Why so few Ios demos ?
category: code [glöplog]
Note that iOS 8 now supports WebGL. The only thing I had to make http://www.pouet.net/prod.php?which=63988 work on it (online version only) was about music (can't play a streamed MP3 if the user does not press a button to start it).
So it's easy to push values and textures to pixel shaders using javascript, and now it runs on PC, Mac, Androis and iOS.
So it's easy to push values and textures to pixel shaders using javascript, and now it runs on PC, Mac, Androis and iOS.
@Soundy hey that's very interesting, it might make me move to iOS8 ! Do you know if the performance is good on an iPhone5 ?
Actually it runs well on my iPhone5... provided there's not too much in Javascript. But shaders run well.
Quote:
In the source for desktop OpenGL PC demo's I've seen where Modern GL features are also used a lot of the legacy/deprecated ways of doing thing are also still present, which leads me to wonder whether any 'fully' modern GL things are out there?
With OpenGL ES, you *need* to use modern stuff, because the legacy stuff was removed from version 2.0 (similar to core profiles on desktop OpenGL). And you don't want to use any pre-2.0 stuff, because that's fixed-function only.
I've done a simple port of my OpenGL stuff to iOS and Android... then back to core profile on the desktop :P
Oh yeah I know that. I started back using GL after long absence on GL ES2.0, and ulimately still feel like i'm being a masochist deciding to stick to core profile style GL on the desktop for now. Its a whole world away from the old accessible GL, and more frustrating to find why things aren't working.
When I say modern GL, a the very basic level I mean having to use VAOs/VBOs not Glbegin / glvertex / glend. This means there's little less on the fly plugging of values, and requires a different coding style.
There's a level of what we'd consider basic programming knowledge needed to use Modern GL that a lot of real-world programmers out there just don't have. I used to work in a role where we had developers as customers, and believe me a lot of people don't know about simple stuff like buffers and pointers, yet get great bits of software going. You have to be very comfortable with that kind of thing to use modern GL effectively.
I feel a lot of the time i'm fighting the API, and not getting things done, which probably explains 1) the low take up 2) why people who want to be game devs are, these days, using stuff like unity.
When I say modern GL, a the very basic level I mean having to use VAOs/VBOs not Glbegin / glvertex / glend. This means there's little less on the fly plugging of values, and requires a different coding style.
There's a level of what we'd consider basic programming knowledge needed to use Modern GL that a lot of real-world programmers out there just don't have. I used to work in a role where we had developers as customers, and believe me a lot of people don't know about simple stuff like buffers and pointers, yet get great bits of software going. You have to be very comfortable with that kind of thing to use modern GL effectively.
I feel a lot of the time i'm fighting the API, and not getting things done, which probably explains 1) the low take up 2) why people who want to be game devs are, these days, using stuff like unity.
Quote:
When I say modern GL, a the very basic level I mean having to use VAOs/VBOs not Glbegin / glvertex / glend. This means there's little less on the fly plugging of values, and requires a different coding style.
Yup, that's OpenGL ES 2.0+/core profiles for you...
Funny, people were always complaining that DX required so much 'boilerplate' to get things working... But now that OpenGL is cleaned up, it needs about the same amount of boilerplate, and unlike DX, the APIs often don't quite make sense, because they are legacy API functions that behave differently once you use a VAO/VBO extension (which is no longer an extension, but part of the core now).
Going sorta back on topic... I have not seen a single Windows Phone (or even Windows RT) demo yet. And that should actually be quite easy to pull off, since you can use the exact same DX11 code as you do on the desktop (the main difference is that you have to connect it to a CoreWindow rather than a HWND, just one function call to change).
Which imho is a lot better than OpenGL, where ES is a different flavour than the desktop OpenGL (annoying that GLSL is not 100% interchangeable between the two... why!?).
I've modified my DX11 codebase to work on WinRT/WP last year, but never did anything with it. I would have half expected some of the people releasing DX11 demos to at least release a Windows Phone port for teh lulz by now. But no, afaik, no Windows Phone prods yet.
Problem is not many have windows phone of course.
What are some cheap good windows phone btw? Don't think I'll buy, but I am curious to see how affordable is one for demos.
