We contract companies to write software for us, but then they retain full ownership of that software.
It means we often don't get to see the source code, and even if we do, we are not permitted to modify it or to get any other company to extend the software, etc.
I have to admit, I find it foolish to enter into such agreements. When I pay for work I expect to receive IP, but such situations clearly do exist. Other people must just not care so much...
Only those people paying attention known they gamed the process. The executives making calls over the next decade that their organisation requires all documents to be stored in an ISO certified format, they don't know MS gamed the process.
This result, if it gets ratified, shoots down the most visible failing in Microsoft's software in years. It doesn't matter if it gets shot down using foul means, because the people who hear about that (us) don't matter enough to Microsof'ts bottom line.
Interesting idea, they're very different branches of mathematics so I'm having trouble working out how to combine them.
The idea in regression is to transform the IVs such that there is a linear relationship between them and the DVs. The transformations people make of IVs to make the relationship simple are a bit of a black art. Most people just get by with log since it solves any polynomial. It's easy enough to just keeping adding higher order polynomials until regression predicts a coefficient of 0 to tell you that you've gone too far, but that doesn't help much if the equation you're trying to predict parameters for is not efficiently modelled by a polynomial.
As for using differentiation to find the optimal transformation function, it's an interesting idea and I'm too tired now to work out if you could use it. A google search doesn't turn up anything quickly, with the following at least trying to determine the transformations automatically instead of guessing. http://pubs.acs.org/cgi-bin/abstract.cgi/ancham/1991/63/i20/f-pdf/f_ac00020a022.pdf?sessid=6006l3
The problem I keep coming back to is predicting the height of a ball bobbing on the surface of the sea given all the appropriate parameters. The see is best modelled as the combination of waves (duh), but I can't think of anything except the intuition of the modeller that would detect this case and suggest log was not an appropriate general function. I recall covering hmm, was it Maclaurin expansion, in my senior numerical methods course, but I never thought of applying that to regression at the time. Relying on google again turns up http://www.springerlink.com/content/44teyxefx9pa8058/ which looks relevant but I'm not subscribed to springer so can't view any more than the abstract.
They've looked for a linear correlation, so if what you've said is true then the analysis they used wouldn't find it.
In order to find a correlation where the input IV (beer consumption) has an optimal value, you would have to do the regression on a transformation of the variable. Perhaps a quadratic would suffice, or else abs(X - k) for some unknown value of k.
I agree. It is the job of the vendor to know what about this sort of thing and turn off the features that I won't ever want so that the sysadmin doesn't have to concern himself with such details.
Saying 'it is the user's problem' just isn't good enough. It is everyone's problem, and if everybody tries to fix it then maybe it will go away.
You could well be right, and for entirely selfish reasons I hope so.
I went through undergrad a few years before you. When I went through C was drilled into every student until they spoke it fluently. I also covered some OO concepts - polymorphic types and whatnot, and I remember the lecturer commenting on how hard it was for the students who he'd taught programming ten years ago as they were all having to come back and train themselves in OO, much like old Cobol programmers had had to train themselves in C a few years before that.
Apart from a little OO, I covered architecture. But unlike the generation of students before me, the deepest we got was theoretical discussions on nand gates and the like. As a result I memorised enough about gates to pass the exam but didn't develop the mastery I did of C. Now I can barely remember what a FPGA is, and I wouldn't be able to even start designing one.
I don't know how the curriculum changed after I left in any great detail. I know they shifted from C to Java, and they ran design paterns, specification writing, and a few other bits and bobs. Out in the 'real world' I gain quite a bit of benefit from the OO stuff I did, and I gain even more from C. However, I cannot believe that whatever they've replaced the hours saved by dumping C is likely of equal benefit to students coming out now.
Perhaps it is how to approach a really big class library (Java,.NET, QT, whatever) and get your head around it? Perhaps it is a deeper understanding of user interface psychology? Whatever it happens to be, I have a hard time believing that uni has got any less useful for a comparable kid going through today.
But, as I said, I hope you're right - if so it has to be good for my job security!
Actually, I see no reason why FUSE wouldn't be perfectly acceptable for zfs - at least until people start seriously considering zfs for root filesystems. FUSE is fast enough, stable enough, and flexible enough. We don't have X in the kernel, I don't see that we _need_ to have the filesystem in the kernel either.
