While you're right that the ppl don't own major TV networks, I think you'll find people genuinely interested in detailed debates of the issues are rather scarce on the ground. Just look at the minor elections (like school boards). For the consistantly successful candidates you see exactly the same style-over-substance approach.
Heh. Windows might be a pile of crap, but it is amazing what hoops someone "skilled in the art" can jump it through. When was the last time you saw hotmail go down, for instance? And there are windows ISPs that advertise four nines reliability.
If you have a reliable windows dealer around, with years of experience of making things work right. And you also have a new linux dealer around, fresh out of college and eager for their first contract, who do you go for?
Now, I'm not comparing apples with oranges, but people rarely have the choice of equally experienced linux and windows vendors. And for many people that the experienced windows operators are a better choice than the inexperienced linux operators. Like the article said, they swapped ISP and they got greater reliability -- well, neither linux nor windows are unreliable -- so what's the bet their old linux ISP was a shoddy operation?
I got quite a suprise the other day hearing a linux advocate describing going linux as having more lock-in than windows. You see, where I live there are plenty of windows firms you could hop between if one goes out of business or starts acting unreasonably. But if you go with linux then there is nobody else you can go to if your operator starts gouging you. Ergo, vendor lock-in! Of course, this is a short term position and in theory Linux has less vendor lock-in. But the real world is made up of short term positions, and customers must choose a vendor for now.
It doesn't matter where the problems came from, you're now the sysadmin and that makes them your problems. Finding the solution to them is also your problem.
If you continue to run around in 'emergency mode', dealing with problems as they come up then sooner or later you will get one too many emergencies at once and the whole stack of cards will come crashing down. Possibly you will have quit before the stack comes crashing down. Possibly you'll get a better job shortly after being fired for the crash.
So, you have a choice. Run around until then fixing problems as they come up and hope that you can get a new job either before or after the stack falls down. Or you could spend some time now and get on top of the situation.
How long would it take to learn? perhaps an hour each day of (additional?) overtime for a month if you're a bit slow or way over your head? Perhaps just for a week if you're a fast learner.
At least at this workplace, I find before the day starts a much better time to put in overtime than after five. Between 6 and 8 AM, very few people will interrupt you; whereas between 5 and 7 PM seems to be fair game.
What you do is your call. I personally like to have day to day business very cruisy, spending my time preventing problems _before_ they arise. This means that when things do screw up I have the time to deal with it. But, I know other people who view getting and staying on top of things as poor use of time -- they claim they can deal with problems that arise in less time than it takes me to organise everything so as to prevent the problem arising. Certainly, I spend a lot of time setting up safety nets that never catch anything.
But I don't think there is much point complaining to slashdot. You know the choices and you know the implications of the choices. All I've done in this post is repeat stuff you already know.
If you ever do any of your idea on company time, e.g. chat about it to colleagues in lunch time, then the company owns a share of it. My limited experience is they own a fairly big share of it.
As an employee of a sub-contractor, I _believe_ the contract you're currently working on would be irrelevant, i.e. did you develop it while working on the sub-contractor's time.
If you are actually working as a contractor then you're not an employee and so contracts can be more severe -- unless your work is treated as an employee (in which case you are treated as an employee by the court).
No. US contract law is very different to contract law in other countries. Outside the land of the free we have these things called 'inalienable rights', and no contract may interfere with them. For instance, no contract can say 'you may not have children while employed here', or 'you may not work for a compeditor after you leave', or 'we own what you produce in your free time'.
Any contract stupid enough to interfere with his free time would be thrown out of court within minutes, and IBM forced to pay all of his costs, as well as damages. I believe you don't have laws allowing the judge to do that in America either.
When I was a poor student at the end of '93, I remember a 486 DX2/66 was barely affordable (roughly $1k, with a 14" monitor). And in '95 I remeber watching an intel promotional video from the release of the pentium -- talking about such things as running two processors in paralell so that if either produces an incorrect output the other catches it. And the DEC Alpha was a blazing 300MHz! But I digress...
I'm not disputing your timeline, but for the average guy walking into a computer shop you would still be being sold a 486 in '94, and maybe in 95 too. The machine I bought in '97 was 166MHz which was considered standard to a little slow for a new machine (max was twice that IIRC).
Hi, I tried the google link you provided but didn't get very far. For a game client server architecture is it possible to check if the client has been modified? (I'm thinking about go, but it would apply to chess too).
Specifically, the current go servers are quite anti open source for fear that somebody will modify their client to add (e.g. offline editing) when they're playing as well as when they're just watching games.
But it would be really cool if people could modify their clients but just not connect to the official server with a modified client. Even better if the unmodified client could be built on different archs and still connect to the official server.
I considered that -- and cursed not doing it when a power cut meant it didn't just keep ticking along -- but with the screen already thretening to fall off and the CD only working if I apply pressure on the base of the case, I was worried I'd break it.
