I found this list to be particularly interesting reading. I happened to pick PostgreSQL rather than MySQL when I needed a DB; this list made me feel very glad that I did.
My point is that "originally closed source" is irrelevant. It wasn't at some point, and Linus made that decision. When it became open source a licence was chose and Linux made that decision. They were NOT technical decisions so the argument that Linus never makes any decisions that aren't on technical merit is wrong.
If that were true, he would never have made Linux. Let's not go mad and pretend that Linus has no political leanings. He absolutely does. If he didn't explain the following:
Starting Linux when minix was about
Choosing the GPL as a licence
Both these decisions show that there was more to Linux than mere technical merit. It has been argued (and I think successfully) that Linux would not exist in the form it does now had it not been for the selection of the GPL as its licence. Whether Linus has strong political leanings or not, many of the other kernel developers do; without some form of politicisim then, large portions of the kernel would not exist.
If (and that's a huge if) Microsoft were right that it's only a few hundred people working on Linux / distribution / favourite app of your choice, then they should be thoroughly ashamed of themselves. Linux is serious competition to Windows in every space. They are having their lunch eaten for them.
If they want to say that there multi-thousand employee empire is being kicked about by a few geeks with spare time at the weekend, then who are we to argue?
Regardless of what the nit-pickers say, I think this is quite a clever idea. The author isn't suggesting this is the best HA solution in the world; but it's certainly simple and effective.
Object relational databases allow you to create systesm with table inheritance. For example:
People
/ \ Employees Customers
| Managers
Now, every record in manager is automatically an employee and every employee is automatically a person. References from other tables to the People table will include all employees and customers.
Additionally, the Employees table might have extra fields - salary and employee number and when viewed as an employee, these fields would show.
Having said all that; I've never been able to find a really good use for an Object Relational database...
IF there were some key genetic difference between men and women with regards to maths. This one datapoint you supply tells us nothing. There is no way to compare your friend emily with her equivlent male version. Who can say that Emaley wouldn't have gotten 40 ACT; 2000 SAT and 1st place at the intel science talent search?
It would be equally impossible to run the equivalent but necessarily different education and raising of Emaley and Emily so that nurture could be ruled out as a factor.
In short: is it possible that men and women are genetically pre-disposed to be better/worse/different at maths - of course it is. Is it possible to prove it? Almost certainly not.
Given that, I'd suggest that it would be better to try to correct the wooly minded thinking that goes on in both genders that lead people to say such illogical things as "I would go so far as to say she is not genetically pre-disposed to being bad at mathematics."
Depends -- the trade you are making is start up time, for performance once running. It's not unlike the trade between bandwidth and latency. A truck full of DVD's will beat anything in bandwidth terms; but it's no use to you.
On the other hand, if the human race had another planet, would it not become that little bit easier for the bigwigs to justify blowing the other one up when things get tricky.
The only thing that stops the use of big giant bombs is that it destroys everybody's planet. When it isn't everybody's planet, that problem goes away.
On an asteroid? WTF? How exactly are these examples of life you speak of hopping a ride on an asteroid? "Hello Microbe Houston, we have a problem".
Your point isn't wrong however, as was mentioned in another thread, life could easily have been carried from Earth to Mars on one of our own probes. Given the variety and ingenuity of life; I would have thought it is more than likely.
I mean, I see your point; but just saying "we're going to want this planet.." doesn't really address the parent's orginal point -- once we've got somewhere else to live, why would we need this planet? Apart from sentimental value, of course, there is no reason. And I wouldn't say that careless is the right word, it wouldn't be done without thought. Mercenary, uncaring, arrogant are all reasonable descriptions for that attitude. But none of them is absolutely "wrong" per se. I think you need a better argument than "just because"
I don't agree, it was certainly buggy, it certainly crashed. But the interface was second to none. All the formatting dialogs where in one modeless window. When you deviated from style the item got a small red dot next to it. It successfully distinguished between styles in the template, styles in the document and styles in the paragraph and made moving them about very easy. All WordPro needed was some bug fixes and it would have been perfect. At the time, MS Word was only slightly more reliable - certainly not to an acceptable level.