It would be easy enough to test, grab a whole lot of students and get them to rate the music they're listening to. Use a standard double blind test with the clean music and the watermarked music, and look for a significant increase in rank for the clean version.
Yeah, that's the same argument I used to convince my wife to buy a 20" imac a few years ago when they first came out. I forget the details but it was something like $2000 for the computer at the time that a decent 20" screen was $1000.
The imac was not designed for price, it was designed for all-in-one ergonomics, fitting into a lounge setting and all of that. As in, if it isn't all-in-one it really isn't in the running.
In your particular case, that would be EnterpriseDB. You didn't say whether you're running SE or EE, and I can't remember whether BI is emulated by EnterpriseDB or if they only emulate AS, but if you want to talk about an open source strategy it is worth at least mentioning. The other big money saver is moving from EE back to SE - partitioning is all well and good, but you can afford about a terabyte of solid-state drive for the saving in licence fees, which would more than make up for it in many cases.
Browser based terminals... Give silverlight a whirl - Win, Mac and Linux with reasonable.net integration. I personally went with Dundas, but I suspect my needs are simpler to yours there.
Yeah, I gave up on USB a couple years ago when power was cut to my server. The server carried on fine (UPS) but I had the logs on USB (foolish notions of reduced risk of losing critical data). Linux could not handle losing power to a critical part of the filesystem and all but hosed the system - I ended up having to hard power off and recover the filesystem.
Having said that, NDAS or whatever they're going to be called has been a more than adequate replacement (though I don't quite think the technology is up to write access from multiple machines). You get around 10MB/s, so a little slower than USB2, but the drive can be as far away from your computer as you like and you can get (readonly) access from every machine at full speed.
After all, can you think of a single project Dvorak has claimed as a failure that didn't succeed spectacularly? His criticism is a strong hint that OLPC is no longer a niche player and is about to make major inroads.
On a more insidious note, Dvorak is an analyst-for-hire. He only comes out with an opinion when somebody pays him to have that opinion. That means one of the big players has decided they want bad PR about OLPC. I wonder if it was Microsoft, Intel, or somebody else?
10 to 100 times more accurate than existing systems means that for every 10 to 100 mistakes that existing systems make, this system will make just one.
For instance if they say current technology is 80% accurate then out of ten thousand emails coming in, 2000 will be incorrectly classified. 100 times more accurate than that means 20 errors, or 99.8% accuracy.
Now, it happens that TFA is peddling snake oil. The top spam blocking programs make one mistake per ten thousand emails processed or 99.99% accuracy. To obtain ten times that accuracy is well beyond belief, and if physically possible, would require a much more detailed data source than is currently being used (perhaps simulating an email client to work out what is actually displayed and running OCR/analysis on that), and/or running semantic/pragmatic understanding of the text rather than just lexical. That's if you care - even with an adequate filter like gmail uses (99.5%), you only spend a minute a day on spam.
Hmm, clients, servers, mobile and embedded devices.
You're right, I misunderstood you. Since this discussion started with apple I interpreted everywhere as 'windows, macos, linux'.
One word: Facebook. Actually, Google is proof as well.
*sigh*, no.
Facebook works for everyone, as does Google and a few other high-profile examples. Do you know how many people they have working on making their systems cross-platform? By that logic, MFC is cross-platform too - look at MS word on a mac!
To me, cross platform means that it is easy to make an application run on multiple platforms, not just possible with lots of skill and effort.
If computer programming languages are so hard to learn why is it that I can code... but can't speak Japanese
Because, in case you hadn't noticed yet, you're not normal. I know people who can speak half a dozen different languages yet can't write in even the simplest scripting languages. Look around, how many people down your street can program in one programming language? How many can program in more than one? What I think you'll find is that most people down your street can use a computer but only a tiny fraction can program. Of that tiny number, only a small fraction of those can program in more than one language and learn another language easily.
Yes, some people learn multiple programming languages without effort. For most it is a daunting, unrealistic prospect. Try to think of the people down the street when you say something is easy, or think of you parents learning a programming language, or... your experiences (or mine, or most of slashdot's) are just not useful for judging how easy things are....blah....blah...I'm so smart....
Actually, my point was the opposite. I tried to make this clear by including how little work I had to do, but obviously I was not clear enough. My point is that.net makes pretty cool things easy - it brings stuff to the masses (ie. me)
Ever notice how practically everyone thinks they are so smart and everyone else is stupid.
Yup
Is it that I said ".NET is chasing the pipe dream of write once, run everywhere"?
