web IDE

classic Classic list List threaded Threaded
57 messages Options
123
Reply | Threaded
Open this post in threaded view
|

web IDE

tong-2
hi,

how about starting a community project for creating an open source IDE
for hx (and other languages) targeting the browser ?
something like bespin[0], etherpad[1] , .. [2] ..
.. a client server architecture with a html/js frontend and a
compiler/data backend.


i see some benefits targeting the browser ...

.. for developing (the IDE itself):

* haXe code
* most of us are familiar with html, css, js (,flash) (no complicated
eclipse-plugin-development setup or something)
* easy plugin development (all haXe).
* it might be easy to integrate other tools like haxelib, automatic unit
tests.
* we already have tools and libs at our hands we can use, they would
improve as reflected.

 .. in using:
 
* zero install
* cross platform (even mobile devices, ipad, chromeOS ...)
* embeddable into existing IDEs since some provide a internal web
browser (eclipse, netbeans,..)
* embeddable into browser, no task switching anymore


 .. some ideas:

* group editing (?)
* if webkit gets added to SWHX (got mentioned on the list) we could
build a standalone desktop version out of it.
* create a desktop version anyhow using air, titanium, ...
* create browsers extensions to add additional functionality, (which is
pretty easy these days, at least for firefox[3] and chrome[4] )
* research into new architectures of IDEs like codebubbles[5], which
looks like a fit for hxs OOP style
* integrate into wave (?)


drawbacks ?:

- performance



what do you think ?


/tong


[0] https://bespin.mozilla.com/
[1] http://etherpad.com/
[2] http://www.catswhocode.com/blog/10-useful-online-code-editors
[3] https://jetpack.mozillalabs.com/
[4] http://code.google.com/chrome/extensions/
[5] http://www.cs.brown.edu/people/acb/codebubbles_site.htm



--
[)   |   5   |<   †   |2   3   3


--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

tom rhodes
have you seen minibuilder?


works in the browser and as an air app, i've done some very simple stuff in ubuntu with the air variety. 

On 17 March 2010 14:36, tong <[hidden email]> wrote:
hi,

how about starting a community project for creating an open source IDE
for hx (and other languages) targeting the browser ?
something like bespin[0], etherpad[1] , .. [2] ..
.. a client server architecture with a html/js frontend and a
compiler/data backend.


i see some benefits targeting the browser ...

.. for developing (the IDE itself):

* haXe code
* most of us are familiar with html, css, js (,flash) (no complicated
eclipse-plugin-development setup or something)
* easy plugin development (all haXe).
* it might be easy to integrate other tools like haxelib, automatic unit
tests.
* we already have tools and libs at our hands we can use, they would
improve as reflected.

 .. in using:

* zero install
* cross platform (even mobile devices, ipad, chromeOS ...)
* embeddable into existing IDEs since some provide a internal web
browser (eclipse, netbeans,..)
* embeddable into browser, no task switching anymore


 .. some ideas:

* group editing (?)
* if webkit gets added to SWHX (got mentioned on the list) we could
build a standalone desktop version out of it.
* create a desktop version anyhow using air, titanium, ...
* create browsers extensions to add additional functionality, (which is
pretty easy these days, at least for firefox[3] and chrome[4] )
* research into new architectures of IDEs like codebubbles[5], which
looks like a fit for hxs OOP style
* integrate into wave (?)


drawbacks ?:

- performance



what do you think ?


/tong


[0] https://bespin.mozilla.com/
[1] http://etherpad.com/
[2] http://www.catswhocode.com/blog/10-useful-online-code-editors
[3] https://jetpack.mozillalabs.com/
[4] http://code.google.com/chrome/extensions/
[5] http://www.cs.brown.edu/people/acb/codebubbles_site.htm



--
[)   |   5   |<   †   |2   3   3


--
haXe - an open source web programming language
http://haxe.org


--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

Gamehaxe
In reply to this post by tong-2
Hi,
Sort of an interesting idea, but I think the file-system integration
may be a bit of a downer.

The killer feature for me would be to have an on-line compiler, to get
true zero-install.  The user creates an account (or uses a "guest" account)
on a server that manages the project on-line, and can compile swf/js
in the background and provide a live "test" button.
You could have various libraries installed too, so it would be
easy to create a simple game/web-banner/whatever.  You could also
host the resulting swf with a good url, eg:  
http://serverurl.com/user_name/game_name.html.

So I like the idea, but add a server and you could get a killer app.

Hugh


> hi,
>
> how about starting a community project for creating an open source IDE
> for hx (and other languages) targeting the browser ?
> something like bespin[0], etherpad[1] , .. [2] ..
> .. a client server architecture with a html/js frontend and a
> compiler/data backend.
>
>
> i see some benefits targeting the browser ...
>
> .. for developing (the IDE itself):
>
> * haXe code
> * most of us are familiar with html, css, js (,flash) (no complicated
> eclipse-plugin-development setup or something)
> * easy plugin development (all haXe).
> * it might be easy to integrate other tools like haxelib, automatic unit
> tests.
> * we already have tools and libs at our hands we can use, they would
> improve as reflected.
>
>  .. in using:
> * zero install
> * cross platform (even mobile devices, ipad, chromeOS ...)
> * embeddable into existing IDEs since some provide a internal web
> browser (eclipse, netbeans,..)
> * embeddable into browser, no task switching anymore
>
>
>  .. some ideas:
>
> * group editing (?)
> * if webkit gets added to SWHX (got mentioned on the list) we could
> build a standalone desktop version out of it.
> * create a desktop version anyhow using air, titanium, ...
> * create browsers extensions to add additional functionality, (which is
> pretty easy these days, at least for firefox[3] and chrome[4] )
> * research into new architectures of IDEs like codebubbles[5], which
> looks like a fit for hxs OOP style
> * integrate into wave (?)
>
>
> drawbacks ?:
>
> - performance
>
>
>
> what do you think ?
>
>
> /tong
>
>
> [0] https://bespin.mozilla.com/
> [1] http://etherpad.com/
> [2] http://www.catswhocode.com/blog/10-useful-online-code-editors
> [3] https://jetpack.mozillalabs.com/
> [4] http://code.google.com/chrome/extensions/
> [5] http://www.cs.brown.edu/people/acb/codebubbles_site.htm
>
>
> 


