It's not just about fancy effects. If you can hold the contents of each window as a surface and get the GPU to render it, the CPU is freed up to be doing other things. It makes the entire experience fast as well as allowing useful things like thumbnails, scaled windows, transparency to happen cheaply. It may be it makes the experience more flashy which I don't see as a bad thing assuming its done with some restraint. Dumping X will make things even faster because there won't be an obstinately 2D, largely obsolete, context switching bottle neck that everything has to run through.
I think that's a disadvantage of how Groovy works which is as a largely dynamically compiled language. It's going to have a greater runtime overhead than would the equivalent Java code. Dynamic compilation & type checking would hurt runtime speed, and might also prevent it benefiting from some of the optimizations that the JVM could offer it. On the flip side however, Java code is rarely performance critical and if you absolutely needed speed in one particular chunk of the program you could code it in Java. There is a Groovy++ which ironically gives Groovy some of the characteristics of vanilla Java such as static types.
I would envisage that any Java++ would be a hybrid which supported static compilation of Java code plus extensions but also had dynamic runtime functionality for things like domain specific languages. So your code for the most part would still be compiled byte code but if you wanted you could drop a bit of declarative UI, or Groovy, or Scala, or HQL or something straight in the middle of the Java class and the runtime would run it seamlessly.
Well Groovy fixes a lot of egregious problems in Java. I think if something appeared with Groovy like terseness but as a true superset of Java (i.e. a Java++) that it could do very well.
Problem as I see it is Oracle is too petrified to do anything to the language for fear of breaking it and at this point the best hope is someone else takes over. That or Oracle splits it's language development out a la Fedora vs Redhat with a stable runtime for enterprises and more frequent unstable releases with new useful stuff.
Java is fine to develop with but it hasn't kept up with either.NET, nor developments which demand a more modern language. It doesn't support domain specific languages, it's got half baked solutions for generics and other language extensions too fearful to embed stuff in class files. Java 7 is massively overdue and gimped.
Biggest issue for the amount of boiler plate crap. Things like anonymous classes where proper closures would make the code a lot cleaner. Eclipse takes care of a lot of refactoring and cleanup but it's still dealing with a lot of bloat. Other issues would be the heavy reliance on XML for control & configuration of apps. Often times you'll spend more time worrying about configuration than code.
In summary I like Java but it's not improving fast enough.
Just like COBOL is not dead. Sure, it's not the hip language, but so many legacy systems are built on it that it's basically guaranteed to live for quite a while longer. I suspect Java will have the same fate.
Java is not remotely in anywhere the same situation as Cobol. Java jobs are plentiful as is the development scene which covers everything from Android all the way up to big iron. There really isn't much to challenge the language at present though given Oracle's pathetic stewardship perhaps there should be.
Possibly because figuring out if you're infringing on 10,000+ patents is not an easy thing to do, and if you believe you are not doesn't mean you won't find yourself in a shitstorm of legal briefs while people strongarm you for money. Look at the stuff going on with Oracle and Google. Oracle are really scraping the bottom of the barrel yet they're going on for the pay day. Now imagine Microsoft (and new best friend Nokia), Apple, HP and whoever else doing the same.
Having apps modular enough to port part of an app to.net and leave part in com objects is a fucking _feature_ of com and.net.
It's a feature of Java too, JNI allows you to call any arbitrary thing you like, which is how SWT (for example) taps Win32 in Eclipse. it's just not handed to people on a plate. You have to define an interface, you have to drop into C and write a DLL, and then package it properly. As a consequence it tends to be used less frequently and the long term result is a cleaner more portable code, or at least code where the "native" bits are cleanly stuck behind an interface.