Debatable, unless by everywhere you mean on 'XP, 2000, Vista, pocket edition, etc.' Their moderately friendly attutude to Mono and also silverlight implies they have some interest in cross-compatibility. But to me.net is about building a rich culture of library use, not about cross-platform compatibility.
Or maybe it's not true that if you want a truly cross-platform application you should write a web app?.
Web apps are more cross compatible than java? Try writing decent portable web-apps sometime. I have never worked anywhere where the intranet web apps ran on anything except windows, even when linux on the desktop was tolerated.
Or maybe programming languages are not that easy to learn, are they?
Clearly not, look how many people fail. It doesn't matter if you, I, or even 99% of slashdot find it easy. Most people learn one programming language (with trouble) and don't generalise at all. I've recently given up teaching SQL to a PL/SQL 'programmer'.
There is a very good reason why the companies that employ hundreds of code monkeys pay them so little, and their ability to learn a new programming language ties in.
Or could it be that Objective-C isn't really very good at OO. Cause it's not like Java was influenced by Objective-C and by extension.NET.
Not sure what you're getting at here, but I'm not enough of an OO programmer to comment anyway.
Oh wait, it must be that Apple is dying and they need.NET to save them from themselves. That must be it.
It could be, though I'll wait for NetCraft's confirmation;-).net is a class library.
Recently I purchased some really cool dashboard widgets. They're written in C# and come with C# examples but because it's all.net, I could write IronPython code that created them them perfectly - I doubt the people who created the library had even heard of ironpython.
Before that, I got a mapping program with an API designed for.net and the ADO.net addons for oracle - in all of a dozen lines of code I had oracle and a mapping program talking to each other. Not only that, I can use the.net add-on for oracle to write extensions (user designed functions), along with a C# class library for Markov modelling and a bit of iron python glue (since I only write C# if someone forces me to) to tie them all together. Voila - markov modelling as a UDF in Oracle. Essentially then,.net has provided me with the fabled code resue that I remember OO was supposed to.
Coming back to Apple, I don't know how.net affects them. I only use macs at home, so I don't really care that much. Maybe future programmers will all be trained in.net and apple is missing an opportunity by not having their computers accessible to these programmers, but that doesn't really ring true does it? Maybe future applications will be written in.net and with better.net support it would be easier to port them to the mac, thereby keeping commercial software more up to date - that rings more true.
You know, I'm pretty sure you weren't trying to troll but I don't see a single true statement in what you wrote.
Lets start with what Mac has already. It has Java 5, as in a.jnlp will 'just work' on OSX. It kinda has python - unless you want to do GUI stuff. Ruby is in much the same boat as python. However, this is all irrelevant..NET is a new paradigm - learn a decent class library and keep using it from then on, whether that's in J#, IronPython, IronRuby, or whatever language is popular next year. Java claimed much the same thing, which isn't too surprising since.net is basically a java knockoff with some design faults smoothed out.
Since you brought up web apps, have you ever done any webapp development on windows? The connection between IIS and.NET is such that you are crazy to develop web apps without using.net. It's so much better than everything else out there that apache has added support for mono to achieve a similar effect (if a little crude at the moment).
As for colleges turning out programmers. Sadly, get real. Maybe one in ten to one in thirty of their graduates can program - as in, given this hot new language, pick it up and start churning out code. This has always been the case and has nothing to do with C#, Java or whatnot. Incidentally, at least around here they mainly seem to be targeting Java at the moment. Of course, when employing you will do your best to get the one person from the class that can actually program but sometimes that just isn't realistic - what if that person is unable to interact with the business? Conversely, if I had a job applicant from a objective-C developer, I wouldn't hold it against them for a C# job - anybody interesting enough to have picked up an obscure language is going to pick up C# just fine.
As for OO done right, neither C# or Objective C are OO done right, and unlike you I'm not convinced that C# is any worse at OO than Objective C, but I don't know enough Objective C to really say. Ruby is probably your best bet if you're looking for a good OO language. Personally I just write glue programs now and take advantage of.NET to keep my programs small enough to make OO irrelevant, but for people writing libraries or writing code unique enough that the majority can't be libraries... well, you'd have to ask them.
Oh, and I'm deliberately being slack with terminology - substitute C#,.net and CLR where appropriate. From a marketing perspective they're all the same anyway;-)
Yeah, sorry about that. It was partially deliberate as I was trying to give the point of view of the person hearing proposals all day, but probably they've permeated my consciousness such that I can now say them with a straight face...