--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

tong-2
On Wed, 2010-03-17 at 22:20 +0800, Hugh Sanderson wrote:

> Hi,
> Sort of an interesting idea, but I think the file-system integration
> may be a bit of a downer.
>
> The killer feature for me would be to have an on-line compiler, to get
> true zero-install.  The user creates an account (or uses a "guest" account)
> on a server that manages the project on-line, and can compile swf/js
> in the background and provide a live "test" button.
> You could have various libraries installed too, so it would be
> easy to create a simple game/web-banner/whatever.  You could also
> host the resulting swf with a good url, eg:  
> http://serverurl.com/user_name/game_name.html.
>
> So I like the idea, but add a server and you could get a killer app.
>
> Hugh


yes. exactly.
thats what i meant with client-server architecture.
.. a remote (or local) running compiler and data backend server.

just optional:
.. with a well designed API, so we can embed the server into a desktop
application by replacing the network stuff with local filesystem and
compiler access, haxe multitarget style.



--
[)   |   5   |<   †   |2   3   3


--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

tong-2
In reply to this post by tom rhodes
On Wed, 2010-03-17 at 15:20 +0100, tom rhodes wrote:
> have you seen minibuilder?
>
>
> http://code.google.com/p/minibuilder/
>

no.
nice.
thats what i mean.


>
> works in the browser and as an air app, i've done some very simple
> stuff in ubuntu with the air variety.
>
> On 17 March 2010 14:36, tong <[hidden email]> wrote:
>         hi,
>        
>         how about starting a community project for creating an open
>         source IDE
>         for hx (and other languages) targeting the browser ?
>         something like bespin[0], etherpad[1] , .. [2] ..
>         .. a client server architecture with a html/js frontend and a
>         compiler/data backend.
>        
>        
>         i see some benefits targeting the browser ...
>        
>         .. for developing (the IDE itself):
>        
>         * haXe code
>         * most of us are familiar with html, css, js (,flash) (no
>         complicated
>         eclipse-plugin-development setup or something)
>         * easy plugin development (all haXe).
>         * it might be easy to integrate other tools like haxelib,
>         automatic unit
>         tests.
>         * we already have tools and libs at our hands we can use, they
>         would
>         improve as reflected.
>        
>          .. in using:
>        
>         * zero install
>         * cross platform (even mobile devices, ipad, chromeOS ...)
>         * embeddable into existing IDEs since some provide a internal
>         web
>         browser (eclipse, netbeans,..)
>         * embeddable into browser, no task switching anymore
>        
>        
>          .. some ideas:
>        
>         * group editing (?)
>         * if webkit gets added to SWHX (got mentioned on the list) we
>         could
>         build a standalone desktop version out of it.
>         * create a desktop version anyhow using air, titanium, ...
>         * create browsers extensions to add additional functionality,
>         (which is
>         pretty easy these days, at least for firefox[3] and
>         chrome[4] )
>         * research into new architectures of IDEs like codebubbles[5],
>         which
>         looks like a fit for hxs OOP style
>         * integrate into wave (?)
>        
>        
>         drawbacks ?:
>        
>         - performance
>        
>        
>        
>         what do you think ?
>        
>        
>         /tong
>        
>        
>         [0] https://bespin.mozilla.com/
>         [1] http://etherpad.com/
>         [2]
>         http://www.catswhocode.com/blog/10-useful-online-code-editors
>         [3] https://jetpack.mozillalabs.com/
>         [4] http://code.google.com/chrome/extensions/
>         [5] http://www.cs.brown.edu/people/acb/codebubbles_site.htm
>        
>        
>         
>         --
>         [)   |   5   |<   †   |2   3   3
>        
>        
>         --
>         haXe - an open source web programming language
>         http://haxe.org
>
>



--
[)   |   5   |<   †   |2   3   3


--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

Gamehaxe
In reply to this post by tong-2
Hi,

> yes. exactly.
> thats what i meant with client-server architecture.
> .. a remote (or local) running compiler and data backend server.

Cool - wasn't sure.
You could start with some existing js control and see what
sort of server load/architecture is required.  You might be able to get a  
tech
demo going in a week or two.  Much more fun discussing these  things
with something concrete to look at (also easier to rope in helpers).

Someone needs to do a haxe port of haxe.  Then you could run it all
in the browser/air app.

Hugh

--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

Michiel Crefcoeur
I've thought of the idea before but I have no time to work on it but here's a tip:

First step should be to write a language/syntax definition for one of these readily available web-based code editors:

Porting one of these to haXe first would make it a bit more interesting.
After that, it shouldn't be too hard to set up a haXe based serverside backend.

2010/3/17 Hugh Sanderson <[hidden email]>
Hi,


yes. exactly.
thats what i meant with client-server architecture.
.. a remote (or local) running compiler and data backend server.

