World: r3wp
[!Cheyenne] Discussions about the Cheyenne Web Server
older newer | first last |
Dockimbel 9-Jul-2010 [8397x3] | Graham, without money, it's just not worth it, you need a lot of libraries to make a programming language useful, no way one man only can build all the required ones...but if it can generate enough incomes, you can pay some developers for that. |
Terry: when he issue is either related to the brower or the OS | |
*then | |
Terry 9-Jul-2010 [8400] | so it connects, C (short for Cheyenne.. don't know why I didn't think of THAT before) prints to the console "connected" .. client disconnects and no response from C |
Graham 9-Jul-2010 [8401] | But if you run on the JVM or .net, then won't you have access to the libraries then? |
Dockimbel 9-Jul-2010 [8402x2] | Terry: enable verbose mode in C starting it with -vvvvv to see if C see the request |
Graham: sure, but either you build an abstraction layer other each java lib (using a scheme or a dialect), or you'll just use these libs with the java syntax, so better use java or scala directly. | |
Terry 9-Jul-2010 [8404] | oops my bad, C does notice the disconnect.. the function was empty :( |
Dockimbel 9-Jul-2010 [8405] | Graham: without a tight and higher level integration of lower level libs, the benefits of a higher language like REBOL will be reduced greatly while you'll pay the performance penality. |
Terry 9-Jul-2010 [8406] | The world has enough languages. |
Graham 9-Jul-2010 [8407] | So, why do they have languages such as JRuby which in some instances can be faster than C Ruby? |
Dockimbel 9-Jul-2010 [8408] | I guess it's edge cases, not performances from usual cases (got any URL for those benchmarks?) |
Graham 9-Jul-2010 [8409] | http://programmingzen.com/2008/12/09/the-great-ruby-shootout-december-2008/ |
Terry 9-Jul-2010 [8410] | Same function, different syntax. Performance is hardly an issue anymore unless you're gaming or modelling |
Dockimbel 9-Jul-2010 [8411x2] | Graham: thanks, interesting reading :-) |
Terry: interpreted (REBOL) vs compiled (java) language will still have a huge performance gap | |
Terry 9-Jul-2010 [8413] | I have a PHP library that I've been using personally for 5 or 6 years now. I'm sure it's of value to someone, but I just can't be bothered to market it to the 500 interested individuals |
Dockimbel 9-Jul-2010 [8414] | Graham: my performance mention was for REBOL vs java, not cREBOL vs jREBOL |
Terry 9-Jul-2010 [8415] | Yeah Doc, but who cares.. what's 2 ms these days? I/O is the bottleneck. |
Dockimbel 9-Jul-2010 [8416] | still, I believe that it would be very hard (if even possible) to make a jREBOL with performances matching the C version. |
Graham 9-Jul-2010 [8417] | the time gained is in writing .. not execution :) |
Terry 9-Jul-2010 [8418x2] | I would (and do) focus on the future.. HTML5, websockets.. less on the tools, more on the results. |
Develop mobile apps rather than low level languages.. the market is just too small for the latter. | |
Graham 9-Jul-2010 [8420] | And Moore's law solves the speed problem |
Terry 9-Jul-2010 [8421x2] | Lik e Cocoa.. I've tried looking at that noise a few times now.. syntax boggles my mind.. but much of it (iphone apps etc).can be done with JS and HTML5.. |
http://www.sencha.com/ | |
Graham 9-Jul-2010 [8423] | I think Google use Java to write their JS/HTML |
Dockimbel 9-Jul-2010 [8424] | Sorry, but Moore's law doesn't allow REBOL to be used as a generic programming language, due to poor performances compared to compiled ones (like C or java). You can't even write a decent compression lib in REBOL (would be too slow). |
Terry 9-Jul-2010 [8425] | which is why rebol is dying.. it's a dinosaur.. 10 years ago it was hot, but the ball was dropped, and ruby took it's place. Stupid license / closed source killed it. The only thing is for a few folk here who prefer to use it for low end development / back office. |
Dockimbel 9-Jul-2010 [8426x2] | Google use their GWT library. |
*uses | |
Graham 9-Jul-2010 [8428] | So, just call a compression library from the jvm |
Dockimbel 9-Jul-2010 [8429] | write in java and compiles to JS/HTML for client side |
Graham 9-Jul-2010 [8430] | that's what I said ! |
Dockimbel 9-Jul-2010 [8431] | Graham: that was just an example of forbidden usage to *REBOL (as long as it is interpreted), performance is still relevant despite of current CPU speed. |
Maxim 9-Jul-2010 [8432] | yes especially on web servers.... the number of users of a site can quickly slam moore`s law. |
Graham 9-Jul-2010 [8433] | Sure .. I understand that for interpreted languages, there are also performance downsides .. but the gain is in the time to deliver product |
Maxim 9-Jul-2010 [8434] | exactly. |
Graham 9-Jul-2010 [8435x2] | That's why people use interpreted languages/scripting languages .. to speed up the development cycle not for writing time critical apps |
If I write a WP ... it's not computationally expensive ... vs a statistical package | |
Dockimbel 9-Jul-2010 [8437] | WP: WordProcessor? |
Graham 9-Jul-2010 [8438] | yes |
Dockimbel 9-Jul-2010 [8439] | REBOL is not even fast enough to write a code editor with syntax coloring...a WP is out of reach ;-) |
Graham 9-Jul-2010 [8440x3] | My CRM apps .. the bottleneck is I/O not Rebol |
That's a fault with the view implementation | |
So, what advantages do you see being lost with a port to the JVM or .Net ? | |
Dockimbel 9-Jul-2010 [8443] | compactness :) |
Maxim 9-Jul-2010 [8444] | actually, REBOL is fast enough... its the R2 integration to AGG is which is quite poor. |
Graham 9-Jul-2010 [8445] | doesn't sound like a big loss .. most people I see already have .net or the jre already installed |
Dockimbel 9-Jul-2010 [8446] | but Graham, I'm not against a jvm and .net port, it would be a good thing |
older newer | first last |