Right, this is what I did. I got a not so old laptop where the screen connection was starting to break so while it works well enough, it has almost zero resale value.
I have found the built in 3GB disk a bit small (running webserver and mail server). I could fit everything but I found I needed to do too much maintainence deleting old files. So I added a USB drive -- the laptop only supports 1.1 so it is painfully slow (500kB/s), but that's fine for what is essentially backups.
Were I richer, I'd go for an ITX solution instead.
For political reasons, the LSB standardised on RPM. Most (all?) of the community have moved onto something better than RPM and so do not comply in that regard.
Er, a cup is 250mL, of course. Using standard sizes is very common in professional recipies where you want results to be exactly reproducable.
Odds are the person you got your banana recipe from is so used to professional recipes that it was more natural to say say 1C than to 1.5 medium bananas. Another possiblility is the recipe was particularly intolerant to variations; I've read recipes where the amount of emulsifier (egg yolk) is calculated to be just enough to bind and so adding even a tiny bit more of something will cause the whole thing to fail -- think about mayonnaise (though I can't imagine banana in such finely balanced recipe).
As an example, say a recipe calls for six eggs. If you live somewhere with big eggs, you'll get a totally different cake to somebody with small eggs. But if you specify 420g of eggs then you'll get the same cake. Likewise, saying a cup of yolks would likewise enable much higher accuracy than saying *shrug* 18? egg yolks.
Not only is it significantly more consistant, with decent electronic scales you'll likely find that it is faster to specify every single ingredient by mass than some by mass, some by volume and some by enumeration.
Of course, if the recipe is tolerant of variations (i.e. just about anything except baking) then this is all a waste of time because anything will work and it is up to the cook to decide the proportions.
When I'm at work I cope fine with deadlines, but I love to cook and I just can't stand doing it to a deadline. So, if I get home at 6 and was just cooking for myself, chances are I wouldn't get dinner to the table until 9.
Unfortunately that doesn't go down so well with my wife, who tends to get hungry very shortly after work (like an hour or two). So, I do my share of cooking but unless it is the weekend, I don't enjoy it much.
Sounds an interesting idea -- I've got perhaps a dozen recipies in his format. Working with someone else would help a lot, and I would guess some photo software could be 'borrowed' to handle the ratings.
I found the layout of the recipe very nice, but it just doesn't scale if the steps are particularly complex -- look at how creme brulee was described if you don't believe me. However, something very similar that does scale is the latex style cooking by Axel Reichert (CTAN link: http://www.ctan.org/tex-archive/macros/latex/contr ib/cooking/)
The essential difference is that instead of nesting columns, Axel's style uses only two columns which enables the second column to be very large if necessary. Though I've got to admit that for simple recipies, the cooking for engineer's site looks very good.
PS: Cooking is a great way to unwind after spending all day coding, especially if you don't mind the meal taking a few hours (and glasses of wine) to prepare...
A cheaper option, is to use vnc. Sure, you can't have two people running the same application at the same time (unless you set up two windows boxes) but the money you save on software enables you to buy an extra box.
Firstly, linux programs tend to be smaller than windows programs (do one thing, and do it well). Even a huge beast like tetex is 'only' 14.4MB -- compare to SP2... This has reduced the demand for delta compression.
Secondly, in the windows world people release rarely. However, the opposite is true in the linux world -- projects with daily releases are not unheard of, and weekly releases are fairly common. This means enumerating patches (v 3.4 -> v. 3.7) is infeasible in Linux where it is feasible in Windows.
More sophisticated algorithms than delta checksums do exist (as I guess you know if your thesis is on them) -- rolling checksums have been used in several projects I know of. However, there is a widespread rumour that these techniques are patented. I have never seen any evidence, but it puts a damper on any implementations.
There is a semi-vapourware project implementing all of this (part of the apache project IIRC). However the project fizzled away several years ago.
The CPU has a blisteringly fast L1 cache. If the data being processed can fit into the L1 cache then the CPU is faster than the GPU (at least at matrix multiplications, and probably for most other tasks).
Yes, google's stats are not totally accurate. For instance, my web site says that someone keeps browsing it using Opera on NetBSD, and as everyone knows, BSD is dying... so no
However, they're not going to be far off and most importantly they are unbiased. That makes them the best stats we have (had).
Slashdot Mirror
That is of course the libertarian argument.
Some of us think it is incorrect. Rather than repeat the arguments here, any google for libertarian should come up with a decent argument.
I disagree.
While you're right that the ppl don't own major TV networks, I think you'll find people genuinely interested in detailed debates of the issues are rather scarce on the ground. Just look at the minor elections (like school boards). For the consistantly successful candidates you see exactly the same style-over-substance approach.
Well, it is a purely software thing. Because I have a powerbook that works fine with its lid closed... running linux.
