Actually, Java 2 (JDK 1.2) included *A LOT* of new graphical features that cannot be used in current applets because they would only be supported by pepole who insalled Sun's plugin, and not the default MS/Netscape installations.
Without the MS JVM holding people at 1.1, you can go ahead and use your "regular" JVM from Sun/IBM with your browser, making it compatible with the new bells and whistles. (If you're in for that sort of thing, anyway).
I wasn't questioning MS's motivation, just indicating that in the long run, the move might prove to be beneficial, or at least insignificant to Java as a language.
And I think (based on a story posted about a week ago here on slashdot) that MS's OEM deals for XP will allow manufacturers to preinstall things they choose. (The article mentions this as well)However this being a manual operation might make it too costly for OEMs to pracice it widely.
Your arguments are all true, but do not really add up to the death of Java. What they add up to is the death of Java in the browser, which has arguably been a bad idea to begin with, and has probably become a worse idea once Flash came out. Other than playing web based games, java in the browser is just plain annoying. (I should mention here that I am a big fan of the Java language, and get paid to develop mostly in Java)
Java used the client side applets to gain prominence, but that was mostly a gimmick. Where Java really comes into its own is on the server, where the developer can control the JVM version, and doesn't have to use Java 1.0 for compatibility's sake.
Since server based Java never relied on the MS JVM, it remains unefective and just as viable as it ever was.
To all of the people using Java out there, this announcement seems like the equivalent of "XP will no longer encorporate a web browser".
Java right now is still a moving target with a new verison coming out every half a year or so, and I don't think that bundling a JVM with an OS would have been beneficial to the language's development, as it would force developers to only use the XP version for fear of losing audience.
But what if I'm a freelancer who wants to exclusively sell an article to a paper printing in NY and one coming out in LA. Can I insist that if the NY paper has an LA edition my article not appear there because I have signed an exclusive contract with a different paper in that city? Is that a more similar case?
How is this different from say a MPAA member studio's right to control their motion picture content? If I as a freelance writer can decide the mediums for which I lisence my article to a publisher, why can't a studio decide what platform my DVD should play on? Or the RIAA decide that while it's ok to listen to a tune on CD, ripping it's not allowed.
I think we should try to steer clear of legalized content control for both corporate and private content producers.
I think that the reason so much troubles are plaguing independent sites is that their revenue model, dependent on ad revenue, is backed by inefficient and ineffective ad display systems.
I believe this topic was previously discussed/., but no real solution arrived at.
I personally think that an ad-serving system should be developed in the open source community that would facilitate easy ad sales/link exchanges, that would be full featured enough to become somwhat of an industry standard. The Apache of ad-serving so to speak.
The existance of this "Ad-pache" would allow smaller sites to have an easier time selling ads and attracting revenue. And of course it would have to provide all the bells and whistles like time/geo targeting, etc, and possibly some of the other features not offered by anyone else (like ad feedback, etc)
With such a system in place and showing tangible results, the playing field might be leveled for the small guy.
I am a Java programmer, and I agree that OOP is not the best thing for anyone to jump into. But not because they have to know how things work on the low level. Actually I think teaching somebody assembler as a first language is an even worse idea than teaching them OOP.
I think that the best way to get people started on programming is by introducing concepts such as variables, data types, flow of control, functions and references. This is why I think C is what should be taught in the (High School) class room. Having come from a Pascal background I have really benefitted from being taught procedural programming before moving on to C++/Java's OOP models.
Teaching people assembler you are forcing them into the deep end of the Marianas trench right away. What's the point of explaining memory segments and offsets to someone who doesn't know such basic concepts as variables and functions?
Why should I concern myself with the difference between near and far jumps before knowing the standard loop constructs that such jumps can be used for?
I think that a licens by the GNU should not stop at allowing the employee to develope open source code. I think that a lot of programmers are in an unfavorable position when it comes to signing an employment contract. Some employees have very strict and prohibitive documents that they make you sign.
