Now take an abitrary failure condition in which your program core dumps every so often for no reason, and you need to find the cause...
Consider the following two algorithms (written for Australia, modify as appropriate).
1) Look Right
2) Look Left
3) Look Right
4) Walk half way across the road
5) Look Left
6) Walk the other half of the way
1) blindly walk across the road without looking
Look how many steps you've saved with the second one, isn't it easier to do. Especially when you consider that in the first example I haven't even tried to do error handling (if car which is dangerous to next movement step, wait until car is no longer a danger, skip back to appropriate looks - if too many cars, abort and move to controlled intersection where different rules apply... i.e. wait for green man).
As soon as countries standardise on 00 as the international access code (and that's happening) then we will have a global unique numbering system administered by countries. It's called the phone system.
In the UK, we can already get "personal numbers" which you can have redirected to wherever you are. There's no reason why companies in other countries can't do the same thing.
And meanwhile here in Australia I've moved about 300 metres away, across a suburb and exchange boundary, and I have to change phone numbers. Of course it's only the landline, which is used more as an ADSL conduit than anything else, but still.
Another thing - I work in Medical Data collection, and we always request two things - the UR Number and the Initials of a person - there's a very good reason for this, a single number with no checksum means that you only need a single bit of corruption to render a valid but incorrect value. The initials provide a double check which is good when there are multiple forms per patient which need to be collated.
I certainly wouldn't want only a single number on
anything that isn't checksummed some other way.
Cracking a 4 digit PIN at internet speed is TRIVIAL!
Not if each site that allows passport (single sign-on) also requires a 4 digit PIN (localised information not stored in your passport)
Add to this the ability for each site to use the standard 3 strikes and you're offline for a day method to stop the internet speed problem.
This actually has the major advantage of raising the barrier for denial of service attacks against PIN accepting sites. Without something like passport (or a try-as-many-times-as-you-like password first) it's trivial for someone to DOS
your account by trying 3 PINs every day. To do this in a world with MS Passport or equivalent the attacker must first steal the passport, then use it to gain access to the PIN interface.
Of course, if the passport is trivially stealable without any evidence of it going missing then it doesn't help much, but consider.
If sites that require a PIN report incorrect attempts to you, you will know when your passport has been stolen and someone has attempted to use it.
Based on how Passport works, this means that you just have to contact Microsoft and have your passport disabled, and nobody can use it. Indeed, the merchant could immediately contact Microsoft upon getting 3 incorrect PINs and have your passport disabled until you authenticate as you through some side-channel and obtain a new passport.
<comment type="placate M$ hating weenies>Though of course this assumes that Microsoft provide a timely service for revoking and replacing passports, and that your information hasn't been stolen in between times</comment>
Apparently the update to the emu10k broke building it as a module. I just build it into the kernel itself (not as a module) and it seems to work fine.
I have to admit I patched it with 2.4.8-ac1 (after patching my 2.4.6 with patch-2.4.7 and patch-2.4.8 from my local mirror) and it builds fine as either a module or builtin.
But when I'm programming any interrupt withing a 3-4 hour time span destroys all my concentration.
Which is why I get all my best 'creating' programming work done between midnight and about 3am. I'm still awake enough to think, but there's no phone or direct interuptions, and my girlfriend has gone to sleep. I can get that 3 hour concentration time.
I'm just very glad that my work allows me flexible working hours and the ability to work from home.
I think that's really stretching it. Far more than even the most vengeful court could see fit to prosecute as.
What, you mean like they wouldn't do anything like arrest a kid and hold them in jail for 2 weeks without bail for breaking rot 13^H^H^H^H^H^H^H^H^H^H^H^H^H^H^Hsending an email?
There's absolutely no point having the pulleys on the kite. The only reason for the
pulleys is to place the majority of the force on the tripod thingy that's holding them, and hence through to the ground.
If you placed the pulleys on the kite, then you wouldn't get the force multiplying effect of the pulleys (ok, multiplying is probably not the most accurate term...). You might as well just have a single rope.
The cost of the cars ranged from $30,000 to more than $1-million.
I bet most of that went into top-of-the-line solar cells.
