C++ and threading each introduce high complexity and difficulty, and their combination might even be worse.
I'm going to disagree. Traditionally attempting to do multi-threading in C or C++ is tricky, sure, because the traditional approach to sharing information between threads was to attempt to use shared structures and objects which inevitably introduce synchronisation issues. So don't share data between threads: send lightweight messages instead.
Personally I don't see what's so hard about traditional multi-threading either, but apparently a lot of developers are scared to death at the idea of having to even think about it.
The fact that 25million records were being sent via. post burnt on DVDs should give some idea of the level of technical competency in the public sector. Apparently they were being sent to the Audit Office, but why the Audit Office needed an off line copy of the data, and a complete copy at that, isn't addressed: no doubt some ridiculous bureaucratic idiocy that makes Brazil look sane.
The idea of burning an unencrypted copy of your sensitive data to a DVD and handing it to a random delivery company should horrify even the most incompetent sysadmin or DBA. Apparently no one in HM Customs & Revenue thought anything of it.
These are the sorts of people who want to build a massive database of all our personal details and tie them to ID cards. They tell us the data will be "perfectly safe". I wouldn't trust them to run a mail server.
It's the linear nature of most startup processes that cause the delay. On Syllable for example there is no delay waiting for DHCP, because the DHCP client is left to run on it's own without blocking the rest of the startup. Linux is beginning to move to a more parallel init system too.
That'd be nice if it were true, but the BBC web sites are almost all excellent in any browser you'd care to mention. Even the news ticker on the news.bbc.co.uk front page works in ABrowse on Syllable (Webcore based). I think you'd have to try pretty hard to find something that is IE only through design on the BBC web sites.
In reply to both of the above posters, my attitude that this update has been handled badly. Sometimes things happen which don't fit into the carefully scheduled world you want to create. It's not much use to say that the update will be pushed with the next release, because the update is time-sensitive: NZ switched to DST today, not in two weeks time. It's also all well and good that Debian have a "Volatile" repository, but why is not enabled by default?
I'd question any system distributor who claims that something like a timezone change doesn't need to be pushed out-of-sync with the release cycle. Users quite rightly expect their machines to know what the time is, and if they rely on you to help handle that process, it's your responsibility.
As a person who works with Glibc outside of FreeBSD & Debian, I say: care to explain? Where's the issue with pushing a new TZ database? Obviously pushing it as a "Security Update" is not correct, but other than that I fail to see the issue.
I had a machine running WindowsME for over two years with no more trouble than I have had since I installed WindowsXP in it's place. It was fairly painless, in fact.
I'd like to point out that it's attitudes like yours that have given us the politicians of today who will never, ever, admit to be being wrong. It's like changing your mind, even in the face of new evidence, has become some sort of moral weakness.
Changing your mind and admitting your were wrong is healthy and normal, and we need more people to do it more often.
C++ has had http://www.codesourcery.com/cxx-abi/">a standard ABI for nearly six years now. GCC 3.0 was the first version of g++ to ship with support for the new ABI on IA-64, which was released in June 2001. The first version of GCC to support the cross vendor C++ ABI on other platforms, including IA-32, was GCC 3.2, which was released in April 2003.
it's very hard to make [C++] build across all of the dozens of "standard library" implementations.
In the Open Source world there is exactly one C++ standard library to worry about, and that's the one that ships with GCC. I'm going to go out on a limb here and guess that as Sun use GCC with Solaris, even that ships with libstc++ from GCC.
There is no need to fork Linux into a "desktop" version. Projects like Syllable already exist, and we re-use a fair amount of code from Linux, GNU and other OSS projects.
Yes, you replied just as I was posting myself. Anyway:
Excel: You get a column of cells, in my case "Five, One, Two, Three, Four, Six"
Calc: You get a column of cells, in my case "Five, Two, Three, Four, [Blank], Six"
No surprises there though. In Excel you select "Insert Cut Cells", in Calc you select "Paste". I wonder what happens if you select "Paste" in Excel instead of "Insert Cut Cells"?
You get a column of cells, in my case "Five, Two, Three, Four, [Blank], Six"
Another non-surprise. If you select the same function in both Excel and Calc, you get the same result.
Your complaint is no more than "The context menu in Calc does not contain an 'Insert Cut Cells' item".
