A five-year Haxe vision?

[Andreas Söderlund]

I'm about to embark on a big web project, and considering that it should be viable for at least 5 years, most likely more, I'd like to think that far ahead. For example, five years ago tablets and mobiles weren't that relevant, but is now a must to support.

So this is a two-fold question: Even though I've already decided to use Haxe, I'd love to hear where the maintainers see Haxe five years from now, but also what anyone think will happen to web technology five years from now? (What new tech you think will be introduced, what will people demand, will software development change, etc...!)

/Andreas

[Hugh]

I say put all your effort into an adobe flex solution to run on blackberry.  Make sure you set up your MySpace page first.  Also, use plenty of texture (web-safe colours of course) in your background gif, and animate them because its cool.

Hugh

[David Elahee]

Hugh is right, the virtual boy will be a huge success...oh wait...

--
To post to this group haxe...@googlegroups.com
http://groups.google.com/group/haxelang?hl=en
---
You received this message because you are subscribed to the Google Groups "Haxe" group.
For more options, visit https://groups.google.com/d/optout.

[Andreas Söderlund]

Ok, it will be that kind of thread...! All right guys, knock yourselves out. ;)

[Justin Donaldson]

A large web company will introduce hot new tech, and then slowly piss it away, alienating thousands of developers.

[Juraj Kirchheim]

Well, I think Hugh is pretty much spot on. 

Think back to early 2010. A dark time, before Haxe had macros. When haxe/flash8 was called haxe/flash and haxe/flash was called haxe/flash9. Way before haxe/php and haxe/cpp were as stable as they are now. A time when using NME was pretty much a suicide mission. A time when Haxe was hosted on google code. A time before Simn!

A time before node, before mongo, before express, before angular. Before Dart, before CoffeScript, before TypeScript. A time when less and sass were hot new things. A time before bootstrap. A time before Google ditched GWT. A time before mobile devices dominated the market. A time before the iPad.

How did we live?

But most importantly: What were your projections then about now? How many of them came true? For me it's certainly less than 50%. That is WORSE THAN GUESSING :D

So it's a little unclear what you want to hear. While Haxe is not as shiny and big as what other companies market, that's also a strength. If Adobe hadn't screwed up so hard with flash, then more energy would have gone into interfacing with the flash eco system. Instead we saw the rise of OpenFl and an influx of people from the flash community. 

As far as setting big trends goes, Haxe is not very influential. As far as adaption to changes is concerned, Haxe does really great.

You can choose Haxe based on some belief that the market will favorably change to reward that choice, if you wish. I believe no matter how crazy the market goes, Haxe will always evolve favorably in response, rewarding anyone committed to it.

Best,

Juraj

[Ashiq A.]

One thing I've found over the last ten years is that technologies and
tools change all the time. You should always be prepared to adopt
something new, and pick the best tool for the job.

That being said, I think the real answer you want is:

- Started in 2006 and still around nearly a decade later:
http://en.wikipedia.org/wiki/Haxe
- Under active, heavy development from multiple developers:
https://github.com/HaxeFoundation/haxe/commits/development
- Multiple releases per year: https://github.com/HaxeFoundation/haxe/tags

These are pretty strong indications to the overall health and
long-term stability of Haxe. Ultimately, you have to decide for
yourself.

[Nicolas Cannasse]

Looks like a big-company interview question "how do you see yourself in
five years ?"

Is there any actual correct answer that can be made to this (without
lying) in such an unstable world as technologies ?

Of course, we have big plans for Haxe, such as improving the
performances, completing the std library, making sure it's more easy to
get started for some usages, and doing all the things necessary to
increase the number of users and ensure long term huge success of the
technology, such as better communication, etc etc etc.

Haxe has been there for some time now, and he is here to stay in the
very long term.

Our vision is that Haxe is the best tool for every programmer in the
world, because it frees you from being tied to a given platform, without
restricting you to access this platform capabilities.

People at Haxe Foundation will keep working hard with that in mind.

Hope that helps,

Nicolas

[Jason O'Neil]

I have a Haxe web project that is 2.5 years old.  In that time we moved from Haxe 2.10 to Haxe 3, and slowly moved from old libraries to new ones (thx, ufront, tink etc).  The transitions to new versions have not been too difficult, compile time errors alert you to most problems.  Hopefully unit tests (or BDD tests!) would catch the rest, though sadly the project in question does not have enough to be effective.

At this point there has been some new technology, like React JS, that I think I may have chosen to use had it existed when I began.  But it's not like the technology I was using has been obsoleted - it all still works, and I'm generally happy to work with it.  Sometimes my DOM updates take a little too long to render, that's all :P

In terms of where it will be in 5 years, I agree with Juraj - who knows?  Most predictions are inaccurate.  At this point Haxe's community is diversified enough that I think it would survive if any one developer left (though obviously some developers do a lot currently!)  But it would survive, and I think it'll have a community at least the same size in 5 years, even if it never "takes off".

I also agree that it's well placed to handle new platforms.  You say you are targetting the web, which is a fairly stable platform.  But say "ES6" takes off and browsers stop supporting current version of JS.  (Extremely unlikely, but just imagine).  It would take only a few weeks for Haxe developers to write a new JS output engine that works for ES6.  We'd be fine. 

If it turns out in 5 years we no longer use desktops or laptops or phones, but have implants in our eyes, then I imagine a) they'll support web protocols somehow anyway, and b) someone will write a Haxe target for our favourite eye implant hardware, and you can port to that :)

In short - I'm 2.5 years into a Haxe web project, and so far I have not felt like the technology is at risk of becoming obsoleted, even though some newer (fancy) technology has come out in that time frame.

Jason

[Robin Burrer]

I have been using Haxe for 2 years now. Basically I was looking for a way to use my Flash skills for HTML projects back then. Shortly after discovering Haxe I also started using it for my Adobe Air projects because it was just so convenient to have common language and share code between JS and Flash. Unfortunately Flash is gone now, but I still use Haxe for most of my HTML5 work. Just a few months ago I ported one of my older Flash apps to HTML5 in no time. The client (Bayer) was really happy with the result and could hardly believe, that switching technologies can be so easy (I had to convince them that the app is not Flash by running it in an iPad browser).  So for me Haxe really gives me the freedom to create things in very intuitive way. If Haxe would go away I certainly could do the same in another language – but it’s kind of nice to have this secret weapon and to create your own custom code base that makes you independent from other frameworks that might not be there tomorrow. So I hope Haxe will stay here for good J

Robin

[David Bergman]

As much as I adore Nicolas -- a true pragmatic genius -- and his creation, and have used it in quite a few commercial projects, I must say this: I probably wouldn't bet my business five years from now on Haxe and its ecology. But, five years... Why not create something that works in, say, one year, and then, if a success, convert to whatever ecology and technology that is suitable in, well, five years.