If a standardized contract that protects a programmer's rights to produce his own work on his time is introduced and accepted in the industry, it should give programmers leverage they need when negotiating these contracts.
Because GNU is already a recognized authority in the IT world, it would be significantly easier to make an employer agre to a contract coming from them than it would be to make them accept something drawn up by your lawyer.
Not to mention the extra business Loki will get. As long as 2x more people are willing to buy if it's 50% off, seems like a worthwhile proposition to me.
Thanks for the link. I believe that the very first result on the google search (which was a pretty interesting read) proves my statement.
I didn't say java was developed for its applet functionality. But it's this functionality that was seized upon by both the industry and the media, a fad that allowed Java to gain acceptance and then move to the server and embedded markets where it is now used most.
It was because Sun saw the trend and seized on the opportunity of the largely useless but very hip sounding applet technology that Java is a major player in today's market. This is exactly the sort of novelty that so many new languages lack and that therefore prevents them from making a substantial splash.
I meant to point out the difficulty in maintaining code written in the language.
If your Ruby programmer leaves ofr greener pastures and the project needs an update (either due to spec change or a newfound bug), you're stuck with having to train another guy in Ruby for no other reason than making a few small changes to the existing codebase.
One can argue that java was different, because it was initially released for an in-browser operation of applets, though I don't think that would be a good point considering the direction towards server-side that has really legitimized java.
However in Java's case, I believe that the difference lies in the way that it was introduced. It was hyped (arguably over-hyped) by Sun upon its release, which assured a good corporate backing and thus a developer base. If you look at something like Ruby for aexample, it's a language that was quietly released, with noone hearing much about it. So whatever it's merits, it's a far cry from being established as an industry standard.
Which means it will be more difficult to learn, to get help with, to troubleshoot and to maintain.
It seems to me that there's currently more languages out there right now than there are programming project types. I mean I can see how you would want one language for a web application and another for a 3d game, but do we really need to have a choice of almost a dozen languages to do just a small variety of tasks? Aren't resources wasted by educating programmers to use a language which, even though it may be slightly better suited to a given task, will not save more time than it takes to master?
But if being just a web directory is a good thing, they still have that option, don't they?
Not without significantly damaging their image/brand value, I don't think. And as far as I remember Yahoo became quite a name on the net before they got into all their side ventures, so you can't say that they would have lacked brand power without them.
And much of the problem right now is that high traffic can be a liability if nobody is paying for advertising.
Not as much of a liability as keeping more people on staff because of unprofitable ventures. A skilled programmer on staff can run as much as $90,000/year with all the benefits thrown in. At current prices, you could probably get quite a bit of hosting time with that chunk of change. Consider Yahoo's current staff size and the amount of people it would take to run a Directory only site. I think that the cost would be much
greater than the decrease in revenue.
I don't doubt that they do an excellent job of providing these additional services. I just think that Yahoo should have remained in the web-directory business, and then if they wanted to do a classifieds site, they should have made that a separate venture.
A directory site requires little in the way of backend programming/resources. Or at least less than a classifieds engine. But because they're both "Yahoo", people expect the same pay structure - free.
So they have put themselves in a position where they have all these (arguably excellent) additions to their core service, but which can no longer be supported by the pure banners because of their increased overhead costs. And noone is ever going to pay for anything with the Yahoo brand, simply because it would seem like a rip off. After all, we're used to Yahoo being all about free goodies.
...is trying to grow too big. Yahoo has it's hands in everything right now, portals, search engine, web mail, shopping, auctions - you name it.
If they stayed in the web directory business which started them off (and which is something they are good at) they may not have needed to hire 10,000 people or have huge overhead operating costs, while they would still get much of the traffic and advertisement revenue they do now.
I think it's their greed that got them into the current mess. Low overhead high traffic sites is the way it should to go. How long would "Slashdot free internet" service last i wonder? Stick to what you know.
