Maybe because JS engines are more powerful these days than PHP? Maybe because the language primitives, and base functionality is far more consistent (not talking browser DOM)? Personally, I've been using Node.js for more and more, simply because it does the job so well, and leverages a language, that any currently working developer should probably have knowledge of.
Well, CoffeeScript seems to be the prescription you are looking for... I don't think Dart will take hold so much... this seems a lot like the EcmaScript4/ActionScript3 direction taken as a pre-processor for JS, with some decent UI hinting to go with it... I do like that the *.d.ts files are pretty self documenting, and could be useful in defining object models for use with JSON based web services... the documentation I've been creating is very similar, as it's pretty straight forward, but may actually refactor to a.d.ts file... the specific implementation details are still under way though.
I think this is far more of an extension to JS, than even CoffeeScript, which does a lot of similar things, but changes the language abstraction a lot more.
Because they don't understand how prototype based inheritance chains work, or how to utilize functional programming techniques. When you add event driven callbacks, along with non-class based structures in JS, it can be very elegant, and with Node.js + async, far more performant than other solutions..
Beyond that, a lot of it seems to line up with the EcmaScript6 work in progress spec, and Anders seems to be interested in re-aligning to that spec as it moves forward. What I found really interesting above everything else, is that it's an npm (node.js) module, and runs from node.js, not WSH, or some other bastardized environment. As I am working on publishing a JSON web service currently, I already have my object definitions documented in a JSON structure very similar, and may just adjust them all to *.d.ts, since it's fairly obvious the intent. I'm still testing using the newer WebAPI from MS, or simply going Node.js for the API service itself... leaning towards node + mongo... but the option seems fairly clean, and open.
I just hope it doesn't wind up another also ran.. It would also be nice to get.d.ts files out of coffee-script declarations for classes as well, so that it can be more interoperable. It would also be nice to see other IDEs adopt the idea. I think JS interaction, and intent of code (documentation, and interaction) is the most difficult thing, which this works to resolve... Though it does remind me a lot of ES4/AS3.
In my humble opinion, you shouldn't be able to bring someone in from overseas (H1-B) for less than 8X minimum wage... at that point, you'll find a lot of people *could* be found locally, while avoiding the slave labor offset.
Agreed, the biggest issue was probably the incomplete implementation, as well as the complexity.. say what you will, but the.(doc|xls|*)x versions are at least easier to work with... though not *that* much.
95 OSR2 had really broken USB support.. and I didn't have enough experience with NT3.5x, though NT4 was decent, it really didn't get there until SP3 iirc.
I mean Tortoise on OSX/Linix... Wish there was something that was as nice... I've tried a few plugins in the past for Nautilus etc, and on OSX, just not nearly as smooth of an integration.
Wow, that wasn't flamebait at all... There are plugins for working against SQLite, and iirc mySQL, not sure about PostgreSQL/EnterpriseDB though. MS-SQL is a pretty capable SQL database, I tend to avoid use of the more proprietary aspects of any database though.
I remember a friend of mine rotating about 3 floppies out of one machine, while downloading the rest, and installing on the final destination... this was arount 1996-97 or so.
That's pretty similar to my take... other than specialized distros of BSD or Linux for router/fw or nas solutions... Mint and Debian proper are probably my top 2 in the go to... for servers, I may go for an Ubuntu LTS release... I've been in the Debian based camp since about 2003 or so, prior to that had tried Mandrake, Redhat and SuSE.
I don't think WPF as a concept is dead... XAML is pretty much the same between WPF and Silverlight as an example... Personally, I like shoehorning as much functionality into the browser as a general application environment.. AppJS, Chrome Apps, etc are making this an even friendlier environment moving forward... Yeah, I know JS can be an exercise in masochism, just the same, it's fairly consistent, cross platform by nature, and not tied to the whims of OS changes (for the most part). You can scale from phone to full screen, and to top it off, you can release updates to everyone accessing a web-server driven environment (deployed apps are a bit different). Of course, this has been my mindset for years... it's finally at a point where most applications can be done in-browser.
What's funny, is iirc HP used to contract out the driver development to MS proper, so they could just be included.... That's before HP wanted you to install 200MB of crap, just to get the 2-5MB driver to install...
That's just funny... depending on the hardware, Drivers are one of the things that binary compatibility is nearly impossible... This is true of every OS out there... try loading a 5+ year old driver on linux with a current kernel... Most likely it needs/needed patching to even compile. It's when a kernel/OS update breaks 3/4 of the programs you use that is really irritating.
I would say that web application dev with MS tools can be really nice... as long as you understand web development, and how html/css/js work together... Most.Net web devs struggle to get the square peg of their third party commercial ui library into the round opening of the rest of their web app.
There are a number of email client-server options that compare well to Exchange + Outlook. I don't think it really needs to be integrated in to the software package, and probably better that it isn't. A lot of the time, I simply un-check Outlook from the installs (especially at home). The biggest advantage of MS Office over LO has got to be the existing system of macros in place, from data automation with Word or Excel for example. Not that it can't be done with LO, but there's a lot of existing code, examples and already in-use systems. I don't think LO wins hands down, given that a number of times I've seen MS Office Word documents that simply look too wrong in LO.