WordPro should have won that battle IMHO; I felt the weakness was the awful Lotus-1-2-3 that was in SmartSuite. My goodness that was an apalling application. I think MS Office was saved by Excel, not by Word.
On the plus side, I feel that the places where OpenOffice has deviated from Word in their interface choices are similar to the choices that WordPro made. It's all going swimmingly.
I think the idea of single sign on is a good one. The problem is, it shouldn't be implemented on the server side. KDE's new KWallet system is a very good example of how this should work - I keep all my logons locally, encrypted, and in a trusted place - my privacy is not at any more risk than it ever was. Now, I single sign on to the KWallet system which is then used by konqueror/kopete/kmail/whatever to auto-logon whereever i go.
With a little bit of support server side (perhaps a standard way of passing logon information to HTTP servers - if the existing method is not deemed good enough) this could easily fake the entire passport system with no need for any centralised server.
This is a great argument - except for one thing. Why can't they release all of their code? If there is somebody else's code in there that has to be binary only why can't THOSE be the parts that are binary only? A couple of ".o" files in a mass of ".c" files named "sgi.o", "ms.o", "sun.o", etc to hide the non-disclosable binary bits. At the very least the bugs in the rest of it could be hunted down and squashed. It's just an excuse, they don't want to release it. I'd rather they said so than blame it on some other company.
That would be such a nice theory if all the thirds were equal. They aren't. You can't stick the two front thirds of a note together and claim you've got 2/3 of a whole one... Actually I'm wrong there: you could claim that.
I'd like to make an analogy (despite it being the weakest form of argument) to the concepts of power and energy from phsyics (although the same is true for many physical processes). As I'm sure everyone here knows: Energy in itself is not a lot of use; it only becomes useful when something is done with it, in the case of energy that can only be the changing of the energy from one form to another. i.e the flow of energy is the important thing (power being used to measure that)...
Similarly with society: to a taxed economy, the total amount of cash available is less important than the amount of flow of cash - it is the flow that is taxed, and hence allows governments to do their (supposed) good works. Equally it is the flow of cash that causes anything to be done. (I build you a fence, you mend my car; if the cash exchanged is the same then nothing has changed other than we now have one fixed fence and one mended car)
I think the same is going to become true of software. I have maintained for a long time that if the only thing you have that makes you valuable is your source code, then you are doomed. It is the ability to create the source code that has value; otherwise when something new is needed, there is no way to make it.
If the idea of free software takes off, the software industry won't die, it will become like the legal profession (yuck;-)): the owning of the books is not what gives a lawyer their value, it is their ability to use those books. The owning of source code will be unimportant, every company will find it useful to maintain an programmer's department in the same way that they find it useful to maintain an IT department.
I'm sure the guy is a good coder as you say. However, the first two items in the blog (to me) demonstrate why Windows is so awful:
"What happens if you pass GetDesktopWindow()? If UI does indeed need to be displayed, you hang the system." - why on earth would you build the ability to hang your system into your window manager; worse, with Windows being GUI only, I assume there is no way to recover from this.
"Nothing is explicitly written about this topic, but you can put on your logic cap and figure it out." - why is nothing explicitly written about this? I've found the man pages for Linux system calls have always told me about things like this. Even if it didn't I could always go and look at the source. With Windows - tough luck, using our "logic cap" seems bad to me as without it being explicitly written down, and without seeing the source you can never know if this is just a best guess that will change in the future.
You rightly point out the problems Linux has with backward compatibility; however these breakages are done in the name of improvement, a year from now it won't matter about NPTL and TLS. To those few to whom it does matter, they still have the ultimate recourse - the source, with no forced upgrades.
Slashdot Mirror
Fair enough. You are entirely correct (as is the parent, now I read it how you have suggested). I withdraw my remarks.
Nonsense. Citation is a tool to let the reader know when you are using work that is not your own. You can therefore cite anything you want.
