, and we're looking at 4.4% (behind Python, above, C++, VB, TSQL, etc.
Of course, this is always going to be a point-in-time study. (I found 63 'Ruby' jobs (out of 745), which would put it at 8.4%, above his 7.2%) You really need to look at long-term trends. (and you need to make sure to not count the same jobs from week after week... although the fact that they can't find someone to fill a job might be a sign that's skill's in more demand, but it doesn't necessarily mean it's good to learn that language if employability is your goal)
The really sad thing is that ASP is above Python (33+10=5.8%), and SQL's not on his list but mentioned in 27.8% (155+52) of the jobs. And Postgres (which he didn't check) has more mentions than Hadoop (which he did)
The problem is that it has to be the same or similar every time... so you'd either have to have a fixed routine that's rather similar every time... or what I would do, which is sit there and flip the bird at the computer and/or cuss it out for such a stupid request.
(Of course, some of the answers to the canned 'security questions' that groups try forcing on me are responses such as 'I don't know' 'How should I know?' 'Why would I know that?' and 'I'm an orphan, you bastard'.)
I had a TI-36 back in college... once time, the classroom was so dark for the test (I think the teacher had something on the overhead)... that the calculator wouldn't turn on. It made having a solar calculator kinda lame.
Luckily, the power needed to keep it on was less than the power to turn it on initially, so I was able to shine the flashlight on it once, and it stayed on for the rest of the test. But it was still pretty annoying.
We had outfitted our chase van for the 1995 SunRayce, and had gotten Bell Atlantic (might've been Bell Atlantic-NYNEX at that point) to donate a car phone plus some coverage... and we got a phone that had an RJ11 plug on it.
So... we did some tests in the DC area before heading out to the race. The only place we could hold a decent connection (9600 baud... that was pretty good for the days of 33.6k modems, considering we were on an analog cell phone) was along the BW Parkway... near the NSA.
Which is retrospect seems kinda strange, now that they don't want any portable electronic devices going into secured places. (unless of course it was a rogue cell tower trying to specifically get people from the NSA to route through them)
You also get lots of strange looks from people when driving through Georgetown in a large white van w/ tinted windows and a half dozen antennas on the roof. (GPS, cell phone, 2 xUHF,2 x CB, radio modem (to talk to the solar car), etc.)
ps. by 'browsed the internet' I mean 'FTPed some files'. We might've used gopher, too.
That 200-terabit milestone also surpasses the ten years' worth of data returned via NASA's Deep Space Network from all other missions managed by NASA's Jet Propulsion Laboratory in Pasadena, California
I'm not sure how to read that sentance. JPL manages missions, and they also manage the DSN. But missions managed at APL, GSFC, MSFC and other places *also* use the DSN.
And DSN's much older than 10 years... Voyager uses it, and it was launched in 1977.
It sounds to me like they just picked a convenient time for their 'more than' comparison, and even then didn't even compare it to the whole thing, only some subset.
(disclaimer : I work for some missions that use the DSN, that aren't managed by JPL)
You have to do the numbers in high density 3.5" (1.44MB formatted)... so 18,222,223 disks for 25TB.
And I have no idea how you got 12.8 million (DD = 800kB or 720kB depending on formatting). Your numbers suggest 2.038MB per floppy... some HD 3.5" were marketed as "2MB" but that was *unformatted*. And the rounding error is likely from 1024 vs. 1000 multiples between kB/MB/GB/TB.
Instagram doesn't make things. (I'm a programmer... a website is not a 'thing'). And it doesn't do research anywhere near the level that Kodak used to.
Instagram *might* be a replacement for Kodak's 20 year old way of distributing photos, or maybe the online services they killed last year... but they don't make cameras (as that's been eaten by the smart phone companies), film and paper (now mostly digital) or printers.
Instragram would've been one small department within a larger company... you might compare it to Ofoto.
The Holiday Inn costs money, which they likely didn't budget for.
Due to the rules that kicked in because of sequestration and other agencies wasting money on conferences, there are a ton of new rules regarding attending meetings... and meetings at a government facility don't have the same restrictions as a that at other locations.
