...is less an example of a failed process than it is a testament to the difficulties of debugging feature-rich software on a timetable that meets marketing demands...
Um, except the part where a developer and a tester go round and round about differences between "debug mode" and "release mode" problems. I mean come on, what kind of amatuer windows developer doesn't recognize the gigantic differences between debug and release mode apps on win32??!! If these guys would just learn to use printf() and/or the OutputDebugString() function they might be able to "see" a bit better.
Oh yeah, and the part where they know "max file size" is an issue, but don't stop to figure out what the OS max file limit is before going on... sheesh. It'd be one thing if this bug's lifetime were a few weeks, but years??!!
Q4: When do I need to execute a license with Microsoft?
A4: At this time Microsoft is only aware of pending patent application claims that cover its submission of the Sender ID specification. Because Microsoft is not aware of any issued patent claims, Microsoft does not require any one to sign a license with Microsoft to implement the Sender ID specification or any part of it that is incorporated into IETF working drafts. In conformance with the IETF IPR policy Microsoft has disclosed the existence of those pending patent claims and has provided its assurance that if such claims are granted Microsoft will make licenses available on reasonable and non-discriminatory terms. Microsoft has also gone beyond the IETF's requirements by clarifying that its licenses will require no fees or other royalties, and further, to make a license available to early adopters who wish at their option to clarify their rights with Microsoft with respect to early implementations. Typically patent holders do not make their license terms available until after the standard has been adopted and until after their patent claims have been granted, leaving early implementers to speculate as to the ultimate terms of the license.
I already read it. This appears to be saying only those intereseted in "clarifying their rights" need to sign it currently. (ie: No claims have even been granted yet.) As to the future, I guess anyone doing redistribution would have to sign it... (Though it wasn't entirely clear to me whether that was only for redistribution with modification or all redistribution)
When I did compsci at Uni we had to logically prove each of our programs. This was not easy, but it made it impossible for a program to crash or have a bug
So uh, did your profressors pass these programs off to the smartest hackers they could find, source included, to see if this was really true after the fact?
From the jargon file:
perfect programmer syndrome n.
Arrogance; the egotistical conviction that one is above normal human error. Most frequently found among programmers of some native ability but relatively little experience (especially new graduates; their perceptions may be distorted by a history of excellent performance at solving toy problems). "Of course my program is correct, there is no need to test it." "Yes, I can see there may be a problem here, but I'll never type rm -r / while in root mode."
The only way you can "prove" your program is correct is by mangling the specification one way or another. (perhaps into something long, precise and deterministic) This leaves room for bugs to creep in at the specification level instead of at the coding level. (It also makes specifications a lot bigger and harder to deal with, and is a hallmark of the type of top-down design preferred by certain large computing companies in the 70's and 80's)
In practice this seems to culminate in less then effective communication between developers and their customers... Specifications should be kept simple so customers and developers can both understand them and easily talk about them.
That said, companies like Microsoft can certainly focus on smaller more solid programs instead of large feature sets, but that's hardly going to make software "perfect".
(and by accept, I mean, printed it out, signed it, and faxed it back to Microsoft).
Isn't that only necessary for those wishing to "clear the air" as to their IP rights? (Read folks with possible patent claims related to this particular standard)
then the k3b people use those to make a nice interface for it.
Yes, except the k3b guys are still programmers (UI programmers granted) not "useability experts".
What you're describing is component based pogramming, and it works great. However, having someone who knows squat about development do all the interface design is just as bad as having someone who knows squat about user interfacing do the interface design...
In order to get a nifty/useful interface you need somone who can deal with *both* user needs and technical needs. (You can try to do it with more than one person, but if you don't have a single individual in charge, you wind up with a schizophrenic mess)
I'm having trouble figuring out what these guys think they've observed from that paper, but I'll be darned if that gravitation curve doesn't look like a somewhat less than critically damped oscillation.
Reading the article and the notes it seems like they're talking about a temporary static change, but looking at the graphs it seems a lot more dynamic, and like they've cut off the edges too much for useful analysis. Since when does the azimuth angle of a pendulum gradually reduce linearly over the course of hours while it is being reset every 14 minutes??!!
Lemme get this strait. You're saying this technology is patentable because it is being applied to purposely "crippled" hardware when it has existed for years on normal hardware?
ie: console = a computer purposely crippled into only doing *some* of the normal functions of a computer.
So uh, the selection mask for "crippled system" is patentable? Seems pretty silly to me. Maybe Microsoft can just rename their console and call it a "GamePC" and sidestep the whole patent?
Blech. But after cruddy release after cruddy release, why would anyone want to actually install one of their players on a computer unless they were an ad-ware tester?
I know I wouldn't even take the chance of hosing up a perfectly good windows system with any of their carp...
Gold is pretty darn useful actually. It won't rust/corrode/what have you and makes for great uncovered electrical contacts. Sure it isn't very hard, but you can mix it with other metals to make lots of nifty alloys that are still pretty corrosion proof. Oh ya and: Oooh! it's also shiny!