I see Haxe as a terrific execution environment today, although the language is a bit, well, dusty, frankly. I would welcome a focus on a "virtual execution environment" rather than language; an intermediary language, sort of like LLVM but for all kinds of devices and the web. Actually started to create a ClojureScript-to-Haxe compiler but Real Life came and took my my time away.

Cheers!

/David

[TroyWorks]

Having been through hundreds of projects at this point.

Which technology is rarely the defining success of a product in a marketplace over that time frame.  Sure pick the wrong one and you have to pay the price of converting it towhatever new hotness, that writes code for you.  Rewrite/refactor is going to happen regardless as you code over six months and better understand the problems.

My suggestion is go with whatever tech stack you know best and can hire easiest for, think Minimal Viable Product and get it to market ASAP. You cannot see your competition and you are likely underestimating the indifference of a jaded market or how to effectively reach them. If you do your job right, not many will know or care what it's written in, and couldn't care less about what magic or lack there of that makes it so.    If you do it in the worst language possible and still are a commercial success, you can afford to rewrite it...it's a good problem to have, vs the ultra cool clean tech stack that gets taken out with the trash after the startup closes up shop for lack of funding.

As a armchair "futurist" I have been playing the game of predicting the future with my crystal globe (er clear ice sphere) in my whisky glass.  I have been ahead of the curve, predicting and using wearable computers a decade before they were cool, and now I see zombie cyborgs everywhere with bluetooth headset enabled, smart phones and ipads  near surgically attached to people everywhere I go.  My ipad has eating my VCR, voice recorder, camera, video camera, radio etc.   That device trend is clear, and we are heading towards a next phase of wearables (watches) to act as augments to tablets.  But so what, these are just new targets often written in the same languages, java, c, javascript etc. that we already use.

In 5 years, html5 canvas, webgl and video will likely be on par with whatever flash feature parity that still has it strong places places.  The clients I have currently are all wanting to push away from app centric as they are tired of the plugin ecosystem, having Google/App stores control over their product, and finicky adobe development pipelines.  I can see haxe fitting in this well at least for smaller teams.

In that world it probably given the fast runtime of js, llvm compilers doing c++ to js whatever, and multiple target meta languages I think that it's really irrelvant.

[Confidant]

+1 to Juraj, in that Haxe's ability to adapt and evolve is a superpower. I am constantly in awe of how externs and cross-compilation make it possible for trends to come and go—whether popular libraries or new languages—and a Haxe developer can use them and still remain a Haxe developer.

I will continue to recommend Haxe because it's a low-risk investment especially for new developers: 

• Whatever you learn is highly portable to other languages.
  
• The code can be translated easily or compiled to other languages.
  
• It's open source.

It doesn't matter what the future holds. Haxe will either keep you equipped with good tools, or else will let you escape with minimal fuss.

[Marcelo de Moraes Serpa]

Very interesting perspective you’ve got there, David. Similar to what Elliot Stoneham is doing with TardisGo. This is a powerful concept indeed (focus on the execution env instead of the language, sort of what JVM does for Java or LLVM with its intermediate representation). Perhaps an intermediate language that could then be compiled to several other targets. However, if you see TardisGo, Haxe is already being used as an intermediate language, albeit not being optimized it. Clever and creative. Food for thought!

[Daniel Glazman]

On 05/03/2015 05:20, Confidant wrote:

> +1 to Juraj, in that Haxe's ability to adapt and evolve is a superpower.
> I am constantly in awe of how externs and cross-compilation make it
> possible for trends to come and go—whether popular libraries or new
> languages—and a Haxe developer can use them and still remain a Haxe
> developer.

Hugh answered with humour but Andreas raised a few good questions, even
if they should have been explicited more clearly:

1. some contributors - like Hugh - are super-crucial to the Haxe
ecosystem. I am myself here mostly because of hxcpp. It could be
interesting to know if Hugh (again, that's an example) has enough
time and income to remain that key contributor.

2. Ensuring Haxe's future requires, IMHO, an increased visibility. Haxe
is not a new thing and relying only on interpersonal communication
for outreach is probably not enough in the long run.

3. Haxe remains largely unknown outside of the gaming community.
Increasing its visibility implies two things: marketing/evangelism
and outreach to other communities. My own project, Quaxe, is a
possible solution to the latter issue. The former issue requires
a strategy, and money...

4. the largest coding pool in the world is JavaScript. EcmaScript6 is
changing the game and a convergence with ES should probably be
carefully studied...

Please no flames in answer to the above. I'm an old monkey in the open
source world and I've seen too many superclever and superpowerful OSS
projects ultimately fail because they reached a glass ceiling and could
not go beyond. Asking if Haxe has what it takes or can have in the
future what it takes to break that glass ceiling is just a fair
question, I think.

</Daniel>

[Juraj Kirchheim]

On Thu, Mar 5, 2015 at 8:35 AM, Daniel Glazman <daniel....@gmail.com> wrote:

1. some contributors - like Hugh - are super-crucial to the Haxe
   ecosystem. I am myself here mostly because of hxcpp. It could be
   interesting to know if Hugh (again, that's an example) has enough
   time and income to remain that key contributor.

The more specific your question, the more speculative the answer will be. 

Generally, I think the foundation has made it pretty clear that they are taking charge of the project if necessary. So far the necessity to replace compiler contributors hasn't arisen, but to my knowledge they have the manpower/budget/network to do that, should it ever come to that. Of course if a tornado were to hit the WWX while the vast majority of the foundation, compiler team and key community members are all in one place, that would be a hard blow to recover from, but short of Pythonesque twists of fate, we're on the safe side.

2. Ensuring Haxe's future requires, IMHO, an increased visibility. Haxe
   is not a new thing and relying only on interpersonal communication
   for outreach is probably not enough in the long run.

What exactly does "future" mean here? I think you are implicitly envisioning something, that is not necessarily clear to everybody else. How would you measure "having a future"? If by that you mean "taking off", then let's call it that.

