Simple: People talk... Or HR does something stupid, like accidentally send a copy of someone's offer letter to the whole office... Or send a copy of the promotion package worksheet to a "public" printer, not bothering to use the PIN-to-print function (and not even just walking to the printer after hitting 'print'). I could go on and on... And of course, none of these things have ever happened where I work.
Actually, usually FPGAs are on the bleeding edge of manufacturing processes. Intel may have beat everyone to 28/32nm, but expect to see 28nm FPGAs from Altera and Xilinx (from TSMC and/or Samsung) around the same time as 28/32nm ASICs from AMD or nVidia. Intel rolls their own, but everyone else is using the same foundries...
Ah, yes... I loved Geoworks Ensemble. I wrote all my school papers through high-school with the word processor (GeoWrite? can't remember what it was called...). I agree, it was much better than Word or WordPerfect or the various apps installed on the Macs at school. All around great PC software for the time.
We've got similarly inane requirements here, except the password length is 8+ characters and I don't think there are any foreign-language dictionary-based checks. and we have to change our passwords every 90 days. Coming up with new passwords is all kinds of fun.
What really gets me is the "3 characters different from the previous 10" requirement, which we have, verbatim. It seems like this would require storing the actual password, rather than a salted hash thereof. It'd be phenomenally awesome if the password database was hacked/leaked/etc.
Prior to my Toyota, every other car I've ever had (or even driven) has had the accelerator pivot from the floor. i found that preferable in general, but it wasn't a deal-breaker, so I bought the car. Is this a European/American vs Japanese thing?
Not sure if it's by design, but I've found that the point where the lane-dividing line becomes solid (as you approach the stop line) is a pretty good indicator of how far out you can safely stop if going close to the speed-limit (give or take). If you're well within the sold region, with a reasonably timed light, it's generally best to continue straight ahead.
Supposedly, at my office (roughly 250-300 people, I think), the company spends over $15k a month on coffee. I can't tell if that means those keurig cups are really expensive, people drink tons of coffee, or if the bean counters are full of it.
Hmmm... My experience in MA has been that state cops routinely set up speed traps (1-6 cops waiting around with LIDAR pointed at traffic), at least on I-290. Traffic may be cruising around 80-85mph (speed limit is 65), but at the speed trap locations, everyone slows to 65-75mph. And no matter how fast/slow traffic is going, there seems to always be 1-2 cars pulled over during the morning commute. It's almost like they adjust the threshold for a ticket to maintain a constant rate of cars pulled-over.
What's really great is when traffic is going around 75-80mph in burst "packets" of maybe 20-30 cars, and the cops step out into the middle lane to point to people at the front of the group to pull over. It's a small wonder that they don't cause horrible 20-car pile-ups all the time...
is this thread still going...? probably not... still...
The problem with this RAID memory scheme is: how do you know which memory chip is providing the correct data without some sort of encoded error correction data (eg ECC)? Drives store CRC data, so you can tell which one is providing wrong data (assuming it's not a drive-is-dead sort of issue). Without that error correction data though, you'd need at least a 3-way mirror (and go with the majority "vote") to know what any given byte of data was correctly supposed to be.
I have to think this "mirroring" function is actually using some sort of hamming code, rather than a direct duplication of bits. It wouldn't make any sense to do otherwise. It'd be interesting to see some documentation though.
But systems that require that new passwords contain no more than N characters in common with the previous M passwords must store the passwords somewhere, no? This is an earnest question. If you can tell that "Abcd" is only one character off from "abcd" from just the hashes themselves, those must be some pretty lousy hash functions.
Where I work, N==2 and M==6, and we have to change passwords every 90 days. Creating a new password is an incredibly annoying exercise. But... if your password expires, upon reseting it to some random value, IT seems to clear the record of previous passwords too. Strong security policies always have little loopholes like this...
I was thinking this same thing... Say 50% of packets have to go into a normal/bulk queue, 35% into a medium queue, and the rest into high. Adjust the thresholds as you like. For users that don't know better, default to normal, and maybe promote to higher levels based on the destination port.
Then, if the rolling average (over a few days or so) for packets marked "high" exceeds the threshold, all subsequent "high" packets get demoted until the average gets back to the required level. If high and medium queues both exceed the threshold, demote everything one more level.