Cool - wasn't sure.
You could start with some existing js control and see what
sort of server load/architecture is required.  You might be able to get a tech
demo going in a week or two.  Much more fun discussing these  things
with something concrete to look at (also easier to rope in helpers).

Someone needs to do a haxe port of haxe.  Then you could run it all
in the browser/air app.

Hugh


--
haXe - an open source web programming language
http://haxe.org


--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

Laurent Kappler
In reply to this post by tong-2
tong a écrit :

> hi,
>
> how about starting a community project for creating an open source IDE
> for hx (and other languages) targeting the browser ?
> something like bespin[0], etherpad[1] , .. [2] ..
> .. a client server architecture with a html/js frontend and a
> compiler/data backend.
>
>
> i see some benefits targeting the browser ...
>
> .. for developing (the IDE itself):
>
> * haXe code
> * most of us are familiar with html, css, js (,flash) (no complicated
> eclipse-plugin-development setup or something)
> * easy plugin development (all haXe).
> * it might be easy to integrate other tools like haxelib, automatic unit
> tests.
> * we already have tools and libs at our hands we can use, they would
> improve as reflected.
>
>  .. in using:
>  
> * zero install
> * cross platform (even mobile devices, ipad, chromeOS ...)
> * embeddable into existing IDEs since some provide a internal web
> browser (eclipse, netbeans,..)
> * embeddable into browser, no task switching anymore
>
>
>  .. some ideas:
>
> * group editing (?)
> * if webkit gets added to SWHX (got mentioned on the list) we could
> build a standalone desktop version out of it.
> * create a desktop version anyhow using air, titanium, ...
> * create browsers extensions to add additional functionality, (which is
> pretty easy these days, at least for firefox[3] and chrome[4] )
> * research into new architectures of IDEs like codebubbles[5], which
> looks like a fit for hxs OOP style
> * integrate into wave (?)
>
>
> drawbacks ?:
>
> - performance
>  
I think performance with the V8 javascript engine is not so much trouble
anymore.

On the global idea I think it's very cool. It will help the community to
developpe some strong javascript library and define a GUI design.
I would like to help somehow. I'm not sure I'm qualified enough.
I'm working on javascript right now, making the website working more as
a client application.
We could build components with html/css/js, that only load data and
format it on client side instead of requesting the server for formating
the requested data to display.

I'm following :)

--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

Tony Polinelli
what you really need to do is bootstrap the compiler in haxe so that
you can compile on the clientside ;P hahah ooooh yeah



On Thu, Mar 18, 2010 at 6:09 AM, Laurent Kappler
<[hidden email]> wrote:

> tong a écrit :
>>
>> hi,
>>
>> how about starting a community project for creating an open source IDE
>> for hx (and other languages) targeting the browser ?
>> something like bespin[0], etherpad[1] , .. [2] ..
>> .. a client server architecture with a html/js frontend and a
>> compiler/data backend.
>>
>>
>> i see some benefits targeting the browser ...
>>
>> .. for developing (the IDE itself):
>>
>> * haXe code
>> * most of us are familiar with html, css, js (,flash) (no complicated
>> eclipse-plugin-development setup or something)
>> * easy plugin development (all haXe).
>> * it might be easy to integrate other tools like haxelib, automatic unit
>> tests.
>> * we already have tools and libs at our hands we can use, they would
>> improve as reflected.
>>
>>  .. in using:
>>  * zero install
>> * cross platform (even mobile devices, ipad, chromeOS ...)
>> * embeddable into existing IDEs since some provide a internal web
>> browser (eclipse, netbeans,..)
>> * embeddable into browser, no task switching anymore
>>
>>
>>  .. some ideas:
>>
>> * group editing (?)
>> * if webkit gets added to SWHX (got mentioned on the list) we could
>> build a standalone desktop version out of it.
>> * create a desktop version anyhow using air, titanium, ...
>> * create browsers extensions to add additional functionality, (which is
>> pretty easy these days, at least for firefox[3] and chrome[4] )
>> * research into new architectures of IDEs like codebubbles[5], which
>> looks like a fit for hxs OOP style
>> * integrate into wave (?)
>>
>>
>> drawbacks ?:
>>
>> - performance
>>
>
> I think performance with the V8 javascript engine is not so much trouble
> anymore.
>
> On the global idea I think it's very cool. It will help the community to
> developpe some strong javascript library and define a GUI design.
> I would like to help somehow. I'm not sure I'm qualified enough.
> I'm working on javascript right now, making the website working more as a
> client application.
> We could build components with html/css/js, that only load data and format
> it on client side instead of requesting the server for formating the
> requested data to display.
>
> I'm following :)
>>
>>
>> what do you think ?
>>
>>
>> /tong
>>
>>
>> [0] https://bespin.mozilla.com/
>> [1] http://etherpad.com/
>> [2] http://www.catswhocode.com/blog/10-useful-online-code-editors
>> [3] https://jetpack.mozillalabs.com/
>> [4] http://code.google.com/chrome/extensions/
>> [5] http://www.cs.brown.edu/people/acb/codebubbles_site.htm
>>
>>
>> 
>>
>
>
> --
> haXe - an open source web programming language
> http://haxe.org
>



--
Tony Polinelli
http://touchmypixel.com

--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

MarcWeber
Excerpts from Tony Polinelli's message of Thu Mar 18 01:12:46 +0100 2010:
> what you really need to do is bootstrap the compiler in haxe so that
> you can compile on the clientside ;P hahah ooooh yeah

