The success of haxe

classic Classic list List threaded Threaded
38 messages Options
12
Reply | Threaded
Open this post in threaded view
|

Re: The success of haxe

David Bergman
I always wondered why all languages, by necessity, must work as well  
for a junior developer - getting into this field two years ago to make  
some (what he/she/it thought to be) easy bucks - as for a senior  
developer, breathing, dreaming and sweating code for decades years...

That is what I like about Haskell, they never tried to be a tool for  
the "nurse," but rather focused on the "brain surgeons."

Sorry if this post sounded a bit elitist. Well, actually, I am not  
sorry ;-)

/David

On May 25, 2009, at 8:05 AM, Hugh Sanderson wrote:

> Hi,
> If your goal is to attract novice programmers then
> one very practical thing that can be done is to have an official  
> forum.
> While then anthropic principle ensures all here like/
> love/could not possibly use anything other than a mailing list,
> this is most certainly not the case for a novices, who
> find mailing lists intimidating.
>
> Another thing I think is the "packaging" of the download.
> These should follow the convertions of the target OS. and
> ideally package an IDE of some sort, so on windows, you would
> have start>haxe>haxedevelop (or whichever is chosen).
> On mac, it might be Xcode plugins or similar IDE.
> But whichever, immediately after install, being able to
> single click, choose a flash template and have an html page
> with "hello world" or a bouncing ball - gives you somewhere
> to start.
>
> A comment for Justin on hxcpp:
>
>> in terminal..
>> haxelib install nme
>> haxelib install neash
>> haxelib install hxcpp
>> open textmate... and .... ?
>
> I think most people can modify a working system, but getting
> started is always hard.
>
> Providing a bunch of haxelib libraries in the standard download
> would help too - or maybe a gui tool.  I know the idea is to
> keep things lean, but command line is a real barrier.
>
> And I know maintaining an installer can be a pain (although
> it can be automated once it is working) But it someone from each
> platform were to put something together it would be a
> really positive thing - and it need not come out day & date
> with the main release, so that may take some pressure off.
>
> Maybe what I am saying is get a platform coordinater for
> each platform to package up the release and add any/all goodies
> that go with the platform.  Maybe extend to linux64 too.
>
> 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: The success of haxe

Armén
I think it is a mistake assuming that elite programmers do not welcome
tools that speed up and ease up their work, just as it is a mistake
assuming that lack of such tools propagates mediocrity, sloppyness and
ignorance.

Even John von Neuman would have appreciated a good IDE with
syntax-coloring and compiler integration.

I think you mixing up language grammar with theory of computers and
computation. The former is absolutely irrelevant as long as it lets
one express all those little things the latter covers. The latter is
what differentiates brain surgeons from nurses. A brain surgeon, by
the way, appreciates better tools too, like I said. But a brain
surgeon speaking English, Spanish and/or Esperanto will be able to
express his professional methods using either of those languages.

I AM NOT SORRY! :-)

On Mon, May 25, 2009 at 16:14, David Bergman
<[hidden email]> wrote:

> I always wondered why all languages, by necessity, must work as well for a
> junior developer - getting into this field two years ago to make some (what
> he/she/it thought to be) easy bucks - as for a senior developer, breathing,
> dreaming and sweating code for decades years...
>
> That is what I like about Haskell, they never tried to be a tool for the
> "nurse," but rather focused on the "brain surgeons."
>
> Sorry if this post sounded a bit elitist. Well, actually, I am not sorry ;-)
>
> /David
>
> On May 25, 2009, at 8:05 AM, Hugh Sanderson wrote:
>
>> Hi,
>> If your goal is to attract novice programmers then
>> one very practical thing that can be done is to have an official forum.
>> While then anthropic principle ensures all here like/
>> love/could not possibly use anything other than a mailing list,
>> this is most certainly not the case for a novices, who
>> find mailing lists intimidating.
>>
>> Another thing I think is the "packaging" of the download.
>> These should follow the convertions of the target OS. and
>> ideally package an IDE of some sort, so on windows, you would
>> have start>haxe>haxedevelop (or whichever is chosen).
>> On mac, it might be Xcode plugins or similar IDE.
>> But whichever, immediately after install, being able to
>> single click, choose a flash template and have an html page
>> with "hello world" or a bouncing ball - gives you somewhere
>> to start.
>>
>> A comment for Justin on hxcpp:
>>
>>> in terminal..
>>> haxelib install nme
>>> haxelib install neash
>>> haxelib install hxcpp
>>> open textmate... and .... ?
>>
>> I think most people can modify a working system, but getting
>> started is always hard.
>>
>> Providing a bunch of haxelib libraries in the standard download
>> would help too - or maybe a gui tool.  I know the idea is to
>> keep things lean, but command line is a real barrier.
>>
>> And I know maintaining an installer can be a pain (although
>> it can be automated once it is working) But it someone from each
>> platform were to put something together it would be a
>> really positive thing - and it need not come out day & date
>> with the main release, so that may take some pressure off.
>>
>> Maybe what I am saying is get a platform coordinater for
>> each platform to package up the release and add any/all goodies
>> that go with the platform.  Maybe extend to linux64 too.
>>
>> Hugh
>>
>>
>> --
>> 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: The success of haxe

jlm@justinfront.net
In reply to this post by Gamehaxe
Hugh 

My proposal based on your textmate question.. 

'KISS HaXe' - project

Cross platform*  quick start assuming an editor can use textmate bundles should be a three step process, and the better it is implemented the better the impression of a professional product, this takes the current excellent install process and makes it even more user friendly.