There are a lot of NASA people involved in this particular field... and one of those other restrictions is no more than ~50-60 people from an agency at a given meeting. (it's actually based on dollar amounts, I think)
So... if it was being held off-site, a lot of NASA folks wouldn't be allowed to go. Hold on at a NASA site, and the Chinese and anyone who has a name on a government watch list can't attend.*
* Note that I did not say 'is on' the list. The whole argument about how there are only a few people being watched is crap when some have multiple aliases, and those names match lots of other people... like my neighbor's kid who hasn't been allowed to check in online for flights since he was 2 years old... because he shares a first & last name with an IRA member
Slashdot Mirror
I haven't had to interview too many people at my current job (boo, federal budget cuts), but when I did, on 80% or so of people, I asked the question:
"Star Wars or Star Trek?"
The thing is, I didn't really care which one you picked, so long as you could explain why. And if you picked something else (Firefly, Battlestar Glactica, Dr. Who, Red Dwarf, etc.) and could give a passionate answer, that's even better. The only wrong answer is the 'I'm not going to pick one or the other because I don't want to offend anyone' unless you could really impress me some creativity in the process.
And for anyone who complains that there might be people who haven't seen any of 'em (I still know some people who are almost 30 and qualify) ... I work at a NASA center ... if you haven't seen any of the TV shows I've listed, there's a *really* high probability that you wouldn't fit in.
... then why not have them be the keepers of the data?
Then they'd be able to accurately monitor how it's actually being used.
Having some third party manage it just seems like one of those 'well, technically we're not supposed to, so we found a loophole' ... like how they're not allowed to operate spy satellites over the US, so they have to instead buy imagery from commercial businesses to get those images.
Not to worry, there are forces at hand already working on the rodent problem there.
Back when I lived in DC (late 1990s) there were regular reports of exploding manholes ... with the best guess of the cause being a combination of gas leaks and electrical shorts. Of couse, in the report on the problem blamed PEPCO (electrical) not Washington Gas.
About 10 years ago, they had a solution -- install manholes with vent holes in them, so the gas pressure can't build up as easily. Of course, you instead get extra water underground, which can lead to faster corrosion of pipes.
Last year, when the methane levels were first reported, the estimate was 38 exploding manholes per year ... so I'm guessing the vented covers have been less than successful.
And then you sit there waiting 20+ minutes, and when it gets to 100%, it then goes back to 0%, because that was just the time to complete one file ... but there's no indication of how many files need to be processed, so you're just left hanging.
I wonder why people even other trying to give estimates on time to completion.
http://xkcd.com/612/
I completely agree -- I've got one on the front porch, but nothing on the back deck. If I want to run a line without leaving the back door open, I end up stretching an extension cord from the detached garage behind the house.
Putting them on the sides of the house would be useful for hedge trimmers, the electric chainsaw for when I don't want to drag out the big one, paint sprayer, pressure washer, etc.
I've been places where there was conduit run so they popped up boxes about a foot high near the large trees. It was low enough so it wasn't a huge distraction, but gave the ability to run lights, sound and fans when having a party and decorations for the holidays.
You're making the mistake of assuming that means that you *must* have 'great content' to 'keep users'.
Which is either not true, or 'great content' has been defined as 'any content that keeps users', and then we might be able to say 'this review was great content' with a completely straight face.
It's been so long ago since I had a land-line account, I can't remember if I was paying $8/month for service plus $6/month to not be listed, or the other way around.
Either way, it was possible for pay the phone company so that I did *not* show up in the phone book. Which I happily paid, as I'm one of those people with a unique name, and previously had a stalker.
For example, does the caffeine whole trying to do last-minute cramming for finals overcome losing so much sleep for so many hours?
And are there implications for PTSD? (if remember things better, could that increase the chances of PTSD?)
And for those too lazy to copy & paste it, here's the link.
Personally, I think it's spot-on, but I don't necessarily think that non-technical managers are a problem. If they know what their deficiencies are, and are willing to ask for help at the appropriate time, they might be just fine.
