Back in the late 90s, the company I worked for then wrote a volume scanning application for a large government agency in the UK. This was a volume scanning app i.e. up to 50 ppm for 8 hours a day, so we were using drivers from Kofax. We installed the s/w on machines running NT4 sp4.
The prime contractor (a company called IBM) found out that the OS was not Y2K compliant (in fact the only thing wrong was it was unable to set a password expiry date of 29/2/2000, but they didn't bother to find that out). So they put sp6a on all the machines which broke the Kofax drivers and hence our software.
For this reason, I would never accept an OS configuration where I don't get to say yes or no to a patch before it installs itself. Admittedly I never say no to M$ Windows Update patches, but at least I know I've put them on if something else mysteriously stops working. Note that I'm not necessarily saying M$ patches break things. In the case of the Kofax drivers above, there was a bug in the driver which they could get away with on sp4 but not on sp6a.
Re:The last of the Apple-based OSX machines?
on
G5s Start Shipping
·
· Score: 3, Informative
Not even remotely true. OS X is based on the Mach kernel with a BSD userland. On top of that they put the Aqua windowing system. The only thing it has in common with Linux is that you build it with gcc.
If Apple ported the OS to the x86 platform, their target market would be about 20 times bigger, so I think it's reasonable to assume that their s/w sales would increase by at least ten times.
However, porting the software is non trivial since they would no longer have any cotrol over the hardware. They'd have to cope with many more different hardware configs, probably to the point that the software would no longer "just work".
Well Ok. I cut my programming teeth on a Commodore Pet with 16K later upgraded to 32. Actually that wasn't as bad as it seems because the BASIC interpreter and the "kernel" [were on 16Kb of ROM.
The Commodore 64 also had its BASIC on ROM, but it was shadowed by RAM, so you could replace it with RAM programmatically if you wanted.
There are currently no manned spacecraft available that can reach a geosynchronous orbit. It's a lot further away than where the current space station is.
I heard an astronomer say that sodium lighting was less bad from their point of view because it only swamps one wavelength as opposed to white light which emits on all human visible wavelengths and some others besides.
Except that the authenticate function should be in its own try block so that the code that handles the failed authentication stays near it
e.g
try {
authenticate () } catch (AuthException e) {/* Log in the security log + tell the user */
return; }/* do other stuff which might be several pages long */
Why not have the distance between the mirrors as (1/(n^2) + 2w) where w is the wavelength of the beam of light?
Because then the path length is greater than n * 2w where n is the number of mirrors and w is the wavelength of the photon (which is a constant). Thus your path length is now infinite and takes an infinite time for a photon to traverse it.
But that would mean that the path would be infinitely long (each mirror adds at least a d to it) and there is no paradox because the photon can take forever to get to the end of an infinitely long path.
Anyway, in real life, quantum effects mean that the photon will stop behaving in the intuitive-to-humans way fairly quickly.
That's after the digits have been read to them only once. Most long distance telephone numbers in the UK are 11 digits long (well 10 as the first digit is always "0") and I can remember several of them purely through the fact that I've dialled them a lot. I can also remember several IP addresses that I use regularly and once I even got to know a 25 aphanumeric licence key because I had to install the relevant piece of software about 8 times due to various configurational and compatibility problems.
In Mail.app, I separate mail addresses with commas, and I get a drop-down to choose from.
I learn something new every day. I doubt if I'll be moving off Mac mail though. Mozilla mail doesn't integrate with my Mac address book and when you click a link it ignores my choie of browser in system prefs. I only keep it around because it IMAP synchronisation support is better and Apple mail seems to have no support at all for mail signing and encryption or client certs on TLS.
I'll probably download the Thunderbird client because it looks slightly less bloated than Mozilla (10Mb download against 16Mb), but I suspect the poor OS integration is always going to be an issue.
By that argument, Windows is the standard for desktop computing and Word is the standard for text documents. Microsoft is the big dog, Linux is an also ran along for the ride.
I fail to see how. gets uses stdin for its input stream. There's no way the programmer can control what comes in on that without closing stdin and reopening to a file with known and trusted contents (in which case why not use fgets directly?). I could pipe anything I like to stdin so there is absolutely no way for the programmer to prevent a buffer overflow.
People who cannot be taught to shut down a server properly must be kept away from them under all circumstances. It's crazy to let people just switch off a production server no matter how convenient it might seem to be or how many times it might have worked in the past.
I think that's actually because Finder is single threaded and waiting for the TCP connection to time out. I've seen the same problem, but a) other applications continue to work normally and b) eventually the Finder figures out that the share has gone and wakes up.
Slashdot Mirror
Back in the late 90s, the company I worked for then wrote a volume scanning application for a large government agency in the UK. This was a volume scanning app i.e. up to 50 ppm for 8 hours a day, so we were using drivers from Kofax. We installed the s/w on machines running NT4 sp4.
The prime contractor (a company called IBM) found out that the OS was not Y2K compliant (in fact the only thing wrong was it was unable to set a password expiry date of 29/2/2000, but they didn't bother to find that out). So they put sp6a on all the machines which broke the Kofax drivers and hence our software.
