Why do you assume that the desktops are stateful? If they're set up properly, everything except the executables is server side, and the executables are installed as part of an install image.
If some idiot working on the mains shorted something out, and fried an entire building full of dumb terminals, you're in the same spot.
A final question - how on earth do DWP recover 60,000 unbootable PCs?
The same way you replace 60,000 blown FRUs; you go to each one, pop in either a) a boot floppy or b) an install CD, and reboot from your base image. Done.
Michael buys a car, not realizing that, like everybody else, he must renew his license plates every year. He decides to fake a renewal sticker, gets caught by the cops, and gets a fine. Boo hoo.
So do something like GTA3, and many other games (that's the first one I can think of off the top my head, though) do; format the textures for your card during the install, or during the first time you play. Check to see if the video card has changed on run, and if not, well, you've already got the textures sitting on the harddrive ready to go.
Re:Testing, Dog Food and Open Source
on
Joel On Software
·
· Score: 1
Proper QA testing is vastly different from actually sitting down and using the product; the two are no where near being close.
If a developer is 'testing' a product, they're suppying known inputs to subroutines to see what gets returned, putting in known data to see if calculations return appropriate answers, and so on.
If they're actually 'using' the product, they notice that that little popup that happens for result X really should just be a status message on the bottom of the window, or that this data column is superfluous, (have you ever seen something which is subfluous, or even just fluous?) and so on.
Think of it as encrypting something such that if you give out one password, you get your top secret data, and if you give out a different password, you get something else; say, fake top secret data. Then, making that encryption scheme work such that you are utterly unable to prove that there's more data than was unencrypted, multiple sets, multiple passwords, or whatever.
Then, when you're being beaten with a rubber hose, you can 'safely' spill your guts.
Actually, his parents did, and I think they have a case, myself; they claim to have ICQ (or some other IM) logs showing that the guys who found the tape planned to distributed it across the Internet specifically to hurt SWK.
And isn't it interesting that they could very easily also disable modded Xboxes that aren't on Live, yet choose not to do so?
They're not interfering with anything you might want to do, they're just not letting you (potentially) interfere with what other people (on Live) are doing.
For "dropping in an extra server when needed without a lot of reconfiguring", what do you mean by "a lot of reconfiguring"? Obviously you need to get the machine, install the os, set up networking, install the web server, setup the web application, point it at the database, etc. How does the application being "stateless" help? I guess, what are some examples of state that an application can have that will make configuring an additional web server difficult?
When you add a new server, do you need only a) make the server functional, join it to your cluster, and watch it take up slack, or b) rewrite half of your application to reflect the changes?
Concerning the pseudo static data regeneration, what if the thing that was being updated was only accessed once every half-hour on average? I am assuming then that generating the page on demand would be better?
Lets say you get a weather condition update from a sensor on your roof every five minutes. Does it make more sense to a) write a dynamic website which queries the equipment via SNMP every time somebody hits the page, or b) write a scheduled job that, once every five minutes, queries the sensor, then writes out a completely static HTML page which people hit?
If the sensor automatically populates a database every five minutes, should you hit the database each and every request, or hit the database once each five minutes?
I don't really know what you mean by "MAKE YOUR WEB SERVERS STATELESS". I mean, they have to know if a request just came in, where the data is, what time it is etc, and that stuff gives it state. I am assuming you mean something else by stateless but I cannot figure it out.
A connection is either stateful, or stateless. In other words, does a state get maintained between transactions?
With the web, it's stateless. If you want to have the web server remember you, you have to store something somewhere, be it a cookie on the client, a bunch of data in a form or in the URL, or whatever. But, if the state is being dealt with by the server, say, you're putting stuff into a shopping cart, so the webserver is keeping your shopping cart in memory, what happens if your next request gets directed to a different webserver? Ooops.
It might not be that hard to throw in some built in checks; "if LiveSupportingHardware = No Then DisableThisOptionOnTheMenu()" or "if HardDriveIsPresent = Yes Then CacheThisInformationForFasterAccess; else StreamThisInformationOffOfTheDVDAsRequired".
Court: Did you have sexual relations with Ms. Lewinsky?
Bubba: Define 'sexual relations.'
Court: Penile penetration of the vagina.
Bubba: No, I did not have 'sexual relations' with Ms. Lewinkski.