The biggest hindrance to LO is the debt in place of existing documents... I tend to install LO for any friends/family that ask about office, since far more often than not it meets their needs... In business, with more advanced usage of Excel or Word, it's a very hard sell.
I will say that I find TFS's integration to Windows to not compare to that of say SVN + TortoiseSVN. On the flip side, conflict resolution tools with TFS are quite a bit nicer than the out of the box tool with tortoise, or even WinMerge. (I do use WinMerge for directory compares, as it works quite well for that). What's funny is that when working on OSX, or in Linux is when I really miss the niceties of WinMerge. Now with reference to Git, I've yet to see a shell integration tool that is easy enough to use and doesn't suck... with Git, you're pretty much tied to having a console window open to work against.
I wouldn't say that, there's been quite a few Millitary applications of FOSS used in-house. Unfortunately a lot of these in-house systems are displaced by corporate contracts and rolled into MS based solutions, also much of it doesn't ever see the light of day.
Unfortunately, ARM doesn't quite compete at larger scales, depending on the load... A large Multi-CPU XEON or Opteron server combined with virtualization is often a better choice than a small cluster of ARM systems. It's cool, don't get me wrong. I think where an ARM cluster can excel is as front-loaded web servers, and even simple data stores, Where IO is the bottleneck, ARM can really give a lot of bang for the buck. Node.js and MongoDB come to mind, unfortunately neither are in a position where they are well supported on the platform. Another issue, is systems such as Raspberry Pi (imho the most widely available ARM platform, for building a cluster at a decent cost), don't have enough RAM to handle certain types of load well.
Unfortunately the fallout for government action, or inaction usually takes 5-20 years to actually know how the effects play out. Socialism's success can not be "proven" with a few politically popular moves. Ireland went through similar actions for a number of decades before the fallout became too bad to keep up with.
Create a web project in Eclipse... getting a hello world web-app running with integrated debugging. Now, imagine doing that with little/no experience. It's absolutely painful... Do the same in VS... There's no comparison.
I've never been more productive in Eclipse than VS... That includes related developments such as Flex/Flash Builder... It's cumbersome, and there's a ton of crap to get the simplest of projects running, let alone deployed. I'd rather use a straight text editor with Ruby, Python, Perl or Node.js than touch Eclipse if I don't have to.
Slashdot Mirror
Maybe because JS engines are more powerful these days than PHP? Maybe because the language primitives, and base functionality is far more consistent (not talking browser DOM)? Personally, I've been using Node.js for more and more, simply because it does the job so well, and leverages a language, that any currently working developer should probably have knowledge of.
Well, CoffeeScript seems to be the prescription you are looking for... I don't think Dart will take hold so much... this seems a lot like the EcmaScript4/ActionScript3 direction taken as a pre-processor for JS, with some decent UI hinting to go with it... I do like that the *.d.ts files are pretty self documenting, and could be useful in defining object models for use with JSON based web services... the documentation I've been creating is very similar, as it's pretty straight forward, but may actually refactor to a .d.ts file... the specific implementation details are still under way though.
I think this is far more of an extension to JS, than even CoffeeScript, which does a lot of similar things, but changes the language abstraction a lot more.
Because they don't understand how prototype based inheritance chains work, or how to utilize functional programming techniques. When you add event driven callbacks, along with non-class based structures in JS, it can be very elegant, and with Node.js + async, far more performant than other solutions..
Beyond that, a lot of it seems to line up with the EcmaScript6 work in progress spec, and Anders seems to be interested in re-aligning to that spec as it moves forward. What I found really interesting above everything else, is that it's an npm (node.js) module, and runs from node.js, not WSH, or some other bastardized environment. As I am working on publishing a JSON web service currently, I already have my object definitions documented in a JSON structure very similar, and may just adjust them all to *.d.ts, since it's fairly obvious the intent. I'm still testing using the newer WebAPI from MS, or simply going Node.js for the API service itself... leaning towards node + mongo... but the option seems fairly clean, and open.
.d.ts files out of coffee-script declarations for classes as well, so that it can be more interoperable. It would also be nice to see other IDEs adopt the idea. I think JS interaction, and intent of code (documentation, and interaction) is the most difficult thing, which this works to resolve... Though it does remind me a lot of ES4/AS3.
I just hope it doesn't wind up another also ran.. It would also be nice to get
what, like github?
In my humble opinion, you shouldn't be able to bring someone in from overseas (H1-B) for less than 8X minimum wage... at that point, you'll find a lot of people *could* be found locally, while avoiding the slave labor offset.
Agreed, the biggest issue was probably the incomplete implementation, as well as the complexity.. say what you will, but the .(doc|xls|*)x versions are at least easier to work with... though not *that* much.
95 OSR2 had really broken USB support.. and I didn't have enough experience with NT3.5x, though NT4 was decent, it really didn't get there until SP3 iirc.
okay.. and having a ruggedized cart or backpack isn't at all an option...