The good thing about races like this is that effort is put into designing the very best cells possible, and a side effect is (eventually, we hope) better solar cells available to the community market.
A major advance was made (the so called 'green cells' in Australia for the Australian solar challenge a few years ago).
It's not just Microsoft that lies like that to customers. We were running some PABX billing software at an ex-employer, and it had heaps of show-stopper bugs (we were running debug binaries for #$^&*'s sake).
I called - they said nobody else was having these problems. I demanded to know how many installed copies they had.
Turns out there were 2 - one of the techs told me while I was chatting to him in a notepad on the PC-Anywhere session he was fixing something with.
If you were a friend who sent me a neat program that you'd written and wanted to show it off (complete with !!!!!'s in the title), I'd just delete it.
Remember that rule about not running attachments, even those sent to you by friends.
Subject: Check out this great program I wrote
Dood, this is leet. Run this.
Attachment: leetapp.exe
P.S. I wrote this in leetspeak first because it was funny, I thought, and Slashdot said the following:
Lameness filter encountered. Post aborted.
Reason: Junk character post.
Maybe this should be cross linked to all those comments about Microsoft adding McHappyLinks to your web site in their browser, or censorship.
Personally I think it's quite reasonable to use 'leetspeak' sarcastically in an attempt at humour, but obviously the autofilters of slashdot don't agree. Bastion of free fucking speach indeed.
RIT Labs has a product (actually 2 now) that sound like they do what you're looking for, although neither is free, but I've been using the bat for around a year now, and I'm really happy with it
Actually I've had the opposite experience with 'The Bat!'. My work has 10 licences for it, and we've been using it (small company) for internal email. I recently had a 500Mb mail file with only 7Mb of data because it wasn't compressed. On reading the (not very helpful) documentation I found that yes, I could 'compress all folders' from the client, or set specific folders to compress on exit - but couldn't set it as a global option.
'The Bat!' uses a binary custom mailbox format.
'The Bat!' only supports IMAP as a POP-alike download system - IMAP folders on the server are
not supported.
The response I received from their tech support line about these issues was "we're planning to fix that in a few months" to every single question I asked. Vapourware'r'us.
I have decided that I don't have time to wait for a commercial company to solve the problems that other companies (Eudora Light 3.06 anyone) have solved years ago. It feels like they're working their way through all the same mistakes.
So, assuming for the sake of argument the monitor has a roughly 16"x12" viewable area, that gives (16*17000) * (12*17000) = 55,488,000,000 points, or 55.5 gigapoints, as the limit of human eye resolution for a screen of that size. That's several orders of magnitude over the announced 9.2 megapixel display, assuming that pixels are roughly equivalent to points.
I guess I really shouldn't be suprised at Slashdot readers having no understanding of basic optics. I note that the parent post has been modded 'Flamebait' as well.
I don't know if the unit is 1700 point sources, but I would read that as per 'square inch', so your mathematics should be (16 * 12 * 17000) = 3,264,000 - or less than 9.2 megapixels.
I remember reading a few years ago (can't remember the cite) that about 3000 by 3000 pixels was the maximum that a human eye could discern (no matter what the distance) while still seeing the entire screen at once. Of course a bigger screen close up would look blocky, but you couldn't view it all without getting futher away anyway.
This was based on calculations which said that (from memory again) humans can see an arc of about 10 degrees at once, and have a certain number of photo-receptors per degree of arc in the eye. Once you get one pixel per photo-receptor, that's about the limit. Make it one and a half for smoother blending - and take into account that inperfections in the lense of the eye will blur it more anyway.
I think there's a couple differences between this pattern of kernel releases and MS service packs.
These don't seem to break programs that previously worked - at least the ones I've used.
2.4.4 on my machine broke the Realtek 8139 card for DHCP - it would freeze for a few minutes and
then start without network. Reverting to 2.4.3 fixed the problem.
I believe that this may be fixed in 2.4.5 from reading the changelog
-pre2:
- Jeff Garzik: 8139too net drvr fix
Palmtop computers are almost completely unnecessary (though desirable) pieces of consumer electronics, and it's quite understandable that when people start worrying about their financial security, these expensive toys are the first things people stop buying.