I found this list to be particularly interesting reading. I happened to pick PostgreSQL rather than MySQL when I needed a DB; this list made me feel very glad that I did.
MySQL Gotchas
To remain fair, there is a similar list for postgresql, but to me they seem like far less serious problems.
PostgreSQL Gotchas
My point is that "originally closed source" is irrelevant. It wasn't at some point, and Linus made that decision. When it became open source a licence was chose and Linux made that decision. They were NOT technical decisions so the argument that Linus never makes any decisions that aren't on technical merit is wrong.
Both these decisions show that there was more to Linux than mere technical merit. It has been argued (and I think successfully) that Linux would not exist in the form it does now had it not been for the selection of the GPL as its licence. Whether Linus has strong political leanings or not, many of the other kernel developers do; without some form of politicisim then, large portions of the kernel would not exist.
Only a guess, but I imagine that the submitter meant "proof of the opposite" by contra-proof. Which is obviously not the same as "not proof".
If (and that's a huge if) Microsoft were right that it's only a few hundred people working on Linux / distribution / favourite app of your choice, then they should be thoroughly ashamed of themselves. Linux is serious competition to Windows in every space. They are having their lunch eaten for them.
If they want to say that there multi-thousand employee empire is being kicked about by a few geeks with spare time at the weekend, then who are we to argue?
Regardless of what the nit-pickers say, I think this is quite a clever idea. The author isn't suggesting this is the best HA solution in the world; but it's certainly simple and effective.
Additionally, the Employees table might have extra fields - salary and employee number and when viewed as an employee, these fields would show.
Having said all that; I've never been able to find a really good use for an Object Relational database...
What nonsense.
IF there were some key genetic difference between men and women with regards to maths. This one datapoint you supply tells us nothing. There is no way to compare your friend emily with her equivlent male version. Who can say that Emaley wouldn't have gotten 40 ACT; 2000 SAT and 1st place at the intel science talent search?
It would be equally impossible to run the equivalent but necessarily different education and raising of Emaley and Emily so that nurture could be ruled out as a factor.
In short: is it possible that men and women are genetically pre-disposed to be better/worse/different at maths - of course it is. Is it possible to prove it? Almost certainly not.
Given that, I'd suggest that it would be better to try to correct the wooly minded thinking that goes on in both genders that lead people to say such illogical things as "I would go so far as to say she is not genetically pre-disposed to being bad at mathematics."
Who'd a thought it?
Unless x=0. Ooops.
e.g. sin(x)/x is not infinity for x=0
But you're right of course, that limits let you find what the answer really is.
Depends -- the trade you are making is start up time, for performance once running. It's not unlike the trade between bandwidth and latency. A truck full of DVD's will beat anything in bandwidth terms; but it's no use to you.
On the other hand, if the human race had another planet, would it not become that little bit easier for the bigwigs to justify blowing the other one up when things get tricky.
The only thing that stops the use of big giant bombs is that it destroys everybody's planet. When it isn't everybody's planet, that problem goes away.
I think you underestimate the sheer pigheadedness of the human race.
On an asteroid? WTF? How exactly are these examples of life you speak of hopping a ride on an asteroid? "Hello Microbe Houston, we have a problem".
Your point isn't wrong however, as was mentioned in another thread, life could easily have been carried from Earth to Mars on one of our own probes. Given the variety and ingenuity of life; I would have thought it is more than likely.
Why?
I mean, I see your point; but just saying "we're going to want this planet.." doesn't really address the parent's orginal point -- once we've got somewhere else to live, why would we need this planet? Apart from sentimental value, of course, there is no reason. And I wouldn't say that careless is the right word, it wouldn't be done without thought. Mercenary, uncaring, arrogant are all reasonable descriptions for that attitude. But none of them is absolutely "wrong" per se. I think you need a better argument than "just because"
I don't agree, it was certainly buggy, it certainly crashed. But the interface was second to none. All the formatting dialogs where in one modeless window. When you deviated from style the item got a small red dot next to it. It successfully distinguished between styles in the template, styles in the document and styles in the paragraph and made moving them about very easy. All WordPro needed was some bug fixes and it would have been perfect. At the time, MS Word was only slightly more reliable - certainly not to an acceptable level.
