You missed one in the original article (which you did read? "It stands in marked contrast to existing law and prior decisions that have determined that Congress was well within its constitutional authority to adopt legislation that prevented trafficking in copies of unauthorised performances of live music," spokesman Jonathan Lamy said.
Um - come again: "trafficking in copies of unauthorised performances of live music"
Who said those performances were unauthorised? By whom?
I have been programming both for years. The pay is a lot better than for Java / C / whatever, basically because of people like you who don't want to learn 'old technology'.
My problem is, you are half right. Eventually their use will die out. I reckon that I will have a big problem in about 10 years, but then again (quoting John Maynard Keynes): 'In the long run, we are all dead'.
NWA for example. There used to be two competing mainframe packages for airlines, one was essentially written in Fortran. I don't know the other one and have fogotten it's name. afaik, people are now migrating away from the Fortran one towards Unix systems (?), but they ain't got thar yet. (it is not a conversion job, it is a complete rewrite)
2. Undefined behaviour for short-cutting logical ors. e.g. the behaviour of
if(flag.and.function(var)) then... is undefined if flag is false and function() has side effects.
You should be shot for trying something like that. In any language. An optimising compiler is supposed to evaluate 'flag' first.
If you want to pull a stunt like that, try c c *WARNING*: function has side-effects so we have to execute it! c
funcres = function (var)
if (flag.and. funcres) then
You assume everyone programs in C or it's derivatives, programming languages which terminate a string by some fixed character.
I have just finished a 4-year stint programming Cobol. Buffer overflows are possible in Cobol, but you really really have to try.
The new place I am working is a Fortran shop. For character arrays (which is where most buffer overflows happen), the function len(buffer) returns the number of bytes in the buffer - NOT the length of one byte. Assigning one string to another is done as:
target = source If 'source' is longer than 'target', the result is truncated. If 'target' is longer then the excess is space-filled. Those rules apply equally to Cobol, although Cobol's 'length' function only works for local buffers, not to passed parameters in a subroutine. You can do that for integer arrays (example) in Fortran as well, but I would have to look up what happens if they are not of the same size - especially for multi-dimension arrays. Probably the compiler would get abusive.
The way C handles strings was a poor original design decision and one of a number that language suffers from. Many years ago I knew Algol68, *that* was a nice language. The Bash scripting language designers obviously knew it as well.
I have not seen this newest book yet (has it even been released in Europe?) but found the previous one so-so. Could have been better, could have been worse.
The sequels which really unreservedly sucked were the last Dune books from Frank Herbert (Chapterhouse Dune is in a class of it's own) but I suppose there were mitigating circumstances: - his wife (?) had died of cancer - he himself was about to go the same way.
Conquer was correct, the nazis gained mastery of (most of) that territory before it was taken away from them by an external force.
That was the OffTopic bit, I agree with the rest of your posting.
That logic also works for the US Government. They made it clear that they wanted to bring Saddam Hussein down, it was their justification that was made up of half-truths and some outright lies.
The thing that amazes me is that Port 445 has apparently been left open. Switching over to my Firewall screen shows that I block a 445 scan every 10 seconds on average. It is not just one or two IP-Addresses which try it, each Source Address will try 3 times and then move on. Two machines a minute are saying 'Hello' on 445, 95% of my scans are on that Port and it has been left open. Sheesh.
The other unblocked Port where I often saw scans is 135, but the frequency there has dropped almost to zero recently.
That is the logic that had George Soros and Warren Buffett shorting the dollar. So far they have been wrong - probably because other economic zones are also in a slump - but in the long term, the $ will have to drop.
Then again, think John Maynard Keynes: 'In the long term, we are all dead'.
As to the third world countries competing: At the moment it is not happening. One of the reasons is that living to the WTO's rules make it for 'them' hard to stop expensive imports, while allowing 'us' to block 'their' exports. Another reason will be corruption, which is actively supported by banks in the developed world. Those Nigerian scams have a real-life background.
The old place was running NT4 until Q2 of 2002. They then upgraded to Win2K although they left Access at the old level because the new one was incompatible to at least one application. The conversion was accompanied by a rollout from Pentium 166 (it was awful!) to those tiny Compaq Celeron 850 PCs. New people later got HP P4 machines but I think they still ran Win2K.
The new place still uses NT4 but is about to move to XP. Most of the hardware will be replaced, although some of the newer stuff will simply be upgraded. I don't know what the new machines will be but the old ones were a mixture from around the Pentium-III era.