The company I work for does a lot of data collection for medical surveys. It's a lot less load on busy doctors to tick the answers into a palm pilot and upload the data to their desktop rather than tick all the boxes on paper forms (most of the questions are tickbox or small selection of posibilites).
Just because you only play games on your PDA (hey, that's all I do at the moment, until I find time to dig into the programming side of things rather than just sysadmin) doesn't mean that's all they can be used for.
Most people on slashdot are posers. They don't even use linux they just hang out here and pretend they are cool.
Or they use Opera and leave it in its default "Pretend to be IE5" mode.
I use Opera on both Linux and Windows out of preference, and while I prefer the Windows version at the moment (gesture support rocks, and the scroll works better than Linux) I can see how leaving it with IE support would mess up the stats.
"This is a country that can AFFORD to employ threehundred thousand people to sit a PC all day and jot down non-china sites.."
Google says there are about 1.3 billion web sites out there. Even if the average time it takes to read those sites and scan for verboten material is only one minute, it will still take those 300,000 about 7 10-hour shifts to do a complete scan of the web, assuming that no pages change during that time. If, for example, the Free Nepal people figure out their weekly cycle, they can mirror the People's Daily six days of the week, post the good word from the Dalai Lama every Sunday, and still have only a 1 in 7 chance of ever being caught. And, if they are, they change IPs and do it again.
Sure, but they don't _have_ to do a complete scan of the web, all they need to do is have a global 'proxy server' to which requests for new URLS are first routed to a human who checks and approves or otherwise the content. This site is then cached, and cache hits are served without re-checking.
With a decent implementation of 'sticky-redirect', other pages from the same site being first fed to the same person if possible (obviously something like geocities would be split amongst multiple checkers) and a storing of checker-ID against each URL with a smaller number of trusted 'secret police' checking random URLs and shooting people who approved the wrong thing, you'd be set.
Yes it's possible (at the expense of a slower web of course.. first hit would take a little while).
Now, perhaps other titles could be determined by looking at domain registrations
that were done about the same time by the same person. (An exercise left to the reader).
Seriously though, if I was working for Lucas I would have made up a couple of hundred 'likely' sounding names and reserved all the domain names. Come on, considering what a single movie costs, squatting over a pile of namespace is nothing.
This USB trend really is a pain in the ass for now, but since USB will be supported in the next kernel, I'm sure there's gonna be a port eventually. Or maybe it would work with a USB-to-parallel converter cable
From the linked site: It contains USB drivers for Windows(R) 98 and 2000 and there is a serial cable available for use with Windows NT(R) systems.
Shouldn't be too much trouble to interface to Linux through the serial option for now (though it does load the system more than USB - and we'll have USB support soon enough.)
Also from the site: Sony is working with Entrust Technologies and I/O Software Inc., to allow them to develop specific software applications and is also actively looking to work with other software providers in the infosec field
It's unlikely that they will be providing open-sourced drivers at first (Sony haven't really "jumped on the open source bandwagon" yet) but with Linux becoming more popular all the time, it's likely that demand will convince them to build drivers.
They'll almost certainly be building drivers for various UN*X systems because, despite Microsoft's efforts to push NT, there are many large institutions which will pay megabucks to have a more secure way of authenticating users that just works! The weakest link in most security is the users themselves, and the pathetic passwords most people choose.
Fingerprint assisted password protection would be much stronger, and I doubt Sony will restrict themselves to a single OS manufacturer if they're getting so many companies to write drivers.
There's no way that slashdot can veryify the validity of Microsoft's claim without opening the "click-through" licence.
Because of this, I don't see any reason why slashdot should accept the claim that the material held on their servers is indeed copyrighted by Microsoft until they are provided with evidence of this claim (e.g. a copy of the document without the licence).
It is certainly unreasonable to remove a posting just because somebody claims that they own copyright on it without providing evidence. If this was the case, I could have any posting removed just by emailing rob and saying it was mine and I wanted it taken down!
A number of patent related comments on slashdot recently[0] have accused the US Patent Office[1] of approving anything technology related without doing proper research for prior art.
Given the number of stupid patents that get through, this certainly seems to be the case. Sure they can argue lack of funding, and it is an issue, but that isn't the line they're taking.