This way, anyone that marked all their packets as high would quickly find all their connections operating at the lowest QoS level. That would either motivate them to fix their settings, or they'd just have to live with it and not be penalizing anyone else by hogging the available bandwidth.
This is more or less off-topic, but project code names can be an incredibly complicated thing. Engineers like to pick fun names, management types like to quibble and force name changes. Where I work, we've had directors complain that some past internal project names (like "Thor", "Baal", "Shivah") were offensive because they were "pagan". Said person had enough clout that now we have rather uninspiring themes, like trees ("oak", "maple", "pine").
It's also unbelievable the amount of high-level arguing that goes on to choose a theme, or rip some team a new one if they dare ignore the theme.
So Thusnelda and Snuppflog are rather refreshingly original...
Indeed, hyperbole isn't helpful. Scuttlebutt at my company has it that reasonably-experienced employees (say, 6-8 years) in India are getting paid about $40-50k, and those in China are getting paid around $30-40k. So that's basically around 35-50% of the salary of someone in the US with the same experience.
Training your workforce is something to do in india, where there will be none of this "cost of living" stuff.
We've been told that most hiring will be done in India and China, unless someone here (north america) can do the job of 2-3 people over there... So, anyone applying for a job here (large asic-design company) had better be highly-skilled and already-trained. I expect this plan to fall a bit short of expectations in a few years, but by then, even the rats will have left the ship...
Re:Anyone care to speculate about his compensation
on
Alan Cox Leaves Red Hat
·
· Score: 3, Informative
In engineering, it's pretty easy to get $80-90k with relatively little experience, or with a not-so-great track record of performance, just by moving around a little. If you're a star performer, in fact, you'd be lucky to get raises sufficient to make much more than new hires who left their previous job because they didn't get any raises (i.e. not great performers), and the new company wants to pay them "market rate". Typically, you'll only match the new hires with your raises. So what, exactly, is the incentive to be a star performer? There is none. You can be a total slacker instead, just change jobs every few years, and do just as well as the guys putting in 90 hours/week and doing the work of several lesser engineers.
This is absolutely true... where I work, the pay ranges that I know of for all north american sites, as of last year, were:
MTS: 103k-168k SMTS: 115k-187k
I'd say roughly MTS means 7+ years of experience, and SMTS means 9+ years, give or take.
Of the people that were brought in at their current level, most (say around 80%) are around the median of those ranges, tending to be a bit higher. Of the people that were promoted to those levels, most are in the bottom 25% of the salary range.
Now, there is certainly significant correlation between compensation and performance, but only when looking within each of the {hired, promoted} groups. But when looking at the whole group (everyone reporting under our director at least), there are several cases of people ranked in the bottom 25% of performers having salaries 30-40% higher than the top ranked engineers. Just from the salary data, it's very easy to tell who's been around for a while.
And you're right, there's no incentive to be a top performer really, other than having a better shot at escaping the layoff axe when it comes around...
... You do realize that the very first congress passed laws instituting warrantless searches at the borders and the supreme court has upheld them citing something along the lines of a key component of sovereignty is controlling what comes in or across the border. The constitution isn't in jeopardy here, at least as the founding fathers intended it to be used.
by rough estimate, i'd say that probably 85+% of the territory of the original 13 colonies fell within this 100-mile band, and probably well over 95% the population was living within that area. the supposition that "the founding fathers" would have supported warrentless "border" searches almost anywhere within the country seems, well, a bit far-fetched.
O7 starts at $4961 a month, and most of them probably have closer to 30 years of service than 20. so that's almost $100k a year in retirement. i'm not certain, but i think at least some of that may be tax-free too.
but the law isn't that the can tap without a warrant, it's that they can get a warrant without providing the evidence/rationale behind asking for it (e.g. he seems like a bad guy...) until up to 3 days later. they didn't bother getting the warrant.
i agree, i love in the dashboard lighting of my jetta, and would gladly pay to have it transferred to a new car when i get one (probably won't be another jetta, i don't like the new styling).
for me, the indigo blue (which i dim almost all the way) is far less obtrusive at night, so that i can focus on the road without distraction. the red serves to highlight the important info should i need to glance at the gages.
i really would love to be able to get a new car (from another manufacturer) with the same dashboard lighting scheme. for me, it really is the best i've seen.