For this reason, I would never accept an OS configuration where I don't get to say yes or no to a patch before it installs itself. Admittedly I never say no to M$ Windows Update patches, but at least I know I've put them on if something else mysteriously stops working. Note that I'm not necessarily saying M$ patches break things. In the case of the Kofax drivers above, there was a bug in the driver which they could get away with on sp4 but not on sp6a.
Not even remotely true.
OS X is based on the Mach kernel with a BSD userland. On top of that they put the Aqua windowing system. The only thing it has in common with Linux is that you build it with gcc.
If Apple ported the OS to the x86 platform, their target market would be about 20 times bigger, so I think it's reasonable to assume that their s/w sales would increase by at least ten times.
However, porting the software is non trivial since they would no longer have any cotrol over the hardware. They'd have to cope with many more different hardware configs, probably to the point that the software would no longer "just work".
Well Ok. I cut my programming teeth on a Commodore Pet with 16K later upgraded to 32. Actually that wasn't as bad as it seems because the BASIC interpreter and the "kernel" [were on 16Kb of ROM.
The Commodore 64 also had its BASIC on ROM, but it was shadowed by RAM, so you could replace it with RAM programmatically if you wanted.
There are currently no manned spacecraft available that can reach a geosynchronous orbit. It's a lot further away than where the current space station is.
Ariane 5??
It'll probably boost the observatory to an even distribution of small fragments over a 400 mile radius around the launch point on the Earth's surface.
But, it'll be very pretty.
I heard an astronomer say that sodium lighting was less bad from their point of view because it only swamps one wavelength as opposed to white light which emits on all human visible wavelengths and some others besides.
Except that the authenticate function should be in its own try block so that the code that handles the failed authentication stays near it
/* Log in the security log + tell the user */ ; /* do other stuff which might be several pages long */
e.g
try
{
authenticate ()
}
catch (AuthException e)
{
return
}
It's not beta anymore
This article reads like an advert to me
I think the phrase "press release" gives the clue. It is an advert.
Why not have the distance between the mirrors as (1/(n^2) + 2w) where w is the wavelength of the beam of light?
Because then the path length is greater than n * 2w where n is the number of mirrors and w is the wavelength of the photon (which is a constant). Thus your path length is now infinite and takes an infinite time for a photon to traverse it.
But that would mean that the path would be infinitely long (each mirror adds at least a d to it) and there is no paradox because the photon can take forever to get to the end of an infinitely long path.
Anyway, in real life, quantum effects mean that the photon will stop behaving in the intuitive-to-humans way fairly quickly.
It's RFC 2373
An IPv6 address is a sequence of 8 16 bit numbers. It's usually written in hex with the numbers separated by colons e.g:
1080:0:0:0:8:800:200C:417A
Apparently IPv4 only nodes are represented in the IPv6 space with the first 6 numbers being
0:0:0:0:0:FFFF
and the last two numbers containing the IPv4 address e.g.
10.1.2.3 would be
0:0:0:0:0:FFFF:A:1:2:3
and may also be written as
0:0:0:0:0:FFFF:10.1.2.3
That's after the digits have been read to them only once. Most long distance telephone numbers in the UK are 11 digits long (well 10 as the first digit is always "0") and I can remember several of them purely through the fact that I've dialled them a lot. I can also remember several IP addresses that I use regularly and once I even got to know a 25 aphanumeric licence key because I had to install the relevant piece of software about 8 times due to various configurational and compatibility problems.
In Mail.app, I separate mail addresses with commas, and I get a drop-down to choose from.
I learn something new every day. I doubt if I'll be moving off Mac mail though. Mozilla mail doesn't integrate with my Mac address book and when you click a link it ignores my choie of browser in system prefs. I only keep it around because it IMAP synchronisation support is better and Apple mail seems to have no support at all for mail signing and encryption or client certs on TLS.
I'll probably download the Thunderbird client because it looks slightly less bloated than Mozilla (10Mb download against 16Mb), but I suspect the poor OS integration is always going to be an issue.
By that argument, Windows is the standard for desktop computing and Word is the standard for text documents. Microsoft is the big dog, Linux is an also ran along for the ride.
Likewise, gets() can be used safely
I fail to see how. gets uses stdin for its input stream. There's no way the programmer can control what comes in on that without closing stdin and reopening to a file with known and trusted contents (in which case why not use fgets directly?). I could pipe anything I like to stdin so there is absolutely no way for the programmer to prevent a buffer overflow.
I work in tech support 90% of all the people I talk to each day are complete morons.
I often have to call tech support. 90% of the time I get put through to a complete moron.
Well if you knew your Monty Python, you'd know there were at least three films.
People who cannot be taught to shut down a server properly must be kept away from them under all circumstances. It's crazy to let people just switch off a production server no matter how convenient it might seem to be or how many times it might have worked in the past.
Frankly, you deserved what you got.
I think that's actually because Finder is single threaded and waiting for the TCP connection to time out. I've seen the same problem, but a) other applications continue to work normally and b) eventually the Finder figures out that the share has gone and wakes up.
SMTP 101. Everything in an e-mail can be faked easily except the IP address of the client mail server.
You can teach anybody to fake e-mails in about 10 minutes as long as you have access to a telnet client.
Bouncing a message with the SMTP response "550 user unknown" is probably OK.
Isn't it time your company trained somebody else to manage the servers when you're not around?