I'm one of those torn citizens of the UK wedged between crazy euro farm subsidies and the US capitalist pigs;) I like to think a balanced a approach is possible.
I don't think it is a good idea to subsidise a loss making aspect of a country or a company indefinately but obviously the pros and cons should be weighed up and people should be offered alternative roles and retrained where at all feasible.
If for economic reasons IBM is outsourcing jobs to india they should look at what the european work force can offer that india cannot and consider retraining where possible as this will be cheaper than re-recruiting over the next few years. Also in some european countries legal action could be taken if IBM finds themselves trying re-recruit for the same jobs at lower salary levels.
Ultimately if these people don't have suitable skills for the region in which they work then they should retrain. If they stay at IBM and IBM doesn't offer retraining then they get fired in 5 years time they will be in an even worse situation as IBM will have sheltered them from the realities of the job market.
Although it seems incredibly harsh there may be some morale arguments in favour of a large cut if they fear the alternative is gradual cuts over several years. No one works well when they think their job is on the line every month.
I often wonder in these massive union/company disputes - does the company ever approach the unions *first* to discuss these kind of issues?
You mean GPL'd software in closed source products. Open source = many licenses (including LGPL, MIT and BSD which can be used in closed source products.)
The message i originally replied to was talking about cross-platform C. I was just pointing out that if you have a C program and a java program and you want to move them to a different OS or hardware platform the java program can be moved easily and the C can't.
The C program will probably have to be rewritten to make use of some cross-platorm library and then you'll probably find you don't have all the libraries you need. Almost by definition a java program will run on any underlying hardware/OS platform.
I don't disagree about the JVM and marketting issues but that is a whole separate issue as would be portability of c code across C compilers for example.
sure jvms are "A" platform but in this context I was talking about different hardware architecture platforms in that send running the sun JVM across all the hardware platforms actually gives you remarkably good portability these days.
I have, and actually i found it ported well from windows to linux and solaris.
The only thing is the paths on windows as people don't usually use cross platform or relative paths.
The only time i had issues was in the early days of java with applets. But my experience of moving large web applications is that it is relatively easy. In fact i've had more problems switching JVMs within a platform than switching platforms - one thing that got me was some people had started to rely on the collection ordering which happened on the SUN JVM to always stay in the same order but on the IBM one recreating the same collection did not maintain ordering, which was fair enough there were no guarantees in the specs.
On the other hand back in the days i wrote some J++ code for windows my god it was the WORST i have EVER experienced, just getting it to run on all versions of windows took weeks not hours. detecting network connections without triggering a RAS dialup across 95->2k was hilariously difficult.
But if it was a choice between one police man on one out of every 1000 bus trips or a CCTV on every bus, i'd focus on CCTV first. Cos you can be damn sure i won't be robbed on that one lucky bus trip with the cop.
Like on the bus going through peckham when groups of youths from the local youth detention center "enquired" about what type of phone i had. Also outside clubs where fights regularly break out.
I'm not saying we don't need more police or anything like that and i hate the whole ID card scheme but I actually quite like CCTV and brighter lights.
Although, some kind of merit award in being able to recognise detrimental international monopolies and illegal price fixing groups, would also be useful in my book!
And some kind of award for recognising commercial interest lobby groups, dodgy legislation and for uncovering widespread IP extortion and blackmail by said groups.
The information that was classified in the original document and is now leaked looks like it was classified for fair enough reasons.
The fact that it is being leaked by the Italian press conjures up images in my mind of hidden details that were being supressed by the americans - however I don't see this to be the case.
In fact if I was an insurgent with internet access some of the leaked classified details could be quite useful. Some describe how the US deal with incidents and their knowledge of terrorists tactics and specific details about what battalions are trained in and exactly where they are located.
Whilst obviously the major error was in the americans releasing a document with hidden meta-data, I do think it is quite irresponsible of the Italian press to advertise the release of this hidden data, thus further endagering the lives of people in Iraq.
I actually find some logic to this. As long as everyone buys in to the concept. It's almost always better to refactor some code or pull it out into a method.
Most comments i see is for code thats too complicated to understand straight off and almost always it's more complicated than it needs to be, rather than fix it some developer just puts a comment at the top along the lines of:
/* Divide the doohicky index by the number of submaxatrons and use the result as a lookup value for the wibble hashtable, assume null results are bananas - PS not entirely sure why */
I always find good OO code relatively easy to understand and just ignore reading the comments, which somtimes are out of date - if I'm not entirely sure what the purpose of the code is I way prefer unit tests to comments as I know the units tests are more likely to be correct - if they still pass.
Of course none of this applies to perl or regex where i need all the comments I can find.
That is not actually correct. What in fact the RIAA can do and probably are doing is lobbying in europe.
No one in the UK pays any attention to what happens in europe but eventually either the law in the UK is changed to bring it in line with European regulations or there is a european test case and Europe starts fining the UK goverment telling it that it should do what it is told or face the consequences.