Slashdot Mirror
Freedom doesn't have to be zero-sum, does it?
Simple: People talk... Or HR does something stupid, like accidentally send a copy of someone's offer letter to the whole office... Or send a copy of the promotion package worksheet to a "public" printer, not bothering to use the PIN-to-print function (and not even just walking to the printer after hitting 'print'). I could go on and on... And of course, none of these things have ever happened where I work.
Actually, usually FPGAs are on the bleeding edge of manufacturing processes. Intel may have beat everyone to 28/32nm, but expect to see 28nm FPGAs from Altera and Xilinx (from TSMC and/or Samsung) around the same time as 28/32nm ASICs from AMD or nVidia. Intel rolls their own, but everyone else is using the same foundries...
Ah, yes... I loved Geoworks Ensemble. I wrote all my school papers through high-school with the word processor (GeoWrite? can't remember what it was called...). I agree, it was much better than Word or WordPerfect or the various apps installed on the Macs at school. All around great PC software for the time.
We've got similarly inane requirements here, except the password length is 8+ characters and I don't think there are any foreign-language dictionary-based checks. and we have to change our passwords every 90 days. Coming up with new passwords is all kinds of fun.
What really gets me is the "3 characters different from the previous 10" requirement, which we have, verbatim. It seems like this would require storing the actual password, rather than a salted hash thereof. It'd be phenomenally awesome if the password database was hacked/leaked/etc.
Prior to my Toyota, every other car I've ever had (or even driven) has had the accelerator pivot from the floor. i found that preferable in general, but it wasn't a deal-breaker, so I bought the car. Is this a European/American vs Japanese thing?
Not sure if it's by design, but I've found that the point where the lane-dividing line becomes solid (as you approach the stop line) is a pretty good indicator of how far out you can safely stop if going close to the speed-limit (give or take). If you're well within the sold region, with a reasonably timed light, it's generally best to continue straight ahead.
I noticed this very thing (green&red countdown timers) when I visited Shanghai for business last year. Seems like a pretty good idea.
Supposedly, at my office (roughly 250-300 people, I think), the company spends over $15k a month on coffee. I can't tell if that means those keurig cups are really expensive, people drink tons of coffee, or if the bean counters are full of it.
Hmmm... My experience in MA has been that state cops routinely set up speed traps (1-6 cops waiting around with LIDAR pointed at traffic), at least on I-290. Traffic may be cruising around 80-85mph (speed limit is 65), but at the speed trap locations, everyone slows to 65-75mph. And no matter how fast/slow traffic is going, there seems to always be 1-2 cars pulled over during the morning commute. It's almost like they adjust the threshold for a ticket to maintain a constant rate of cars pulled-over.
What's really great is when traffic is going around 75-80mph in burst "packets" of maybe 20-30 cars, and the cops step out into the middle lane to point to people at the front of the group to pull over. It's a small wonder that they don't cause horrible 20-car pile-ups all the time...
is this thread still going...? probably not... still...
The problem with this RAID memory scheme is: how do you know which memory chip is providing the correct data without some sort of encoded error correction data (eg ECC)? Drives store CRC data, so you can tell which one is providing wrong data (assuming it's not a drive-is-dead sort of issue). Without that error correction data though, you'd need at least a 3-way mirror (and go with the majority "vote") to know what any given byte of data was correctly supposed to be.
I have to think this "mirroring" function is actually using some sort of hamming code, rather than a direct duplication of bits. It wouldn't make any sense to do otherwise. It'd be interesting to see some documentation though.
But systems that require that new passwords contain no more than N characters in common with the previous M passwords must store the passwords somewhere, no? This is an earnest question. If you can tell that "Abcd" is only one character off from "abcd" from just the hashes themselves, those must be some pretty lousy hash functions.
Where I work, N==2 and M==6, and we have to change passwords every 90 days. Creating a new password is an incredibly annoying exercise. But... if your password expires, upon reseting it to some random value, IT seems to clear the record of previous passwords too. Strong security policies always have little loopholes like this...
I was thinking this same thing... Say 50% of packets have to go into a normal/bulk queue, 35% into a medium queue, and the rest into high. Adjust the thresholds as you like. For users that don't know better, default to normal, and maybe promote to higher levels based on the destination port.