No doubt there is an app for OS X that does much the same thing. Someone else posted a link to http://www.alxsoft.com/mac/sleepless.html
Heh. Windows might be a pile of crap, but it is amazing what hoops someone "skilled in the art" can jump it through. When was the last time you saw hotmail go down, for instance? And there are windows ISPs that advertise four nines reliability.
If you have a reliable windows dealer around, with years of experience of making things work right. And you also have a new linux dealer around, fresh out of college and eager for their first contract, who do you go for?
Now, I'm not comparing apples with oranges, but people rarely have the choice of equally experienced linux and windows vendors. And for many people that the experienced windows operators are a better choice than the inexperienced linux operators. Like the article said, they swapped ISP and they got greater reliability -- well, neither linux nor windows are unreliable -- so what's the bet their old linux ISP was a shoddy operation?
I got quite a suprise the other day hearing a linux advocate describing going linux as having more lock-in than windows. You see, where I live there are plenty of windows firms you could hop between if one goes out of business or starts acting unreasonably. But if you go with linux then there is nobody else you can go to if your operator starts gouging you. Ergo, vendor lock-in! Of course, this is a short term position and in theory Linux has less vendor lock-in. But the real world is made up of short term positions, and customers must choose a vendor for now.
It doesn't matter where the problems came from, you're now the sysadmin and that makes them your problems. Finding the solution to them is also your problem.
If you continue to run around in 'emergency mode', dealing with problems as they come up then sooner or later you will get one too many emergencies at once and the whole stack of cards will come crashing down. Possibly you will have quit before the stack comes crashing down. Possibly you'll get a better job shortly after being fired for the crash.
So, you have a choice. Run around until then fixing problems as they come up and hope that you can get a new job either before or after the stack falls down. Or you could spend some time now and get on top of the situation.
How long would it take to learn? perhaps an hour each day of (additional?) overtime for a month if you're a bit slow or way over your head? Perhaps just for a week if you're a fast learner.
At least at this workplace, I find before the day starts a much better time to put in overtime than after five. Between 6 and 8 AM, very few people will interrupt you; whereas between 5 and 7 PM seems to be fair game.
What you do is your call. I personally like to have day to day business very cruisy, spending my time preventing problems _before_ they arise. This means that when things do screw up I have the time to deal with it. But, I know other people who view getting and staying on top of things as poor use of time -- they claim they can deal with problems that arise in less time than it takes me to organise everything so as to prevent the problem arising. Certainly, I spend a lot of time setting up safety nets that never catch anything.
But I don't think there is much point complaining to slashdot. You know the choices and you know the implications of the choices. All I've done in this post is repeat stuff you already know.
If you ever do any of your idea on company time, e.g. chat about it to colleagues in lunch time, then the company owns a share of it. My limited experience is they own a fairly big share of it.
As an employee of a sub-contractor, I _believe_ the contract you're currently working on would be irrelevant, i.e. did you develop it while working on the sub-contractor's time.
If you are actually working as a contractor then you're not an employee and so contracts can be more severe -- unless your work is treated as an employee (in which case you are treated as an employee by the court).
No. US contract law is very different to contract law in other countries. Outside the land of the free we have these things called 'inalienable rights', and no contract may interfere with them. For instance, no contract can say 'you may not have children while employed here', or 'you may not work for a compeditor after you leave', or 'we own what you produce in your free time'.
Any contract stupid enough to interfere with his free time would be thrown out of court within minutes, and IBM forced to pay all of his costs, as well as damages. I believe you don't have laws allowing the judge to do that in America either.
Did you check which country he was from? Not all countries have stupid IP laws you know...
When I was a poor student at the end of '93, I remember a 486 DX2/66 was barely affordable (roughly $1k, with a 14" monitor). And in '95 I remeber watching an intel promotional video from the release of the pentium -- talking about such things as running two processors in paralell so that if either produces an incorrect output the other catches it. And the DEC Alpha was a blazing 300MHz! But I digress...
I'm not disputing your timeline, but for the average guy walking into a computer shop you would still be being sold a 486 in '94, and maybe in 95 too. The machine I bought in '97 was 166MHz which was considered standard to a little slow for a new machine (max was twice that IIRC).
Hi, I tried the google link you provided but didn't get very far. For a game client server architecture is it possible to check if the client has been modified? (I'm thinking about go, but it would apply to chess too).
Specifically, the current go servers are quite anti open source for fear that somebody will modify their client to add (e.g. offline editing) when they're playing as well as when they're just watching games.
But it would be really cool if people could modify their clients but just not connect to the official server with a modified client. Even better if the unmodified client could be built on different archs and still connect to the official server.
I guess quake must have similar problems?
I considered that -- and cursed not doing it when a power cut meant it didn't just keep ticking along -- but with the screen already thretening to fall off and the CD only working if I apply pressure on the base of the case, I was worried I'd break it.