I don't think it really matters how that a Linux machine beats Windows on uptime. I'm simply stating the improvement of win2000 over any previous version. I remember when it was announced that iwndows crashes after a 48 day uptime a couple of years back, everyone just laughed it off because we were still restarting all machines every day.
Now I have a machine that I've been running for almost a month, with apache/jserv/mysql on it for development (no, it's not serving 100s of users), plus using it for a game of Unreal now and then, as well as EMACS, Outlook, and god only knows how many browser windows.
So all in all, I would say that the system has become pretty stable if you compare it even to NT4.
I mean the MS line up. I've been running win2k here at work for the last 30 days with no shutdown, and the last time I shut it down was in order to re-partition the hard drive.
You can say what you want, but I believe that this improvement in quality is caused by the DOJ hearings, at least in part. With the world spotlighting their practices and software quality, MS simply has to put its best foot forward.
Would the people benefit if they get split? I think so. After all it's always better to have a company focus on one thing instead of many product lines. I don't really want features going into my OS that will make it easier for Word to run, but not WordPerfect for example.
Let MS::OS make the best possible windows, and then MS::APP try to build the best possible Office Suite on top of it. If the two product lines stant on their own, they will have to be of better quality.
Slashdot Mirror
Without the MS JVM holding people at 1.1, you can go ahead and use your "regular" JVM from Sun/IBM with your browser, making it compatible with the new bells and whistles. (If you're in for that sort of thing, anyway).
And I think (based on a story posted about a week ago here on slashdot) that MS's OEM deals for XP will allow manufacturers to preinstall things they choose. (The article mentions this as well)However this being a manual operation might make it too costly for OEMs to pracice it widely.
Read the article. They can ship it for seven more years if they chose to.
Java used the client side applets to gain prominence, but that was mostly a gimmick. Where Java really comes into its own is on the server, where the developer can control the JVM version, and doesn't have to use Java 1.0 for compatibility's sake.
Since server based Java never relied on the MS JVM, it remains unefective and just as viable as it ever was.
Servlets
Java right now is still a moving target with a new verison coming out every half a year or so, and I don't think that bundling a JVM with an OS would have been beneficial to the language's development, as it would force developers to only use the XP version for fear of losing audience.
Isn't that pretty much a shrink-wrap licence then?
But what if I'm a freelancer who wants to exclusively sell an article to a paper printing in NY and one coming out in LA. Can I insist that if the NY paper has an LA edition my article not appear there because I have signed an exclusive contract with a different paper in that city? Is that a more similar case?
I think we should try to steer clear of legalized content control for both corporate and private content producers.
I believe this topic was previously discussed /., but no real solution arrived at.
I personally think that an ad-serving system should be developed in the open source community that would facilitate easy ad sales/link exchanges, that would be full featured enough to become somwhat of an industry standard. The Apache of ad-serving so to speak.
The existance of this "Ad-pache" would allow smaller sites to have an easier time selling ads and attracting revenue. And of course it would have to provide all the bells and whistles like time/geo targeting, etc, and possibly some of the other features not offered by anyone else (like ad feedback, etc)
With such a system in place and showing tangible results, the playing field might be leveled for the small guy.
I think that the best way to get people started on programming is by introducing concepts such as variables, data types, flow of control, functions and references. This is why I think C is what should be taught in the (High School) class room. Having come from a Pascal background I have really benefitted from being taught procedural programming before moving on to C++/Java's OOP models.
Teaching people assembler you are forcing them into the deep end of the Marianas trench right away. What's the point of explaining memory segments and offsets to someone who doesn't know such basic concepts as variables and functions? Why should I concern myself with the difference between near and far jumps before knowing the standard loop constructs that such jumps can be used for?
You mean that's NOT what really happens?
The site is slashdotted already... That usually happens at around 100 posts :)... Anyone have a mirror URL?