Large companies do not like having to support masses of users on a mixture of platforms if they can avoid it.
Most companies upgrade the SW and HW at the same time, according to my own anecdotal evidence based on a very small sample:-)
Seriously, there is no 'incentive to upgrade'. How many people replace one version of Windows by another? Maybe some do to get away from an unsupported version like Win95, or a version which does not support new peripherals like Win98, but not that many people upgrade. The market has moved on from the days where the choice was between Win 3.11 and Win95.
The market is in new systems. The natural (non-)decision is to get the newest version available, and that will eventually be Longhorn. Microsoft would save themselves a lot of money by not bothering to produce a new version of Windows. Then again, how long has WinXP been on sale? Maybe they are doing exactly that.
I use Mozilla and have been doing so since before they offered a filter. The filter hardly ever has a false positive, but it does miss about 30% of the spam I get - despite training.
Maybe 20 a day is not enough.
Re:Myth: IT Journalists Never Run Out Of Ideas...
on
IT Myths
·
· Score: 1
Bingo.
The critical stuff on which the business runs is usually on high-performance mainframes.
I used to work for a Telco. Their entire cable-network was mapped on our mainframe. Planning the use of existing cables or where to add/upgrade cables was all done there.
The other main application we had was billing of customers. Customers were split into 20 groups and they were each billed once a month. Normally one set of bills went out on any given day, occasionally we had days with zero or two (once even 3 after a major hardware problem) groups.
Maybe some of those Excel tables are important, but that stuff was essential.
Re:Server upgrades _do_ matter
on
IT Myths
·
· Score: 1
In the Intel/AMD world, you can change PCI cards or storage easily. Swapping a processor is far more difficult because the socket layout changes every few months and will be incompatable. Adding memory is somewhere in between. If your server runs a recent version of Windows, major changes may well be rejected anyway.
In the Intel/AMD world, it is usually cheaper to buy a new server than perform a major upgrade on an existing one. We all know that and I suspect the data behind 'Myth 1' - inasfar as it is not just anecdotal - comes from that area.
I work on mainframes and the situation is a bit different there. Processor speed is governed by a 'key tape'. If you want the processor(s) to run faster then you buy a faster tape which removes the software brakes deep in the OS. This costs serious money. Bypassing that coding is a violation of service/licensing agreements.
The advantage for the customer is that it takes seconds and does not require a reboot.
The advantage for the vendor is that they can supply one model for a large range of customers and the customers pay according to their needs.
The lifetime of one of these beasts is around 7-8 years so you will almost certainly want to change something.
The other obvious upgrade on the mainframes is that of storage, Disc or Tape. Virtually all of the upgrades I have been involved with were storage or Key Tape.
They're so top heavy its unbelievable. They charge more then even EDS does, and the service levels you get frankly leave something to be desired.
Unisys have three main markets. The old Univac/Sperry mainframe line, the old Burroughs line and these Intel servers. Two of those three are in decline. That is the reason for the top-heavy structure, the managers get rid of the people who were supporting the Univac or Burroughs lines (most just take early retirement, they are old enough) but they don't get rid of themselves.
I am one of the dinosaurs, but I work for a customer and not for Unisys themselves. The software and hardware are both reliable and the support people have being going down like flies for years.
The Khmer Rouge had legitimate greviances, but that does not mean that they were not evil.
The weird thing about them is that when they finally went too far and annoyed the Vietnamese, the Vietnamese swatted them aside and the US then started backing the Khmer Rouge.
I can understand bin Laden being against the Saudi Royal House, but their religous xenophobia and their methods disqualify them. The Provos only became a negotiating partner when they toned their act down and started looking for peace.
Slashdot Mirror
So what you are essentially saying is: this is not abuse of the DMCA - the DMCA is supposed to be used like this.
It is a large club which is available to be wielded by King Kong-like companies without fear of reprisal if they turn out to be lying.
Sounds about right - the campaign-spending issue means that those King Kongs own (as in the sense of having purchased) the lawmaking machinery.
You missed one in the original article (which you did read? "It stands in marked contrast to existing law and prior decisions that have determined that Congress was well within its constitutional authority to adopt legislation that prevented trafficking in copies of unauthorised performances of live music," spokesman Jonathan Lamy said.
Um - come again: "trafficking in copies of unauthorised performances of live music"
Who said those performances were unauthorised? By whom?