.NET's issue was it almost encouraged people to make their code dirty..NET 1.0 &.2.0 were so inadequate that it was virtually impossible to write a real world app without a few hacks into Win32 (e.g. piss poor Winforms support). And ActiveX / COM was just a wizard away. That might have been great for VB developers used to that kind of thing but it polluted the platform. I'm sure that was Microsoft's intent but it will come back to bite them on the ass. Now we have all these real world apps tied to x86 windows and often a mess of dependencies on COM, OLE, ODBC too. Now MS want to port Windows to tablets and ARM processors and they realise the mess they've made for themselves. It's not surprising they've chosen to encourage HTML. I wouldn't be surprised if.NET on ARM was just a a subset, possibly just what Windows Phone 7.x provides and nothing else.
I don't think GNOME is going for featuritis. If anything they're cutting features and functions, probably more deeply than they should be. GNOME is taking advantage of modern composition and 3D effects but I do not see that as a bad thing at all.
In general I think GNOME 3 is a very attractive desktop but it is broken in some silly ways. The offscreen activity bar is incredibly annoying for example. Switching from one screen to another to launch or change an app is the most stupid example - provide an on screen dock like representation for those who want it. The decision to prevent desktop shortcuts is also bizarre (though you can use tweak to fix that) - reinstate it as a default action.
I don't think anything qualifies as unfixable, and at least in 3.0 they can be forgiven for getting a few things wrongs. But now the devs really need to step back and think. Some of the workflow doesn't work very well and needs to be fixed in 3.2. I don't believe these things are insurmountable but it's a question of whether the devs, or the UI designers are listening which is of more concern. If you're going to change a UI, then you have to supply something as analogous and usable as what came before.
That off screen task bar thing is severely silly. I don't think GNOME 3 is unfixable but it has to realise that it needs some measure of customisation and it does need to offer functional equivalents of some of the things other modern GUIs offer.
Not sure what Fidelity app you're talking about here. Many real world.NET apps were started with the best intention of porting code running in C++ or VB over to.NET to improve the look & feel and perhaps benefit in other ways such as being a more stable development platform than C++. But usually those intentions were never fulfilled for budgetary reasons and apps become frankenstein creations of bits of old C++, ActiveX controls with.NET holding it all together, possibly attached to MSDE / MS SQL Server and other MS technologies.
It was my experience with porting apps and seeing all this stuff which convinced me that Mono was a waste of time..NET is tied to Windows and it will never change. Even if MS deprecate it, I expect it will still be around for a long time to come.
MS is at fault for making it easy to pollute.NET with unmanaged code. Maybe customers demanded it but it could have been done in a manner similar to JNI - causing enough pain to be discouraged. It might have seemed like a Great Idea at the time to tie devs into Windows, but it really tied them to x86 Windows. So.NET could well bite MS on the ass when they try and introduce ARM devices. Perhaps they think it's easier to start with a clean slate than bother sorting out the mess.
Sorry, I'm new to this. So when you develop an app, Jeff Bezos comes round to your house and says "I hear you're planning to use Android Marketplace, pity, this place is dry as tinder... one spark and it could all go up..."?
Really?
You know exactly what I mean.
Otherwise, Amazon is just going to have to offer the chance of higher sales, better margins, better promotion etc. to justify the higher costs. Oh, and they give you a free year to try it out and decide if its worth $100/year to you.
How gracious of them to waive it. I sure hope there is no autorenewal or contractual clauses to catch you out there.
Plus, $100 is chump change - about the equivalent of "serious callers only". If you don't sell enough to easily cover that then (lets break it to you gently) nobody wants your product. You seem to be looking for a vanity publishing service.
a) Not all apps are released to make money b) Not all free apps are useless or not worth using c) $100 is $100 less in a developer's pocket. It's unjustifiable especially considering Marketplace charges $25 for lifetime exposure to 100x the number of users.
If you said that they'd only give you 70% of $2 then I might be more sympathetic, but if they effectively set a lower limit of 20% on your take, meaning that they get nothing on a $2 sale, that sounds more than fair. Their motive for doing that would be to drive sales, so they get more traffic and you stand to make a lot more $2s than you would $7s.