But I've also had 'technical' managers who were from different fields -- one only dealt with mainframes, and when our departments got merged, didn't understand that the unix team oversaw dozens of machines per person and didn't just have a single task; his boss wouldn't bother showing up to meetings and read the white boards afterwards and would report on that to upper management (with the boards sometimes being the 'no, that won't work, I'll show you why after the meeting' diagrams). He also didn't understand why a 35k user mail system change (spread across 15 hosts) wan't just insert a disk and double click. With the two of them together, they'd do things like plan a power outage to service the machine room UPS and not bother telling the sysadmins until the week before, so we had a mad scramble to coordinate between groups on what order the 200+ systems had to come down and back up.
I currently have more than one manager, because I'm a contractor ... my boss (who assigns the tasks) has enough IT skills to be dangerous, but he delegates to the rest of us for the most part, so it's not a big deal; my manager deals with the contracting company's HR / corporate headaches. Both come from the sciences; neither one's business-school trained, but they're not IT, either. I actually think it's better than having washed up IT people in management jobs, who think they know what they're doing. (eg, insisting on specific hardware configurations, but not realizing that the process is single threaded so they would've been better off buying more 2-cpu boxes rather than the beefier ones they got some great deal on)
Golden Retriever?
No, no ... St. Bernard ...
Latency might suck, but the burst rate is fantaastic if you fill the little barrel full.
There are two types -- ones that get paid by the company doing the hiring, and those that you have to pay if you get the job.
In the first case, you'd stand a pretty good chance of getting the job -- they select a few few (maybe 3-4) people who they consider to be qualified for the job, and then pass it up to the company. Of course, they get paid 20-40% of the annual salary to do so, but it's basically the equivalent of the company outsourcing their HR department. (and they're not a company that's got a requirement that they post the jobs publicly, or the job's been open for too long with no good hits)
In the second case, they have absolutely no problem with wasting your time, because they're not wasting theirs, either ... they can just throw a bunch of people at the company until one sticks, and then they get their money. All for the cost of a person making a few phone calls and maybe restructuring your resume. For a couple of days work (for all of the people they process), they get a nice fat check ... but *you* have to pay them for it. (maybe a month or two of salary).
If you get a call from a headhunter, ask them what company the job is with ... if they're in the second group, they'll never tell you, as you can make the end-run around them and not have to pay them. The first one is often the gatekeeper to the job, and so you can't go around them ... but if you can, you might be able to talk the company into a signing bonus (that they'd have otherwise had to pay to the headhunter).
In the worst case I heard of, the headhunter started pressuring someone to take certification classes to make them more 'attractive' to employers. The way I heard the story, I wouldn't be surprised if they were getting kickbacks from the training company.
Year before last I had an hour long conversation with my grandmother, as she had called me up to ask me about the whole '2012' thing, and I would know if the sun was going to blow up or not.
It took a while, but I managed to get her to accept that the weathermen can't even tell us if it's going to rain next week, and we think there was some group hundreds of years ago that knew down to the day when something was going to happen? It'd be like claiming to know years in advance of when an earthquake, volcano eruption or tornado was going to happen.
Imagine being one of the people who read the e-mail address placed on error messages on the STEREO servers.
Then imagine that STEREO had a highly compressed near-real time data stream that was used for space weather forecasting. And that the files were replaced once the full-resolution data was downlinked and processed.
And, by working for the government, you have a duty to respond to requests for information from the public, even when they're being completely abusive in their messages.
I've probably spent weeks of my life responding to people trying to explain that no, NASA is not covering up evidence of UFOs, because we know what those items are-- compression artifacts, internal reflections, SEP hits, etc. (in this particular case, it's pixel bleed on the CCDs).
And there's no reason to view some GIF that someone made when you can just view a slideshow of the images directly. It'll let you speed up / slow down the images (once you've downloaded them). After that, it should move into the field of view of COR2A
But we do have it ... it's called photovoltaic.
It's possible to make a solar panel that's highly efficient at a single wavelength, and then you point a laser of that wavelength at it.
The problem is, there's groups like GDI and Pacific Tech out there.
He took a list of 10 languages, added another list of languages, then looked for those ...
Ie, there could well be other languages that he didn't look for that are more in demand than the one he looked for.
I don't know exactly how he handled only 'programmer' jobs for Dice.com ... but they've got 31 jobs that match 'perl'. Add that to the 3 from the other site, and we're looking at 4.4% (behind Python, above, C++, VB, TSQL, etc.