From the link in the article: Patent officials made the decision after some information came to light that was not considered when the patent was originally granted
They're not admitting that they didn't do anywhere near enough checking the first time, or explaining how such a stupid patent could have happend, just covering their backsides about this second look.
Further down: Experts say decisions by the PTO to review patents that are already approved are rare.
It certainly seems to me that someone along the way has forgotten that patents exist to protect original work that contains unique elements. The Patent Office exists not only to grant patents, but to block them when they clearly don't fit the charter.
[0] Admittedly mainly restating the obvious in moderator friendly format. [1] Disclaimer: I'm in Australia. I rest my case on pointy haired.govs.
Slashdot Mirror
So I guess those news sites haven't been shut down yet then?
char * strdup (char * s) {
return strcpy(malloc(strlen(s) + 1), s);
}
Now write it checking for every possible error...
Now take an abitrary failure condition in which your program core dumps every so often for no reason, and you need to find the cause...
Consider the following two algorithms (written for Australia, modify as appropriate).
1) Look Right
2) Look Left
3) Look Right
4) Walk half way across the road
5) Look Left
6) Walk the other half of the way
1) blindly walk across the road without looking
Look how many steps you've saved with the second one, isn't it easier to do. Especially when you consider that in the first example I haven't even tried to do error handling (if car which is dangerous to next movement step, wait until car is no longer a danger, skip back to appropriate looks - if too many cars, abort and move to controlled intersection where different rules apply... i.e. wait for green man).
And meanwhile here in Australia I've moved about 300 metres away, across a suburb and exchange boundary, and I have to change phone numbers. Of course it's only the landline, which is used more as an ADSL conduit than anything else, but still.
Another thing - I work in Medical Data collection, and we always request two things - the UR Number and the Initials of a person - there's a very good reason for this, a single number with no checksum means that you only need a single bit of corruption to render a valid but incorrect value. The initials provide a double check which is good when there are multiple forms per patient which need to be collated.
I certainly wouldn't want only a single number on anything that isn't checksummed some other way.
Not if each site that allows passport (single sign-on) also requires a 4 digit PIN (localised information not stored in your passport)
Add to this the ability for each site to use the standard 3 strikes and you're offline for a day method to stop the internet speed problem.
This actually has the major advantage of raising the barrier for denial of service attacks against PIN accepting sites. Without something like passport (or a try-as-many-times-as-you-like password first) it's trivial for someone to DOS your account by trying 3 PINs every day. To do this in a world with MS Passport or equivalent the attacker must first steal the passport, then use it to gain access to the PIN interface.
Of course, if the passport is trivially stealable without any evidence of it going missing then it doesn't help much, but consider.
If sites that require a PIN report incorrect attempts to you, you will know when your passport has been stolen and someone has attempted to use it.
Based on how Passport works, this means that you just have to contact Microsoft and have your passport disabled, and nobody can use it. Indeed, the merchant could immediately contact Microsoft upon getting 3 incorrect PINs and have your passport disabled until you authenticate as you through some side-channel and obtain a new passport.
<comment type="placate M$ hating weenies>Though of course this assumes that Microsoft provide a timely service for revoking and replacing passports, and that your information hasn't been stolen in between times</comment>
And if it ends in FF it will give it all back,
oh well, I guess #0000FF is a brighter blue than #0000AA.
I have to admit I patched it with 2.4.8-ac1 (after patching my 2.4.6 with patch-2.4.7 and patch-2.4.8 from my local mirror) and it builds fine as either a module or builtin.
Yay for Alan.
Which is why I get all my best 'creating' programming work done between midnight and about 3am. I'm still awake enough to think, but there's no phone or direct interuptions, and my girlfriend has gone to sleep. I can get that 3 hour concentration time.
I'm just very glad that my work allows me flexible working hours and the ability to work from home.
What, you mean like they wouldn't do anything like arrest a kid and hold them in jail for 2 weeks without bail for breaking rot 13^H^H^H^H^H^H^H^H^H^H^H^H^H^H^Hsending an email?
