We have a so-called "non-directional" day where I work. It's pretty nice; every Wednesday you're supposed to work on some sort of side project. There's a few around you can join, or start your own. The caveat, of course, is that the company owns any outcome from this, but that's fair since it's their time. Of course, if you contribute to a GPL licensed product, then the company is the proud owner of the copyright to a GPL'ed patch, so you can do that if you want. It's also possible to get approval to start a new GPLed project, and people do have independant (non-GPL) projects that they work on in their "real" spare time that the company doesn't have any sort of claim to.
You do have to get the project approved, but that's only to prevent you from starting a "let's blow up the company" type of project. The only one that was ever been turned down was one that would directly compete with us.
The main problem has actually been getting people motivated to start a project, and then keeping them working on it (especially in the face of real deadlines). A few have turned out to be surprisingly interesting, but we haven't had any notable successes like Google has, at least not just yet. There have been a few sizable improvements to internal projects that came from this though. A key factor was moving the day from Friday to Wednesday; when it was on Friday there was just about no motivation to get started on these things.
If you can convince your management to approve this, it's nice and rewarding.
How about it doesn't happen because it breaks completely when you have virtual hosting? When it comes to email, reverse DNS is a useful tool for discovering who's hosting someone's domain, and not much else.
Similarly, my first two university programming courses were in Pascal and then Modula-3 (a very feature-full OO language). We all hated Modula-3 with a passion, but in hindsight it taught me a lot of stuff that I wouldn't have gotten at all out of C++. It's a shame the idiotic licensing kept it from seeing any real use, and that the syntax was so ugly at times.
So clearly, all they need to do is attempt to verify it several times. If each attempt fails, then obviously it backs up the results of the formula. If one of the attempts succeeds, then obviously the formula must have been wrong. So simple! Errr, wait...
I don't see why. An awful lot of geeks (including dear old Slashdot) were on the MySql bandwagon before they had any kind of proper open-source license, and I can't imagine why going back to the way things were would make that many people leave.
There's just something about MySql that makes people want to use it. It certainly isn't features. It used to be (reading) speed, but that's somewhat less relevant these days since Postgres sucks less for that now. Honestly, I think it's just that people like the simpler solution that they understand; either that, or MySql is easier to set up and get going, and people don't like to change once they've learned it.
This is not unique to monorails. For instance, the Montreal subway system uses rubber tires on concrete paths, but it's otherwise identical to a standard subway system (they still have the standard track and wheels on the cars as a backup in case of flat tires). The ride is certainly quieter than metal-on-metal, especially around the corners, but it can be fairly bouncy.
That's the beauty of hotmail - a full account never stops incoming mail! It just bumps out old stuff! You know, the stuff that you actually wanted to keep. Obviously that's much less useful than some brand-new spam.
In Mozilla and Firebird at least there is the very useful, but little-known shortcut Ctrl-L. This highlights the url bar but does *not* copy it to the clipboard. So when I'm in that situation I do Ctrl-L, delete, middle click.
Of course, the middle-click on the page body works too, as long as you don't have to edit the URL. Ctrl-L is still super-handy if you want to type in an URL by hand or something.
The only thing that longhorn claims it will have that ATSUI doesn't have yet is the graphics card rendering support. Ever wonder why resizing a window is so slow on OS X? ATSUI is the reason.
FYI, I have an A7N8X-X (not the Deluxe) at work and I found it would lock up hard every couple of days until I turned off SMP support in my kernel. This is enabled by default for all kernel sources, and I imagine it gets turned on by some vendor kernels as well. You should try that to see if it helps; I haven't had any problems with my machine since I tried, but it's only been a week or so. However, several co-workers reported exactly the same thing, and they've all been fine since they disabled SMP.
It's still a pretty pointless board to run in Linux - why bother getting something with all sorts of fancy onboard crap if there's no good drivers to use it all? They definitely should have gone with a KT600 chipset board.
QNX is designed like a modern os should be. It's straigt out of an Operating Systems 101 textbook.
You're more right than you probably think you are -- QNX is, in fact, the result of 20 years of hacking on a school OS project:) The two founders took the real-time OS course at Waterloo, then took their RTOS at the end of the course, polished it up an awful lot, and then founded QNX.
At least, that's the story they told me when I took the course. The project you write for the class still looks an awful lot like super-primitive QNX.