We mostly want visibility, since as social animals we just love confirmation. But the correlation to Haxe's suitability as a choice for a long term project as Andreas is undertaking is rather low.

I am not saying that making Haxe visible is an unworthy goal. To slightly exaggerate, Haxe is a great tool and we're doing anyone we can get to use it a great favor.

What made Haxe useful as a tool for me was to commit to it. It can be frustrating to not see Haxe become the world dominating solution that it arguably deserves to be. I let that wish go a few years ago and it has been very liberating. Many people have these grand ideas about all sorts of preconditions that must be met, before they can productively work. That's just mental self-mutilation. Trust me, I've been there. Haxe can be used as a highly productive tool in many areas if you're just willing to learn to use it. 

For the past years I have been working with exclusively with Haxe full time, the last two of them with haxe/js. It was probably one of the best choices I made thus far. I did not wait for Haxe to become visible or anything. I used it. Needless to mention that in that time the tool I chose grew and matured. Even if it hadn't, it gave me a chance to grow and mature, because it is simply an amazing language.

3. Haxe remains largely unknown outside of the gaming community.
   Increasing its visibility implies two things: marketing/evangelism
   and outreach to other communities. My own project, Quaxe, is a
   possible solution to the latter issue. The former issue requires
   a strategy, and money...

Strategy is a bit of a strong word at this point. A vaguely directed effort would already be a good start :D

I'm not sure if money is the problem. That's something to discuss with the foundation. But I think it's more of a lack of initiative and actual experience in that area. If you feel up to the task, please do take the initiative. Probably just email Nicolas regarding financial issues.

 

4. the largest coding pool in the world is JavaScript. EcmaScript6 is
   changing the game and a convergence with ES should probably be
   carefully studied...

What exactly do you mean by "convergence"? Shrinkwrapped interfaces to 3rd party JS libraries are no doubt important. Something like DefinitelyTyped for TypeScript. Maybe it would be good to focus some effort on the whole ts2hx avenue to solve this issue. But that seems like something we should be able to accomplish over the next 5 years.

Please no flames in answer to the above. I'm an old monkey in the open
source world and I've seen too many superclever and superpowerful OSS
projects ultimately fail because they reached a glass ceiling and could
not go beyond. Asking if Haxe has what it takes or can have in the
future what it takes to break that glass ceiling is just a fair
question, I think.

What would you use as a metric to determine if Haxe has reached that glass ceiling? What do you mean by that? 

Also how would you characterize project "failure"? Of course project success can be seen as getting everybody to acknowledge that your particular project is the holy grail of programming. Then you're setting yourself up for "failure" right from the start. But I kind of tried to explain that above.

There are so many things that obviously need to be done and even with those lying around undone, there is still a lot of growth in other areas. Other projects typically stall because nothing happens and its also unclear what remains to be done. That's no exactly our scenario ;)

Haxe's evolution is a complex process that unfolds in many dimensions. In almost all of them it plateaus episodically. For example, right now "releases" would be one aspect where Haxe has done a bit poorly in the recent past, but that's about to change. There is no single glass ceiling, unless your look at just one dimension and just one plateau in it. So which one is that?

----

Of course there is nothing wrong with thinking about the future, but ultimately we live in the here and now. And here and now Haxe is a great tool. And like any tool it can only be useful to those who use it. Those who are instead reveling in ideas of how Haxe must be, will never reap its benefits. That is, I believe, the answer to Andreas' question. That doesn't mean we should stop undertaking things to make Haxe better. But beyond talking, that also takes some doing. Which is generally not the outcome of these recurring discussions.

Best,

Juraj

[Simon Krajewski]

Am 05.03.2015 um 11:43 schrieb Juraj Kirchheim:

4. the largest coding pool in the world is JavaScript. EcmaScript6 is

   changing the game and a convergence with ES should probably be
   carefully studied...

What exactly do you mean by "convergence"? Shrinkwrapped interfaces to 3rd party JS libraries are no doubt important. Something like DefinitelyTyped for TypeScript. Maybe it would be good to focus some effort on the whole ts2hx avenue to solve this issue. But that seems like something we should be able to accomplish over the next 5 years.

About that: ts2hx is the strangest project I currently have because I keep hearing from all directions how useful it could be, yet nobody seems to actually use it. I would gladly fix any issues with it, I just don't know about any since I don't personally use it at the moment.

Simon

[Juraj Kirchheim]

I was also just basing this on my observations as an outsider. When I need a 3rd party JavaScript lib I define the 5 methods I am using and move on. But the absence of something like DefinitelyTyped in Haxe is inspiring a certain level of insecurity among prospective haxe/js devs.

I'm not even saying that such an effort has to be using ts2hx or TypeScript as a base, but it seems the most obvious choice for such a massive undertaking. But still, that would need to be driven by somebody who actually is going to use this in their day to day work.

But now that we've successfully lured you into this discussion, I would like to know how hard it would be to have the compiler be able to get externs directly from .d.ts for haxe/js, much like it can get externs from .swc for haxe/flash and .jar files for haxe/java. An IDL such as WebIDL or Thrift might also be worth considering. Not full alternative frontends, that would have to jump through hoops to . Just something that parses signatures and bends them onto the Haxe type system, much like some of the compiler targets do already. Any thoughts on that?

Best,

Juraj

--

[Dan Korostelev]

четверг, 5 марта 2015 г., 16:57:16 UTC+3 пользователь back2dos написал:

But now that we've successfully lured you into this discussion, I would like to know how hard it would be to have the compiler be able to get externs directly from .d.ts for haxe/js, much like it can get externs from .swc for haxe/flash and .jar files for haxe/java. An IDL such as WebIDL or Thrift might also be worth considering. Not full alternative frontends, that would have to jump through hoops to . Just something that parses signatures and bends them onto the Haxe type system, much like some of the compiler targets do already. Any thoughts on that?

That is a very clever idea! Supporting d.ts would be a great advantage in competition with TypeScript! We should definetely look into it. 

[Cauê Waneck]

OKay, I got more excited about this than it would be normal to. BUT PORTING ts2hx TO OCAML AND BAKING IT INTO THE HAXE COMPILER WOULD BE GENIUS!!!!!!!!!

Let's do this.