Then, if the rolling average (over a few days or so) for packets marked "high" exceeds the threshold, all subsequent "high" packets get demoted until the average gets back to the required level. If high and medium queues both exceed the threshold, demote everything one more level.
This way, anyone that marked all their packets as high would quickly find all their connections operating at the lowest QoS level. That would either motivate them to fix their settings, or they'd just have to live with it and not be penalizing anyone else by hogging the available bandwidth.
Seems that he's still at RAD:
http://www.ddj.com/hpc-high-performance-computing/216402188
who said anything about loving Grandma...?
This is more or less off-topic, but project code names can be an incredibly complicated thing. Engineers like to pick fun names, management types like to quibble and force name changes. Where I work, we've had directors complain that some past internal project names (like "Thor", "Baal", "Shivah") were offensive because they were "pagan". Said person had enough clout that now we have rather uninspiring themes, like trees ("oak", "maple", "pine").
It's also unbelievable the amount of high-level arguing that goes on to choose a theme, or rip some team a new one if they dare ignore the theme.
So Thusnelda and Snuppflog are rather refreshingly original...
Indeed, hyperbole isn't helpful. Scuttlebutt at my company has it that reasonably-experienced employees (say, 6-8 years) in India are getting paid about $40-50k, and those in China are getting paid around $30-40k. So that's basically around 35-50% of the salary of someone in the US with the same experience.
Training your workforce is something to do in india, where there will be none of this "cost of living" stuff.
We've been told that most hiring will be done in India and China, unless someone here (north america) can do the job of 2-3 people over there... So, anyone applying for a job here (large asic-design company) had better be highly-skilled and already-trained. I expect this plan to fall a bit short of expectations in a few years, but by then, even the rats will have left the ship...
In engineering, it's pretty easy to get $80-90k with relatively little experience, or with a not-so-great track record of performance, just by moving around a little. If you're a star performer, in fact, you'd be lucky to get raises sufficient to make much more than new hires who left their previous job because they didn't get any raises (i.e. not great performers), and the new company wants to pay them "market rate". Typically, you'll only match the new hires with your raises. So what, exactly, is the incentive to be a star performer? There is none. You can be a total slacker instead, just change jobs every few years, and do just as well as the guys putting in 90 hours/week and doing the work of several lesser engineers.
This is absolutely true... where I work, the pay ranges that I know of for all north american sites, as of last year, were:
MTS: 103k-168k
SMTS: 115k-187k
I'd say roughly MTS means 7+ years of experience, and SMTS means 9+ years, give or take.
Of the people that were brought in at their current level, most (say around 80%) are around the median of those ranges, tending to be a bit higher. Of the people that were promoted to those levels, most are in the bottom 25% of the salary range.
Now, there is certainly significant correlation between compensation and performance, but only when looking within each of the {hired, promoted} groups. But when looking at the whole group (everyone reporting under our director at least), there are several cases of people ranked in the bottom 25% of performers having salaries 30-40% higher than the top ranked engineers. Just from the salary data, it's very easy to tell who's been around for a while.
And you're right, there's no incentive to be a top performer really, other than having a better shot at escaping the layoff axe when it comes around...
... You do realize that the very first congress passed laws instituting warrantless searches at the borders and the supreme court has upheld them citing something along the lines of a key component of sovereignty is controlling what comes in or across the border. The constitution isn't in jeopardy here, at least as the founding fathers intended it to be used.
by rough estimate, i'd say that probably 85+% of the territory of the original 13 colonies fell within this 100-mile band, and probably well over 95% the population was living within that area. the supposition that "the founding fathers" would have supported warrentless "border" searches almost anywhere within the country seems, well, a bit far-fetched.
sure, kobe (wagyu) beef would be much more "marbled". the visual difference, especially uncooked, is quite striking.
more than double that...
http://www.armytimes.com/projects/money/pay_charts/2008/retirement_pay/
O7 starts at $4961 a month, and most of them probably have closer to 30 years of service than 20. so that's almost $100k a year in retirement. i'm not certain, but i think at least some of that may be tax-free too.
but the law isn't that the can tap without a warrant, it's that they can get a warrant without providing the evidence/rationale behind asking for it (e.g. he seems like a bad guy...) until up to 3 days later. they didn't bother getting the warrant.
that product would be so much more useful if it used ECC memory.