( *Ubuntu is only linux version supported in this proposal for sake of argument, it was most recommend for beginners by haXe community to me, other linix distributions should be covered in 'other proposal' below )

Bundles are becoming a sort of standard being used across at least two commercial products and several opensource.
I have assumed that in the future standard haxe will include hxcpp and in future by default an example quickstart folder is added to desktop (or my documents) in the special quick start version.

1) Install Bundle
download haxe2.tmBundle.zip 
double click to expand
double click to install

2) Install Haxe
download haxe
double click to install

3) Open/Use textmate or e-textmate or linux equivalent ( Maybe we could add a menu to allow easy selection of projects with descriptions )
Then select a *.tmproj example from the example folder on the desktop
Assuming the *.hxml was the last selected folder from the tray of files and the project was saved then you don't have to select the *.hxml.
'bundles>haXe2>Build Selected'  ( apple shift B or os equivalent ) 
Then from the tray ( just right click to and make the swf or html executable we can pre-do this ) and then double click on the file to test.


TODO: - To make this possible, and to make this easier.

1 ) Create and maintain Quick start one click installer page on the wiki where we can provide instructions and field common setup issues, ie from the very simple to slightly off topic like how to install basic ubuntu on a mac with parallels, or what ever - I can do this or take responsibility for organizing, we just need to work out a suitable interface with current page.

2) Create quick start one click installer, (this will not be really light weight version as it will include examples) - Nicolas to help with?
Include install of haXe bundle (y/n) option
Include menu to put examples folder on desktop (y/n + select folder)  
Create haxe version with hxcpp in.
Add help stuff

3)  Create desktop app to allow easy selection of a number of example pages.- I can create flash visuals but need help with desktop stuff

4)  Improve haxe2 bundle and work on autocompleation - I have dabbled but I doubt have all the skills. 

5)  Create fun and useful examples - I can create some more flash ones and some suggested workflows with flash IDE ( as3 publish ) and Swish perhaps ( as2 ) to widen user base, complex video server stuff I will leave to others, I can show some basic 3d, use of some haxeLib... but tutorials and which to include can be thought about more once people have made tutorials.  Others will need to create tutorials for the other sections, iPhone, javascript, php etc....

6) Linux with bundles is currently an issue, I believe the current e-texteditor is not easy to install and it may not be brilliant, I think we need to create a ready to run build of one that works on standard ubuntu (mac and or pc based) or ... well its an open question, but quick start, simple to use; is not Eclipse from my point of view, and you will only piss off a designer or html monkey if you insist that it is, people happy to use eclipse can usually set it all up for themselves.  But the idea of textmate clone cross platform even if it is not always the best editor seems sensible for a quick start.

7) Add haxeLib plugin support to haxe2 bundle? - help?

8) Create using textmate/clone general use instructions, a list of common needs and howtos - I can do.

9) If for some user this quick start approach fails they can use the Advanced one click or manual approach

Other Proposal

Ron and others create a second page with Eclipse setup with the current one click installer a enterprise approach with a slightly different approach for more terminal based users and more backend savy to include manual and basic one click installation, call it the 'Enterprise Installation' maybe.   If the Eclipse approach can be improved to the ease of use of the quick install then it could later replace the textmate simple install, but even Flexbuilder scares many actionscript savy flash designers and probably any javascript monkeys, so I am yet to be convinced that fits with the quick up and running and it will scare dreamweaver users etc... Coding haXe is not hard and setting up haXe can be far simplier than setting up the Flexbuilder or the flex sdk which as I found out helping a friend yesterday is a pain, so if enterprise users need better tools that is cool but I don't see that as an entry point to just simple haXe use.


Other Thoughts
While I think flashdevelop is an excellent product it is limited to one platform and I think that we should really be looking for solutions that work across all platforms as I feel this is the haXe philosophy.  So it should not be part of the Simple or Advanced setups I am proposing, but be a separate page for current flashdevelop users.

Anyway this is just a suggestion from a long time self taught flash user so please take it as a proposal to stimulate ideas, I also like the idea already proposed of pages per type of haXe as I would love to use php or javascript haXe but its not so easy if you only know flash, the coding is not hard but the platform specific and approach are. 

Note to Nicholas, I am willing to take responsibility for the "KISS HaXe" pages provided I get community support, so its not a case of I would like others to do it for me, I am very aware that "DO" is more important than talk, it is a case of this is what I would like if I was a new user I can put some of it into effect if others can help fill in the gaps and if it seems a good approach then I will start to "DO", obviously like anything if its good its used if not haxe already has stuff.   

Cheers

;j 



On 25 May 2009, at 13:05, Hugh Sanderson wrote:

Hi,
If your goal is to attract novice programmers then
one very practical thing that can be done is to have an official forum.
While then anthropic principle ensures all here like/
love/could not possibly use anything other than a mailing list,
this is most certainly not the case for a novices, who
find mailing lists intimidating.

Another thing I think is the "packaging" of the download.
These should follow the convertions of the target OS. and
ideally package an IDE of some sort, so on windows, you would
have start>haxe>haxedevelop (or whichever is chosen).
On mac, it might be Xcode plugins or similar IDE.
But whichever, immediately after install, being able to
single click, choose a flash template and have an html page
with "hello world" or a bouncing ball - gives you somewhere
to start.

A comment for Justin on hxcpp:

in terminal..
haxelib install nme
haxelib install neash
haxelib install hxcpp
open textmate... and .... ?

I think most people can modify a working system, but getting
started is always hard.

Providing a bunch of haxelib libraries in the standard download
would help too - or maybe a gui tool.  I know the idea is to
keep things lean, but command line is a real barrier.