I mean Tortoise on OSX/Linix... Wish there was something that was as nice... I've tried a few plugins in the past for Nautilus etc, and on OSX, just not nearly as smooth of an integration.
Wow, that wasn't flamebait at all... There are plugins for working against SQLite, and iirc mySQL, not sure about PostgreSQL/EnterpriseDB though. MS-SQL is a pretty capable SQL database, I tend to avoid use of the more proprietary aspects of any database though.
I remember a friend of mine rotating about 3 floppies out of one machine, while downloading the rest, and installing on the final destination... this was arount 1996-97 or so.
That's pretty similar to my take... other than specialized distros of BSD or Linux for router/fw or nas solutions... Mint and Debian proper are probably my top 2 in the go to... for servers, I may go for an Ubuntu LTS release... I've been in the Debian based camp since about 2003 or so, prior to that had tried Mandrake, Redhat and SuSE.
I don't think WPF as a concept is dead... XAML is pretty much the same between WPF and Silverlight as an example... Personally, I like shoehorning as much functionality into the browser as a general application environment.. AppJS, Chrome Apps, etc are making this an even friendlier environment moving forward... Yeah, I know JS can be an exercise in masochism, just the same, it's fairly consistent, cross platform by nature, and not tied to the whims of OS changes (for the most part). You can scale from phone to full screen, and to top it off, you can release updates to everyone accessing a web-server driven environment (deployed apps are a bit different). Of course, this has been my mindset for years... it's finally at a point where most applications can be done in-browser.
I think that Windows 98, 98SE, 2000 (not ME), and 7 were pretty solid at release... Other releases, maybe not so much.
What's funny, is iirc HP used to contract out the driver development to MS proper, so they could just be included.... That's before HP wanted you to install 200MB of crap, just to get the 2-5MB driver to install...
That's just funny... depending on the hardware, Drivers are one of the things that binary compatibility is nearly impossible... This is true of every OS out there... try loading a 5+ year old driver on linux with a current kernel... Most likely it needs/needed patching to even compile. It's when a kernel/OS update breaks 3/4 of the programs you use that is really irritating.
I would say that web application dev with MS tools can be really nice... as long as you understand web development, and how html/css/js work together... Most .Net web devs struggle to get the square peg of their third party commercial ui library into the round opening of the rest of their web app.
Well, the cars seem to be able to do it faster than 120 feet per hour...
There are a number of email client-server options that compare well to Exchange + Outlook. I don't think it really needs to be integrated in to the software package, and probably better that it isn't. A lot of the time, I simply un-check Outlook from the installs (especially at home). The biggest advantage of MS Office over LO has got to be the existing system of macros in place, from data automation with Word or Excel for example. Not that it can't be done with LO, but there's a lot of existing code, examples and already in-use systems. I don't think LO wins hands down, given that a number of times I've seen MS Office Word documents that simply look too wrong in LO.
The biggest hindrance to LO is the debt in place of existing documents... I tend to install LO for any friends/family that ask about office, since far more often than not it meets their needs... In business, with more advanced usage of Excel or Word, it's a very hard sell.
I will say that I find TFS's integration to Windows to not compare to that of say SVN + TortoiseSVN. On the flip side, conflict resolution tools with TFS are quite a bit nicer than the out of the box tool with tortoise, or even WinMerge. (I do use WinMerge for directory compares, as it works quite well for that). What's funny is that when working on OSX, or in Linux is when I really miss the niceties of WinMerge. Now with reference to Git, I've yet to see a shell integration tool that is easy enough to use and doesn't suck... with Git, you're pretty much tied to having a console window open to work against.
I wouldn't say that, there's been quite a few Millitary applications of FOSS used in-house. Unfortunately a lot of these in-house systems are displaced by corporate contracts and rolled into MS based solutions, also much of it doesn't ever see the light of day.
Unfortunately, ARM doesn't quite compete at larger scales, depending on the load... A large Multi-CPU XEON or Opteron server combined with virtualization is often a better choice than a small cluster of ARM systems. It's cool, don't get me wrong. I think where an ARM cluster can excel is as front-loaded web servers, and even simple data stores, Where IO is the bottleneck, ARM can really give a lot of bang for the buck. Node.js and MongoDB come to mind, unfortunately neither are in a position where they are well supported on the platform. Another issue, is systems such as Raspberry Pi (imho the most widely available ARM platform, for building a cluster at a decent cost), don't have enough RAM to handle certain types of load well.
Unfortunately the fallout for government action, or inaction usually takes 5-20 years to actually know how the effects play out. Socialism's success can not be "proven" with a few politically popular moves. Ireland went through similar actions for a number of decades before the fallout became too bad to keep up with.
Create a web project in Eclipse... getting a hello world web-app running with integrated debugging. Now, imagine doing that with little/no experience. It's absolutely painful... Do the same in VS... There's no comparison.
I've never been more productive in Eclipse than VS... That includes related developments such as Flex/Flash Builder... It's cumbersome, and there's a ton of crap to get the simplest of projects running, let alone deployed. I'd rather use a straight text editor with Ruby, Python, Perl or Node.js than touch Eclipse if I don't have to.