They could sell the title for $2.85 and you'd still only get $2. And why you think you'd make more money on an app store with 1/100th the users (most of whom are simply after the free downloads) is beyond me. What is more likely to happen is devs will wise up and shun the store, or release "Amazon edition" versions of their apps, or rebrand them entirely, or gimp them. Just enough difference that they can contractually justify setting different prices to compensate for the Amazon heavily discounting their titles in the first place.
Except that, if you read between the whines, TFA makes it quite clear that they explicitly write and ask for separate permission before giving away your app for nothing. As does your original statement that they could discount all the way to 20%of the list price.
And if you read the comments you will see they retroactively changed their terms so people were under the impression they got 20% but actually received nothing. So in this instance 100,000 freeloaders got their app and the developer was burdened with supporting them for zero gain.
The way the store works are present is simply broken. Perhaps it's no big deal for the large developers who have lawyers and salesmen and the clout to negotiate their own deals. It certainly is for anyone stuck with the default one sided terms and conditions.
More likely they'd get a slap on the wrist and a large but not disproportionate fine. Not too long ago a UK firm called ACS:Law did try to shake down potential P2P filesharers usually through threats of legal action and summary judgements filed in bulk. Things didn't go too well for them what with being hacked, customer data being leaked, complaints to the Information Commissioner (over the breach), fines and judicial criticism.
Unless the app provided some kind of subscription service and a free trial, or ad revenue, or some other means to collect money people who pay nothing are not customers. They're freeloaders. Who can blame them for grabbing a free app but they're not contributing to the bottom line.
The concept that they'll become regular customers is also as flawed as the one used by Groupon. People who are looking for extreme bargains are not interested in sticking around when the deal is over. They'll just swoop off somewhere else looking for the next one.
I think Marketplace isn't perfect but it is pretty transparent in how it works - you set the price, you split the sale 70:30 in your favour. It's also transparent in how you register, upload apps and everything else - pay a $25 fee and then set the price, wording screenshots etc for yourself. After a few minutes your app appears.
There is no arguing that they should be allowed to compete. The point I expect most devs would make is that 20% of the list price does not represent it's "wholesale" cost.
Amazon is poised to launch a bunch of tablets of their own. A captive audience. Once that happens you can bet that apps will still be deep discounted to give the impression that the app store is cheaper than marketplace. Which it is of course but only by fucking over developers in the process.
And that 20% discount is a killer. You list your app on Android Marketplace for $10, you're compelled to list it for $10 Amazon's App Store. But then Amazon goes and sells it for $2 and you have no input in this. The amount you get paid is 70% of the sale price or 20% of the list price, whichever is higher. So if they sell for $2 you get $2. If they sell for $2.85 or less you still only get $2. If they sell for $5 you get $3.5 etc. In virtually every case you're going to get less revenue than on Marketplace, possibly a lot less revenue.
It's a scam for developers and one could argue it's anti-competitive too since it means Amazon's sale price always matches or undercuts the price on Marketplace. A fairer system which would allow them to undercut would be to allow devs to set a wholesale price and an MSRP. Devs would always get 100% of the wholesale price. And Amazon could discount anywhere between MSRP and wholesale that they wished. It allows devs to set their "bottom line". Chances are Google would reciprocate with a similar scheme so that both stores would be competing against each other rather than stiffing developers.
In the absence of this I strongly expect that devs will produce a "special edition" of their apps on Amazon which prevents like for like comparison and lets them adjust their pricing in a more favourable way. I expect the Amazon edition will find it gimped in some way, or the price bloated by some minimal extra functionality so devs can recoup their costs for using the app store.
Amazon's robbery starts the minute they sign up - a $100 annual fee, graciously waived the first year. What a bargain compared to the $25 one time registration on Android Marketplace. So if you want to write free apps or esoteric niche apps you can fuck off right now because you're going to pay through your own pocket to do it.