Sheesh.
The two problems are:
- the other 6 months (ok, less than that)
- the atmosphere in the 'good' 6 months. All that lovely snow to freeze the lens over.
Does this explain the Userfriendly theme this last week?
btw. Their names are Sarah and Henry.
Well, don't make it available to the world. You (especially now you have read about this problem) deserve all you get.
I think warning people is a good idea, not sure I'll take it up though.
You are correct, a 'Doppelkarte' is an insurance document you need to register a car in Germany. That website has to do with Car Insurance.
I have been programming both for years. The pay is a lot better than for Java / C / whatever, basically because of people like you who don't want to learn 'old technology'.
My problem is, you are half right. Eventually their use will die out. I reckon that I will have a big problem in about 10 years, but then again (quoting John Maynard Keynes): 'In the long run, we are all dead'.
Airlines.
NWA for example.
There used to be two competing mainframe packages for airlines, one was essentially written in Fortran. I don't know the other one and have fogotten it's name.
afaik, people are now migrating away from the Fortran one towards Unix systems (?), but they ain't got thar yet. (it is not a conversion job, it is a complete rewrite)
2. Undefined behaviour for short-cutting logical ors. e.g. the behaviour of
.and. funcres) then
if(flag.and.function(var)) then...
is undefined if flag is false and function() has side effects.
You should be shot for trying something like that. In any language. An optimising compiler is supposed to evaluate 'flag' first.
If you want to pull a stunt like that, try
c
c *WARNING*: function has side-effects so we have to execute it!
c
funcres = function (var)
if (flag
You assume everyone programs in C or it's derivatives, programming languages which terminate a string by some fixed character.
I have just finished a 4-year stint programming Cobol. Buffer overflows are possible in Cobol, but you really really have to try.
The new place I am working is a Fortran shop. For character arrays (which is where most buffer overflows happen), the function len(buffer) returns the number of bytes in the buffer - NOT the length of one byte. Assigning one string to another is done as:
target = source
If 'source' is longer than 'target', the result is truncated. If 'target' is longer then the excess is space-filled. Those rules apply equally to Cobol, although Cobol's 'length' function only works for local buffers, not to passed parameters in a subroutine.
You can do that for integer arrays (example) in Fortran as well, but I would have to look up what happens if they are not of the same size - especially for multi-dimension arrays. Probably the compiler would get abusive.
The way C handles strings was a poor original design decision and one of a number that language suffers from.
Many years ago I knew Algol68, *that* was a nice language. The Bash scripting language designers obviously knew it as well.
I have not seen this newest book yet (has it even been released in Europe?) but found the previous one so-so. Could have been better, could have been worse.
The sequels which really unreservedly sucked were the last Dune books from Frank Herbert (Chapterhouse Dune is in a class of it's own) but I suppose there were mitigating circumstances:
- his wife (?) had died of cancer
- he himself was about to go the same way.
Conquer was correct, the nazis gained mastery of (most of) that territory before it was taken away from them by an external force.
That was the OffTopic bit, I agree with the rest of your posting.
That logic also works for the US Government. They made it clear that they wanted to bring Saddam Hussein down, it was their justification that was made up of half-truths and some outright lies.
Still, at least it is an indication he knows what he is talking about (XP).
I see a case of Domain squatting coming up, backslashdot.com is not registered either.
Umm, I run Linux. I was just pointing out what my firewall is blocking and - imho - what XP SP2 should be addressing.
The thing that amazes me is that Port 445 has apparently been left open. Switching over to my Firewall screen shows that I block a 445 scan every 10 seconds on average. It is not just one or two IP-Addresses which try it, each Source Address will try 3 times and then move on.
Two machines a minute are saying 'Hello' on 445, 95% of my scans are on that Port and it has been left open. Sheesh.
The other unblocked Port where I often saw scans is 135, but the frequency there has dropped almost to zero recently.
That is the logic that had George Soros and Warren Buffett shorting the dollar. So far they have been wrong - probably because other economic zones are also in a slump - but in the long term, the $ will have to drop.
Then again, think John Maynard Keynes: 'In the long term, we are all dead'.
As to the third world countries competing: At the moment it is not happening.
One of the reasons is that living to the WTO's rules make it for 'them' hard to stop expensive imports, while allowing 'us' to block 'their' exports.
Another reason will be corruption, which is actively supported by banks in the developed world. Those Nigerian scams have a real-life background.