So, in other words, you're saying it'd be nice if the book covered a few security basics, so that developers would be less likely to make those common mistakes. Hmm, sounds familiar.
I've been looking into PHP, and I love working with it, but I'm still concerned about writing secure scripts. To do a decent job of securing a script, you at least have to worry about several configuration directives, which may or may not be under your control. Basically, I want to see a "PHP Security Best Practices" document, and it's a shame that this book doesn't look like it includes anything on the subject.
Re:making it *slower* than (upgraded) 802.11b
on
802.11g Slows Down
·
· Score: 1
Just don't be fooled by any claims of "256 bit WEP" without closer inspection. I don't know about this particular version, but the problems with the original WEP went much deeper than just the crappy keysize. "256 bits" (some of that is actually for an initialization vector) may be a bit better, maybe even against some common exploits, but you should still get the security heebie-jeebies if you're thinking about putting sensitive data across a wireless link, and unless they say otherwise it's probably got all WEP's other problems with authentication, replay attacks, etc.
Then, in turn, commericial software could not be sold without a licensed programmer's 'signature'. (This could work much like the current engineer scenario of a single engineer watching over draftsmen - the real programmers (people that hvae been programming for years, with many languages, etc - programming managers, basically, instead of the clueless IT Managers we have now) look over, debug, and LART the 'coders'. Granted, there'd probably be a higher ratio of programmers/coders than there is of engineers/draftsman, simply because it takes a lot more man hours to review code than it does to look over a blueprint.
I don't know if this would be quite as effective as you think. As it stands, the software development tools and environments we have are not robust enough to be able to be as sure about the reliability of a program (as opposed to knowing, say, a bridge or a building is reliable and won't fall down). On a large software project, basically any out of millions of lines of code can interact badly with any other line of code, causing a (potentially critical) bug.
As an example, one company I worked for had a policy of doing effectively what you promote; before anyone could check in any code, the changes had to be reviewed by two out of about 7 or 8 of the senior coders. This review occasionally caught bugs before they went in, and it was certainly a big help in keeping things stylistically in line, but the software was so complicated that it was impossible to be sure that any given change didn't introduce a whole pile of bugs.
I mean, sure it's a nice idea, and it was certainly implemented well (and to good effect) at that company, but overall it doesn't make as big a difference to reliability as you might think. To get the changes you want, you need to take the idea to its extreme, a la NASA, but that ends up being prohibitively expensive for everyone except, well, NASA. For further evidence, see Mozilla; they do basically the same thing as my old company, and they still let in plenty of new bugs.
In my experience, the primary indicator for the quality of the code is the talent and experience of the actual code writers; the designers have a much lesser effect. Until this reverses, the engineer/drafter approach won't solve much of anything.
As much as I hate Word, its scripting once saved me a whole pile of work. I was trying to un-mangle some imported WordStar files, where Word had ignored all end-underline markers and treated the next start of underlining as the end of the previous one. A simple VBA macro to invert the underline status of some selected text made life a lot easier.
So there you go - Word's scripting made up for its horrible import filters!
Well, they say it'll shrink to about the size of a PS2. And the PS2 has a normal-size DVD drive and an empty 5.25" expansion slot. So I don't imagine they'll have too much trouble.
If they had said they were making it the size of the GameCube, I'd start to worry.
"[...] the 48-bit RGBA provides the highest level of precision available on any desktop system today"
Oh? Quick, everyone with Radeon 9700 PRO graphics boards in your PCs, make sure you have them in tower cases, or something!
For reference, the ATI specs page states:
Pixel shaders up to 160 instructions with 128-bit floating point precision
You're comparing apples and oranges here. SGIs can handle 48bpp (12 bit per channel RGBA) integer colours all the way through the pipeline. The ATI card you mention, as well as nVidia's new chip, can use 128 bit floating point numbers for some internal calculations, but that's about it. The only point of this is that you have less round-off error if you run a complicated shader program, plus you get a bigger number to put on the box:) Like you say, ATI can also do 10:10:10:2 RGBA integer mode, but that's pretty useless for most people looking for this sort of thing.
Bottom line, if you need high precision integer colours, you still need an SGI. Of course, there's not many people who do, and someone will probably be doing it on the PC in a couple of years, so it's looking pretty grim for SGI as that's one of their few remaining technical advantages in the graphics workstation market.