Some work has been done :)
http://lib.haxe.org/p/hscript

Marc Weber

--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

tong-2
In reply to this post by Michiel Crefcoeur
On Wed, 2010-03-17 at 17:15 +0100, Michiel Crefcoeur wrote:

> I've thought of the idea before but I have no time to work on it but
> here's a tip:
>
>
> First step should be to write a language/syntax definition for one of
> these readily available web-based code editors:
> http://sourceforge.net/projects/editarea/
> http://sourceforge.net/projects/codepress
> http://labs.mozilla.com/projects/bespin/
>
>
> Porting one of these to haXe first would make it a bit more
> interesting.
> After that, it shouldn't be too hard to set up a haXe based serverside
> backend.


i don't think having a haxe-haxe compiler is needed.
every device i can input code does have internet access.
also compile times wouldn't be that good running the compiler on a
javascript vm, i guess.
for the web based IDE .. supporting different (compileable) targets
would require to have remote access, otherwise we would have to embed
javac, gcc, whatever into the client application.


as i am using XMPP in my every days work i would tend to use it as
underlying architecture to power the IDE system.
this would strip out a lot of headaches and would let concentrate on the
application level since a lot of required infrastructure is provided by
the jabber system out of box:
* HTTP push system
* server load (handle by the jabber server)
* authentication model
* contact and group management
* security
* shared editing
* ...

it would also allow everybody to host a IDE backend without any access
to a remote system, since there exists many free jabber services to
which a client account could be registerd with.
we would end up having a realtime haxe-IDE-cloud hosted for free by
google or facebook ;)


my time is limited too, i don't think such a project makes much sense
without great support from the community and a core team with at least a
handful of members.

lets ci. ...

/tong



--
[)   |   5   |<   †   |2   3   3


--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

Lee Sylvester
Really, we only need a web based IDE that understands the haXe syntax  
and has file management. There's no reason why it then couldn't run  
the standard haXe compiler running on the server as a process and  
launch the result in the browser.

Regards,
Lee




On 18 Mar 2010, at 10:33, tong <[hidden email]> wrote:

> On Wed, 2010-03-17 at 17:15 +0100, Michiel Crefcoeur wrote:
>> I've thought of the idea before but I have no time to work on it but
>> here's a tip:
>>
>>
>> First step should be to write a language/syntax definition for one of
>> these readily available web-based code editors:
>> http://sourceforge.net/projects/editarea/
>> http://sourceforge.net/projects/codepress
>> http://labs.mozilla.com/projects/bespin/
>>
>>
>> Porting one of these to haXe first would make it a bit more
>> interesting.
>> After that, it shouldn't be too hard to set up a haXe based  
>> serverside
>> backend.
>
>
> i don't think having a haxe-haxe compiler is needed.
> every device i can input code does have internet access.
> also compile times wouldn't be that good running the compiler on a
> javascript vm, i guess.
> for the web based IDE .. supporting different (compileable) targets
> would require to have remote access, otherwise we would have to embed
> javac, gcc, whatever into the client application.
>
>
> as i am using XMPP in my every days work i would tend to use it as
> underlying architecture to power the IDE system.
> this would strip out a lot of headaches and would let concentrate on  
> the
> application level since a lot of required infrastructure is provided  
> by
> the jabber system out of box:
> * HTTP push system
> * server load (handle by the jabber server)
> * authentication model
> * contact and group management
> * security
> * shared editing
> * ...
>
> it would also allow everybody to host a IDE backend without any access
> to a remote system, since there exists many free jabber services to
> which a client account could be registerd with.
> we would end up having a realtime haxe-IDE-cloud hosted for free by
> google or facebook ;)
>
>
> my time is limited too, i don't think such a project makes much sense
> without great support from the community and a core team with at  
> least a
> handful of members.
>
> lets ci. ...
>
> /tong
>
>
> 
> --
> [)   |   5   |<   †   |2   3   3
>
>
> --
> haXe - an open source web programming language
> http://haxe.org

--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

Michiel Crefcoeur
In reply to this post by tong-2
I wasn't suggesting porting haXe to haXe! :-)
The links are just browser-based code editors with syntax-highlighting support.
Someone could write a syntax definition for one of these (a .js file).
To make it more practical for the haXe community, it would be nice if someone would port the editor code from JavaScript to haXe.
A clientside compiler wouldn't make sense at all! :-)

2010/3/18 tong <[hidden email]>
On Wed, 2010-03-17 at 17:15 +0100, Michiel Crefcoeur wrote:
> I've thought of the idea before but I have no time to work on it but
> here's a tip:
>
>
> First step should be to write a language/syntax definition for one of
> these readily available web-based code editors:
> http://sourceforge.net/projects/editarea/
> http://sourceforge.net/projects/codepress
> http://labs.mozilla.com/projects/bespin/
>
>
> Porting one of these to haXe first would make it a bit more
> interesting.
> After that, it shouldn't be too hard to set up a haXe based serverside
> backend.


i don't think having a haxe-haxe compiler is needed.
every device i can input code does have internet access.
also compile times wouldn't be that good running the compiler on a
javascript vm, i guess.
for the web based IDE .. supporting different (compileable) targets
would require to have remote access, otherwise we would have to embed
javac, gcc, whatever into the client application.


as i am using XMPP in my every days work i would tend to use it as
underlying architecture to power the IDE system.
this would strip out a lot of headaches and would let concentrate on the
application level since a lot of required infrastructure is provided by
the jabber system out of box:
* HTTP push system
* server load (handle by the jabber server)
* authentication model
* contact and group management
* security
* shared editing
* ...

it would also allow everybody to host a IDE backend without any access
to a remote system, since there exists many free jabber services to
which a client account could be registerd with.
we would end up having a realtime haxe-IDE-cloud hosted for free by
google or facebook ;)


my time is limited too, i don't think such a project makes much sense
without great support from the community and a core team with at least a
handful of members.

lets ci. ...