If you placed the pulleys on the kite, then you wouldn't get the force multiplying effect of the pulleys (ok, multiplying is probably not the most accurate term...). You might as well just have a single rope.
I bet most of that went into top-of-the-line solar cells.
The good thing about races like this is that effort is put into designing the very best cells possible, and a side effect is (eventually, we hope) better solar cells available to the community market.
A major advance was made (the so called 'green cells' in Australia for the Australian solar challenge a few years ago).
A good resource on PV cells (notice my Australian bias!) is http://acre.murdoch.edu.au/refiles/pv/text.html.
I called - they said nobody else was having these problems. I demanded to know how many installed copies they had.
Turns out there were 2 - one of the techs told me while I was chatting to him in a notepad on the PC-Anywhere session he was fixing something with.
Stupid vendors.
I guess nobody else here has ever installed Activestate Perl then...
I'm wondering if it's just because Windows Installer requires a 'clickthrough licence' bit that they show it - it's the real GPL though.
Remember that rule about not running attachments, even those sent to you by friends.
Subject: Check out this great program I wrote
Dood, this is leet. Run this.
Attachment: leetapp.exe
P.S. I wrote this in leetspeak first because it was funny, I thought, and Slashdot said the following:
Lameness filter encountered. Post aborted.
Reason: Junk character post.
Maybe this should be cross linked to all those comments about Microsoft adding McHappyLinks to your web site in their browser, or censorship.
Personally I think it's quite reasonable to use 'leetspeak' sarcastically in an attempt at humour, but obviously the autofilters of slashdot don't agree. Bastion of free fucking speach indeed.
So that's where she got to. She taught me first year CompSci at Tas Uni (Hobart) in 1996. Small world.
Actually I've had the opposite experience with 'The Bat!'. My work has 10 licences for it, and we've been using it (small company) for internal email. I recently had a 500Mb mail file with only 7Mb of data because it wasn't compressed. On reading the (not very helpful) documentation I found that yes, I could 'compress all folders' from the client, or set specific folders to compress on exit - but couldn't set it as a global option.
'The Bat!' uses a binary custom mailbox format.
'The Bat!' only supports IMAP as a POP-alike download system - IMAP folders on the server are not supported.
The response I received from their tech support line about these issues was "we're planning to fix that in a few months" to every single question I asked. Vapourware'r'us.
I have decided that I don't have time to wait for a commercial company to solve the problems that other companies (Eudora Light 3.06 anyone) have solved years ago. It feels like they're working their way through all the same mistakes.
I guess I really shouldn't be suprised at Slashdot readers having no understanding of basic optics. I note that the parent post has been modded 'Flamebait' as well.
I don't know if the unit is 1700 point sources, but I would read that as per 'square inch', so your mathematics should be (16 * 12 * 17000) = 3,264,000 - or less than 9.2 megapixels.
I remember reading a few years ago (can't remember the cite) that about 3000 by 3000 pixels was the maximum that a human eye could discern (no matter what the distance) while still seeing the entire screen at once. Of course a bigger screen close up would look blocky, but you couldn't view it all without getting futher away anyway.
This was based on calculations which said that (from memory again) humans can see an arc of about 10 degrees at once, and have a certain number of photo-receptors per degree of arc in the eye. Once you get one pixel per photo-receptor, that's about the limit. Make it one and a half for smoother blending - and take into account that inperfections in the lense of the eye will blur it more anyway.
These don't seem to break programs that previously worked - at least the ones I've used.
2.4.4 on my machine broke the Realtek 8139 card for DHCP - it would freeze for a few minutes and then start without network. Reverting to 2.4.3 fixed the problem.
I believe that this may be fixed in 2.4.5 from reading the changelog
-pre2: - Jeff Garzik: 8139too net drvr fix
The company I work for does a lot of data collection for medical surveys. It's a lot less load on busy doctors to tick the answers into a palm pilot and upload the data to their desktop rather than tick all the boxes on paper forms (most of the questions are tickbox or small selection of posibilites).
Just because you only play games on your PDA (hey, that's all I do at the moment, until I find time to dig into the programming side of things rather than just sysadmin) doesn't mean that's all they can be used for.
Or they use Opera and leave it in its default "Pretend to be IE5" mode.