And I know maintaining an installer can be a pain (although
it can be automated once it is working) But it someone from each
platform were to put something together it would be a
really positive thing - and it need not come out day & date
with the main release, so that may take some pressure off.

Maybe what I am saying is get a platform coordinater for
each platform to package up the release and add any/all goodies
that go with the platform.  Maybe extend to linux64 too.

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: The success of haxe

David Bergman
In reply to this post by Armén
On May 25, 2009, at 10:39 AM, [hidden email] wrote:

> I think it is a mistake assuming that elite programmers do not welcome
> tools that speed up and ease up their work, just as it is a mistake
> assuming that lack of such tools propagates mediocrity, sloppyness and
> ignorance.

Totally agree, and I do not consider raw machine code hacking to the  
crown of coding ;-)

My comment was somewhat tangential, and concerned the language per se.  
I.e., I am pretty happy with haXe being a language not particularly  
tailored for the novice programmer.

> Even John von Neuman would have appreciated a good IDE with
> syntax-coloring and compiler integration.

I am not sure about that one. I have read a lot of stuff by von  
Neumann - primarily in the field of physics - and he does not come  
through as the "syntax-coloring" kind of guy ;-)

> I think you mixing up language grammar with theory of computers and
> computation.

No, I am not mixing up anything. I rarely do in this field.

> The former is absolutely irrelevant as long as it lets
> one express all those little things the latter covers. The latter is
> what differentiates brain surgeons from nurses.

You mean the *insight* into the latter? If so, perhaps. Even though  
very few experienced developers have a firm grasp on that latter field.

> A brain surgeon, by
> the way, appreciates better tools too, like I said. But a brain
> surgeon speaking English, Spanish and/or Esperanto will be able to
> express his professional methods using either of those languages.

If those languages are at an equal level of abstraction and  
expression, yes. This is not the case here. I.e. even though Forth is  
- obviously - Turing complete, the choice between Forth and haXe -  
even disregarding the access to code libraries - is quite relevant,  
and will affect the productivity of the developer.

Grammar - and syntax - does matter. Again, even disregarding code  
libraries at one's disposal.

What I was saying is that there is a tendency to always make one's  
favorite language accessible to novice programmers, and I do take that  
notion as an a priori.

This sentiment of mine has nothing to do with trying to build a better  
environment, such as trying hard to make the IDE's better for haXers -  
which I am doing right now. BUT: I primarily want to make life better  
for the senior developer.

> I AM NOT SORRY! :-)

Good :-)

/David




>
> On Mon, May 25, 2009 at 16:14, David Bergman
> <[hidden email]> wrote:
>> I always wondered why all languages, by necessity, must work as  
>> well for a
>> junior developer - getting into this field two years ago to make  
>> some (what
>> he/she/it thought to be) easy bucks - as for a senior developer,  
>> breathing,
>> dreaming and sweating code for decades years...
>>
>> That is what I like about Haskell, they never tried to be a tool  
>> for the
>> "nurse," but rather focused on the "brain surgeons."
>>
>> Sorry if this post sounded a bit elitist. Well, actually, I am not  
>> sorry ;-)
>>
>> /David
>>
>> On May 25, 2009, at 8:05 AM, Hugh Sanderson wrote:
>>
>>> Hi,
>>> If your goal is to attract novice programmers then
>>> one very practical thing that can be done is to have an official  
>>> forum.
>>> While then anthropic principle ensures all here like/
>>> love/could not possibly use anything other than a mailing list,
>>> this is most certainly not the case for a novices, who
>>> find mailing lists intimidating.
>>>
>>> Another thing I think is the "packaging" of the download.
>>> These should follow the convertions of the target OS. and
>>> ideally package an IDE of some sort, so on windows, you would
>>> have start>haxe>haxedevelop (or whichever is chosen).
>>> On mac, it might be Xcode plugins or similar IDE.
>>> But whichever, immediately after install, being able to
>>> single click, choose a flash template and have an html page
>>> with "hello world" or a bouncing ball - gives you somewhere
>>> to start.
>>>
>>> A comment for Justin on hxcpp:
>>>
>>>> in terminal..
>>>> haxelib install nme
>>>> haxelib install neash
>>>> haxelib install hxcpp
>>>> open textmate... and .... ?
>>>
>>> I think most people can modify a working system, but getting
>>> started is always hard.
>>>
>>> Providing a bunch of haxelib libraries in the standard download
>>> would help too - or maybe a gui tool.  I know the idea is to
>>> keep things lean, but command line is a real barrier.
>>>
>>> And I know maintaining an installer can be a pain (although
>>> it can be automated once it is working) But it someone from each
>>> platform were to put something together it would be a
>>> really positive thing - and it need not come out day & date
>>> with the main release, so that may take some pressure off.
>>>
>>> Maybe what I am saying is get a platform coordinater for
>>> each platform to package up the release and add any/all goodies
>>> that go with the platform.  Maybe extend to linux64 too.
>>>
>>> Hugh
>>>
>>>
>>> --
>>> 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: The success of haxe

jlm@justinfront.net
In reply to this post by Gamehaxe
Hugh

I never got the hxcpp running, but I can give it another go, I think I  
got half way through the stage of having to build my own haXe build.  
But happy to write up some successful use, I decided might be easier  
to wait for it to be fully implemented into the oneclick.

Depending on timescales I may give it another go, but I am one of  
thoughts that  only ever learnt how to do dos and try to avoid too  
much terminal :), there are some issues on the mac and I wonder if I  
might try it on linux first as I was finding pure linux terminal  
slightly easier to use?!  Or maybe I am just learning more.