Then the robbery continues on the terms and conditions. You can set the list price of your app but it cannot be any higher on Amazon Appstore than it is on other app stores. So if the list price is $9.99 on Marketplace it has to be $9.99 (or less) on App Store. And like Marketplace you take a 70% of the sale price. But... the sale price and list price are not the same thing. You set the list price, Amazon sets the sale price. Amazon can deep discount your app all the way down to 20% of its list price if they like. So instead of getting $7 revenue from $9.99 you get $2 from $2. Or Amazon could even give the app away and you get nothing at all. Basically they're using your app to fuck you over.
Frankly it's daylight robbery. I'm quite certain that Amazon will launch their own tablets before long and some app writers will be compelled to serve this market, but I wonder if there will be a general revolt over what is going on. They're not in a position to dictate this stuff as Apple might. Perhaps Amazon tablet users can enjoy gimped "Amazon edition" apps where functionality has to be unlocked with micropayments or similar. I certainly don't see the situation as being tolerable even for the more popular apps.
A more fair system if Amazon preferred to compete with other marketplaces would be to allow the developer to set the wholesale price and an MSRP. Amazon would discount within those two bounds, presumably by comparing the price on Marketplace and undercutting it but within agreed limits. I expect Google would reciprocate with its own version of the same and the two markets would compete much like stores do for real world goods. That would be fair. The current situation absolutely isn't.
Peaceful protest my ass. They're children taking down sites and then engaging in some post hoc ergo propter hoc justification for doing it. There is no point believing there is anything more to it than that because their isn't.
Slashdot Mirror
Pray tell what difference it would make please.
We will miss X greatly. Why this push lately to screw up the Linux desktop, anyway?
Run X over Wayland as a client. In exactly the same way that X is supported Windows or OS X.
It's not just about fancy effects. If you can hold the contents of each window as a surface and get the GPU to render it, the CPU is freed up to be doing other things. It makes the entire experience fast as well as allowing useful things like thumbnails, scaled windows, transparency to happen cheaply. It may be it makes the experience more flashy which I don't see as a bad thing assuming its done with some restraint. Dumping X will make things even faster because there won't be an obstinately 2D, largely obsolete, context switching bottle neck that everything has to run through.
Half the fun of Nethack was in finding exploits.
I would envisage that any Java++ would be a hybrid which supported static compilation of Java code plus extensions but also had dynamic runtime functionality for things like domain specific languages. So your code for the most part would still be compiled byte code but if you wanted you could drop a bit of declarative UI, or Groovy, or Scala, or HQL or something straight in the middle of the Java class and the runtime would run it seamlessly.
Well Groovy fixes a lot of egregious problems in Java. I think if something appeared with Groovy like terseness but as a true superset of Java (i.e. a Java++) that it could do very well. Problem as I see it is Oracle is too petrified to do anything to the language for fear of breaking it and at this point the best hope is someone else takes over. That or Oracle splits it's language development out a la Fedora vs Redhat with a stable runtime for enterprises and more frequent unstable releases with new useful stuff.
Biggest issue for the amount of boiler plate crap. Things like anonymous classes where proper closures would make the code a lot cleaner. Eclipse takes care of a lot of refactoring and cleanup but it's still dealing with a lot of bloat. Other issues would be the heavy reliance on XML for control & configuration of apps. Often times you'll spend more time worrying about configuration than code.
In summary I like Java but it's not improving fast enough.
Just like COBOL is not dead. Sure, it's not the hip language, but so many legacy systems are built on it that it's basically guaranteed to live for quite a while longer. I suspect Java will have the same fate.
Java is not remotely in anywhere the same situation as Cobol. Java jobs are plentiful as is the development scene which covers everything from Android all the way up to big iron. There really isn't much to challenge the language at present though given Oracle's pathetic stewardship perhaps there should be.