The question isn't whether we'll need/want gold, the question is merely how available it will be, or if even more useful minerals (like platinum, titanium etc) will become cheaper and more common.
For instance, no one will really care about owning gold if they can simply transmute whatever substance they have on hand into it (or something better) cheaply and effectively. Or more near term, no one really cares about owning gold now because we know how to dig it out of low-grade ore mines at a certain cost and that cost is only likely to go down in the future... (Lots of places with low grade ore, technology for digging gets better every year)
[LSP or Layered Service Provider is a piece of software that can be inserted into the Windows TCP/IP handler like a link in a chain. However, due to bugs in the LSP software or deletion of the software, this chain can get broken, rendering the user unable to access the Internet. Spyware is good at this, and some cleaners leave a broken LSP behind.
With the correct tool, the fix takes seconds. Without the tool, you need to uninstall and re-install the winsocket, or else the same with the entire network support. Otherwise you fall into the trap this poor bloke got into.]
Would you prefer gold, some other metal dug out of the ground or rare sea shells?
A lot of people do... Putting money in gold and other precious metals has been the hallmark of "un-patriotic" money management for years.
Basically your choices are: a) put faith in a completely made up system run by "the authorities". b) put faith in continued scarcity of certain once-precious resources.
Neither is really all that appealing. Personally I put faith in my ability to make computers do stupid tricks. Other than that, I figure ride the tides of fortune that come your way.
GRC's utility is useful for people who are too afraid to figure out how to turn the service off themselves, or can't be bothered to, or turning on the firewall is too difficult. I've seriously met users like this. Download the program and push the button are words they can understand.
Ya, this works great until GRC.com gets hacked and someone throws a trojan up there. There's no substitute for user education, or proper network configuration/adminstration.
like say a small text display that shows the serial number of the device
That's not likely to get checked in the middle of a crisis where it looks like someone's heart just quit...
Better yet... there should be hard-wired identifiers built into the harnesses so that it is impossible (in typical operation, with bright red displays in "atypical" operation) to set up a monitoring device in one doc that receives data from another. Basically... "room docs" should be incapable of receiving remote input. "nurse docs" should be much more configurable.
Sometimes the features you leave out are as important as the features you put in.
I always figured it was because he'd heard of/used keyboards, but that they were some old oddity in his time. (which he was obviously still familiar with)
It won't display right in large font, making it unusable in IE, and doesn't appear to work in Mozilla at all... (Though mozilla can get around the large font issue)
The truth is, html-based e-mail represents more of a security risk than phishing scams.
Really... HTML email is insecure. Not MS LookOut email, but any HTML email using any browser? Come on, Thunderbird is insecure?
What about browsing the web in general? Should we all just go home and hide in a cave 'cause we might catch viruses from exploits that haven't been discovered yet in products with relatively good track records?
Slashdot Mirror
oh yeah... and the thought never occured to them to clean up extraneous links ON EVERY SAVE instead of just waiting until they run out of memory.
...is less an example of a failed process than it is a testament to the difficulties of debugging feature-rich software on a timetable that meets marketing demands...
Um, except the part where a developer and a tester go round and round about differences between "debug mode" and "release mode" problems. I mean come on, what kind of amatuer windows developer doesn't recognize the gigantic differences between debug and release mode apps on win32??!! If these guys would just learn to use printf() and/or the OutputDebugString() function they might be able to "see" a bit better.
Oh yeah, and the part where they know "max file size" is an issue, but don't stop to figure out what the OS max file limit is before going on... sheesh. It'd be one thing if this bug's lifetime were a few weeks, but years??!!
So uh, did your profressors pass these programs off to the smartest hackers they could find, source included, to see if this was really true after the fact?
From the jargon file:The only way you can "prove" your program is correct is by mangling the specification one way or another. (perhaps into something long, precise and deterministic) This leaves room for bugs to creep in at the specification level instead of at the coding level. (It also makes specifications a lot bigger and harder to deal with, and is a hallmark of the type of top-down design preferred by certain large computing companies in the 70's and 80's)
In practice this seems to culminate in less then effective communication between developers and their customers... Specifications should be kept simple so customers and developers can both understand them and easily talk about them.
That said, companies like Microsoft can certainly focus on smaller more solid programs instead of large feature sets, but that's hardly going to make software "perfect".
(and by accept, I mean, printed it out, signed it, and faxed it back to Microsoft).
Isn't that only necessary for those wishing to "clear the air" as to their IP rights? (Read folks with possible patent claims related to this particular standard)
then the k3b people use those to make a nice interface for it.
Yes, except the k3b guys are still programmers (UI programmers granted) not "useability experts".
What you're describing is component based pogramming, and it works great. However, having someone who knows squat about development do all the interface design is just as bad as having someone who knows squat about user interfacing do the interface design...
In order to get a nifty/useful interface you need somone who can deal with *both* user needs and technical needs. (You can try to do it with more than one person, but if you don't have a single individual in charge, you wind up with a schizophrenic mess)
But the patent this article is about is enhancing the console to do other stuff, which is completly differnt.