--

[Skial Bainn]

But now that we've successfully lured you into this discussion, I would like to know how hard it would be to have the compiler be able to get externs directly from .d.ts for haxe/js, much like it can get externs from .swc for haxe/flash and .jar files for haxe/java. An IDL such as WebIDL or Thrift might also be worth considering. Not full alternative frontends, that would have to jump through hoops to . Just something that parses signatures and bends them onto the Haxe type system, much like some of the compiler targets do already. Any thoughts on that?

This just reminded me of a project I came across some time ago but forgot to mention. When I found it I thought it might be able to do exactly this - https://github.com/brownplt/LambdaS5, how much trouble would it take to hook this into haxe?

--

[Juraj Kirchheim]

Yeah, I've always wanted to try that with Context.onTypeNotFound. Unfortunately ts2hx doesn't have positions and hxparse won't work at macro mode. But I figured it shouldn't be too hard to add that into the compiler directly.

It would indeed be freaking awesome. I just thought it's one of the easy additions that don't make it for obscure reasons. Like short lambdas :D

--

[Franco Ponticelli]

Did I hear `short lambdas`? My heart bleeds :(

[Philippe Elsass]

Re: convergence to ES6

It is very real as Microsoft, Google and Facebook are all standardizing on ES6 and typescript-y definitions - and I have first-hand confirmations that JS devs are jumping on the bandwagon.

I've tried ts2hx on a few externs and I found it awkward to use and, just like you don't want to generate externs for JARs or .NET DLLs, supporting natively the .d.ts in the Haxe compiler would be dramatically better. Especially if it could have a bit of flexibility that doesn't require the API torture needed to get some JS patterns into externs (ie. module containing not only classes but also consts and functions).

--

Philippe

[clemos]

I totally agree with Daniel here: increased visibility, targetting the JS community, and more ES6 features (and yes Franco, short lambdas :D)

I believe Haxe has a huge yet underrated potential as a web (non-game) language / toolset.

I find Scala nice but too convoluted, node.js powerful but too wild, 

not to mention server/client convergence (react.js and stuff),

Haxe just fits perfectly there.

Best,

Clément

On Thu, Mar 5, 2015 at 8:35 AM, Daniel Glazman <daniel....@gmail.com> wrote:

[Dan Korostelev]

четверг, 5 марта 2015 г., 17:31:21 UTC+3 пользователь Philippe Elsass написал:

Microsoft, Google and Facebook are all standardizing on ES6 and typescript-y definitions

What's those typescript-y definitions? Any link? How different are they from current d.ts? 

[Philippe Elsass]

AtScript is apparently building on Typescript's d.ts:

http://blogs.msdn.com/b/typescript/archive/2014/10/22/typescript-and-the-road-to-2-0.aspx

But there are a few subtleties between Typescript and Flow:

https://github.com/Microsoft/TypeScript/issues/1265

--

To post to this group haxe...@googlegroups.com
http://groups.google.com/group/haxelang?hl=en
---
You received this message because you are subscribed to the Google Groups "Haxe" group.
For more options, visit https://groups.google.com/d/optout.

--

Philippe

[Daniel Glazman]

On 05/03/2015 15:55, Philippe Elsass wrote:

> AtScript is apparently building on Typescript's d.ts:
> http://blogs.msdn.com/b/typescript/archive/2014/10/22/typescript-and-the-road-to-2-0.aspx
>
> But there are a few subtleties between Typescript and Flow:
> https://github.com/Microsoft/TypeScript/issues/1265

Honestly, I am not sure diving into the technical details is worth it.
On the contrary, let's look at the market situation:

- a few years ago, nobody would have bet a single penny on a script-like
language to build native apps. In fact everyone was laughing at me
when I mentioned it.
- then Haxe came
- then others came
- then Swift came
- then, then, then

Remember Gandhi? "First they ignore you, then they laugh at you, then
they fight you, then you win". Haxe's model has clearly won. But as I
said earlier, it remains unknown outside of the gaming community.

The only thing I know for sure - because I live with these people all
year long - is the following one: offer a html/css/ecmascript developer
a way to develop a native app on both desktop and mobile from the
technologies he already knows w/o having to learn new ones and you'll
win a new and devout user. Worth a discussion in the Haxe community
IMHO...

</Daniel>

[Heinz Hölzer]

short lambdas would indeed be a nice new feature!

[David Elahee]

Haxe has been around under the mtasc, motiontype or whatever for around 6-7 years now...and those disucssions ahve been around for that long now. the reality is that if you don't have a big google/sun/ms company still backing haxe, there will not be this haxe trend coming and that's actually cool.

Haxe aim is to get past trends and crazyness, it is meant to be a reliable, stable language that will always be here. Sicne we at Motion Twin use it a -lot-..I think haxe dynamic will stay the same, growing slowly and reliably... i mean, who want to go back to c++, another cosed source swift or to go back to unity3d after the big engine will get stability ? How do one justify swift crazyness when there are rust , ocaml, python doing same but better ( less integrated, i grant one...)

So there will still be trends and effusion around the next js thing etc but this is not haxe way of evoluting.  

We have no GAFA and no trend and maybe it 's just ok that we do not have that effusion here so that we can focus on improving rather than unrolling feature that makes good communication but pain the reliability (yes DCE that's is you i am talking of :) )

So in 5 years : ufront gets the documentation it deserves outside github, haxelib gets a new site, Simn finishes genc and Heaps Studio become the next open source unity3d. What else :) trollololololo !

Bisous !

--

David Elahee

[Rafael Oliveira]

In five years people will be using haxe to create mobile apps (not games), using native speed, and will forget phonegap :-)

[Mike Robinson]

I'd like to weigh-in here as a total newcomer . . .

I think that it should be "obvious to everyone in the computer software business" that "JavaScript (of course ...™) applications, running on [unpredictable] web browsers, on who-knows-what platform" ... has always been ridiculous.

The only reason why we ever saw fit to put up with such a thing, is that it was the only way that we had.  At the time.  But then, tablets and smart-phones came along, and their computing and (especially ...) graphics power made our "desktop" computers look silly.  The vendors of those portable devices offered us a Hobson's Choice:  "accept that We have Won The Game™, [pay us money to ...] become One of Our Developers, and spend The Rest Of Your Career™ writing Applications For Us™ using Our One-Of-A-Kind [of course™ ...] Tools and Techniques."

Projects such as Haxe (and its now-many offshoots ...) are just the sort of "brilliant, and obvious in retrospect" challenges to such a status-quo that one day will look so very brilliant, and so obvious:

1. "Necessarily," the language that your program winds up in, must be one that is "native to the platform in question."
2. However, the language that your program is written in, must be "platform agnostic," because you cannot afford to pay for individual development efforts "to do the same thing."