Possibly because figuring out if you're infringing on 10,000+ patents is not an easy thing to do, and if you believe you are not doesn't mean you won't find yourself in a shitstorm of legal briefs while people strongarm you for money. Look at the stuff going on with Oracle and Google. Oracle are really scraping the bottom of the barrel yet they're going on for the pay day. Now imagine Microsoft (and new best friend Nokia), Apple, HP and whoever else doing the same.
Ah good, not the app I was working on then :)
Having apps modular enough to port part of an app to .net and leave part in com objects is a fucking _feature_ of com and .net.
It's a feature of Java too, JNI allows you to call any arbitrary thing you like, which is how SWT (for example) taps Win32 in Eclipse. it's just not handed to people on a plate. You have to define an interface, you have to drop into C and write a DLL, and then package it properly. As a consequence it tends to be used less frequently and the long term result is a cleaner more portable code, or at least code where the "native" bits are cleanly stuck behind an interface.
I don't think GNOME is going for featuritis. If anything they're cutting features and functions, probably more deeply than they should be. GNOME is taking advantage of modern composition and 3D effects but I do not see that as a bad thing at all.
I don't think anything qualifies as unfixable, and at least in 3.0 they can be forgiven for getting a few things wrongs. But now the devs really need to step back and think. Some of the workflow doesn't work very well and needs to be fixed in 3.2. I don't believe these things are insurmountable but it's a question of whether the devs, or the UI designers are listening which is of more concern. If you're going to change a UI, then you have to supply something as analogous and usable as what came before.
That off screen task bar thing is severely silly. I don't think GNOME 3 is unfixable but it has to realise that it needs some measure of customisation and it does need to offer functional equivalents of some of the things other modern GUIs offer.
It was my experience with porting apps and seeing all this stuff which convinced me that Mono was a waste of time. .NET is tied to Windows and it will never change. Even if MS deprecate it, I expect it will still be around for a long time to come.
MS is at fault for making it easy to pollute .NET with unmanaged code. Maybe customers demanded it but it could have been done in a manner similar to JNI - causing enough pain to be discouraged. It might have seemed like a Great Idea at the time to tie devs into Windows, but it really tied them to x86 Windows. So .NET could well bite MS on the ass when they try and introduce ARM devices. Perhaps they think it's easier to start with a clean slate than bother sorting out the mess.
Sorry, I'm new to this. So when you develop an app, Jeff Bezos comes round to your house and says "I hear you're planning to use Android Marketplace, pity, this place is dry as tinder... one spark and it could all go up..."? Really?
You know exactly what I mean.
Otherwise, Amazon is just going to have to offer the chance of higher sales, better margins, better promotion etc. to justify the higher costs. Oh, and they give you a free year to try it out and decide if its worth $100/year to you.
How gracious of them to waive it. I sure hope there is no autorenewal or contractual clauses to catch you out there.
Plus, $100 is chump change - about the equivalent of "serious callers only". If you don't sell enough to easily cover that then (lets break it to you gently) nobody wants your product. You seem to be looking for a vanity publishing service.
a) Not all apps are released to make money b) Not all free apps are useless or not worth using c) $100 is $100 less in a developer's pocket. It's unjustifiable especially considering Marketplace charges $25 for lifetime exposure to 100x the number of users.
If you said that they'd only give you 70% of $2 then I might be more sympathetic, but if they effectively set a lower limit of 20% on your take, meaning that they get nothing on a $2 sale, that sounds more than fair. Their motive for doing that would be to drive sales, so they get more traffic and you stand to make a lot more $2s than you would $7s.
They could sell the title for $2.85 and you'd still only get $2. And why you think you'd make more money on an app store with 1/100th the users (most of whom are simply after the free downloads) is beyond me. What is more likely to happen is devs will wise up and shun the store, or release "Amazon edition" versions of their apps, or rebrand them entirely, or gimp them. Just enough difference that they can contractually justify setting different prices to compensate for the Amazon heavily discounting their titles in the first place.
Except that, if you read between the whines, TFA makes it quite clear that they explicitly write and ask for separate permission before giving away your app for nothing. As does your original statement that they could discount all the way to 20%of the list price.