/tong



--
[)   |   5   |<   †   |2   3   3


--
haXe - an open source web programming language
http://haxe.org


--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

tong-2
In reply to this post by Lee Sylvester
On Thu, 2010-03-18 at 12:37 +0000, Lee McColl Sylvester wrote:
> Really, we only need a web based IDE that understands the haXe syntax  
> and has file management. There's no reason why it then couldn't run  
> the standard haXe compiler running on the server as a process and  
> launch the result in the browser.
>
> Regards,
> Lee
>


no, a simple send-compile-response isn't secure.
imagine you want to compile code including passwords or other sensitive
information .. you need authentication and strong security.
also it think it isn't cool EVERYONE can compile executeables on my
server.

/tong



--
[)   |   5   |<   †   |2   3   3


--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

Andy Li
In reply to this post by Michiel Crefcoeur
Actually this topic has quite a few ideas:
  • Haxe based web IDE
  • Haxe based IDE
  • Haxe based Haxe compiler
I think three of them can be seen as a separated project and interesting enough with its own possibility. They can just exist without each others but multiply the power when they're all combined.

Andy

On Thu, Mar 18, 2010 at 8:42 PM, Michiel Crefcoeur <[hidden email]> wrote:
I wasn't suggesting porting haXe to haXe! :-)
The links are just browser-based code editors with syntax-highlighting support.
Someone could write a syntax definition for one of these (a .js file).
To make it more practical for the haXe community, it would be nice if someone would port the editor code from JavaScript to haXe.
A clientside compiler wouldn't make sense at all! :-)

2010/3/18 tong <[hidden email]>

On Wed, 2010-03-17 at 17:15 +0100, Michiel Crefcoeur wrote:
> I've thought of the idea before but I have no time to work on it but
> here's a tip:
>
>
> First step should be to write a language/syntax definition for one of
> these readily available web-based code editors:
> http://sourceforge.net/projects/editarea/
> http://sourceforge.net/projects/codepress
> http://labs.mozilla.com/projects/bespin/
>
>
> Porting one of these to haXe first would make it a bit more
> interesting.
> After that, it shouldn't be too hard to set up a haXe based serverside
> backend.


i don't think having a haxe-haxe compiler is needed.
every device i can input code does have internet access.
also compile times wouldn't be that good running the compiler on a
javascript vm, i guess.
for the web based IDE .. supporting different (compileable) targets
would require to have remote access, otherwise we would have to embed
javac, gcc, whatever into the client application.


as i am using XMPP in my every days work i would tend to use it as
underlying architecture to power the IDE system.
this would strip out a lot of headaches and would let concentrate on the
application level since a lot of required infrastructure is provided by
the jabber system out of box:
* HTTP push system
* server load (handle by the jabber server)
* authentication model
* contact and group management
* security
* shared editing
* ...

it would also allow everybody to host a IDE backend without any access
to a remote system, since there exists many free jabber services to
which a client account could be registerd with.
we would end up having a realtime haxe-IDE-cloud hosted for free by
google or facebook ;)


my time is limited too, i don't think such a project makes much sense
without great support from the community and a core team with at least a
handful of members.

lets ci. ...

/tong



--
[)   |   5   |<   †   |2   3   3


--
haXe - an open source web programming language
http://haxe.org


--
haXe - an open source web programming language
http://haxe.org


--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

Lee Sylvester
I think a haXe based haXe compiler is a bad idea. A subset, sure (such
as hscript), but a full blown compiler in haXe would be slow and would
serve no point whatsoever.

Lee





Andy Li wrote:

> Actually this topic has quite a few ideas:
>
>     * Haxe based web IDE
>     * Haxe based IDE
>     * Haxe based Haxe compiler
>
> I think three of them can be seen as a separated project and
> interesting enough with its own possibility. They can just exist
> without each others but multiply the power when they're all combined.
>
> Andy
>
> On Thu, Mar 18, 2010 at 8:42 PM, Michiel Crefcoeur <[hidden email]
> <mailto:[hidden email]>> wrote:
>
>     I wasn't suggesting porting haXe to haXe! :-)
>     The links are just browser-based code editors with
>     syntax-highlighting support.
>     Someone could write a syntax definition for one of these (a .js file).
>     To make it more practical for the haXe community, it would be nice
>     if someone would port the editor code from JavaScript to haXe.
>     A clientside compiler wouldn't make sense at all! :-)
>
>     2010/3/18 tong <[hidden email] <mailto:[hidden email]>>
>
>         On Wed, 2010-03-17 at 17:15 +0100, Michiel Crefcoeur wrote:
>         > I've thought of the idea before but I have no time to work
>         on it but
>         > here's a tip:
>         >
>         >
>         > First step should be to write a language/syntax definition
>         for one of
>         > these readily available web-based code editors:
>         > http://sourceforge.net/projects/editarea/
>         > http://sourceforge.net/projects/codepress
>         > http://labs.mozilla.com/projects/bespin/
>         >
>         >
>         > Porting one of these to haXe first would make it a bit more
>         > interesting.
>         > After that, it shouldn't be too hard to set up a haXe based
>         serverside
>         > backend.
>
>
>         i don't think having a haxe-haxe compiler is needed.
>         every device i can input code does have internet access.
>         also compile times wouldn't be that good running the compiler on a
>         javascript vm, i guess.
>         for the web based IDE .. supporting different (compileable)
>         targets
>         would require to have remote access, otherwise we would have
>         to embed
>         javac, gcc, whatever into the client application.
>
>
>         as i am using XMPP in my every days work i would tend to use it as
>         underlying architecture to power the IDE system.
>         this would strip out a lot of headaches and would let
>         concentrate on the
>         application level since a lot of required infrastructure is
>         provided by
>         the jabber system out of box:
>         * HTTP push system
>         * server load (handle by the jabber server)
>         * authentication model
>         * contact and group management
>         * security
>         * shared editing
>         * ...
>
>         it would also allow everybody to host a IDE backend without
>         any access
>         to a remote system, since there exists many free jabber
>         services to
>         which a client account could be registerd with.
>         we would end up having a realtime haxe-IDE-cloud hosted for
>         free by
>         google or facebook ;)
>
>
>         my time is limited too, i don't think such a project makes
>         much sense
>         without great support from the community and a core team with
>         at least a
>         handful of members.
>
>         lets ci. ...
>
>         /tong
>
>
>         
>         --
>         [)   |   5   |<   †   |2   3   3
>
>
>         --
>         haXe - an open source web programming language
>         http://haxe.org
>
>
>
>     --
>     haXe - an open source web programming language
>     http://haxe.org
>
>