And while I agree about IT why projects fail, how do you get the competence necessary to ensure project success? The number of skilled senior IT manages I have met is... well, not quite zero. Even if you manage to get one, how do you hold on to them? And if you have ordinary management, should you avoid embarking on IT projects?
We're getting a long way from the original issue, but something that came up today was a project IBM offered to manage. Now my initial reaction looking at it was - there is no way we should be paying $100 to $450/hr for all of those people - they're not doing anything I couldn't do! But to be honest if I had set out and hired a whole team to do the same job, what are the odds I would've screwed up and hired a couple duds that end up wrecking the whole project? You end up paying exorbitant rates because IBM has a team that works pretty well together, and you haven't. (There's other factors like the extra respect that outsiders can get, but now we're on to a digression from a digression)...
Returning back to the topic. The reason I chose IT projects failing is that I wanted to point out that for C level executives, they have to choose between dozens of different projects and one thing they have to weigh up for IT projects is the odds of it turning into a white elephant. They don't know enough about the projects to accurately ascertain which ones are going to blow up, their job is to weigh up the alternative strategic directions and choose projects which fit the strategic direction of the company. Essentially, while an IT middle manager will push what's best for IT, a sales manager will push what's best for sales, and a marketing manager will push what's best for marketing, it's up to senior management to decide whre the company is headed and which projects they can afford to shelve.
Slashdot Mirror
I think he means like SQL Server Reporting Services, MS Access and SQL Server Analytic services.
Hard to be sure though.
The tools like Mondrian, JasperSoft, Petaho, Navicat, etc. They're all okay, but nothing like as polished as Microsoft's.
It seems to be that way where I work.
We contract companies to write software for us, but then they retain full ownership of that software.
It means we often don't get to see the source code, and even if we do, we are not permitted to modify it or to get any other company to extend the software, etc.
I have to admit, I find it foolish to enter into such agreements. When I pay for work I expect to receive IP, but such situations clearly do exist. Other people must just not care so much...
Only those people paying attention known they gamed the process. The executives making calls over the next decade that their organisation requires all documents to be stored in an ISO certified format, they don't know MS gamed the process.
This result, if it gets ratified, shoots down the most visible failing in Microsoft's software in years. It doesn't matter if it gets shot down using foul means, because the people who hear about that (us) don't matter enough to Microsof'ts bottom line.
Interesting idea, they're very different branches of mathematics so I'm having trouble working out how to combine them.
The idea in regression is to transform the IVs such that there is a linear relationship between them and the DVs. The transformations people make of IVs to make the relationship simple are a bit of a black art. Most people just get by with log since it solves any polynomial. It's easy enough to just keeping adding higher order polynomials until regression predicts a coefficient of 0 to tell you that you've gone too far, but that doesn't help much if the equation you're trying to predict parameters for is not efficiently modelled by a polynomial.
As for using differentiation to find the optimal transformation function, it's an interesting idea and I'm too tired now to work out if you could use it. A google search doesn't turn up anything quickly, with the following at least trying to determine the transformations automatically instead of guessing. http://pubs.acs.org/cgi-bin/abstract.cgi/ancham/1991/63/i20/f-pdf/f_ac00020a022.pdf?sessid=6006l3
The problem I keep coming back to is predicting the height of a ball bobbing on the surface of the sea given all the appropriate parameters. The see is best modelled as the combination of waves (duh), but I can't think of anything except the intuition of the modeller that would detect this case and suggest log was not an appropriate general function. I recall covering hmm, was it Maclaurin expansion, in my senior numerical methods course, but I never thought of applying that to regression at the time. Relying on google again turns up http://www.springerlink.com/content/44teyxefx9pa8058/ which looks relevant but I'm not subscribed to springer so can't view any more than the abstract.
They've looked for a linear correlation, so if what you've said is true then the analysis they used wouldn't find it.
In order to find a correlation where the input IV (beer consumption) has an optimal value, you would have to do the regression on a transformation of the variable. Perhaps a quadratic would suffice, or else abs(X - k) for some unknown value of k.
I'd bet it is fine.
The reason most people have terrible passport photos is they're taken by disinterested photographers (or even machines) using cheap equipment.
Having said that, the new laws about not smiling and so on sure don't help.
There is also that weird xml viewer thing that loads every time I try to open a configuration file. Omni Outliner or somesuch.