Compromise solutions ... such as "JavaScript plus-plus™," or "let's wrap a JavaScript + web-browser solution in an 'App' wrapper and hope that nobody notices" ... good and sincere and well-intentioned though they might be ... are still "compromise™ solutions."

And, in comparison to these, "Haxe is revolutionary."

It is the technology that we've been looking for.  It should come as a surprise to no one that the world will take a little more time to catch up.

[Raoul Duke]

Somebody please send me a shipment of the Haxe flavor of Kool Aid
y'all got there ;-)

[Mike Robinson]

PhoneGap's designers (et al ...) should be pleased with themselves that they offered "a finger thrust into the dike at a most-opportune time."  (For this, they should be richly praised.)  However, they should graciously concede that their moment of prominence was always intended to be fleeting.  "Aye, you may be John the Baptist, but when Jesus shows up, get lost."  ;-)

[David Elahee]

Just a word about haxe for applications and adoption problems.

After discussing at Fosdem and with many local entrepreneurs that thought about haxe but skipped, haxe has some serious issues for them to adopt, namely :

- No established native toolkit access (cocoa,android toolkit, etc)
- Web toolchain lack integration and is hard to start with.
- No out of the box debugging solutions for those.
- No support company serious about (non media) applications

These are their arguments. I understand them since most of us have answered these problematics with custom or not documented or not standard solutions. Since the mass of haxe users is not appealed by having these issues solved ( cocoa...sigh...) I won t make any rant at all...

...But I am fairly confident Hf will help anyone that want to work on these aspects.

Please realize that if you think haxe should be usable by app developers ( not games or multi media ), you are to take actions because game and multimedia devs simply are not likely to invest in this out of the blue if they won't use it actively.

Thanks.

Ps: short lambda ftw !

[Andreas Söderlund]

Thank you guys for great feedback! The question is surely hard to answer, but the fact that many of us are making our living in such a fluctuating area, isn't that a bit worrying? Is our fate to always be on our toes, ready to target the next technology that will suddenly appear and claim to be as useful as all the previous obsolete ones?

It may be fun a while to delve into the latest, but I don't like it anymore. I want more stability than worrying about what will happen in 5 years from now (imagine that in any other area). I want to use well-proven ideas and timeless concepts to build great software. For reasons usually involving money I don't think this will happen to the industry, but with Haxe we'll at least have stability in a really nice and flexible language. I'm convinced by your answers that won't change, so thanks again!

/Andreas

[Daniel Glazman]

On 05/03/2015 23:54, David Elahee wrote:

> Haxe has been around under the mtasc, motiontype or whatever for around
> 6-7 years now...and those disucssions ahve been around for that long
> now. the reality is that if you don't have a big google/sun/ms company
> still backing haxe, there will not be this haxe trend coming and that's
> actually cool.

Wait. "that's actually cool" ?!? You can't be serious.

> Haxe aim is to get past trends and crazyness, it is meant to be a
> reliable, stable language that will always be here. Sicne we at Motion
> Twin use it a -lot-..I think haxe dynamic will stay the same, growing
> slowly and reliably... i mean, who want to go back to c++, another cosed
> source swift or to go back to unity3d after the big engine will get
> stability ? How do one justify swift crazyness when there are rust ,
> ocaml, python doing same but better ( less integrated, i grant one...)

Slowly and past "trends and crazyness", eh? Let me give a few
interesting figures:

- Number of mentions of Haxe on CNet, the largest IT-related documents
database in the world, since epoch: ONE. And that document is about
Apple Swift with one single mention of Haxe and not even a hyperlink.

- Number of search results for Haxe on Google on the first page of
results outside of the Haxe core community: ZERO.

- Number of Press Releases (in French, communiqués de presse) sent by
the Haxe Foundation since epoch: apparently ZERO. I searched on all
known PR databases and found nothing. Let me remind you a PR sent
through for instance PRWeb to *all* IT magazines in the world costs
between $150 and $250, so next to nothing for potential big wins.

- Number of news mentions of the fact Tivo switched to Haxe,
quite a _major_ breathrough that should have triggered intense
communication: ZERO. Only a few community blogs and SilexLabs have
discussed it.

Let me output a loud wow, really. You're perfectly right, no trend!

Right now as we chat, developers in the Apple ecosystem are
experimenting with Swift more and more every day, and I know gazillions
of them who declare their next project will be based on Swift instead of
Obj-C. That trend is everywhere, and the other major vendors also
follow it, sometimes moved before Apple. I bet a box of cookies it will
happen to those script-like languages what happened to .Net. An OSS
cross-platform version of it will appear, just like Mono for .Net, and
that will a bell ringing for market rationalization.
That day, Haxe will be in trouble because of his structure: the core
of the technology (I'm not speaking of libs here) is maintained by only
a handful of people, part-time only.

Speaking of "How do one justify swift crazyness", I have a trivial
answer: 800,000,000 iOS devices sold (20140601), $10B sales on the
App Store in 2013, incredibly powerful and mature environment offered
by Xcode. The other languages you're quoting (rust, ocaml, python) are
nowhere near this and that's very unlikely to change in ANY
foreseeable future.

> So there will still be trends and effusion around the next js thing etc
> but this is not haxe way of evoluting.
>
> We have no GAFA and no trend and maybe it 's just ok that we do not have

If you have no GAFA, build one. I heard the Haxe Foundation is not a
french association but a corporation, is that right? (for people outside
of France, "Foundations" as they exist in the anglo-saxon world do not
exist in French law). If that choice was made, that's probably because
of the expectation of a revenue stream for the Haxe Foundation, right?
I'd love to read your explanations here and I'd love to have the
financial report of the Haxe Foundation, please. This is the kind of
data an investor absolutely needs to make sure the technology he could
bet on is reliable. Please make that financial report public.

> that effusion here so that we can focus on improving rather than
> unrolling feature that makes good communication but pain the reliability
> (yes DCE that's is you i am talking of :) )
>
> So in 5 years : ufront gets the documentation it deserves outside
> github, haxelib gets a new site, Simn finishes genc and Heaps Studio
> become the next open source unity3d. What else :) trollololololo !