Of course, this is always going to be a point-in-time study. (I found 63 'Ruby' jobs (out of 745), which would put it at 8.4%, above his 7.2%) You really need to look at long-term trends. (and you need to make sure to not count the same jobs from week after week ... although the fact that they can't find someone to fill a job might be a sign that's skill's in more demand, but it doesn't necessarily mean it's good to learn that language if employability is your goal)
The really sad thing is that ASP is above Python (33+10=5.8%), and SQL's not on his list but mentioned in 27.8% (155+52) of the jobs. And Postgres (which he didn't check) has more mentions than Hadoop (which he did)
The problem is that it has to be the same or similar every time ... so you'd either have to have a fixed routine that's rather similar every time ... or what I would do, which is sit there and flip the bird at the computer and/or cuss it out for such a stupid request.
(Of course, some of the answers to the canned 'security questions' that groups try forcing on me are responses such as 'I don't know' 'How should I know?' 'Why would I know that?' and 'I'm an orphan, you bastard'.)
I had a TI-36 back in college ... once time, the classroom was so dark for the test (I think the teacher had something on the overhead) ... that the calculator wouldn't turn on. It made having a solar calculator kinda lame.
Luckily, the power needed to keep it on was less than the power to turn it on initially, so I was able to shine the flashlight on it once, and it stayed on for the rest of the test. But it was still pretty annoying.
Well, I wasn't driving at the same time.
We had outfitted our chase van for the 1995 SunRayce, and had gotten Bell Atlantic (might've been Bell Atlantic-NYNEX at that point) to donate a car phone plus some coverage ... and we got a phone that had an RJ11 plug on it.
So ... we did some tests in the DC area before heading out to the race. The only place we could hold a decent connection (9600 baud ... that was pretty good for the days of 33.6k modems, considering we were on an analog cell phone) was along the BW Parkway ... near the NSA.
Which is retrospect seems kinda strange, now that they don't want any portable electronic devices going into secured places. (unless of course it was a rogue cell tower trying to specifically get people from the NSA to route through them)
You also get lots of strange looks from people when driving through Georgetown in a large white van w/ tinted windows and a half dozen antennas on the roof. (GPS, cell phone, 2 xUHF,2 x CB, radio modem (to talk to the solar car), etc.)
ps. by 'browsed the internet' I mean 'FTPed some files'. We might've used gopher, too.
I'm not sure how to read that sentance. JPL manages missions, and they also manage the DSN. But missions managed at APL, GSFC, MSFC and other places *also* use the DSN.
And DSN's much older than 10 years ... Voyager uses it, and it was launched in 1977.
It sounds to me like they just picked a convenient time for their 'more than' comparison, and even then didn't even compare it to the whole thing, only some subset.
(disclaimer : I work for some missions that use the DSN, that aren't managed by JPL)
You have to do the numbers in high density 3.5" (1.44MB formatted) ... so 18,222,223 disks for 25TB.
And I have no idea how you got 12.8 million (DD = 800kB or 720kB depending on formatting). Your numbers suggest 2.038MB per floppy ... some HD 3.5" were marketed as "2MB" but that was *unformatted*. And the rounding error is likely from 1024 vs. 1000 multiples between kB/MB/GB/TB.
302 is a 'We've moved somewhere else'.
At the very least if they were going to use a redirect, they should've recommended a 307 ('We're at some other place, but check back later')
100% agreed.
Instagram doesn't make things. (I'm a programmer ... a website is not a 'thing'). And it doesn't do research anywhere near the level that Kodak used to.
Instagram *might* be a replacement for Kodak's 20 year old way of distributing photos, or maybe the online services they killed last year ... but they don't make cameras (as that's been eaten by the smart phone companies), film and paper (now mostly digital) or printers.
Instragram would've been one small department within a larger company ... you might compare it to Ofoto.
A few issues here:
So ... if it was being held off-site, a lot of NASA folks wouldn't be allowed to go. Hold on at a NASA site, and the Chinese and anyone who has a name on a government watch list can't attend.*
* Note that I did not say 'is on' the list. The whole argument about how there are only a few people being watched is crap when some have multiple aliases, and those names match lots of other people ... like my neighbor's kid who hasn't been allowed to check in online for flights since he was 2 years old ... because he shares a first & last name with an IRA member