;j

On 25 May 2009, at 13:05, Hugh Sanderson wrote:

> Hi,
> If your goal is to attract novice programmers then
> one very practical thing that can be done is to have an official  
> forum.
> While then anthropic principle ensures all here like/
> love/could not possibly use anything other than a mailing list,
> this is most certainly not the case for a novices, who
> find mailing lists intimidating.
>
> Another thing I think is the "packaging" of the download.
> These should follow the convertions of the target OS. and
> ideally package an IDE of some sort, so on windows, you would
> have start>haxe>haxedevelop (or whichever is chosen).
> On mac, it might be Xcode plugins or similar IDE.
> But whichever, immediately after install, being able to
> single click, choose a flash template and have an html page
> with "hello world" or a bouncing ball - gives you somewhere
> to start.
>
> A comment for Justin on hxcpp:
>
>> in terminal..
>> haxelib install nme
>> haxelib install neash
>> haxelib install hxcpp
>> open textmate... and .... ?
>
> I think most people can modify a working system, but getting
> started is always hard.
>
> Providing a bunch of haxelib libraries in the standard download
> would help too - or maybe a gui tool.  I know the idea is to
> keep things lean, but command line is a real barrier.
>
> And I know maintaining an installer can be a pain (although
> it can be automated once it is working) But it someone from each
> platform were to put something together it would be a
> really positive thing - and it need not come out day & date
> with the main release, so that may take some pressure off.
>
> Maybe what I am saying is get a platform coordinater for
> each platform to package up the release and add any/all goodies
> that go with the platform.  Maybe extend to linux64 too.
>
> 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: The success of haxe

David Bergman
In reply to this post by David Bergman
Ugh, I wrote:

> What I was saying is that there is a tendency to always make one's  
> favorite language accessible to novice programmers, and I do take  
> that notion as an a priori.

I obviously meant *not take* that notion as an a priori...

/David

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

Re: The success of haxe

Gamehaxe
In reply to this post by David Bergman
Well, I understand your point.
I started with "If your goal is to attract novice programmers then",
but if your goal is to attract "brain surgeons" you could take a
different approach - probably via powerful tech demos - not sure.
There is room for both - I was just offering opinions for getting
people started.

Hugh

> I always wondered why all languages, by necessity, must work as well for  
> a junior developer - getting into this field two years ago to make some  
> (what he/she/it thought to be) easy bucks - as for a senior developer,  
> breathing, dreaming and sweating code for decades years...
>  That is what I like about Haskell, they never tried to be a tool for  
> the "nurse," but rather focused on the "brain surgeons."
>  Sorry if this post sounded a bit elitist. Well, actually, I am not sorry
>  /David
>  On May 25, 2009, at 8:05 AM, Hugh Sanderson wrote:
>  Hi,
> If your goal is to attract novice programmers then
...


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

Re: The success of haxe

Gamehaxe
In reply to this post by jlm@justinfront.net
Justin
Once an integrated version comes out, some of the issues
will go away.  Since that was the first "wider" release
of hxcpp, there were some issues with makefiles etc, that
I could only really find out via some testing in
the wider community.
I have been messing about with XCode, and I can see
the "ultimate" way on mac might be an XCode project template
with a haxe code plugin for syntax highlighting (and completion),
with an automatic build step that compiles the hx->cpp and then
the cpp->exe.  This should allow you to put breakpoints in the
cpp code as well.
I'm hoping to squeeze out another round of hxcpp on haxelib
before the next offical haxe release to see if I can make
some of these things easier, and tune it up for the first
offical release - don't know if I have the time.
I use tcsh on linux, mac and windows, and this may have
created some of the assumptions in my code.

Hugh

> Hugh
>
> I never got the hxcpp running, but I can give it another go, I think I  
> got half way through the stage of having to build my own haXe build.  
> But happy to write up some successful use, I decided might be easier to  
> wait for it to be fully implemented into the oneclick.
>
> Depending on timescales I may give it another go, but I am one of  
> thoughts that  only ever learnt how to do dos and try to avoid too much  
> terminal :), there are some issues on the mac and I wonder if I might  
> try it on linux first as I was finding pure linux terminal slightly  
> easier to use?!  Or maybe I am just learning more.
>
> ;j
>
> On 25 May 2009, at 13:05, Hugh Sanderson wrote:
>
>> Hi,
>> If your goal is to attract novice programmers then
>> one very practical thing that can be done is to have an official forum.
>> While then anthropic principle ensures all here like/
>> love/could not possibly use anything other than a mailing list,
>> this is most certainly not the case for a novices, who
>> find mailing lists intimidating.
>>
>> Another thing I think is the "packaging" of the download.
>> These should follow the convertions of the target OS. and
>> ideally package an IDE of some sort, so on windows, you would
>> have start>haxe>haxedevelop (or whichever is chosen).
>> On mac, it might be Xcode plugins or similar IDE.
>> But whichever, immediately after install, being able to
>> single click, choose a flash template and have an html page
>> with "hello world" or a bouncing ball - gives you somewhere
>> to start.
>>
>> A comment for Justin on hxcpp:
>>
>>> in terminal..
>>> haxelib install nme
>>> haxelib install neash
>>> haxelib install hxcpp
>>> open textmate... and .... ?
>>
>> I think most people can modify a working system, but getting
>> started is always hard.
>>
>> Providing a bunch of haxelib libraries in the standard download
>> would help too - or maybe a gui tool.  I know the idea is to
>> keep things lean, but command line is a real barrier.
>>
>> And I know maintaining an installer can be a pain (although
>> it can be automated once it is working) But it someone from each
>> platform were to put something together it would be a
>> really positive thing - and it need not come out day & date
>> with the main release, so that may take some pressure off.
>>
>> Maybe what I am saying is get a platform coordinater for
>> each platform to package up the release and add any/all goodies
>> that go with the platform.  Maybe extend to linux64 too.
>>
>> 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: The success of haxe