--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

Andy Li
I can imagine it wouldn't be as fast as the current one. But it should serve a few use cases like possibility of making haXe able to be interpreted(well, interpreted language has its own pros and cons). Also it let the haXe users being able to contribute, or at least have their own fork of the complier without learning another language(haXe is decided to let you stay in one language, right?).

But yes, I think the benefit may not outweigh the effort needed to develop it. So I think if we work on the three one-by-one, the order should be IDE > web IDE > haXe compiler?

Cheers,
Andy

On Thu, Mar 18, 2010 at 9:10 PM, Lee McColl Sylvester <[hidden email]> wrote:
I think a haXe based haXe compiler is a bad idea. A subset, sure (such as hscript), but a full blown compiler in haXe would be slow and would serve no point whatsoever.

Lee





Andy Li wrote:
Actually this topic has quite a few ideas:

   * Haxe based web IDE
   * Haxe based IDE
   * Haxe based Haxe compiler

I think three of them can be seen as a separated project and interesting enough with its own possibility. They can just exist without each others but multiply the power when they're all combined.

Andy

On Thu, Mar 18, 2010 at 8:42 PM, Michiel Crefcoeur <[hidden email] <mailto:[hidden email]>> wrote:

   I wasn't suggesting porting haXe to haXe! :-)
   The links are just browser-based code editors with
   syntax-highlighting support.
   Someone could write a syntax definition for one of these (a .js file).
   To make it more practical for the haXe community, it would be nice
   if someone would port the editor code from JavaScript to haXe.
   A clientside compiler wouldn't make sense at all! :-)

   2010/3/18 tong <[hidden email] <mailto:[hidden email]>>


       On Wed, 2010-03-17 at 17:15 +0100, Michiel Crefcoeur wrote:
       > I've thought of the idea before but I have no time to work
       on it but
       > here's a tip:
       >
       >
       > First step should be to write a language/syntax definition
       for one of
       > these readily available web-based code editors:
       > http://sourceforge.net/projects/editarea/
       > http://sourceforge.net/projects/codepress
       > http://labs.mozilla.com/projects/bespin/
       >
       >
       > Porting one of these to haXe first would make it a bit more
       > interesting.
       > After that, it shouldn't be too hard to set up a haXe based
       serverside
       > backend.


       i don't think having a haxe-haxe compiler is needed.
       every device i can input code does have internet access.
       also compile times wouldn't be that good running the compiler on a
       javascript vm, i guess.
       for the web based IDE .. supporting different (compileable)
       targets
       would require to have remote access, otherwise we would have
       to embed
       javac, gcc, whatever into the client application.


       as i am using XMPP in my every days work i would tend to use it as
       underlying architecture to power the IDE system.
       this would strip out a lot of headaches and would let
       concentrate on the
       application level since a lot of required infrastructure is
       provided by
       the jabber system out of box:
       * HTTP push system
       * server load (handle by the jabber server)
       * authentication model
       * contact and group management
       * security
       * shared editing
       * ...

       it would also allow everybody to host a IDE backend without
       any access
       to a remote system, since there exists many free jabber
       services to
       which a client account could be registerd with.
       we would end up having a realtime haxe-IDE-cloud hosted for
       free by
       google or facebook ;)


       my time is limited too, i don't think such a project makes
       much sense
       without great support from the community and a core team with
       at least a
       handful of members.

       lets ci. ...

       /tong


       
       --
       [)   |   5   |<   †   |2   3   3


       --
       haXe - an open source web programming language
       http://haxe.org



   --
   haXe - an open source web programming language
   http://haxe.org




--
haXe - an open source web programming language
http://haxe.org


--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

David Bergman
In reply to this post by Lee Sylvester
On Mar 18, 2010, at 9:10 AM, Lee McColl Sylvester wrote:

> I think a haXe based haXe compiler is a bad idea. A subset, sure (such as hscript), but a full blown compiler in haXe would be slow and would serve no point whatsoever.

Why would it not do it; first of all, it gives (more) credibility to the language for potential users - it is often used as a litmus test for a language's expressivity - and would actually allow for some interesting bootstrapping and generative scenarios, besides allowing you to compile code in an AIR-based IDE.