But the main change I make is dumping garage band, that program uses _way_ too much disk space.
I agree. It is the job of the vendor to know what about this sort of thing and turn off the features that I won't ever want so that the sysadmin doesn't have to concern himself with such details.
Saying 'it is the user's problem' just isn't good enough. It is everyone's problem, and if everybody tries to fix it then maybe it will go away.
You could well be right, and for entirely selfish reasons I hope so.
.NET, QT, whatever) and get your head around it? Perhaps it is a deeper understanding of user interface psychology? Whatever it happens to be, I have a hard time believing that uni has got any less useful for a comparable kid going through today.
I went through undergrad a few years before you. When I went through C was drilled into every student until they spoke it fluently. I also covered some OO concepts - polymorphic types and whatnot, and I remember the lecturer commenting on how hard it was for the students who he'd taught programming ten years ago as they were all having to come back and train themselves in OO, much like old Cobol programmers had had to train themselves in C a few years before that.
Apart from a little OO, I covered architecture. But unlike the generation of students before me, the deepest we got was theoretical discussions on nand gates and the like. As a result I memorised enough about gates to pass the exam but didn't develop the mastery I did of C. Now I can barely remember what a FPGA is, and I wouldn't be able to even start designing one.
I don't know how the curriculum changed after I left in any great detail. I know they shifted from C to Java, and they ran design paterns, specification writing, and a few other bits and bobs. Out in the 'real world' I gain quite a bit of benefit from the OO stuff I did, and I gain even more from C. However, I cannot believe that whatever they've replaced the hours saved by dumping C is likely of equal benefit to students coming out now.
Perhaps it is how to approach a really big class library (Java,
But, as I said, I hope you're right - if so it has to be good for my job security!
Think of how much deeper the understanding of Objects, ER diagrams, ... will be for those students who started with Java.
Which will be more useful in the end, a bit more understanding of low level programming or a bit more understanding of ERP?
Actually, I see no reason why FUSE wouldn't be perfectly acceptable for zfs - at least until people start seriously considering zfs for root filesystems. FUSE is fast enough, stable enough, and flexible enough. We don't have X in the kernel, I don't see that we _need_ to have the filesystem in the kernel either.
Perhaps.
It would be easy enough to test, grab a whole lot of students and get them to rate the music they're listening to. Use a standard double blind test with the clean music and the watermarked music, and look for a significant increase in rank for the clean version.
Or, more accurately...
"...is to protect planes from manportable systems (which used to be IR guided way back in '07)..."
If this technology is put into place, terrorists will use something else - maybe a camcorder on their rocket, maybe an AA gun...
Use the digital audio output on your macbook, it sounds GREAT!
Cables are pretty cheap too, at least not too bad
Yeah, that's the same argument I used to convince my wife to buy a 20" imac a few years ago when they first came out. I forget the details but it was something like $2000 for the computer at the time that a decent 20" screen was $1000.
The imac was not designed for price, it was designed for all-in-one ergonomics, fitting into a lounge setting and all of that. As in, if it isn't all-in-one it really isn't in the running.
Remember OSS isn't just Linux.
.net integration. I personally went with Dundas, but I suspect my needs are simpler to yours there.
In your particular case, that would be EnterpriseDB. You didn't say whether you're running SE or EE, and I can't remember whether BI is emulated by EnterpriseDB or if they only emulate AS, but if you want to talk about an open source strategy it is worth at least mentioning. The other big money saver is moving from EE back to SE - partitioning is all well and good, but you can afford about a terabyte of solid-state drive for the saving in licence fees, which would more than make up for it in many cases.
Browser based terminals... Give silverlight a whirl - Win, Mac and Linux with reasonable
Yeah, I gave up on USB a couple years ago when power was cut to my server. The server carried on fine (UPS) but I had the logs on USB (foolish notions of reduced risk of losing critical data). Linux could not handle losing power to a critical part of the filesystem and all but hosed the system - I ended up having to hard power off and recover the filesystem.
Having said that, NDAS or whatever they're going to be called has been a more than adequate replacement (though I don't quite think the technology is up to write access from multiple machines). You get around 10MB/s, so a little slower than USB2, but the drive can be as far away from your computer as you like and you can get (readonly) access from every machine at full speed.
Oh, and SMART etc all works.
After all, can you think of a single project Dvorak has claimed as a failure that didn't succeed spectacularly? His criticism is a strong hint that OLPC is no longer a niche player and is about to make major inroads.