To answer your question directly, no. He didn't 'concede' anything; he called up Bush and said 'hey, according to the exit polls, looks like you won. Good game.'
As far as I can tell, by doing this, if he winds up losing, he looks far more gracious than Bush did last time around, getting SCOTUS to call the election for him. If he winds up winning, again, it makes it difficult for Bush to start trying to lay in the shennanigans.
Sure, but once again, either it works, or it doesn't work; you'll either get errors or no errors. Regardless of what Monster would like you to believe, your digital audio stream isn't going to sound any more full just because you're using their Uber Super Duper coax.
I remember tests being run where a DVD player was connected to a DD Receiver with a bit rate/error counter via, get this, a metal coat hanger. It moved the bits just fine.
I believe that there's also a thing where if you use gold plated connectors, but your jacks arne't gold plated, you get faster corrosion.
Also, if you're going to be passing a digital signal of any form (coax or toslink) cable quality doesn't matter a whit. Period. It either works or it doesn't work.
Also, a crappy s-video cable will always be better than the finest RCA composite cable. A crappy component cable will always be better than the finest s-video.
So on one hand, it's good to track customers for the purposes of offering them extra service in reward for long and profitable custom, but not to prevent users from screwing you over?
If this guy made a habit of buying a book, reading it, then returning it, do you think that Amazon would extend him such courtesies? No.
Somebody else made some comment about the world used to be a smaller place, shopkeepers knew you, and didn't do what the computer told them. Well, the thing there is that the shopkeepers knew who the idiots and the deadbeats are. And they generally told the other shopkeepers, too. End result: the same.
Personally, I moved from MRTG/RRD to Cricket with RRDtool, but aye, nothing says 'I know what I'm talking about' like having historical data to point at.
Well, there you go, then. Will TiVo be working with content providers to allow this flexibility? I imagine they'd have to, as they need some way to distinguish between PPV and non-PPV channels.
Here's the question: Can you sign away your fair use rights?
By signing up with a content provider, THEN signing up for PPV (might be a second step for you, esp. if you sign up for the kind of service where the thing automagically phones home), AND agree to the contract by giving them money, aren't you agreeing to not do what you're now doing?
I agree, TiVo shouldn't really be caring one way or the other; it's the owner who infringes or not infringes. Never the less, I think this is pretty cut and dried.
Slashdot Mirror
Why do you assume that the desktops are stateful? If they're set up properly, everything except the executables is server side, and the executables are installed as part of an install image.
If some idiot working on the mains shorted something out, and fried an entire building full of dumb terminals, you're in the same spot.
The same way you replace 60,000 blown FRUs; you go to each one, pop in either a) a boot floppy or b) an install CD, and reboot from your base image. Done.
NSA isn't allowed to spy on US Citizens.
Would the profit not be that they now have a functional piece of software that they didn't before they performed the illegal act?
Michael buys a car, not realizing that, like everybody else, he must renew his license plates every year. He decides to fake a renewal sticker, gets caught by the cops, and gets a fine. Boo hoo.
So do something like GTA3, and many other games (that's the first one I can think of off the top my head, though) do; format the textures for your card during the install, or during the first time you play. Check to see if the video card has changed on run, and if not, well, you've already got the textures sitting on the harddrive ready to go.
Proper QA testing is vastly different from actually sitting down and using the product; the two are no where near being close.
If a developer is 'testing' a product, they're suppying known inputs to subroutines to see what gets returned, putting in known data to see if calculations return appropriate answers, and so on.
If they're actually 'using' the product, they notice that that little popup that happens for result X really should just be a status message on the bottom of the window, or that this data column is superfluous, (have you ever seen something which is subfluous, or even just fluous?) and so on.
It just means that the actual feedback is processed and recorded by underlings; much like staffers intercept the mail going to a congresscritter.
Think of it as encrypting something such that if you give out one password, you get your top secret data, and if you give out a different password, you get something else; say, fake top secret data. Then, making that encryption scheme work such that you are utterly unable to prove that there's more data than was unencrypted, multiple sets, multiple passwords, or whatever.
Then, when you're being beaten with a rubber hose, you can 'safely' spill your guts.
In other words, if you've got an AWACS up, put some of these puppies up, too, and suddenly you've got theater missile defense.
Actually, his parents did, and I think they have a case, myself; they claim to have ICQ (or some other IM) logs showing that the guys who found the tape planned to distributed it across the Internet specifically to hurt SWK.