If a standardized contract that protects a programmer's rights to produce his own work on his time is introduced and accepted in the industry, it should give programmers leverage they need when negotiating these contracts.
Because GNU is already a recognized authority in the IT world, it would be significantly easier to make an employer agre to a contract coming from them than it would be to make them accept something drawn up by your lawyer.
Besides, Slashdot is like a LUG anyway.
It was because Sun saw the trend and seized on the opportunity of the largely useless but very hip sounding applet technology that Java is a major player in today's market. This is exactly the sort of novelty that so many new languages lack and that therefore prevents them from making a substantial splash.
I meant to point out the difficulty in maintaining code written in the language. If your Ruby programmer leaves ofr greener pastures and the project needs an update (either due to spec change or a newfound bug), you're stuck with having to train another guy in Ruby for no other reason than making a few small changes to the existing codebase.
However in Java's case, I believe that the difference lies in the way that it was introduced. It was hyped (arguably over-hyped) by Sun upon its release, which assured a good corporate backing and thus a developer base. If you look at something like Ruby for aexample, it's a language that was quietly released, with noone hearing much about it. So whatever it's merits, it's a far cry from being established as an industry standard.
Which means it will be more difficult to learn, to get help with, to troubleshoot and to maintain.
It seems to me that there's currently more languages out there right now than there are programming project types. I mean I can see how you would want one language for a web application and another for a 3d game, but do we really need to have a choice of almost a dozen languages to do just a small variety of tasks? Aren't resources wasted by educating programmers to use a language which, even though it may be slightly better suited to a given task, will not save more time than it takes to master?
Not without significantly damaging their image/brand value, I don't think. And as far as I remember Yahoo became quite a name on the net before they got into all their side ventures, so you can't say that they would have lacked brand power without them.
And much of the problem right now is that high traffic can be a liability if nobody is paying for advertising.
Not as much of a liability as keeping more people on staff because of unprofitable ventures. A skilled programmer on staff can run as much as $90,000/year with all the benefits thrown in. At current prices, you could probably get quite a bit of hosting time with that chunk of change. Consider Yahoo's current staff size and the amount of people it would take to run a Directory only site. I think that the cost would be much greater than the decrease in revenue.
A directory site requires little in the way of backend programming/resources. Or at least less than a classifieds engine. But because they're both "Yahoo", people expect the same pay structure - free.
So they have put themselves in a position where they have all these (arguably excellent) additions to their core service, but which can no longer be supported by the pure banners because of their increased overhead costs. And noone is ever going to pay for anything with the Yahoo brand, simply because it would seem like a rip off. After all, we're used to Yahoo being all about free goodies.
If they stayed in the web directory business which started them off (and which is something they are good at) they may not have needed to hire 10,000 people or have huge overhead operating costs, while they would still get much of the traffic and advertisement revenue they do now.
I think it's their greed that got them into the current mess. Low overhead high traffic sites is the way it should to go. How long would "Slashdot free internet" service last i wonder? Stick to what you know.
Now I have a machine that I've been running for almost a month, with apache/jserv/mysql on it for development (no, it's not serving 100s of users), plus using it for a game of Unreal now and then, as well as EMACS, Outlook, and god only knows how many browser windows.
So all in all, I would say that the system has become pretty stable if you compare it even to NT4.
I remember netscape 1.1n, and I hated it, used it only when required. I remember 2.0 was worse.
So were you running IE back then? Or was it Konqueror perhaps?
You can say what you want, but I believe that this improvement in quality is caused by the DOJ hearings, at least in part. With the world spotlighting their practices and software quality, MS simply has to put its best foot forward.
Would the people benefit if they get split? I think so. After all it's always better to have a company focus on one thing instead of many product lines. I don't really want features going into my OS that will make it easier for Word to run, but not WordPerfect for example.
Let MS::OS make the best possible windows, and then MS::APP try to build the best possible Office Suite on top of it. If the two product lines stant on their own, they will have to be of better quality.