Perhaps we could just fork nautilus. That way you can use spatial nautilus and the rest of us can stop wasting time navigating round hundreds of randomly located windows.
I'm not convinced it's X rendering that makes linux distros seems slower than windows. Applications and windows usually take longer to appear. For example gedit takes nearly as long as word to load on my PC and is about 10times longer than notepad.
After massive improvements and work to nautilus is about launches at a reasonable speed if you aren't using the file explorer version (which of course most people do once they've worked out how to re-enable it!). Maybe it's got something to do with library caching/toolkits being preloaded on windows, but whatever it is linux needs it.
This doesn't really matter. For ad networks using CPA (cost-per-acquisition) rather than CPC the tracking is done using cookies which will record a sign up or purchase even on repeat visits.
Although CPA based advertising is a lot less prone to fraud there are issues of which the biggest is in deploying pixels to do tracking on all sign up/purchase pages there are also more fixable issues with double counting (crediting to more than one ad network). CPA may be vulnerable to publisher fraud of course though this is usually less of an issue as you have a contractual relationship with them.
CPC click fraud was a problem long before google starting selling space on a CPC basis. I believe commission junction used to have a whole fraud detection team (they may still do).
I agree it sounds mainly Larry's fault. But I kind of understand why Torvalds got so cross, he's only human and I'm sure he must realise that he has himself triggered the chain of events by switching to bitkeeper in the first place.
When making a decision about which are the best tools to use for development you have to look at the "development process" who's involved and what people are likely to "buy into" this is part of the challenge of managing multi person projects. I reckon Torvalds should read more of Dilberts management hints:
Boss: You need to socialize your idea with the rest of your group. Dilbert: Socialize? Is that the same as getting buy-in? Boss: It's one step below buy-in. Its more like dialoging for feedback. Dilbert: Wait...I thought that building a consensus was one step below buy-in. Boss: Just run it up the flagpole and see who salutes. Dilbert: Wouldn't it be better to do a temperature check with a straw man? Boss: Maybe...but is that going to inoculate the stakeholders?
Slashdot Mirror
I'm one of those torn citizens of the UK wedged between crazy euro farm subsidies and the US capitalist pigs ;) I like to think a balanced a approach is possible.
I don't think it is a good idea to subsidise a loss making aspect of a country or a company indefinately but obviously the pros and cons should be weighed up and people should be offered alternative roles and retrained where at all feasible.
If for economic reasons IBM is outsourcing jobs to india they should look at what the european work force can offer that india cannot and consider retraining where possible as this will be cheaper than re-recruiting over the next few years. Also in some european countries legal action could be taken if IBM finds themselves trying re-recruit for the same jobs at lower salary levels.
Ultimately if these people don't have suitable skills for the region in which they work then they should retrain. If they stay at IBM and IBM doesn't offer retraining then they get fired in 5 years time they will be in an even worse situation as IBM will have sheltered them from the realities of the job market.
Although it seems incredibly harsh there may be some morale arguments in favour of a large cut if they fear the alternative is gradual cuts over several years. No one works well when they think their job is on the line every month.
I often wonder in these massive union/company disputes - does the company ever approach the unions *first* to discuss these kind of issues?
You mean GPL'd software in closed source products. Open source = many licenses (including LGPL, MIT and BSD which can be used in closed source products.)
possibly because there are so many successfull apache jakarta projects.
still doesn't warrant posting it twice, zonk is obviously having a bad day.
The message i originally replied to was talking about cross-platform C. I was just pointing out that if you have a C program and a java program and you want to move them to a different OS or hardware platform the java program can be moved easily and the C can't.
The C program will probably have to be rewritten to make use of some cross-platorm library and then you'll probably find you don't have all the libraries you need. Almost by definition a java program will run on any underlying hardware/OS platform.
I don't disagree about the JVM and marketting issues but that is a whole separate issue as would be portability of c code across C compilers for example.
sure jvms are "A" platform but in this context I was talking about different hardware architecture platforms in that send running the sun JVM across all the hardware platforms actually gives you remarkably good portability these days.
Atleast in my experience.
I have, and actually i found it ported well from windows to linux and solaris.
The only thing is the paths on windows as people don't usually use cross platform or relative paths.
The only time i had issues was in the early days of java with applets. But my experience of moving large web applications is that it is relatively easy. In fact i've had more problems switching JVMs within a platform than switching platforms - one thing that got me was some people had started to rely on the collection ordering which happened on the SUN JVM to always stay in the same order but on the IBM one recreating the same collection did not maintain ordering, which was fair enough there were no guarantees in the specs.
On the other hand back in the days i wrote some J++ code for windows my god it was the WORST i have EVER experienced, just getting it to run on all versions of windows took weeks not hours. detecting network connections without triggering a RAS dialup across 95->2k was hilariously difficult.
Well I would like more police about yes.
But if it was a choice between one police man on one out of every 1000 bus trips or a CCTV on every bus, i'd focus on CCTV first. Cos you can be damn sure i won't be robbed on that one lucky bus trip with the cop.
i do feel more safe if there is CCTV.