Zenko Klapko Jr.
In reply to this post by Nicolas Cannasse
Look below:

On Mon, May 25, 2009 at 12:09:20PM +0200, Nicolas Cannasse wrote:

> >I'll tell how I feel we should promote it in my next post ;)
>
> Second part ;)
>
> Since I often go to Flash-related conferences, I know quite well about
> how haXe is received by the people using Flash. First, we are at an
> obvious disadvantage because we are not the "official" solution for
> creating Flash content. It will be quite impossible to convince the
> average Flash developer, but we can instead focus on two kind of
> developers :
>
> a) hardcore Flash developers : they want to push the Flash Player beyond
> its limits. I think they can also have great impact on other developers
> as well since they work if often well known. Of course, haXe already
> helps a lot with that, thanks to the language itself and thanks to its
> Flash9/10 speed improvements.
>
> But that's maybe not enough : have a read at http://tinyurl.com/pejzmm
> A lot of people have been complaining about AS3 recently, but haXe is
> still not seen as a viable alternative to it. There are a few things we
> can do to improve that :
>
>  - show that haXe is not something that I'm doing alone in my garage.
> We need more people to speak out about haXe, write articles or blog
> posts, etc. I'm very interested in developing the haXe community and I
> have several ideas for it.
>
>  - work an on an Eclipse-based IDE, so people that are used to either
> FlexBuilder or FDT feel much more "at-home" and can test haXe on
> small-or-medium projects more easily
>
> b) new Flash developers : we often see people that don't want to mess
> with Adobe tools coming and asking questions on the mailing list. For
> one guy that ask question, you can be sure that 10 others simply didn't
> go further.
>
Bingo! This is where I fit in. My main experience comes from PASCAL,
C/C++, perl, and have used a smathering of other languages when the
situation called for it: Java, Javascript, and scheme.

I've been against Flash for a very long time, mostly due to the
proprietary nature and high costs involved in developing for the
language. Not to mention that most flash I've seen are terribly annoying
advertisements or web developers who think a website is a single flash
file. My browser uses FlashBlock. My perception is... I don't take flash
seriously.

I found out about haxe because it made it to the front page of reddit.
A lot has changed after visiting the haxe website. It has opened my eyes
and I've updated my opinions. The people here know how to nicely blend
several web technologies into a cohesive and amazing website. The people
here are also fiercly intelligent when it comes to knowing about the
flash platform and how to bend it to their will. Now that Adobe is
taking Linux seriously as well as opening their standards. It's a very
interesting time. Especially for the haxe community. I bet there's a lot
of people like me out there, who still aren't aware of the changes Adobe
has made as well as the benefits haxe provides.

Haxe has my full attention but a certain level of flash knowledge is
assumed. I've been making my own dictionary of acronyms and separating
the different sub-technologies. Where does AIR fit into flash? Is
ColdFushion an IDE? Is FLEX a framework in the same manner as django is
a framework? Is the Flash client a vm? Binary interpreter? Or something
completely different? I hear the MovieClip used a lot, is Flash really
an animation technology with interactivity & logic thrown on top? I'm
looking for meat & potato explanations, not the marketing calorie filler
on Adobe's website. My ignorance is a result of ignoring this technology
for almost a decade.

I'm going to continue my gap filling because I don't want to slow down
the craziness on the list: Additional language features, bug finding,
iPhone experimentation, etc. I'm looking forward to contributing in some
way once I have a competent understanding.

Oh, I ordered the Haxe/Neko book. It should arrive this week!

-Zenko

> I think that in order to promote haXe as an alternative platform for
> Flash development - without any prior AS3 knowledge - we need to write
> more Flash "get started" tutorials and documentation : how to use
> SWFMill and haXe together, how to integrate SVG graphics, how to handle
> key/mouse events, network, UI, ... every often-asked question should be
> referenced somewhere on the Wiki.
>
> There are two things that are in the works for that :
>
>  - The "Pratique d'AS3" book by Thibault Imbert is being rewritten by
> Michal Romecki to adapt it to haXe. When done, we will have an online
> book with complete Flash9 apis comprehensive walkthrough.... in French.
> Having it translated to English would be great as well.
>
>  - We started discussing with a few people about a new "haXe Manual"
> that would be much more structured than the current organic Wiki. So far
> only an introduction to haXe is available on http://haxe.org/manual . If
> some of you are interested in taking on a part of the Manual, don't
> hesitate !
>
> For other platforms, I guess that what applies to Flash also applies to
> them : better IDE support, more documentation and tutorials, more
> communication.
>
> The later is quite important : so far we were able to put haXe on the
> map inside the Flash world. People have good or bad reasons not to use
> it, but at least they know about it.
>
> For all other platforms, haXe is almost invisible. We need to fix that
> by communicating more to the outside world. High traffic community
> websites such as Reddit.com, Ajaxian.com (for JS) or other PHP-related
> portals might be good targets.
>
> Don't hesitate to contact the website writers and introduce haXe, or ask
> to write an article for it. This is definitely an area were the more
> people will get involved the better the results will be.
>
> From what I hear when I talk about haXe, we already have a very good
> technology with a lot of possibilities.
>
> What is needed now is :
>  - more dev tools (IDE)
>  - more documentation
>  - more communication
>
> The good news is that it is something that can be easily distributed
> among a lot of people - unlike compiler writing. So everyone interested
> in further promoting haXe can give a hand ;)
>
> Sorry for the long post,
>
> Best,
> Nicolas
>
>
> --
> haXe - an open source web programming language
> http://haxe.org
--
I would rather be exposed to the inconveniences attending too much liberty than to those attending too small a degree of it.
 - Thomas Jefferson