Um... enhancing them to do stuff computers already do? Isn't that just uncrippling them a bit? (eg: adjusting the "cripple mask")
I'm having trouble figuring out what these guys think they've observed from that paper, but I'll be darned if that gravitation curve doesn't look like a somewhat less than critically damped oscillation.
Reading the article and the notes it seems like they're talking about a temporary static change, but looking at the graphs it seems a lot more dynamic, and like they've cut off the edges too much for useful analysis. Since when does the azimuth angle of a pendulum gradually reduce linearly over the course of hours while it is being reset every 14 minutes??!!
...because it wasn't designed to.
Lemme get this strait. You're saying this technology is patentable because it is being applied to purposely "crippled" hardware when it has existed for years on normal hardware?
ie: console = a computer purposely crippled into only doing *some* of the normal functions of a computer.
So uh, the selection mask for "crippled system" is patentable? Seems pretty silly to me. Maybe Microsoft can just rename their console and call it a "GamePC" and sidestep the whole patent?
Meanwhile you've just set in motion as nasty little email war that's bad for every spammer.
It's also bad for every email user, and spills over into web-browsing. Not a fun notion.
how many of those dial-up users are still AoLers who are creating the majority of the problems ...
And how many of those broadband users are firewall-less windows users causing the other half of the problems on the internet...
Blech. But after cruddy release after cruddy release, why would anyone want to actually install one of their players on a computer unless they were an ad-ware tester?
I know I wouldn't even take the chance of hosing up a perfectly good windows system with any of their carp...
Gold is pretty darn useful actually. It won't rust/corrode/what have you and makes for great uncovered electrical contacts. Sure it isn't very hard, but you can mix it with other metals to make lots of nifty alloys that are still pretty corrosion proof. Oh ya and: Oooh! it's also shiny!
The question isn't whether we'll need/want gold, the question is merely how available it will be, or if even more useful minerals (like platinum, titanium etc) will become cheaper and more common.
For instance, no one will really care about owning gold if they can simply transmute whatever substance they have on hand into it (or something better) cheaply and effectively. Or more near term, no one really cares about owning gold now because we know how to dig it out of low-grade ore mines at a certain cost and that cost is only likely to go down in the future... (Lots of places with low grade ore, technology for digging gets better every year)
[LSP or Layered Service Provider is a piece of software that can be inserted into the Windows TCP/IP handler like a link in a chain. However, due to bugs in the LSP software or deletion of the software, this chain can get broken, rendering the user unable to access the Internet. Spyware is good at this, and some cleaners leave a broken LSP behind.
With the correct tool, the fix takes seconds. Without the tool, you need to uninstall and re-install the winsocket, or else the same with the entire network support. Otherwise you fall into the trap this poor bloke got into.]
And the "proper tool" would be...?
Would you prefer gold, some other metal dug out of the ground or rare sea shells?
A lot of people do... Putting money in gold and other precious metals has been the hallmark of "un-patriotic" money management for years.
Basically your choices are:
a) put faith in a completely made up system run by "the authorities".
b) put faith in continued scarcity of certain once-precious resources.
Neither is really all that appealing. Personally I put faith in my ability to make computers do stupid tricks. Other than that, I figure ride the tides of fortune that come your way.
GRC's utility is useful for people who are too afraid to figure out how to turn the service off themselves, or can't be bothered to, or turning on the firewall is too difficult. I've seriously met users like this. Download the program and push the button are words they can understand.
Ya, this works great until GRC.com gets hacked and someone throws a trojan up there. There's no substitute for user education, or proper network configuration/adminstration.
like say a small text display that shows the serial number of the device
That's not likely to get checked in the middle of a crisis where it looks like someone's heart just quit...
Better yet... there should be hard-wired identifiers built into the harnesses so that it is impossible (in typical operation, with bright red displays in "atypical" operation) to set up a monitoring device in one doc that receives data from another. Basically... "room docs" should be incapable of receiving remote input. "nurse docs" should be much more configurable.
Sometimes the features you leave out are as important as the features you put in.
Ya, you might lose the options, but no one can ever "force" you to exercise them...
He could have used: alt-58, alt-41, and alt-63 on the number pad.
Oh wait. You must be a javaDeveloper eh?
As long as you're not using wrdHungarianNotation.
You know. Like VI. (NOT emacs)
Then you'll be doing mostly cut/paste/search/replace etc anyways...
I always figured it was because he'd heard of/used keyboards, but that they were some old oddity in his time. (which he was obviously still familiar with)
;-)
You know... like uh... the Unix command line
Dontcha.
Or IRC. Sheesh, I only got 68 adjusted with 90% accuracy. (It helped that they let me backspace within words...)
It won't display right in large font, making it unusable in IE, and doesn't appear to work in Mozilla at all... (Though mozilla can get around the large font issue)
The truth is, html-based e-mail represents more of a security risk than phishing scams.
Really... HTML email is insecure. Not MS LookOut email, but any HTML email using any browser? Come on, Thunderbird is insecure?
What about browsing the web in general? Should we all just go home and hide in a cave 'cause we might catch viruses from exploits that haven't been discovered yet in products with relatively good track records?