/David

>
> Lee
>
>
>
>
>
> Andy Li wrote:
>> Actually this topic has quite a few ideas:
>>
>>    * Haxe based web IDE
>>    * Haxe based IDE
>>    * Haxe based Haxe compiler
>>
>> I think three of them can be seen as a separated project and interesting enough with its own possibility. They can just exist without each others but multiply the power when they're all combined.
>>
>> Andy
>>
>> On Thu, Mar 18, 2010 at 8:42 PM, Michiel Crefcoeur <[hidden email] <mailto:[hidden email]>> wrote:
>>
>>    I wasn't suggesting porting haXe to haXe! :-)
>>    The links are just browser-based code editors with
>>    syntax-highlighting support.
>>    Someone could write a syntax definition for one of these (a .js file).
>>    To make it more practical for the haXe community, it would be nice
>>    if someone would port the editor code from JavaScript to haXe.
>>    A clientside compiler wouldn't make sense at all! :-)
>>
>>    2010/3/18 tong <[hidden email] <mailto:[hidden email]>>
>>
>>        On Wed, 2010-03-17 at 17:15 +0100, Michiel Crefcoeur wrote:
>>        > I've thought of the idea before but I have no time to work
>>        on it but
>>        > here's a tip:
>>        >
>>        >
>>        > First step should be to write a language/syntax definition
>>        for one of
>>        > these readily available web-based code editors:
>>        > http://sourceforge.net/projects/editarea/
>>        > http://sourceforge.net/projects/codepress
>>        > http://labs.mozilla.com/projects/bespin/
>>        >
>>        >
>>        > Porting one of these to haXe first would make it a bit more
>>        > interesting.
>>        > After that, it shouldn't be too hard to set up a haXe based
>>        serverside
>>        > backend.
>>
>>
>>        i don't think having a haxe-haxe compiler is needed.
>>        every device i can input code does have internet access.
>>        also compile times wouldn't be that good running the compiler on a
>>        javascript vm, i guess.
>>        for the web based IDE .. supporting different (compileable)
>>        targets
>>        would require to have remote access, otherwise we would have
>>        to embed
>>        javac, gcc, whatever into the client application.
>>
>>
>>        as i am using XMPP in my every days work i would tend to use it as
>>        underlying architecture to power the IDE system.
>>        this would strip out a lot of headaches and would let
>>        concentrate on the
>>        application level since a lot of required infrastructure is
>>        provided by
>>        the jabber system out of box:
>>        * HTTP push system
>>        * server load (handle by the jabber server)
>>        * authentication model
>>        * contact and group management
>>        * security
>>        * shared editing
>>        * ...
>>
>>        it would also allow everybody to host a IDE backend without
>>        any access
>>        to a remote system, since there exists many free jabber
>>        services to
>>        which a client account could be registerd with.
>>        we would end up having a realtime haxe-IDE-cloud hosted for
>>        free by
>>        google or facebook ;)
>>
>>
>>        my time is limited too, i don't think such a project makes
>>        much sense
>>        without great support from the community and a core team with
>>        at least a
>>        handful of members.
>>
>>        lets ci. ...
>>
>>        /tong
>>
>>
>>        
>>        --
>>        [)   |   5   |<   †   |2   3   3
>>
>>
>>        --
>>        haXe - an open source web programming language
>>        http://haxe.org
>>
>>
>>
>>    --
>>    haXe - an open source web programming language
>>    http://haxe.org
>>
>>
>
>
> --
> haXe - an open source web programming language
> http://haxe.org


--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

clemos
In reply to this post by Andy Li
Hi

I agree with that.
There is also another idea : simply, an Haxe IDE.

I tried minibuilder and it looks quite good. At least it seems
interesting as a basis for a cross-platform haXe IDE (or editor, to
keep it simple).
Maybe we could start by trying to connect it with the haXe compiler,
not only for compilation but also for code completion.

Then, there could be lots of cool features we could add to it, like
hxml parsing/generation, haxelib/haxed integration, whatever...
I'm not really an expert in Air/Flex/MXML, but since it is "based on
Flash", maybe we could write these extensions in haXe...
This would make it closer to an "haXe based Haxe IDE", which looks
like a cool idea, even if it's bigger.

Actually, I think the problem with writing an "haXe base Haxe IDE" is
really about the GUI, and using minibuilder / Flex components seems to
solve this problem.
So the idea of using a minibuilder-like Flex based (it is, right?)
GUI, together with or inside an haXe/Flash layer looks promising.
AFAIK, if we use haXe remoting and stuff, this would eventually allow
it to run in several different flavors:
- standalone, either with an Air backend or with a Neko/SWHX backend.
- web-based, with a server-side backend that could also be written in
haXe/neko or haXe/PHP.
The "Haxe based web IDE" would be there...

Further developements could include growing the server-side backend to
a full-featured community coding website, something between github and
bespin.
This would really demonstrate the power of haXe targets, since every
part of the application would be written in and extendable with haXe.

As a Linux user, I'm still not very satisified with the haXe
compatible editors available.
I'm still using Geany with the code completion plugin, but it's still
a little buggy, and the big problem is it's difficult to make it
evolve (at least, for me, and I guess for a majority of haXe devs).
It would be so great, even if it's minimal, to have a cross-platform
haXe IDE that would be extendable by haXe code.

Then about using Javascript/HTML for the GUI, I think it's at the same
time very complex and very limiting, since the editor could only run
in a browser environnement.
Bespin is really a good project, but to me it's still very
experimental, and I don't see myself writing code in a browser, or in
any Javascript based environnement (I already find Java-based editors
to be so heavy and slow...).
Moreover, the haXe/JS target is, IMHO, the one that's the less mature.