On a more insidious note, Dvorak is an analyst-for-hire. He only comes out with an opinion when somebody pays him to have that opinion. That means one of the big players has decided they want bad PR about OLPC. I wonder if it was Microsoft, Intel, or somebody else?
Er, no.
10 to 100 times more accurate than existing systems means that for every 10 to 100 mistakes that existing systems make, this system will make just one.
For instance if they say current technology is 80% accurate then out of ten thousand emails coming in, 2000 will be incorrectly classified. 100 times more accurate than that means 20 errors, or 99.8% accuracy.
Now, it happens that TFA is peddling snake oil. The top spam blocking programs make one mistake per ten thousand emails processed or 99.99% accuracy. To obtain ten times that accuracy is well beyond belief, and if physically possible, would require a much more detailed data source than is currently being used (perhaps simulating an email client to work out what is actually displayed and running OCR/analysis on that), and/or running semantic/pragmatic understanding of the text rather than just lexical. That's if you care - even with an adequate filter like gmail uses (99.5%), you only spend a minute a day on spam.
That would be cool.
I noticed Cocoa# the other day too, but I haven't looked into it at all.
This is ridiculous
... but can't speak Japanese
... your experiences (or mine, or most of slashdot's) are just not useful for judging how easy things are. ...blah....blah...I'm so smart....
.net makes pretty cool things easy - it brings stuff to the masses (ie. me)
Feel free to stop replying any time.
Hmm, clients, servers, mobile and embedded devices.
You're right, I misunderstood you. Since this discussion started with apple I interpreted everywhere as 'windows, macos, linux'.
One word: Facebook. Actually, Google is proof as well.
*sigh*, no.
Facebook works for everyone, as does Google and a few other high-profile examples. Do you know how many people they have working on making their systems cross-platform? By that logic, MFC is cross-platform too - look at MS word on a mac!
To me, cross platform means that it is easy to make an application run on multiple platforms, not just possible with lots of skill and effort.
If computer programming languages are so hard to learn why is it that I can code
Because, in case you hadn't noticed yet, you're not normal. I know people who can speak half a dozen different languages yet can't write in even the simplest scripting languages. Look around, how many people down your street can program in one programming language? How many can program in more than one? What I think you'll find is that most people down your street can use a computer but only a tiny fraction can program. Of that tiny number, only a small fraction of those can program in more than one language and learn another language easily.
Yes, some people learn multiple programming languages without effort. For most it is a daunting, unrealistic prospect. Try to think of the people down the street when you say something is easy, or think of you parents learning a programming language, or
Actually, my point was the opposite. I tried to make this clear by including how little work I had to do, but obviously I was not clear enough. My point is that
Ever notice how practically everyone thinks they are so smart and everyone else is stupid.
.net is about building a rich culture of library use, not about cross-platform compatibility.
.NET.
.NET to save them from themselves. That must be it.
;-) .net is a class library.
.net, I could write IronPython code that created them them perfectly - I doubt the people who created the library had even heard of ironpython.
.net and the ADO.net addons for oracle - in all of a dozen lines of code I had oracle and a mapping program talking to each other. Not only that, I can use the .net add-on for oracle to write extensions (user designed functions), along with a C# class library for Markov modelling and a bit of iron python glue (since I only write C# if someone forces me to) to tie them all together. Voila - markov modelling as a UDF in Oracle. Essentially then, .net has provided me with the fabled code resue that I remember OO was supposed to.
.net affects them. I only use macs at home, so I don't really care that much. Maybe future programmers will all be trained in .net and apple is missing an opportunity by not having their computers accessible to these programmers, but that doesn't really ring true does it? Maybe future applications will be written in .net and with better .net support it would be easier to port them to the mac, thereby keeping commercial software more up to date - that rings more true.
Yup
Is it that I said ".NET is chasing the pipe dream of write once, run everywhere"?
Debatable, unless by everywhere you mean on 'XP, 2000, Vista, pocket edition, etc.' Their moderately friendly attutude to Mono and also silverlight implies they have some interest in cross-compatibility. But to me
Or maybe it's not true that if you want a truly cross-platform application you should write a web app?.
Web apps are more cross compatible than java? Try writing decent portable web-apps sometime. I have never worked anywhere where the intranet web apps ran on anything except windows, even when linux on the desktop was tolerated.