There's exceptions to this of course. One time, we had switched away our long distance service to another company, but later our calling pattern changed and we wanted to switch back. However, before we could call the old phone company again, they called us one afternoon asking if we wanted to switch back. So I played hard-to-get and asked if we got any sort of deal for it; the response was that I could get a $100 credit against our account.
Needless to say, I accepted, and the moral of the story is that their telemarketing cost them $100 they didn't need to pay us:) Just another case where new customers are more valued than old and disloyalty is rewarded....
Slashdot Mirror
We have a so-called "non-directional" day where I work. It's pretty nice; every Wednesday you're supposed to work on some sort of side project. There's a few around you can join, or start your own. The caveat, of course, is that the company owns any outcome from this, but that's fair since it's their time. Of course, if you contribute to a GPL licensed product, then the company is the proud owner of the copyright to a GPL'ed patch, so you can do that if you want. It's also possible to get approval to start a new GPLed project, and people do have independant (non-GPL) projects that they work on in their "real" spare time that the company doesn't have any sort of claim to.
You do have to get the project approved, but that's only to prevent you from starting a "let's blow up the company" type of project. The only one that was ever been turned down was one that would directly compete with us.
The main problem has actually been getting people motivated to start a project, and then keeping them working on it (especially in the face of real deadlines). A few have turned out to be surprisingly interesting, but we haven't had any notable successes like Google has, at least not just yet. There have been a few sizable improvements to internal projects that came from this though. A key factor was moving the day from Friday to Wednesday; when it was on Friday there was just about no motivation to get started on these things.
If you can convince your management to approve this, it's nice and rewarding.
How about it doesn't happen because it breaks completely when you have virtual hosting? When it comes to email, reverse DNS is a useful tool for discovering who's hosting someone's domain, and not much else.
Similarly, my first two university programming courses were in Pascal and then Modula-3 (a very feature-full OO language). We all hated Modula-3 with a passion, but in hindsight it taught me a lot of stuff that I wouldn't have gotten at all out of C++. It's a shame the idiotic licensing kept it from seeing any real use, and that the syntax was so ugly at times.
So clearly, all they need to do is attempt to verify it several times. If each attempt fails, then obviously it backs up the results of the formula. If one of the attempts succeeds, then obviously the formula must have been wrong. So simple! Errr, wait...
I don't see why. An awful lot of geeks (including dear old Slashdot) were on the MySql bandwagon before they had any kind of proper open-source license, and I can't imagine why going back to the way things were would make that many people leave.
There's just something about MySql that makes people want to use it. It certainly isn't features. It used to be (reading) speed, but that's somewhat less relevant these days since Postgres sucks less for that now. Honestly, I think it's just that people like the simpler solution that they understand; either that, or MySql is easier to set up and get going, and people don't like to change once they've learned it.
Using wire to boost your wireless. Oh, the irony.
She sells cshs by the cshore.
This is not unique to monorails. For instance, the Montreal subway system uses rubber tires on concrete paths, but it's otherwise identical to a standard subway system (they still have the standard track and wheels on the cars as a backup in case of flat tires). The ride is certainly quieter than metal-on-metal, especially around the corners, but it can be fairly bouncy.
That's the beauty of hotmail - a full account never stops incoming mail! It just bumps out old stuff! You know, the stuff that you actually wanted to keep. Obviously that's much less useful than some brand-new spam.
Hey, cool, it works in Linux too. Ctrl-L works in Windows as well, though it's less relevant because it wouldn't copy it to the clipboard anyway.
What mozilla are you using? And on what platform? For me, with Mozilla 1.6 on Linux, Ctrl-U is View Source.
In Mozilla and Firebird at least there is the very useful, but little-known shortcut Ctrl-L. This highlights the url bar but does *not* copy it to the clipboard. So when I'm in that situation I do Ctrl-L, delete, middle click.
Of course, the middle-click on the page body works too, as long as you don't have to edit the URL. Ctrl-L is still super-handy if you want to type in an URL by hand or something.
Gesundheit.
FYI, I have an A7N8X-X (not the Deluxe) at work and I found it would lock up hard every couple of days until I turned off SMP support in my kernel. This is enabled by default for all kernel sources, and I imagine it gets turned on by some vendor kernels as well. You should try that to see if it helps; I haven't had any problems with my machine since I tried, but it's only been a week or so. However, several co-workers reported exactly the same thing, and they've all been fine since they disabled SMP.