I use Opera on both Linux and Windows out of preference, and while I prefer the Windows version at the moment (gesture support rocks, and the scroll works better than Linux) I can see how leaving it with IE support would mess up the stats.
Remember this when reading your logs, webmasters!
Google says there are about 1.3 billion web sites out there. Even if the average time it takes to read those sites and scan for verboten material is only one minute, it will still take those 300,000 about 7 10-hour shifts to do a complete scan of the web, assuming that no pages change during that time. If, for example, the Free Nepal people figure out their weekly cycle, they can mirror the People's Daily six days of the week, post the good word from the Dalai Lama every Sunday, and still have only a 1 in 7 chance of ever being caught. And, if they are, they change IPs and do it again.
Sure, but they don't _have_ to do a complete scan of the web, all they need to do is have a global 'proxy server' to which requests for new URLS are first routed to a human who checks and approves or otherwise the content. This site is then cached, and cache hits are served without re-checking.
With a decent implementation of 'sticky-redirect', other pages from the same site being first fed to the same person if possible (obviously something like geocities would be split amongst multiple checkers) and a storing of checker-ID against each URL with a smaller number of trusted 'secret police' checking random URLs and shooting people who approved the wrong thing, you'd be set.
Yes it's possible (at the expense of a slower web of course.. first hit would take a little while).
So the search engine converts all urls to:
/etc/resolv.conf entry with
http://server.aol.newtlds.com/index.html
Where newtlds.com is the "oldspace" root - yes
it means that each vhost needs a line:
ServerAlias server.aol server.aol.newtlds.com
But that's probably the safest compatibility hack. In fact combine this with a
search newtlds.com
on your squid proxy box, and you don't even need to fix the clients for pure web stuff - just point them through the proxy!
Seriously though, if I was working for Lucas I would have made up a couple of hundred 'likely' sounding names and reserved all the domain names. Come on, considering what a single movie costs, squatting over a pile of namespace is nothing.
From the linked site: It contains USB drivers for Windows(R) 98 and 2000 and there is a serial cable available for use with Windows NT(R) systems.
Shouldn't be too much trouble to interface to Linux through the serial option for now (though it does load the system more than USB - and we'll have USB support soon enough.)
Also from the site: Sony is working with Entrust Technologies and I/O Software Inc., to allow them to develop specific software applications and is also actively looking to work with other software providers in the infosec field
It's unlikely that they will be providing open-sourced drivers at first (Sony haven't really "jumped on the open source bandwagon" yet) but with Linux becoming more popular all the time, it's likely that demand will convince them to build drivers.
They'll almost certainly be building drivers for various UN*X systems because, despite Microsoft's efforts to push NT, there are many large institutions which will pay megabucks to have a more secure way of authenticating users that just works! The weakest link in most security is the users themselves, and the pathetic passwords most people choose.
Fingerprint assisted password protection would be much stronger, and I doubt Sony will restrict themselves to a single OS manufacturer if they're getting so many companies to write drivers.
Because of this, I don't see any reason why slashdot should accept the claim that the material held on their servers is indeed copyrighted by Microsoft until they are provided with evidence of this claim (e.g. a copy of the document without the licence).
It is certainly unreasonable to remove a posting just because somebody claims that they own copyright on it without providing evidence. If this was the case, I could have any posting removed just by emailing rob and saying it was mine and I wanted it taken down!
IANAL and all that
Given the number of stupid patents that get through, this certainly seems to be the case. Sure they can argue lack of funding, and it is an issue, but that isn't the line they're taking.
From the link in the article: Patent officials made the decision after some information came to light that was not considered when the patent was originally granted
They're not admitting that they didn't do anywhere near enough checking the first time, or explaining how such a stupid patent could have happend, just covering their backsides about this second look.
Further down: Experts say decisions by the PTO to review patents that are already approved are rare.
It certainly seems to me that someone along the way has forgotten that patents exist to protect original work that contains unique elements. The Patent Office exists not only to grant patents, but to block them when they clearly don't fit the charter.
[0] Admittedly mainly restating the obvious in moderator friendly format. .govs.
[1] Disclaimer: I'm in Australia. I rest my case on pointy haired