I feel like all the pieces are here to start a good project "haXe
based haXe web/standalone IDE" or whatever, and it could start with a
very simple basis (minibuilder haXe editor with haxe highlighting and
code completion looks not so complicated), and evolve to something
very interresting...
I'd definitely like to contribute to such a project.

+++++
Clément

On Thu, Mar 18, 2010 at 2:00 PM, Andy Li <[hidden email]> wrote:

> Actually this topic has quite a few ideas:
>
> Haxe based web IDE
> Haxe based IDE
> Haxe based Haxe compiler
>
> I think three of them can be seen as a separated project and interesting
> enough with its own possibility. They can just exist without each others but
> multiply the power when they're all combined.
> Andy
> On Thu, Mar 18, 2010 at 8:42 PM, Michiel Crefcoeur <[hidden email]> wrote:
>>
>> I wasn't suggesting porting haXe to haXe! :-)
>> The links are just browser-based code editors with syntax-highlighting
>> support.
>> Someone could write a syntax definition for one of these (a .js file).
>> To make it more practical for the haXe community, it would be nice if
>> someone would port the editor code from JavaScript to haXe.
>> A clientside compiler wouldn't make sense at all! :-)
>>
>> 2010/3/18 tong <[hidden email]>
>>>
>>> On Wed, 2010-03-17 at 17:15 +0100, Michiel Crefcoeur wrote:
>>> > I've thought of the idea before but I have no time to work on it but
>>> > here's a tip:
>>> >
>>> >
>>> > First step should be to write a language/syntax definition for one of
>>> > these readily available web-based code editors:
>>> > http://sourceforge.net/projects/editarea/
>>> > http://sourceforge.net/projects/codepress
>>> > http://labs.mozilla.com/projects/bespin/
>>> >
>>> >
>>> > Porting one of these to haXe first would make it a bit more
>>> > interesting.
>>> > After that, it shouldn't be too hard to set up a haXe based serverside
>>> > backend.
>>>
>>>
>>> i don't think having a haxe-haxe compiler is needed.
>>> every device i can input code does have internet access.
>>> also compile times wouldn't be that good running the compiler on a
>>> javascript vm, i guess.
>>> for the web based IDE .. supporting different (compileable) targets
>>> would require to have remote access, otherwise we would have to embed
>>> javac, gcc, whatever into the client application.
>>>
>>>
>>> as i am using XMPP in my every days work i would tend to use it as
>>> underlying architecture to power the IDE system.
>>> this would strip out a lot of headaches and would let concentrate on the
>>> application level since a lot of required infrastructure is provided by
>>> the jabber system out of box:
>>> * HTTP push system
>>> * server load (handle by the jabber server)
>>> * authentication model
>>> * contact and group management
>>> * security
>>> * shared editing
>>> * ...
>>>
>>> it would also allow everybody to host a IDE backend without any access
>>> to a remote system, since there exists many free jabber services to
>>> which a client account could be registerd with.
>>> we would end up having a realtime haxe-IDE-cloud hosted for free by
>>> google or facebook ;)
>>>
>>>
>>> my time is limited too, i don't think such a project makes much sense
>>> without great support from the community and a core team with at least a
>>> handful of members.
>>>
>>> lets ci. ...
>>>
>>> /tong
>>>
>>>
>>> 
>>> --
>>> [)   |   5   |<   †   |2   3   3
>>>
>>>
>>> --
>>> haXe - an open source web programming language
>>> http://haxe.org
>>
>>
>> --
>> haXe - an open source web programming language
>> http://haxe.org
>
>
> --
> haXe - an open source web programming language
> http://haxe.org
>

--
haXe - an open source web programming language
http://haxe.org
Reply | Threaded
Open this post in threaded view
|

Re: web IDE

go2ghana
In reply to this post by tong-2
tong schrieb:
On Wed, 2010-03-17 at 17:15 +0100, Michiel Crefcoeur wrote:
  
I've thought of the idea before but I have no time to work on it but
here's a tip:


First step should be to write a language/syntax definition for one of
these readily available web-based code editors:
http://sourceforge.net/projects/editarea/
http://sourceforge.net/projects/codepress
http://labs.mozilla.com/projects/bespin/


Porting one of these to haXe first would make it a bit more
interesting.
After that, it shouldn't be too hard to set up a haXe based serverside
backend.
    


i don't think having a haxe-haxe compiler is needed.
every device i can input code does have internet access.
also compile times wouldn't be that good running the compiler on a
javascript vm, i guess.
for the web based IDE .. supporting different (compileable) targets
would require to have remote access, otherwise we would have to embed
javac, gcc, whatever into the client application.


as i am using XMPP in my every days work i would tend to use it as
underlying architecture to power the IDE system.
this would strip out a lot of headaches and would let concentrate on the
application level since a lot of required infrastructure is provided by
the jabber system out of box:
* HTTP push system
* server load (handle by the jabber server)
* authentication model
* contact and group management
* security
* shared editing
* ...

it would also allow everybody to host a IDE backend without any access
to a remote system, since there exists many free jabber services to
which a client account could be registerd with.
we would end up having a realtime haxe-IDE-cloud hosted for free by
google or facebook ;)


my time is limited too, i don't think such a project makes much sense 
without great support from the community and a core team with at least a
handful of members.

lets ci. ...

/tong



  
Hi,
I found the idea of a server side haxe nice too - however will auto completion be possible fast enough over the network?

Axel

--
haXe - an open source web programming language
http://haxe.org
123