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

signature.asc (196 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: The success of haxe

David Bergman
In reply to this post by Gamehaxe
On May 25, 2009, at 11:23 AM, Hugh Sanderson wrote:

> Justin
> Once an integrated version comes out, some of the issues
> will go away.  Since that was the first "wider" release
> of hxcpp, there were some issues with makefiles etc, that
> I could only really find out via some testing in
> the wider community.

I think you misspelt "wilder" ;-)

> I have been messing about with XCode, and I can see
> the "ultimate" way on mac might be an XCode project template
> with a haxe code plugin for syntax highlighting (and completion),

We have one of those already, right? I think Yariv (Sadan) created one  
such plugin.

> with an automatic build step that compiles the hx->cpp and then
> the cpp->exe.  This should allow you to put breakpoints in the
> cpp code as well.

That would be very nice!

Somehow I feel that this hxcpp target of yours has triggered almost  
too much attention - since it is genius at play - to the point of  
squeezing you thin. Hmm, I wonder how we could delegate this a bit...

/David

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

Re: The success of haxe

David Bergman
In reply to this post by Gamehaxe
On May 25, 2009, at 11:14 AM, Hugh Sanderson wrote:

> Well, I understand your point.
> I started with "If your goal is to attract novice programmers then",
> but if your goal is to attract "brain surgeons" you could take a
> different approach - probably via powerful tech demos - not sure.
> There is room for both - I was just offering opinions for getting
> people started.

Definitely. My post had little - if anything - to do with your post,  
except for being triggered by "novice programmer." I am sorry about  
that tangential post of mine.

PS
I only know two actual brain surgeons and they are computer  
illiterate ;-)

/David

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

Re: The success of haxe

Chris Hecker
In reply to this post by Nicolas Cannasse

> we need to write more Flash "get started" tutorials and documentation
> : how to use SWFMill and haXe together, how to integrate SVG
> graphics,

I was going to mention this, but I think a) the swf target for haxe is
the most important one, and b) swfmill is the weak link in using haxe to
make swfs.  I'm on the swfmill mailing list as well, and it's a mess.
There's no official build process or binaries, there are multiple
branches, unanswered calls for help, it's a simple tool that requires a
ton of libraries to build, etc.  The best thing you could do for the
haxe and the swf target is to take over swfmill (or better yet, fork it)
and make it part of the haxe distribution and make it work right.

My philosophy on products and marketing is you should try to be the best
at something, not just mediocre at a lot of things, and I think haxe is
close to being the best for "hardcore flash programming", but there are
still some weak links, most notably swfmill and the ability to do a
working preloader.  I'm sure there are some other issues; I'm not
actively trying to use haxe right now.  I think the other targets are
nice, and I'm sure an IDE would be nice for a lot of people (I don't
personally care, although a visual debugger would be good and presumably
that would come with any IDE work), but if I were you I'd get rid of all
the weak links and make it clear that haxe is simply the best way to
make programmed swfs (as opposed to flash animations, which you'll never
compete on).  The compiler is 100x faster, the code runs faster, the
swfs are smaller, the language is more advanced and cleaner, etc.  Post
metrics backing up those claims, tutorials for making the tests, make it
clear the swfs are 100% compatible, etc.

Chris


Nicolas Cannasse wrote:

>> I'll tell how I feel we should promote it in my next post ;)
>
> Second part ;)
>
> Since I often go to Flash-related conferences, I know quite well about
> how haXe is received by the people using Flash. First, we are at an
> obvious disadvantage because we are not the "official" solution for
> creating Flash content. It will be quite impossible to convince the
> average Flash developer, but we can instead focus on two kind of
> developers :
>
> a) hardcore Flash developers : they want to push the Flash Player beyond
> its limits. I think they can also have great impact on other developers
> as well since they work if often well known. Of course, haXe already
> helps a lot with that, thanks to the language itself and thanks to its
> Flash9/10 speed improvements.
>
> But that's maybe not enough : have a read at http://tinyurl.com/pejzmm
> A lot of people have been complaining about AS3 recently, but haXe is
> still not seen as a viable alternative to it. There are a few things we
> can do to improve that :
>
>  - show that haXe is not something that I'm doing alone in my garage. We
> need more people to speak out about haXe, write articles or blog posts,
> etc. I'm very interested in developing the haXe community and I have
> several ideas for it.
>
>  - work an on an Eclipse-based IDE, so people that are used to either
> FlexBuilder or FDT feel much more "at-home" and can test haXe on
> small-or-medium projects more easily
>
> b) new Flash developers : we often see people that don't want to mess
> with Adobe tools coming and asking questions on the mailing list. For
> one guy that ask question, you can be sure that 10 others simply didn't
> go further.
>
> I think that in order to promote haXe as an alternative platform for
> Flash development - without any prior AS3 knowledge - we need to write
> more Flash "get started" tutorials and documentation : how to use
> SWFMill and haXe together, how to integrate SVG graphics, how to handle
> key/mouse events, network, UI, ... every often-asked question should be
> referenced somewhere on the Wiki.
>
> There are two things that are in the works for that :
>
>  - The "Pratique d'AS3" book by Thibault Imbert is being rewritten by
> Michal Romecki to adapt it to haXe. When done, we will have an online
> book with complete Flash9 apis comprehensive walkthrough.... in French.
> Having it translated to English would be great as well.
>
>  - We started discussing with a few people about a new "haXe Manual"
> that would be much more structured than the current organic Wiki. So far
> only an introduction to haXe is available on http://haxe.org/manual . If
> some of you are interested in taking on a part of the Manual, don't
> hesitate !
>
> For other platforms, I guess that what applies to Flash also applies to
> them : better IDE support, more documentation and tutorials, more
> communication.
>
> The later is quite important : so far we were able to put haXe on the
> map inside the Flash world. People have good or bad reasons not to use
> it, but at least they know about it.
>
> For all other platforms, haXe is almost invisible. We need to fix that
> by communicating more to the outside world. High traffic community
> websites such as Reddit.com, Ajaxian.com (for JS) or other PHP-related
> portals might be good targets.
>
> Don't hesitate to contact the website writers and introduce haXe, or ask
> to write an article for it. This is definitely an area were the more
> people will get involved the better the results will be.
>
>  From what I hear when I talk about haXe, we already have a very good
> technology with a lot of possibilities.
>
> What is needed now is :
>  - more dev tools (IDE)
>  - more documentation
>  - more communication
>
> The good news is that it is something that can be easily distributed
> among a lot of people - unlike compiler writing. So everyone interested
> in further promoting haXe can give a hand ;)
>
> Sorry for the long post,
>
> Best,
> Nicolas
>
>