Or maybe programming languages are not that easy to learn, are they?
Clearly not, look how many people fail. It doesn't matter if you, I, or even 99% of slashdot find it easy. Most people learn one programming language (with trouble) and don't generalise at all. I've recently given up teaching SQL to a PL/SQL 'programmer'.
There is a very good reason why the companies that employ hundreds of code monkeys pay them so little, and their ability to learn a new programming language ties in.
Or could it be that Objective-C isn't really very good at OO. Cause it's not like Java was influenced by Objective-C and by extension
Not sure what you're getting at here, but I'm not enough of an OO programmer to comment anyway.
Oh wait, it must be that Apple is dying and they need
It could be, though I'll wait for NetCraft's confirmation
Recently I purchased some really cool dashboard widgets. They're written in C# and come with C# examples but because it's all
Before that, I got a mapping program with an API designed for
Coming back to Apple, I don't know how
You know, I'm pretty sure you weren't trying to troll but I don't see a single true statement in what you wrote.
.jnlp will 'just work' on OSX. It kinda has python - unless you want to do GUI stuff. Ruby is in much the same boat as python. However, this is all irrelevant. .NET is a new paradigm - learn a decent class library and keep using it from then on, whether that's in J#, IronPython, IronRuby, or whatever language is popular next year. Java claimed much the same thing, which isn't too surprising since .net is basically a java knockoff with some design faults smoothed out.
.NET is such that you are crazy to develop web apps without using .net. It's so much better than everything else out there that apache has added support for mono to achieve a similar effect (if a little crude at the moment).
.NET to keep my programs small enough to make OO irrelevant, but for people writing libraries or writing code unique enough that the majority can't be libraries... well, you'd have to ask them.
.net and CLR where appropriate. From a marketing perspective they're all the same anyway ;-)
Lets start with what Mac has already. It has Java 5, as in a
Since you brought up web apps, have you ever done any webapp development on windows? The connection between IIS and
As for colleges turning out programmers. Sadly, get real. Maybe one in ten to one in thirty of their graduates can program - as in, given this hot new language, pick it up and start churning out code. This has always been the case and has nothing to do with C#, Java or whatnot. Incidentally, at least around here they mainly seem to be targeting Java at the moment. Of course, when employing you will do your best to get the one person from the class that can actually program but sometimes that just isn't realistic - what if that person is unable to interact with the business? Conversely, if I had a job applicant from a objective-C developer, I wouldn't hold it against them for a C# job - anybody interesting enough to have picked up an obscure language is going to pick up C# just fine.
As for OO done right, neither C# or Objective C are OO done right, and unlike you I'm not convinced that C# is any worse at OO than Objective C, but I don't know enough Objective C to really say. Ruby is probably your best bet if you're looking for a good OO language. Personally I just write glue programs now and take advantage of
Oh, and I'm deliberately being slack with terminology - substitute C#,
Yeah, sorry about that. It was partially deliberate as I was trying to give the point of view of the person hearing proposals all day, but probably they've permeated my consciousness such that I can now say them with a straight face...
... well, not quite zero. Even if you manage to get one, how do you hold on to them? And if you have ordinary management, should you avoid embarking on IT projects?
And while I agree about IT why projects fail, how do you get the competence necessary to ensure project success? The number of skilled senior IT manages I have met is
We're getting a long way from the original issue, but something that came up today was a project IBM offered to manage. Now my initial reaction looking at it was - there is no way we should be paying $100 to $450/hr for all of those people - they're not doing anything I couldn't do! But to be honest if I had set out and hired a whole team to do the same job, what are the odds I would've screwed up and hired a couple duds that end up wrecking the whole project? You end up paying exorbitant rates because IBM has a team that works pretty well together, and you haven't. (There's other factors like the extra respect that outsiders can get, but now we're on to a digression from a digression)...
Returning back to the topic. The reason I chose IT projects failing is that I wanted to point out that for C level executives, they have to choose between dozens of different projects and one thing they have to weigh up for IT projects is the odds of it turning into a white elephant. They don't know enough about the projects to accurately ascertain which ones are going to blow up, their job is to weigh up the alternative strategic directions and choose projects which fit the strategic direction of the company. Essentially, while an IT middle manager will push what's best for IT, a sales manager will push what's best for sales, and a marketing manager will push what's best for marketing, it's up to senior management to decide whre the company is headed and which projects they can afford to shelve.
PS: And I didn't even mention pet projects!