WordPro should have won that battle IMHO; I felt the weakness was the awful Lotus-1-2-3 that was in SmartSuite. My goodness that was an apalling application. I think MS Office was saved by Excel, not by Word.
On the plus side, I feel that the places where OpenOffice has deviated from Word in their interface choices are similar to the choices that WordPro made. It's all going swimmingly.
I think the idea of single sign on is a good one. The problem is, it shouldn't be implemented on the server side. KDE's new KWallet system is a very good example of how this should work - I keep all my logons locally, encrypted, and in a trusted place - my privacy is not at any more risk than it ever was. Now, I single sign on to the KWallet system which is then used by konqueror/kopete/kmail/whatever to auto-logon whereever i go.
With a little bit of support server side (perhaps a standard way of passing logon information to HTTP servers - if the existing method is not deemed good enough) this could easily fake the entire passport system with no need for any centralised server.
This is a great argument - except for one thing. Why can't they release all of their code? If there is somebody else's code in there that has to be binary only why can't THOSE be the parts that are binary only? A couple of ".o" files in a mass of ".c" files named "sgi.o", "ms.o", "sun.o", etc to hide the non-disclosable binary bits. At the very least the bugs in the rest of it could be hunted down and squashed. It's just an excuse, they don't want to release it. I'd rather they said so than blame it on some other company.
Here's a suggestion: learn to code HTML without use of a nice IDE. Sure it's harder, but it is free.
WHAT? Am I missing something - surely stereo is two mono channels to start with?
That would be such a nice theory if all the thirds were equal. They aren't. You can't stick the two front thirds of a note together and claim you've got 2/3 of a whole one... Actually I'm wrong there: you could claim that.
I'd like to make an analogy (despite it being the weakest form of argument) to the concepts of power and energy from phsyics (although the same is true for many physical processes). As I'm sure everyone here knows: Energy in itself is not a lot of use; it only becomes useful when something is done with it, in the case of energy that can only be the changing of the energy from one form to another. i.e the flow of energy is the important thing (power being used to measure that)...
;-)): the owning of the books is not what gives a lawyer their value, it is their ability to use those books. The owning of source code will be unimportant, every company will find it useful to maintain an programmer's department in the same way that they find it useful to maintain an IT department.
Similarly with society: to a taxed economy, the total amount of cash available is less important than the amount of flow of cash - it is the flow that is taxed, and hence allows governments to do their (supposed) good works. Equally it is the flow of cash that causes anything to be done. (I build you a fence, you mend my car; if the cash exchanged is the same then nothing has changed other than we now have one fixed fence and one mended car)
I think the same is going to become true of software. I have maintained for a long time that if the only thing you have that makes you valuable is your source code, then you are doomed. It is the ability to create the source code that has value; otherwise when something new is needed, there is no way to make it.
If the idea of free software takes off, the software industry won't die, it will become like the legal profession (yuck
I'm sure the guy is a good coder as you say. However, the first two items in the blog (to me) demonstrate why Windows is so awful:
"What happens if you pass GetDesktopWindow()? If UI does indeed need to be displayed, you hang the system." - why on earth would you build the ability to hang your system into your window manager; worse, with Windows being GUI only, I assume there is no way to recover from this.
"Nothing is explicitly written about this topic, but you can put on your logic cap and figure it out." - why is nothing explicitly written about this? I've found the man pages for Linux system calls have always told me about things like this. Even if it didn't I could always go and look at the source. With Windows - tough luck, using our "logic cap" seems bad to me as without it being explicitly written down, and without seeing the source you can never know if this is just a best guess that will change in the future.
You rightly point out the problems Linux has with backward compatibility; however these breakages are done in the name of improvement, a year from now it won't matter about NPTL and TLS. To those few to whom it does matter, they still have the ultimate recourse - the source, with no forced upgrades.