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

haxe interpretor

Ian Martins
In reply to this post by Ian Liu Rodrigues
Ian Liu wrote:
> One thing that speed the learning
> process is an interpretor, like python.
> Oh my, I would love a haXe interpretor ;)
I was thinking about this the other day, since I've been doing more
python.  haxe compiles so fast that it could be used in an application that:
- appends a line of input to the past lines of input that didn't cause
an error
- write all input to a file (wrapped in "class T{ function main() { .. }}"
- compile
- run
- display output

every time the user hits enter.

this wouldn't make sense for swf's or js, but would be nice for playing
with the crossplatform api in neko.  maybe php also?
is there a reason this wouldn't work?  would it be worthwhile?



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

Re: haxe interpretor

laurence taylor
You could probably do this fairly easily with hxScript. By easily, I mean of course "I don't know how to do it". But once you''ve got the command line to accept input, all you've got to do is feed that into hxScript and return any errors or traces back to the command line. A loop parsing for ; then to nekoserver->hxscript->return. Nekoserver because you need persistent memory, I guess.

On Tue, May 26, 2009 at 7:31 PM, Ian Martins <[hidden email]> wrote:
Ian Liu wrote:
One thing that speed the learning
process is an interpretor, like python.
Oh my, I would love a haXe interpretor ;)
I was thinking about this the other day, since I've been doing more python.  haxe compiles so fast that it could be used in an application that:
- appends a line of input to the past lines of input that didn't cause an error
- write all input to a file (wrapped in "class T{ function main() { .. }}"
- compile
- run
- display output

every time the user hits enter.

this wouldn't make sense for swf's or js, but would be nice for playing with the crossplatform api in neko.  maybe php also?
is there a reason this wouldn't work?  would it be worthwhile?



--
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: haxe interpretor

Ian Martins
thanks. I hadn't looked at that before.  it looks really good, but it
doesn't allow "object declarations such as { x: 0, y: -1}."  maybe I'll
try your approach anyway.  I like that it doesn't need the tempfile or
system call.

-ianxm

laurence taylor wrote:

> You could probably do this fairly easily with hxScript. By easily, I
> mean of course "I don't know how to do it". But once you''ve got the
> command line to accept input, all you've got to do is feed that into
> hxScript and return any errors or traces back to the command line. A
> loop parsing for ; then to nekoserver->hxscript->return. Nekoserver
> because you need persistent memory, I guess.
>
> On Tue, May 26, 2009 at 7:31 PM, Ian Martins <[hidden email]
> <mailto:[hidden email]>> wrote:
>
>     Ian Liu wrote:
>
>         One thing that speed the learning
>         process is an interpretor, like python.
>         Oh my, I would love a haXe interpretor ;)
>
>     I was thinking about this the other day, since I've been doing
>     more python.  haxe compiles so fast that it could be used in an
>     application that:
>     - appends a line of input to the past lines of input that didn't
>     cause an error
>     - write all input to a file (wrapped in "class T{ function main()
>     { .. }}"
>     - compile
>     - run
>     - display output
>
>     every time the user hits enter.
>
>     this wouldn't make sense for swf's or js, but would be nice for
>     playing with the crossplatform api in neko.  maybe php also?
>     is there a reason this wouldn't work?  would it be worthwhile?
>
>
>
>     --
>     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: haxe interpretor

Robert Sköld
In reply to this post by laurence taylor
It would be nice to have a kind of interactive console "ihx", like irb  
for ruby. And it could use the compiler code-completion feature on tab-
press too.

A session could be something like:

# ihx -format -neko # Using "neko" target and "format" haxelib
 > var zip = neko.io.File.read( "hello.zip" );
 > var files = new format.zip.Reader( zip ).read();
 > for( f in files ) {
 >> neko.Lib.printLn( f.fileName ); <-- Indented once since a block is  
started
 > } <-- Un-indented when "}" is typed, and the block is then executed  
when pressing enter.
text.txt
other.txt
image.png
blabla.swf
 >
And so on...