I just changed jobs.
:-)
The old place was running NT4 until Q2 of 2002. They then upgraded to Win2K although they left Access at the old level because the new one was incompatible to at least one application. The conversion was accompanied by a rollout from Pentium 166 (it was awful!) to those tiny Compaq Celeron 850 PCs. New people later got HP P4 machines but I think they still ran Win2K.
The new place still uses NT4 but is about to move to XP. Most of the hardware will be replaced, although some of the newer stuff will simply be upgraded. I don't know what the new machines will be but the old ones were a mixture from around the Pentium-III era.
Large companies do not like having to support masses of users on a mixture of platforms if they can avoid it.
Most companies upgrade the SW and HW at the same time, according to my own anecdotal evidence based on a very small sample
Win98SE? This is a Linux site :-)
Seriously, there is no 'incentive to upgrade'. How many people replace one version of Windows by another? Maybe some do to get away from an unsupported version like Win95, or a version which does not support new peripherals like Win98, but not that many people upgrade. The market has moved on from the days where the choice was between Win 3.11 and Win95.
The market is in new systems. The natural (non-)decision is to get the newest version available, and that will eventually be Longhorn. Microsoft would save themselves a lot of money by not bothering to produce a new version of Windows. Then again, how long has WinXP been on sale? Maybe they are doing exactly that.
The prophet has no honour in his own land? They are probably right.
:-)
Did you see this other link on that page? Sometimes Reading TFA has unexpected benefits
I use Mozilla and have been doing so since before they offered a filter. The filter hardly ever has a false positive, but it does miss about 30% of the spam I get - despite training.
Maybe 20 a day is not enough.
Bingo.
The critical stuff on which the business runs is usually on high-performance mainframes.
I used to work for a Telco. Their entire cable-network was mapped on our mainframe. Planning the use of existing cables or where to add/upgrade cables was all done there.
The other main application we had was billing of customers. Customers were split into 20 groups and they were each billed once a month. Normally one set of bills went out on any given day, occasionally we had days with zero or two (once even 3 after a major hardware problem) groups.
Maybe some of those Excel tables are important, but that stuff was essential.
In the Intel/AMD world, you can change PCI cards or storage easily. Swapping a processor is far more difficult because the socket layout changes every few months and will be incompatable. Adding memory is somewhere in between. If your server runs a recent version of Windows, major changes may well be rejected anyway.
In the Intel/AMD world, it is usually cheaper to buy a new server than perform a major upgrade on an existing one. We all know that and I suspect the data behind 'Myth 1' - inasfar as it is not just anecdotal - comes from that area.
I work on mainframes and the situation is a bit different there. Processor speed is governed by a 'key tape'. If you want the processor(s) to run faster then you buy a faster tape which removes the software brakes deep in the OS. This costs serious money. Bypassing that coding is a violation of service/licensing agreements.
The advantage for the customer is that it takes seconds and does not require a reboot.
The advantage for the vendor is that they can supply one model for a large range of customers and the customers pay according to their needs.
The lifetime of one of these beasts is around 7-8 years so you will almost certainly want to change something.
The other obvious upgrade on the mainframes is that of storage, Disc or Tape. Virtually all of the upgrades I have been involved with were storage or Key Tape.
My sister works for them. She is on holiday so I can't ask her but I thought they had cheapo contractors in to do the support anyway.
Newham is a cash-strapped E London borough where the Labour Party always have the majority on the council.
They're so top heavy its unbelievable. They charge more then even EDS does, and the service levels you get frankly leave something to be desired.
Unisys have three main markets. The old Univac/Sperry mainframe line, the old Burroughs line and these Intel servers. Two of those three are in decline. That is the reason for the top-heavy structure, the managers get rid of the people who were supporting the Univac or Burroughs lines (most just take early retirement, they are old enough) but they don't get rid of themselves.
I am one of the dinosaurs, but I work for a customer and not for Unisys themselves. The software and hardware are both reliable and the support people have being going down like flies for years.
Let's go back in time 25 years.
The Khmer Rouge had legitimate greviances, but that does not mean that they were not evil.
The weird thing about them is that when they finally went too far and annoyed the Vietnamese, the Vietnamese swatted them aside and the US then started backing the Khmer Rouge.
I can understand bin Laden being against the Saudi Royal House, but their religous xenophobia and their methods disqualify them.
The Provos only became a negotiating partner when they toned their act down and started looking for peace.