I had to reread this several times to believe it. As you said yourself,
Haxe is 6-7 years old and is still totally unknown outside of a small
bubble of early adopters. The number of paid core contributors is so
restricted it could be at risk if only one was ill or forced to work
fulltime on something else just for financial reasons. So even with
smileys and humour, I find your answer not only depressing but also
scary, sorry.

I have the gut feeling the Haxe Foundation focuses strictly on the
technical side, neglecting entirely the mandatory steps of
communication, evangelism and ecosystem maintainance and support.
And I fail finding any hint at a strategy plan beyond the technical
level.

From my POV, Andreas's (tjena!) questions are only fair, and carry
strong industrial value. They remain, at least to me, too largely
unaddressed. I am currently looking for investors on a Haxe-based
project and I now have the same questions Andreas had. I won't embark
on a large investment - and no VC would embark either - w/o
understanding better how Hf and Motion Twin see and will organize the
future of Haxe, from a non-technical point of view. I suggest we
dedicate some time at WWX to sit around a table and have a frank
discussion on the topic, it's definitely worth it. Thanks.

</Daniel>

[David Elahee]

@daniel don't shoot me, we simply had this discussion internally many _many_  times and we simply got to the pragmatic stage. The haxe foundation is not currently capable of doing the right amount of  evangelism and generating the need/hype you evoke. Because it requires marketing force.

So I actually said you are perfectly right in you remarks AND that at the moment the haxe ecosystem cannot ( mechanically) overcome it.

It is not a matter of good will but of having the good people to do that job. 

"And I fail finding any hint at a strategy plan beyond the technical

level."

"I have the gut feeling the Haxe Foundation focuses strictly on the

technical side"

-> That sums it all. you are right and that is exactly what I said.

If you know some, we will be happy to discuss that :) 

If you think that makes haxe no less serious about being a great tech, you are wrong.

Please understand that making a technology mainstream have strong requirements and I am simply pragmatic about it.

" I find your answer not only depressing but also

scary, sorry."

No offense meant really, I don't want to depress you, I got past that stage myself and feel confortable in the haxe scope. I would really actually want you to say "Yeah ok i ll help evangelize"...

Of course you are doing it on a daily basis and me too but if one asks me for the next five years, I have to be honest and that is what I always do. Call me Cassandra if need be.

Haxe need more arms and insights about how to proceed and who can help efficient promotion, this we never shied away. I'll gladly discuss with you at wwx ;)

</Daniel>

--
To post to this group haxe...@googlegroups.com
http://groups.google.com/group/haxelang?hl=en
--- You received this message because you are subscribed to the Google Groups "Haxe" group.
For more options, visit https://groups.google.com/d/optout.

--

David Elahee

[Daniel Glazman]

On 06/03/2015 15:49, David Elahee wrote:
>
> @daniel don't shoot me, we simply had this discussion internally many
> _many_ times and we simply got to the pragmatic stage. The haxe
> foundation is not currently capable of doing the right amount of
> evangelism and generating the need/hype you evoke. Because it requires
> marketing force.

Yes. And to do that, Haxe needs to go beyond its bubble of early
adopters. Expressed as you did, it's a chicken and egg issue.
It's not.

> Of course you are doing it on a daily basis and me too but if one asks

No we don't, and we don't at all. Evangelism is a specific job and
talent. We do talk about what we do, that's entirely different.

The only fact you're saying «I would really actually want you to say
"Yeah ok i ll help evangelize"» shows the depth of the problem. I
cannot evangelize first because I am not an evangelist, second because
I am working on a product and have only 24hrs per day, third because
that's the Haxe Foundation that gets revenue from corporations using
Haxe, not my company. This is Hf's role to expand Haxe's ecosystem.

The simple question is are you ready to change of scale factor or not.
Your previous message sounded like not only you're not, but you're also
not welcoming that change, hence my reply.

Note: I still want the financial report of the Hf.

</Daniel>

[David Elahee]

You should talk with Nicolas. We're all in the same team here.

</Daniel>

--
To post to this group haxe...@googlegroups.com
http://groups.google.com/group/haxelang?hl=en
--- You received this message because you are subscribed to the Google Groups "Haxe" group.
For more options, visit https://groups.google.com/d/optout.

--

David Elahee

[Daniel Glazman]

On 06/03/2015 16:12, David Elahee wrote:

> You should talk with Nicolas. We're all in the same team here.

Alright then. I suppose he receives and reads all the messages in
this forum so I'm waiting here for his answers :-)

</Daniel>

[David Elahee]

Daniel,

Mailing list are not exactly the place to discuss such complex issues extensively, I dare say you know it. Hf will gladly hand you reports, and financially support initiatives aiming at widening haxe scope and even pay people for hard labour, we said that tons of times.

So, please consider mailing Nicolas or skyping him, ( and deliver your conclusions later on, no problem ) but mailing list is really not nice for that.

</Daniel>

--
To post to this group haxe...@googlegroups.com
http://groups.google.com/group/haxelang?hl=en
--- You received this message because you are subscribed to the Google Groups "Haxe" group.
For more options, visit https://groups.google.com/d/optout.

--

David Elahee

[Daniel Glazman]

On 06/03/2015 16:46, David Elahee wrote:

> Daniel,
>
> Mailing list are not exactly the place to discuss such complex issues

Ah. This goes against all my personal experience. I must have been - and
with me thousands of people I've been working with during all these
years of OSS activity - mistaken.

> extensively, I dare say you know it. Hf will gladly hand you reports,
> and financially support initiatives aiming at widening haxe scope and
> even pay people for hard labour, we said that tons of times.
>
> So, please consider mailing Nicolas or skyping him, ( and deliver your
> conclusions later on, no problem ) but mailing list is really not nice
> for that.

Wait. Hf, a "foundation" about Open Source Software, wants me to send a
_private_ message to get a _private_ answer and a financial report for
my eyes only? Seriously? Wow. You just don't get it, do you? Hf must
*publish* its financial report so anyone interested in Haxe can read it
and place a safe bet on the fate of the technology.

IMHO, everyone here in this forum is potentially interested, now or
later, by those "complex issues". So why not answer here?

</Daniel>

[Juraj Kirchheim]

On Fri, Mar 6, 2015 at 4:01 PM, Daniel Glazman <daniel....@gmail.com> wrote:

The only fact you're saying «I would really actually want you to say
"Yeah ok i ll help evangelize"» shows the depth of the problem. I
cannot evangelize first because I am not an evangelist, second because
I am working on a product and have only 24hrs per day, third because
that's the Haxe Foundation that gets revenue from corporations using
Haxe, not my company. This is Hf's role to expand Haxe's ecosystem.