Like on the bus going through peckham when groups of youths from the local youth detention center "enquired" about what type of phone i had. Also outside clubs where fights regularly break out.
I'm not saying we don't need more police or anything like that and i hate the whole ID card scheme but I actually quite like CCTV and brighter lights.
Wrong.I downloaded it, my c program still isn't cross-compatible.
Oh i see, you want me to rewrite my entire program to use this library!
Heh.
Or put another way "C" is theoretically cross platform but in practice it isn't.
sure.
Although, some kind of merit award in being able to recognise detrimental international monopolies and illegal price fixing groups, would also be useful in my book!
And some kind of award for recognising commercial interest lobby groups, dodgy legislation and for uncovering widespread IP extortion and blackmail by said groups.
Oh now, wait, THEY'RE JUST KIDS.
The information that was classified in the original document and is now leaked looks like it was classified for fair enough reasons.
The fact that it is being leaked by the Italian press conjures up images in my mind of hidden details that were being supressed by the americans - however I don't see this to be the case.
In fact if I was an insurgent with internet access some of the leaked classified details could be quite useful. Some describe how the US deal with incidents and their knowledge of terrorists tactics and specific details about what battalions are trained in and exactly where they are located.
Whilst obviously the major error was in the americans releasing a document with hidden meta-data, I do think it is quite irresponsible of the Italian press to advertise the release of this hidden data, thus further endagering the lives of people in Iraq.
I would do only it turns out all my programs only run on windows. why is that. oh yes cos they have a MONOPOLY.
i forgot that for a second.
lol
Sadly not. More likely this means someone else can trust your computer but you can't.
Most comments i see is for code thats too complicated to understand straight off and almost always it's more complicated than it needs to be, rather than fix it some developer just puts a comment at the top along the lines of:I always find good OO code relatively easy to understand and just ignore reading the comments, which somtimes are out of date - if I'm not entirely sure what the purpose of the code is I way prefer unit tests to comments as I know the units tests are more likely to be correct - if they still pass.
Of course none of this applies to perl or regex where i need all the comments I can find.
That is not actually correct. What in fact the RIAA can do and probably are doing is lobbying in europe.
No one in the UK pays any attention to what happens in europe but eventually either the law in the UK is changed to bring it in line with European regulations or there is a european test case and Europe starts fining the UK goverment telling it that it should do what it is told or face the consequences.
I agree.
Also I'm not sure laws are inherently bad. I quite like the idea of having a right to use content i buy in whatever way I see fit.
I was in fact being sarcastic.
:)
Remembering where you left things when they aren't visible is hard, particularly when you regularly use many folders.
why not make it 3 clicks to turn spatial browsing on, since most people don't like it
Perhaps we could just fork nautilus. That way you can use spatial nautilus and the rest of us can stop wasting time navigating round hundreds of randomly located windows.
probably not kidding, but i expect the real reason is that first it's the OSS client, next it's the server, then it's bye bye bitkeeper...
I'm not convinced it's X rendering that makes linux distros seems slower than windows. Applications and windows usually take longer to appear. For example gedit takes nearly as long as word to load on my PC and is about 10times longer than notepad.
After massive improvements and work to nautilus is about launches at a reasonable speed if you aren't using the file explorer version (which of course most people do once they've worked out how to re-enable it!). Maybe it's got something to do with library caching/toolkits being preloaded on windows, but whatever it is linux needs it.
This doesn't really matter. For ad networks using CPA (cost-per-acquisition) rather than CPC the tracking is done using cookies which will record a sign up or purchase even on repeat visits.
Although CPA based advertising is a lot less prone to fraud there are issues of which the biggest is in deploying pixels to do tracking on all sign up/purchase pages there are also more fixable issues with double counting (crediting to more than one ad network). CPA may be vulnerable to publisher fraud of course though this is usually less of an issue as you have a contractual relationship with them.
CPC click fraud was a problem long before google starting selling space on a CPC basis. I believe commission junction used to have a whole fraud detection team (they may still do).
Reminds me about the akamai speedera merger. We liked Spakamai.
Why have the overhead of using a database then? the application could just be allocated 15GB and hold it in memory.
I agree it sounds mainly Larry's fault. But I kind of understand why Torvalds got so cross, he's only human and I'm sure he must realise that he has himself triggered the chain of events by switching to bitkeeper in the first place.
When making a decision about which are the best tools to use for development you have to look at the "development process" who's involved and what people are likely to "buy into" this is part of the challenge of managing multi person projects. I reckon Torvalds should read more of Dilberts management hints:
Boss: You need to socialize your idea with the rest of your group.
Dilbert: Socialize? Is that the same as getting buy-in?
Boss: It's one step below buy-in. Its more like dialoging for feedback.
Dilbert: Wait...I thought that building a consensus was one step below buy-in.
Boss: Just run it up the flagpole and see who salutes.
Dilbert: Wouldn't it be better to do a temperature check with a straw man?
Boss: Maybe...but is that going to inoculate the stakeholders?