Typing "var zip = nek" and then pressing tab should show a list of  
possible alternatives or just finishing the package name with "neko."  
if it's only one so it's easy to just tab again to get the sub-
packages of the neko package etc.

It could be a pretty nice way to learn the syntax, and having a ecma-
ish interpreter seems like it's missing (or i didn't search enough).

I'd use it ;)

/bob

On May 26, 2009, at 23:30, laurence taylor wrote:

> You could probably do this fairly easily with hxScript. By easily, I  
> mean of course "I don't know how to do it". But once you''ve got the  
> command line to accept input, all you've got to do is feed that into  
> hxScript and return any errors or traces back to the command line. A  
> loop parsing for ; then to nekoserver->hxscript->return. Nekoserver  
> because you need persistent memory, I guess.
>
> On Tue, May 26, 2009 at 7:31 PM, Ian Martins <[hidden email]> wrote:
> Ian Liu wrote:
> One thing that speed the learning
> process is an interpretor, like python.
> Oh my, I would love a haXe interpretor ;)
> I was thinking about this the other day, since I've been doing more  
> python.  haxe compiles so fast that it could be used in an  
> application that:
> - appends a line of input to the past lines of input that didn't  
> cause an error
> - write all input to a file (wrapped in "class T{ function main()  
> { .. }}"
> - compile
> - run
> - display output
>
> every time the user hits enter.
>
> this wouldn't make sense for swf's or js, but would be nice for  
> playing with the crossplatform api in neko.  maybe php also?
> is there a reason this wouldn't work?  would it be worthwhile?
>
>
>
> --
> 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: haxe interpretor

David Bergman
In reply to this post by Ian Martins
I created an interactive shell for C++, mainly for some junior  
developers in my team to learn and understand C++ constructs. Perhaps  
I could generalize it to work with any compiling language, including  
haXe?

If you still want an interactive shell for haXe in a month, let me  
know and I dig into my old code at http://code.google.com/p/cpsh/

/David

On May 26, 2009, at 6:20 PM, Ian Martins wrote:

> thanks. I hadn't looked at that before.  it looks really good, but  
> it doesn't allow "object declarations such as { x: 0, y: -1}."  
> maybe I'll try your approach anyway.  I like that it doesn't need  
> the tempfile or system call.
>
> -ianxm
>
> laurence taylor wrote:
>> You could probably do this fairly easily with hxScript. By easily,  
>> I mean of course "I don't know how to do it". But once you''ve got  
>> the command line to accept input, all you've got to do is feed that  
>> into hxScript and return any errors or traces back to the command  
>> line. A loop parsing for ; then to nekoserver->hxscript->return.  
>> Nekoserver because you need persistent memory, I guess.
>>
>> On Tue, May 26, 2009 at 7:31 PM, Ian Martins <[hidden email] <mailto:[hidden email]
>> >> wrote:
>>
>>    Ian Liu wrote:
>>
>>        One thing that speed the learning
>>        process is an interpretor, like python.
>>        Oh my, I would love a haXe interpretor ;)
>>
>>    I was thinking about this the other day, since I've been doing
>>    more python.  haxe compiles so fast that it could be used in an
>>    application that:
>>    - appends a line of input to the past lines of input that didn't
>>    cause an error
>>    - write all input to a file (wrapped in "class T{ function main()
>>    { .. }}"
>>    - compile
>>    - run
>>    - display output
>>
>>    every time the user hits enter.
>>
>>    this wouldn't make sense for swf's or js, but would be nice for
>>    playing with the crossplatform api in neko.  maybe php also?
>>    is there a reason this wouldn't work?  would it be worthwhile?
>>
>>
>>
>>    --     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: haxe interpretor

Ian Liu Rodrigues
Is there an eval like function for neko? If so, doing an interactive haxe would be easy I guess.

On Tue, May 26, 2009 at 7:29 PM, David Bergman <[hidden email]> wrote:
I created an interactive shell for C++, mainly for some junior developers in my team to learn and understand C++ constructs. Perhaps I could generalize it to work with any compiling language, including haXe?

If you still want an interactive shell for haXe in a month, let me know and I dig into my old code at http://code.google.com/p/cpsh/

/David


On May 26, 2009, at 6:20 PM, Ian Martins wrote:

thanks. I hadn't looked at that before.  it looks really good, but it doesn't allow "object declarations such as { x: 0, y: -1}."  maybe I'll try your approach anyway.  I like that it doesn't need the tempfile or system call.

-ianxm

laurence taylor wrote:
You could probably do this fairly easily with hxScript. By easily, I mean of course "I don't know how to do it". But once you''ve got the command line to accept input, all you've got to do is feed that into hxScript and return any errors or traces back to the command line. A loop parsing for ; then to nekoserver->hxscript->return. Nekoserver because you need persistent memory, I guess.

On Tue, May 26, 2009 at 7:31 PM, Ian Martins <[hidden email] <mailto:[hidden email]>> wrote:

  Ian Liu wrote:

      One thing that speed the learning
      process is an interpretor, like python.
      Oh my, I would love a haXe interpretor ;)

  I was thinking about this the other day, since I've been doing
  more python.  haxe compiles so fast that it could be used in an
  application that:
  - appends a line of input to the past lines of input that didn't
  cause an error
  - write all input to a file (wrapped in "class T{ function main()
  { .. }}"
  - compile
  - run
  - display output

  every time the user hits enter.

  this wouldn't make sense for swf's or js, but would be nice for
  playing with the crossplatform api in neko.  maybe php also?
  is there a reason this wouldn't work?  would it be worthwhile?



  --     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
12