"I cut and paste entire columns on a regular basis and Excel actually removes those columns and pastes the columns where I tell it to. Calc on the other hand simply cuts the data out and places the data where I tell it to, overwriting whatever data was originally there."
So let's just get this straight:
Excel: Removes ("Cuts") the column and then pastes ("Places") it "where I tell it to"
Calc: Cuts ("Removes") the data and places ("Pastes") it "where I tell it to"
Man, that's shocking: both Excel and Calc do the same thing. I can see how that might be confusing.
..couldn't it send some kind of error, indicating that the MS client would have to try a simpler version of the protocol?
No. DHCP is not a complex protocol. All the server can do is send back a DHCPNAK response to the client, or send nothing (in which case the client will timeout).
It's nothing to do with Ubuntu. It's your general attitude that you're better than other users because you've chosen a "better" distribution. If you'd taken the most cursory look at the link in my signature you'd now I don't particularly care for any Linux distributions.
Let's neither of us pretend you were joking. Attitudes like yours are posted on Slashdot all the time. People actually believe that sort of rubbish.
Quite a lot. A lot of the other developers who work with me run Ubuntu, too. Before I switched to Ubuntu, I ran Mandrake.
Now I'm not saying that Ubuntu is "leet" or something, but silly rubbish like "Ubuntu is for n00bs LOLROTFLCOPTER!!!1111!!eleven!" is demonstratively wrong. Just run whatever you like and be happy with your choice, but your choice doesn't elevate you above other users.
Unless you choose Syllable of course. The chicks love that.
I never really understood why Commodore chose to use the (by then 'obsolete') 6502 for their computers
Because Commodore owned MOS Technology, along with the designers and fabrication facilities to make all their own chips. They didn't need to pay anyone for the 6510's they were putting into all of their hardware; during the high-point of Commodore they litterly cost them a couple of cents each.
The real question is how Commodore failed to improve on the basic 6502 design in any significant way. The short answer is of course "Bad management", but then this is Commodore we're talking about.
Slashdot Mirror
C++ and threading each introduce high complexity and difficulty, and their combination might even be worse.
I'm going to disagree. Traditionally attempting to do multi-threading in C or C++ is tricky, sure, because the traditional approach to sharing information between threads was to attempt to use shared structures and objects which inevitably introduce synchronisation issues. So don't share data between threads: send lightweight messages instead.
Personally I don't see what's so hard about traditional multi-threading either, but apparently a lot of developers are scared to death at the idea of having to even think about it.
The fact that 25million records were being sent via. post burnt on DVDs should give some idea of the level of technical competency in the public sector. Apparently they were being sent to the Audit Office, but why the Audit Office needed an off line copy of the data, and a complete copy at that, isn't addressed: no doubt some ridiculous bureaucratic idiocy that makes Brazil look sane.
The idea of burning an unencrypted copy of your sensitive data to a DVD and handing it to a random delivery company should horrify even the most incompetent sysadmin or DBA. Apparently no one in HM Customs & Revenue thought anything of it.
These are the sorts of people who want to build a massive database of all our personal details and tie them to ID cards. They tell us the data will be "perfectly safe". I wouldn't trust them to run a mail server.
My brane hurts.
It's the linear nature of most startup processes that cause the delay. On Syllable for example there is no delay waiting for DHCP, because the DHCP client is left to run on it's own without blocking the rest of the startup. Linux is beginning to move to a more parallel init system too.
That'd be nice if it were true, but the BBC web sites are almost all excellent in any browser you'd care to mention. Even the news ticker on the news.bbc.co.uk front page works in ABrowse on Syllable (Webcore based). I think you'd have to try pretty hard to find something that is IE only through design on the BBC web sites.
Sporks, the Pancake Eating Ninjas, Trolltalk...
In reply to both of the above posters, my attitude that this update has been handled badly. Sometimes things happen which don't fit into the carefully scheduled world you want to create. It's not much use to say that the update will be pushed with the next release, because the update is time-sensitive: NZ switched to DST today, not in two weeks time. It's also all well and good that Debian have a "Volatile" repository, but why is not enabled by default?
I'd question any system distributor who claims that something like a timezone change doesn't need to be pushed out-of-sync with the release cycle. Users quite rightly expect their machines to know what the time is, and if they rely on you to help handle that process, it's your responsibility.