The Haxe Foundation's role is explained here: http://haxe.org/foundation/

Nowhere does it even mention evangelism, expansion of the eco system or anything like it.

The Haxe Foundation gets money through "paid support plans". Full stop. I don't know how that would make them responsible to ANYONE to either make financial reports or fund whatever crusade they are planning to undertake.

You are making the Haxe Foundation look like they don't do their jobs. That borders on defamation.

At the same time I do agree that it would be nice if the foundation would tackle evangelism as well. But for a sensible discussion it might be good to see the status quo for what it is.

And forgive me for not sugar coating this, but seriously, dial it down. You come off as EXTREMELY full of yourself and insensitive to anyone who doesn't share your particular view. You're not even trying to understand other positions. That's not going to lead anywhere.

Best,

Juraj

[David Elahee]

"Seriously? Wow. You just don't get it, do you?"

It took me a big time to speak clearfully and explain, I tried to keep it civil. But it is time for me to bail out. I'll return to my gpu's.

I'll let Nicolas handle this.

--

David Elahee

[Daniel Glazman]

I look forward to reading him here, about strategy, communication,
evangelism, public financial reports and more. And all in all, about Andreas's question : where will be Haxe in a few years from now?
And yes, as in any other Open Source community, all of that should
happen in public.

</Daniel>

[Nicolas Cannasse]

Wow, this whole subthread went a bit far, and of course it happens when
I'm on vacation :)

A few comments:

- NO, being "under the radar" is not cool in any way. As Haxe original
creator, I want it to get the credit it deserves because it's a great
technology that all of the people here enjoy using on a daily basis.
David is talking in his own name and not in the name of Haxe Foundation.

- YES, we have hard time dealing with communication, doing PR, etc.
That's because we have no-one dedicated to this and I the current tasks
I'm handling are taking all of my time already. We have some budget from
our partners but it's not enough to hire someone with a high profile
full time either. I've been looking for some people that could help part
time but with no success so far.

I'm usually presenting our current state and strategy at yearly WWX.
I'll make sure to include all the financial data this year since we have
a better overview after a bit more than 2 years since I started HF, and
we have to find ways together to improve the skills we're lacking.

Best,
Nicolas

[white...@gmail.com]

Hi Nicolas,

in order to fully understand your point what would it take to help HT part-time? Do you want to hire contractors on specific PR issues? Or something else? Would you have a todo list ready ?

 

(No need to rush to answer though. Enjoy your vacation  )

 

++

 

François Nicaise
http://www.linkedin.com/in/fnicaise
www.francoisnicaise.fr

Freelance Game Developer / Designer 
Business Relations @ FrenchCows
Gaming & Business @ Bordeaux Games

"Nicolas Cannasse" <ncan...@gmail.com> wrote:

--

[Daniel Glazman]

On 09/03/2015 14:04, Nicolas Cannasse wrote:

> Wow, this whole subthread went a bit far, and of course it happens when
> I'm on vacation :)
>
> A few comments:
>
> - NO, being "under the radar" is not cool in any way. As Haxe original
> creator, I want it to get the credit it deserves because it's a great
> technology that all of the people here enjoy using on a daily basis.
> David is talking in his own name and not in the name of Haxe Foundation.

Aaaah. Thanks for bringing a bit of sanity here. I started wondering if
sense of market-driven strategy was a curse word in the Haxe community.

> - YES, we have hard time dealing with communication, doing PR, etc.
> That's because we have no-one dedicated to this and I the current tasks
> I'm handling are taking all of my time already. We have some budget from
> our partners but it's not enough to hire someone with a high profile
> full time either. I've been looking for some people that could help part
> time but with no success so far.

My own company is in the same position. Too little to hire a comm
person, too isolated to avoid using PRs. I ended up writing myself my
PRs and using a web service like PRWeb.com. A PR sent to all magazines
in the world is rather cheap, for immense return value if you follow a
few rules. Again, the TiVo achievement should end up in a PR, probably
co-published by Hf and TiVo. That would be huge, and is definitely
worth the time away from code.
Again, there are a few simples rules to follow, and some formatting
habits to get; but apart from that, writing and releasing a PR is waaaay
simpler than it was 15 years ago.

> I'm usually presenting our current state and strategy at yearly WWX.
> I'll make sure to include all the financial data this year since we have
> a better overview after a bit more than 2 years since I started HF, and
> we have to find ways together to improve the skills we're lacking.

Ok, cool. I suggest you don't wait WWX and make these data public ASAP
so we have enough time to read them, understand them and prepare
questions or even suggestions for WWX. As an example, W3C sends its
yearly financial report to all W3C Members a while before the plenary
meeting so we can review it and be ready for the session.
If we discover the data during the session, that drastically reduces
attendees' ability to deal with it AND it does not touch everyone.
Thanks.

</Daniel>

[Nicolas Cannasse]

> My own company is in the same position. Too little to hire a comm
> person, too isolated to avoid using PRs. I ended up writing myself my
> PRs and using a web service like PRWeb.com. A PR sent to all magazines
> in the world is rather cheap, for immense return value if you follow a
> few rules. Again, the TiVo achievement should end up in a PR, probably
> co-published by Hf and TiVo. That would be huge, and is definitely
> worth the time away from code.
> Again, there are a few simples rules to follow, and some formatting
> habits to get; but apart from that, writing and releasing a PR is waaaay
> simpler than it was 15 years ago.

Yes, you can be sure we tried that, but TiVo is a big company, so as
soon as it gets into a PR, you need agreement not only from the
engineering dpt but also marketing, etc. We should definitely have
pushed more in that direction.

>> I'm usually presenting our current state and strategy at yearly WWX.
>> I'll make sure to include all the financial data this year since we have
>> a better overview after a bit more than 2 years since I started HF, and
>> we have to find ways together to improve the skills we're lacking.
>
> Ok, cool. I suggest you don't wait WWX and make these data public ASAP
> so we have enough time to read them, understand them and prepare
> questions or even suggestions for WWX. As an example, W3C sends its
> yearly financial report to all W3C Members a while before the plenary
> meeting so we can review it and be ready for the session.
> If we discover the data during the session, that drastically reduces
> attendees' ability to deal with it AND it does not touch everyone.