Right, this is what I did. I got a not so old laptop where the screen connection was starting to break so while it works well enough, it has almost zero resale value.
I have found the built in 3GB disk a bit small (running webserver and mail server). I could fit everything but I found I needed to do too much maintainence deleting old files. So I added a USB drive -- the laptop only supports 1.1 so it is painfully slow (500kB/s), but that's fine for what is essentially backups.
Were I richer, I'd go for an ITX solution instead.
Coming from a UK background, I would have used a semicolon myself. As in:
..."
"Its official; I'm the
But if I were to use something other than a semicolon, I would choose a comma rather than a full stop.
Much as we bash microsoft here... Se for example:f .pdf
http://research.microsoft.com/~joshuago/smartproo
For political reasons, the LSB standardised on RPM. Most (all?) of the community have moved onto something better than RPM and so do not comply in that regard.
Er, a cup is 250mL, of course. Using standard sizes is very common in professional recipies where you want results to be exactly reproducable.
Odds are the person you got your banana recipe from is so used to professional recipes that it was more natural to say say 1C than to 1.5 medium bananas. Another possiblility is the recipe was particularly intolerant to variations; I've read recipes where the amount of emulsifier (egg yolk) is calculated to be just enough to bind and so adding even a tiny bit more of something will cause the whole thing to fail -- think about mayonnaise (though I can't imagine banana in such finely balanced recipe).
As an example, say a recipe calls for six eggs. If you live somewhere with big eggs, you'll get a totally different cake to somebody with small eggs. But if you specify 420g of eggs then you'll get the same cake. Likewise, saying a cup of yolks would likewise enable much higher accuracy than saying *shrug* 18? egg yolks.
Not only is it significantly more consistant, with decent electronic scales you'll likely find that it is faster to specify every single ingredient by mass than some by mass, some by volume and some by enumeration.
Of course, if the recipe is tolerant of variations (i.e. just about anything except baking) then this is all a waste of time because anything will work and it is up to the cook to decide the proportions.
My only problem with this is time...
When I'm at work I cope fine with deadlines, but I love to cook and I just can't stand doing it to a deadline. So, if I get home at 6 and was just cooking for myself, chances are I wouldn't get dinner to the table until 9.
Unfortunately that doesn't go down so well with my wife, who tends to get hungry very shortly after work (like an hour or two). So, I do my share of cooking but unless it is the weekend, I don't enjoy it much.
*shrug*
Sounds an interesting idea -- I've got perhaps a dozen recipies in his format. Working with someone else would help a lot, and I would guess some photo software could be 'borrowed' to handle the ratings.
Corrin
I found the layout of the recipe very nice, but it just doesn't scale if the steps are particularly complex -- look at how creme brulee was described if you don't believe me. However, something very similar that does scale is the latex style cooking by Axel Reichert (CTAN link: http://www.ctan.org/tex-archive/macros/latex/contr ib/cooking/)
The essential difference is that instead of nesting columns, Axel's style uses only two columns which enables the second column to be very large if necessary. Though I've got to admit that for simple recipies, the cooking for engineer's site looks very good.
PS: Cooking is a great way to unwind after spending all day coding, especially if you don't mind the meal taking a few hours (and glasses of wine) to prepare...
A cheaper option, is to use vnc. Sure, you can't have two people running the same application at the same time (unless you set up two windows boxes) but the money you save on software enables you to buy an extra box.
Ideal up to about 10 clients, anyway.
Firstly, linux programs tend to be smaller than windows programs (do one thing, and do it well). Even a huge beast like tetex is 'only' 14.4MB -- compare to SP2... This has reduced the demand for delta compression.
Secondly, in the windows world people release rarely. However, the opposite is true in the linux world -- projects with daily releases are not unheard of, and weekly releases are fairly common. This means enumerating patches (v 3.4 -> v. 3.7) is infeasible in Linux where it is feasible in Windows.
More sophisticated algorithms than delta checksums do exist (as I guess you know if your thesis is on them) -- rolling checksums have been used in several projects I know of. However, there is a widespread rumour that these techniques are patented. I have never seen any evidence, but it puts a damper on any implementations.
There is a semi-vapourware project implementing all of this (part of the apache project IIRC). However the project fizzled away several years ago.
The CPU has a blisteringly fast L1 cache. If the data being processed can fit into the L1 cache then the CPU is faster than the GPU (at least at matrix multiplications, and probably for most other tasks).
Between evil companies and evil patents, it's pretty easy to see which is worse.
Yes, google's stats are not totally accurate. For instance, my web site says that someone keeps browsing it using Opera on NetBSD, and as everyone knows, BSD is dying... so no
However, they're not going to be far off and most importantly they are unbiased. That makes them the best stats we have (had).