It's still a pretty pointless board to run in Linux - why bother getting something with all sorts of fancy onboard crap if there's no good drivers to use it all? They definitely should have gone with a KT600 chipset board.
Really, I'd hope that any system would pipe all 2D graphics through the graphics card - where else would they go? :)
My pet peeve when I'm doing remote troubleshooting is 'ifconfig eth0 down'...oops. At least NASA is smarter than that.
Peter.
You're more right than you probably think you are -- QNX is, in fact, the result of 20 years of hacking on a school OS project
At least, that's the story they told me when I took the course. The project you write for the class still looks an awful lot like super-primitive QNX.
So, in other words, you're saying it'd be nice if the book covered a few security basics, so that developers would be less likely to make those common mistakes. Hmm, sounds familiar.
I've been looking into PHP, and I love working with it, but I'm still concerned about writing secure scripts. To do a decent job of securing a script, you at least have to worry about several configuration directives, which may or may not be under your control. Basically, I want to see a "PHP Security Best Practices" document, and it's a shame that this book doesn't look like it includes anything on the subject.
Just don't be fooled by any claims of "256 bit WEP" without closer inspection. I don't know about this particular version, but the problems with the original WEP went much deeper than just the crappy keysize. "256 bits" (some of that is actually for an initialization vector) may be a bit better, maybe even against some common exploits, but you should still get the security heebie-jeebies if you're thinking about putting sensitive data across a wireless link, and unless they say otherwise it's probably got all WEP's other problems with authentication, replay attacks, etc.
I don't know if this would be quite as effective as you think. As it stands, the software development tools and environments we have are not robust enough to be able to be as sure about the reliability of a program (as opposed to knowing, say, a bridge or a building is reliable and won't fall down). On a large software project, basically any out of millions of lines of code can interact badly with any other line of code, causing a (potentially critical) bug.
As an example, one company I worked for had a policy of doing effectively what you promote; before anyone could check in any code, the changes had to be reviewed by two out of about 7 or 8 of the senior coders. This review occasionally caught bugs before they went in, and it was certainly a big help in keeping things stylistically in line, but the software was so complicated that it was impossible to be sure that any given change didn't introduce a whole pile of bugs.
I mean, sure it's a nice idea, and it was certainly implemented well (and to good effect) at that company, but overall it doesn't make as big a difference to reliability as you might think. To get the changes you want, you need to take the idea to its extreme, a la NASA, but that ends up being prohibitively expensive for everyone except, well, NASA. For further evidence, see Mozilla; they do basically the same thing as my old company, and they still let in plenty of new bugs.
In my experience, the primary indicator for the quality of the code is the talent and experience of the actual code writers; the designers have a much lesser effect. Until this reverses, the engineer/drafter approach won't solve much of anything.
As much as I hate Word, its scripting once saved me a whole pile of work. I was trying to un-mangle some imported WordStar files, where Word had ignored all end-underline markers and treated the next start of underlining as the end of the previous one. A simple VBA macro to invert the underline status of some selected text made life a lot easier.
So there you go - Word's scripting made up for its horrible import filters!
Well, they say it'll shrink to about the size of a PS2. And the PS2 has a normal-size DVD drive and an empty 5.25" expansion slot. So I don't imagine they'll have too much trouble.
If they had said they were making it the size of the GameCube, I'd start to worry.
Peter.
Bottom line, if you need high precision integer colours, you still need an SGI. Of course, there's not many people who do, and someone will probably be doing it on the PC in a couple of years, so it's looking pretty grim for SGI as that's one of their few remaining technical advantages in the graphics workstation market.
There's exceptions to this of course. One time, we had switched away our long distance service to another company, but later our calling pattern changed and we wanted to switch back. However, before we could call the old phone company again, they called us one afternoon asking if we wanted to switch back. So I played hard-to-get and asked if we got any sort of deal for it; the response was that I could get a $100 credit against our account.
:) Just another case where new customers are more valued than old and disloyalty is rewarded....
Needless to say, I accepted, and the moral of the story is that their telemarketing cost them $100 they didn't need to pay us
Peter.
Er, try again. Smoke is also IRIX-based; it's the Octane2 version of Fire, like Flame is the Octane2 version of Inferno.
:)
I don't think you can call anything that makes you buy an Octane2 low-end