What are some cheap good windows phone btw? Don't think I'll buy, but I am curious to see how affordable is one for demos.
Quote:
Problem is not many have windows phone of course.
You don't even need one :) There's a Windows Phone emulator in the SDK.
Quote:
What are some cheap good windows phone btw? Don't think I'll buy, but I am curious to see how affordable is one for demos.
I think you can get a cheap Lumia for around 100-120 euros these days.
Quote:
I've modified my DX11 codebase to work on WinRT/WP last year, but never did anything with it. I would have half expected some of the people releasing DX11 demos to at least release a Windows Phone port for teh lulz by now. But no, afaik, no Windows Phone prods yet.
WP has the same problem as IOS - appstore access only.
Awwww
Quote:
WP has the same problem as IOS - appstore access only.
Is that actually a problem though?
I know that Kewlers has their demos in the Android app store. Perhaps Apple is too picky about what they want in the store, but I doubt that Microsoft would make a problem of it.
And as already said, you can always 'disguise' it as a benchmark ;)
I'm sure there's a way through, perhaps as an "art experience" or suchlike.
If Mountain can do it, it certainly should be doable.
Besides that it will be pretty hard to get it through the review process you also need to pay 99 dollars a year to keep it in the store.
Blackpawn got this through the review process though:
https://itunes.apple.com/us/app/fire-skull/id476328452?mt=8
https://itunes.apple.com/us/app/plasma-effect/id321096615?mt=8
https://itunes.apple.com/us/app/fire-skull/id476328452?mt=8
https://itunes.apple.com/us/app/plasma-effect/id321096615?mt=8
I guess the thing they might like about those two is interactivity?
Quote:
Besides that it will be pretty hard to get it through the review process you also need to pay 99 dollars a year to keep it in the store.
Hum, but if you have one account, you can publish as many apps as you like, right?
So, an idea would be to have a 'pouet.net'/'scene.org' account, which will publish iOS releases for all groups. That way it's only 99 dollars for the whole scene, rather than for each group.
I think Android and Windows Phone have considerably lower fees for publishing though.
Windows Phone is also 99 dollars. And I think Android too but I am not sure.
A shared account might work but there is a limit to the number of developers assigned to a team. Also someone will need to maintain all profiles, submission etc. You don't want to give everyone admin rights.
A shared account might work but there is a limit to the number of developers assigned to a team. Also someone will need to maintain all profiles, submission etc. You don't want to give everyone admin rights.
(and pure demos/intros won't get accepted anyway)
Quote:
Windows Phone is also 99 dollars. And I think Android too but I am not sure.
Ah yes, for a full company account.
Individual account is only 19 dollars: http://msdn.microsoft.com/en-us/library/windows/apps/jj206719(v=vs.105).aspx
But I'm not sure if an individual account would be allowed for this.
Of course without the app store, there's always the jailbroken route, at least for wild entries?
Way too many comments for me to say anything new but basically:
1) $100/year
2) Apple won't approve your demo app (and nobody jailbreaks their phones anymore), in both cases nobody will run your demo
1) $100/year
2) Apple won't approve your demo app (and nobody jailbreaks their phones anymore), in both cases nobody will run your demo
Quote:
2) Apple won't approve your demo app (and nobody jailbreaks their phones anymore), in both cases nobody will run your demo
Does that even matter though?
The challenge is to write a demo that makes the most of the iPhone/iPad hardware and compete with it.
As Canopy says... you can treat it as a 'wild demo', which can be installed on jailbroken devices, and just put a YouTube up for people who don't want to install it.
Whether other people actually run the product on a physical device, or just watch a video of it, doesn't matter imho.
an android developers account is 25 dollar a year. you can release as much as you want and apps don't require pre-approval, but your account can get flagged by users for releasing crap :)
Regarding iOS, alternatively you can do enterprise builds for apps so you can install non-appstore apps over iTunes, but afaik that requires listing endorsed device IDs in the app manifest itself, so not really convenient for demoscene purposes :) at least, that's how it was ~3 years ago.
Regarding iOS, alternatively you can do enterprise builds for apps so you can install non-appstore apps over iTunes, but afaik that requires listing endorsed device IDs in the app manifest itself, so not really convenient for demoscene purposes :) at least, that's how it was ~3 years ago.