And isn't it interesting that they could very easily also disable modded Xboxes that aren't on Live, yet choose not to do so?
They're not interfering with anything you might want to do, they're just not letting you (potentially) interfere with what other people (on Live) are doing.
And I seem to recall that the entire POINT behind HR163 was that the Republicans would never ever send their own kids off to war.
When you add a new server, do you need only a) make the server functional, join it to your cluster, and watch it take up slack, or b) rewrite half of your application to reflect the changes?
Lets say you get a weather condition update from a sensor on your roof every five minutes. Does it make more sense to a) write a dynamic website which queries the equipment via SNMP every time somebody hits the page, or b) write a scheduled job that, once every five minutes, queries the sensor, then writes out a completely static HTML page which people hit?
If the sensor automatically populates a database every five minutes, should you hit the database each and every request, or hit the database once each five minutes?
A connection is either stateful, or stateless. In other words, does a state get maintained between transactions?
With the web, it's stateless. If you want to have the web server remember you, you have to store something somewhere, be it a cookie on the client, a bunch of data in a form or in the URL, or whatever. But, if the state is being dealt with by the server, say, you're putting stuff into a shopping cart, so the webserver is keeping your shopping cart in memory, what happens if your next request gets directed to a different webserver? Ooops.
It might not be that hard to throw in some built in checks; "if LiveSupportingHardware = No Then DisableThisOptionOnTheMenu()" or "if HardDriveIsPresent = Yes Then CacheThisInformationForFasterAccess; else StreamThisInformationOffOfTheDVDAsRequired".
True, true.
Didn't it go something like:
Court: Did you have sexual relations with Ms. Lewinsky?
Bubba: Define 'sexual relations.'
Court: Penile penetration of the vagina.
Bubba: No, I did not have 'sexual relations' with Ms. Lewinkski.
To answer your question directly, no. He didn't 'concede' anything; he called up Bush and said 'hey, according to the exit polls, looks like you won. Good game.'
As far as I can tell, by doing this, if he winds up losing, he looks far more gracious than Bush did last time around, getting SCOTUS to call the election for him. If he winds up winning, again, it makes it difficult for Bush to start trying to lay in the shennanigans.
Yes, and untold thousands of women typed on them, but very few men.
Sure, but once again, either it works, or it doesn't work; you'll either get errors or no errors. Regardless of what Monster would like you to believe, your digital audio stream isn't going to sound any more full just because you're using their Uber Super Duper coax.
I remember tests being run where a DVD player was connected to a DD Receiver with a bit rate/error counter via, get this, a metal coat hanger. It moved the bits just fine.
I believe that there's also a thing where if you use gold plated connectors, but your jacks arne't gold plated, you get faster corrosion.
Also, if you're going to be passing a digital signal of any form (coax or toslink) cable quality doesn't matter a whit. Period. It either works or it doesn't work.
Also, a crappy s-video cable will always be better than the finest RCA composite cable. A crappy component cable will always be better than the finest s-video.
So on one hand, it's good to track customers for the purposes of offering them extra service in reward for long and profitable custom, but not to prevent users from screwing you over?
If this guy made a habit of buying a book, reading it, then returning it, do you think that Amazon would extend him such courtesies? No.
Somebody else made some comment about the world used to be a smaller place, shopkeepers knew you, and didn't do what the computer told them. Well, the thing there is that the shopkeepers knew who the idiots and the deadbeats are. And they generally told the other shopkeepers, too. End result: the same.
Personally, I moved from MRTG/RRD to Cricket with RRDtool, but aye, nothing says 'I know what I'm talking about' like having historical data to point at.
Is the Monolithic Kernal stronger than the Microkernal?
No. Faster, easier, more seductive.
Well, there you go, then. Will TiVo be working with content providers to allow this flexibility? I imagine they'd have to, as they need some way to distinguish between PPV and non-PPV channels.
Here's the question: Can you sign away your fair use rights?
By signing up with a content provider, THEN signing up for PPV (might be a second step for you, esp. if you sign up for the kind of service where the thing automagically phones home), AND agree to the contract by giving them money, aren't you agreeing to not do what you're now doing?
I agree, TiVo shouldn't really be caring one way or the other; it's the owner who infringes or not infringes. Never the less, I think this is pretty cut and dried.