And if you read the comments you will see they retroactively changed their terms so people were under the impression they got 20% but actually received nothing. So in this instance 100,000 freeloaders got their app and the developer was burdened with supporting them for zero gain.
The way the store works are present is simply broken. Perhaps it's no big deal for the large developers who have lawyers and salesmen and the clout to negotiate their own deals. It certainly is for anyone stuck with the default one sided terms and conditions.
More likely they'd get a slap on the wrist and a large but not disproportionate fine. Not too long ago a UK firm called ACS:Law did try to shake down potential P2P filesharers usually through threats of legal action and summary judgements filed in bulk. Things didn't go too well for them what with being hacked, customer data being leaked, complaints to the Information Commissioner (over the breach), fines and judicial criticism.
The concept that they'll become regular customers is also as flawed as the one used by Groupon. People who are looking for extreme bargains are not interested in sticking around when the deal is over. They'll just swoop off somewhere else looking for the next one.
I think Marketplace isn't perfect but it is pretty transparent in how it works - you set the price, you split the sale 70:30 in your favour. It's also transparent in how you register, upload apps and everything else - pay a $25 fee and then set the price, wording screenshots etc for yourself. After a few minutes your app appears.
There is no arguing that they should be allowed to compete. The point I expect most devs would make is that 20% of the list price does not represent it's "wholesale" cost.
Amazon is poised to launch a bunch of tablets of their own. A captive audience. Once that happens you can bet that apps will still be deep discounted to give the impression that the app store is cheaper than marketplace. Which it is of course but only by fucking over developers in the process.
It's a scam for developers and one could argue it's anti-competitive too since it means Amazon's sale price always matches or undercuts the price on Marketplace. A fairer system which would allow them to undercut would be to allow devs to set a wholesale price and an MSRP. Devs would always get 100% of the wholesale price. And Amazon could discount anywhere between MSRP and wholesale that they wished. It allows devs to set their "bottom line". Chances are Google would reciprocate with a similar scheme so that both stores would be competing against each other rather than stiffing developers.
In the absence of this I strongly expect that devs will produce a "special edition" of their apps on Amazon which prevents like for like comparison and lets them adjust their pricing in a more favourable way. I expect the Amazon edition will find it gimped in some way, or the price bloated by some minimal extra functionality so devs can recoup their costs for using the app store.
Then the robbery continues on the terms and conditions. You can set the list price of your app but it cannot be any higher on Amazon Appstore than it is on other app stores. So if the list price is $9.99 on Marketplace it has to be $9.99 (or less) on App Store. And like Marketplace you take a 70% of the sale price. But... the sale price and list price are not the same thing. You set the list price, Amazon sets the sale price. Amazon can deep discount your app all the way down to 20% of its list price if they like. So instead of getting $7 revenue from $9.99 you get $2 from $2. Or Amazon could even give the app away and you get nothing at all. Basically they're using your app to fuck you over.
Frankly it's daylight robbery. I'm quite certain that Amazon will launch their own tablets before long and some app writers will be compelled to serve this market, but I wonder if there will be a general revolt over what is going on. They're not in a position to dictate this stuff as Apple might. Perhaps Amazon tablet users can enjoy gimped "Amazon edition" apps where functionality has to be unlocked with micropayments or similar. I certainly don't see the situation as being tolerable even for the more popular apps.
A more fair system if Amazon preferred to compete with other marketplaces would be to allow the developer to set the wholesale price and an MSRP. Amazon would discount within those two bounds, presumably by comparing the price on Marketplace and undercutting it but within agreed limits. I expect Google would reciprocate with its own version of the same and the two markets would compete much like stores do for real world goods. That would be fair. The current situation absolutely isn't.
Peaceful protest my ass. They're children taking down sites and then engaging in some post hoc ergo propter hoc justification for doing it. There is no point believing there is anything more to it than that because their isn't.
Don't be ridiculous