As a person who works with Glibc outside of FreeBSD & Debian, I say: care to explain? Where's the issue with pushing a new TZ database? Obviously pushing it as a "Security Update" is not correct, but other than that I fail to see the issue.
I had a machine running WindowsME for over two years with no more trouble than I have had since I installed WindowsXP in it's place. It was fairly painless, in fact.
Once you've read this, read Where Wizards Stay Up Late. Then read Hackers.
I'd like to point out that it's attitudes like yours that have given us the politicians of today who will never, ever, admit to be being wrong. It's like changing your mind, even in the face of new evidence, has become some sort of moral weakness.
Changing your mind and admitting your were wrong is healthy and normal, and we need more people to do it more often.
No standard ABI
C++ has had http://www.codesourcery.com/cxx-abi/">a standard ABI for nearly six years now. GCC 3.0 was the first version of g++ to ship with support for the new ABI on IA-64, which was released in June 2001. The first version of GCC to support the cross vendor C++ ABI on other platforms, including IA-32, was GCC 3.2, which was released in April 2003.
it's very hard to make [C++] build across all of the dozens of "standard library" implementations.
In the Open Source world there is exactly one C++ standard library to worry about, and that's the one that ships with GCC. I'm going to go out on a limb here and guess that as Sun use GCC with Solaris, even that ships with libstc++ from GCC.
Have you looked at projects such as Syllable?
There is no need to fork Linux into a "desktop" version. Projects like Syllable already exist, and we re-use a fair amount of code from Linux, GNU and other OSS projects.
No surprises there though. In Excel you select "Insert Cut Cells", in Calc you select "Paste". I wonder what happens if you select "Paste" in Excel instead of "Insert Cut Cells"?
- You get a column of cells, in my case "Five, Two, Three, Four, [Blank], Six"
Another non-surprise. If you select the same function in both Excel and Calc, you get the same result.Your complaint is no more than "The context menu in Calc does not contain an 'Insert Cut Cells' item".
"I cut and paste entire columns on a regular basis and Excel actually removes those columns and pastes the columns where I tell it to. Calc on the other hand simply cuts the data out and places the data where I tell it to, overwriting whatever data was originally there."
So let's just get this straight:
- Excel: Removes ("Cuts") the column and then pastes ("Places") it "where I tell it to"
- Calc: Cuts ("Removes") the data and places ("Pastes") it "where I tell it to"
Man, that's shocking: both Excel and Calc do the same thing. I can see how that might be confusing.Is there a particular reason you need the driver to be open-source?
Yes
Unless you are a driver developer yourself..
I am.
..couldn't it send some kind of error, indicating that the MS client would have to try a simpler version of the protocol?
No. DHCP is not a complex protocol. All the server can do is send back a DHCPNAK response to the client, or send nothing (in which case the client will timeout).
Yes, you've hit the nail on the head. This is exactly why I gave up and moved to Ubuntu instead of upgrading from Mandrake 10.1.
It's nothing to do with Ubuntu. It's your general attitude that you're better than other users because you've chosen a "better" distribution. If you'd taken the most cursory look at the link in my signature you'd now I don't particularly care for any Linux distributions.
Let's neither of us pretend you were joking. Attitudes like yours are posted on Slashdot all the time. People actually believe that sort of rubbish.
What did you write?
Quite a lot. A lot of the other developers who work with me run Ubuntu, too. Before I switched to Ubuntu, I ran Mandrake.
Now I'm not saying that Ubuntu is "leet" or something, but silly rubbish like "Ubuntu is for n00bs LOLROTFLCOPTER!!!1111!!eleven!" is demonstratively wrong. Just run whatever you like and be happy with your choice, but your choice doesn't elevate you above other users.
Unless you choose Syllable of course. The chicks love that.
A binary Linux driver doesn't help much if you're not running Linux.
It's nice to see someone who actually understands the problem. I've never seen it described so eloquently, either.
I never really understood why Commodore chose to use the (by then 'obsolete') 6502 for their computers
Because Commodore owned MOS Technology, along with the designers and fabrication facilities to make all their own chips. They didn't need to pay anyone for the 6510's they were putting into all of their hardware; during the high-point of Commodore they litterly cost them a couple of cents each.
The real question is how Commodore failed to improve on the basic 6502 design in any significant way. The short answer is of course "Bad management", but then this is Commodore we're talking about.