That's a good reason, although i'm not sure that interests a lot of
people, I'll do my best to publish things out before WWX.

Best,
Nicolas

[Nicolas Cannasse]

Le 09/03/2015 14:14, white...@gmail.com a écrit :
> Hi Nicolas,
>
> in order to fully understand your point what would it take to help
> HT part-time? Do you want to hire contractors on specific PR issues? Or
> something else? Would you have a todo list ready ?

We have elaborated some strategy already but it needs to be implemented
by someone, most likely part time, in a global manner. Wwe don't want to
hire 10 different people because we don't have much time allocated for
management so we are looking for someone which can works independently,
reports on a regular basis, and asks for input when necessary.

Best,
Nicolas

[white...@gmail.com]

ok! Thanks for sharing 

 

François Nicaise
http://www.linkedin.com/in/fnicaise
www.francoisnicaise.fr

Freelance Game Developer / Designer 
Business Relations @ FrenchCows
Gaming & Business @ Bordeaux Games

"Nicolas Cannasse" <ncan...@gmail.com> wrote:

--

[Nicolas Cannasse]

Actually to be precise, Josefiene Pertosa has started to help us out in
that matter just a few weeks ago, so it's not like we're doing nothing
in that matter, it just takes more time than we first expected. We still
welcome any help/suggestions in that area.

Best,
Nicolas

[Daniel Glazman]

On 09/03/2015 14:43, Nicolas Cannasse wrote:

> Yes, you can be sure we tried that, but TiVo is a big company, so as
> soon as it gets into a PR, you need agreement not only from the
> engineering dpt but also marketing, etc. We should definitely have
> pushed more in that direction.

Yeah, well. There are multiple ways of annoucing such things. First,
if TiVo's usage of Haxe is public information, you don't need their
approval if you don't speak for them. There are ways to present a PR
so it does not express any second-party's POV and will be considered
harmless by their marketing department. And then if it's not, you can
tell TiVo they're already supporting the Hf and that not building above
it in terms of communications would be counter-productive for both.

> That's a good reason, although i'm not sure that interests a lot of
> people, I'll do my best to publish things out before WWX.

The "Rapport Annuel de la Cour des Comptes" in France is read by roughly
two thousand people people only in a country of 68 millions. But if it's
not published, it's read by zero.
Hf is a corporation because of the french law and because you have
revenue streams. But you called it a Foundation. A Foundation
publishes financial information. You still have to publish them on a
yearly basis at the Tribunal de Commerce and they can be obtained
through the Greffe, right?

</Daniel>

[Mike Robinson]

I'd cautiously weigh-in here with the frank opinion that:  

"you have: what nobody else has, and what everybody is looking for."

You have a genuine, viable, cross-compiler technology which can deploy a native application to multiple platforms ... including ones that are still emerging ... from one code-base.

You also have "a strongly-typed compiler for JavaScript," which enables many errors to be detected "at compile time" which otherwise must be caught "at runtime" using some kind of debugger.  Even if your target is JavaScript, this greatly reduces the likelihood that a "silly mistake" will not be caught.  (And, "silly mistakes" are expensive.)  The same is true of other targets such as PHP.

Right now, developers are caught in a conundrum that they don't want to be in.  They have one of four [bad] choices:

1. Pretend that "other markets just don't matter."  Pretend that everyone has an iPhone® or an iPad® in their pocket, or that they soon will.  Develop native applications in the native application-development environment of one particular vendor:  an environment that is intended to be unique to that platform and thereby to "lock-in" developers to it.
2. Launch multiple parallel efforts, targeting multiple platforms, and doing each one in the native application-development environment for that platform.  Now, you are writing and maintaining n entirely-different source code bases.  Your development costs have risen by [more than ...] n times.
   
3. Use Java, somehow, because Java claims to be cross-platform.  Hope that you can find a JVM that actually runs fast enough to get out of its own way.
4. Pretend that "an 'app' is 'a web site.'"  Pretend that, if you put a web-site into a box with an icon on it, you've actually written an 'app.'  Pretend that JavaScript really is "okay."  Pretend that a mobile version of "Internet Explorer hell" really is just "a toasty, warm fire that you like to sit beside."  Pretend that "the 'right way' has been found, and that 'right way' is PhoneGap / Cordova."
   

The haxe, NME, and OpenFL teams ... each in their own way, and working cooperatively instead of being at-loggerheads ... identified a very practical strategy and have spent years perfecting it, such that, "today, it actually works."  It works for more than just "game" development.  The resulting application is native, and multiple targets can be hit using just one source-code base.  The amount of "platform-specific" code that you have to write and to maintain is only a small fraction of the total.

Most, if not all, of the Really Good Things™ that all of us rely on, every day, did not come from big corporations who spent a bunch o' money marketing it.  (Per contra, many of the things that were hyper-marketed ... no longer exist.)  The Really Good Things in our world were all things that someone wrote, and that someone else discovered, and slowly it caught fire.  Only then did the reporters, sniffing a story, show up and start to write about it.

I repeat:  "you have what nobody else has, and what everybody is looking for."  Soon enough, they will find it.  In fact, they are finding it now.

[Alex Hoyau]

Hello

Sorry to bloat this thread a little more but I feel like it is a most important subject on which Silex Labs must have a word (once again ;). Thank you @andreas for this opportunity!

In 2011 in Bordeaux, we sat and we talked about 

- how to make Haxe mainstream, 

- how to ensure the sustainability of Haxe, 

- how we wanted people to see us as a community,

We never got to agree but many of us have expressed a strong interest in these matters. And following these discussions the Haxe Foundation and Silex Labs WWX have been created.

These are very good first steps. But we can do more. The first time was good? Let's do it again! We sit face to face and ask the same questions again. But this time we set goals for the upcoming 2 years. 

The thaxe force discussion group will talk about this to make it happen at the WWX2015. 

So what now? Wait and see? No! Please do something!

Why bother? This is about having more fun: more 100% Haxe jobs, more interesting Haxe projects for freelancers, less barriers for companies to say yes to Haxe for their projects. And maybe girls in the community one day?!

How to help? If you do not want to get involved in the thaxe force nor in the evangelization efforts of Silex Labs, we need to know that you still care about this. Please share with us in this thread, or just say "+1".

And if you are a new comer (especially one with the background of @daniel ;) please do not feel like you do not have your word to say, we need you too, this is an opportunity to be useful!

See you in 2 months in Paris!

Alex Hoyau