Fwd: Gamepad API [Was: New feature flag proposal: Joystick API]
Forwarding two messages from Jerry Seeger, whose e-mails aren't getting to the list yet. Answering Jerry's comment, I think that targeting the spec to explicitly address gaming use case is better than the current situation. A very large group of devices is undoubtedly built with gaming and only gaming in mind. Jerry's proposed line of thinking certainly seems very promising, too. - WBR, Alexey Proskuryakov Начало переадресованного сообщения:
От: Jerry Seeger <vikingjs@me.com> Тема: Ответ: [webkit-dev] Gamepad API [Was: New feature flag proposal: Joystick API] Дата: 12 октября 2011 г. 18:10:05 Тихоокеанское летнее время Кому: Alexey Proskuryakov <ap@webkit.org> Копия: Scott Graham <scottmg@chromium.org>, WebKit Development <webkit-dev@lists.webkit.org>
'Game' is a use case, not a description of device characteristics. GameController and Gamepad both carry the same ambiguity. Alexey, I have to admit that maybe I didn't understand your original objections, if a swapping -pad for -Controller makes the spec OK.
Sorry Alexey to beat the horse, but I'm including here my previous message, while I figure out why my messages aren't going to the list:
I haven't been a direct contributor to Webkit, but I think Alexey has a valid point. A gamepad is some sort of amalgam of controls for gaming that probably has joysticks and multiple buttons (and inertial controllers?). A joystick is a piece of hardware with specific characteristics, which may or may not be used for gaming. Confusion about the name in this case is a flag that indicates that the scope of the spec is ambiguous.
I read the scope, and it seems pretty clear, but it's worth asking: If you stated what you DO cover better, is it no longer necessary to state what you DON'T cover. If the scope is clear, why is the name so difficult? In this case a 'gamepad' has been defined as a collection of analog inputs with input values between either 0..1 or -1..1.From my (very brief) scanning of the spec, it seems like this is the 'bounded (or finite?) analog input collection' spec. And there you have a description that describes what you cover, and intrinsically excludes what you don't.
Which begs a larger question not really for this list: is that really two specs - one for bounded linear analog devices, and one for collections of devices.
The good people writing the spec have probably gone round and round with this, but I think the places I'm not convinced mirror Alexey's, so I thought I'd try my hand at articulating those objections a different way.
And now I will shut up and watch with interest.
Jerry
On Oct 12, 2011, at 5:07 PM, Alexey Proskuryakov wrote:
12.10.2011, в 16:43, Scott Graham написал(а):
- Spec development process that ignores feedback. Multiple people who want this feature (including one of spec editors) were aware of feedback but didn't act on it. We had consensus on webkit-dev that the name was bad
I was indeed aware of your feedback (that you felt the name was unclear, as far as I understood it), but after discussing during a WG meeting, we haven't thought of a better name. If you have any suggestions, I'm sure all would be open to an improved name.
One way to come up with a good name for something already in existence is to look how other people call it. For example, a Wikipedia article on USB HID devices puts everything this spec cares about into "game controllers" section. How about GameController?
You could look at other classifications of input devices. Maybe USB HID spec itself has a good classification?
- WBR, Alexey Proskuryakov
_______________________________________________ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev
Hi,
Alexey Proskuryakov wrote:
One way to come up with a good name for something already in existence is to look how other people call it. For example, a Wikipedia article on USB HID devices puts everything this spec cares about into "game controllers" section. How about GameController?
Yes, agreed, GameController was the third candidate that was discussed in the WG (other than Joystick and Gamepad). Gamepad was favoured, because "Controller" seemed to imply too much generality (cameras, et al.). The discussion is in the minutes here for those interested http://www.w3.org/2011/09/20-webevents-minutes.html. Personally, I see Gamepad as a representative name rather than necessarily all encompassing. Analogously we use "Mouse" in other specs. Many people use trackballs, trackpads, joystick-like devices, etc. to control "some indexed buttons with a 2d location + scrolling indications". This is because there's no particularly better word for that input concept than "mouse". I believe this is the case for "gamepad" too. Jerry Seeger <vikingjs@me.com> wrote:
I read the scope, and it seems pretty clear, but it's worth asking: If you stated what you DO cover better, is it no longer necessary to state what you DON'T cover. If the scope is clear, why is the name so difficult? In this case a 'gamepad' has been defined as a collection of analog inputs with input values between either 0..1 or -1..1.From my (very brief) scanning of the spec, it seems like this is the 'bounded (or finite?) analog input collection' spec.
Thanks for the feedback. That does seem like a good way of summarizing the scope, and I think we should include something like that in the spec to help make things more explicit. As a name though, boundedAnalogInputDeviceCollection seems a bit unwieldy. scott
13.10.2011, в 7:15, Scott Graham написал(а):
Personally, I see Gamepad as a representative name rather than necessarily all encompassing. Analogously we use "Mouse" in other specs. Many people use trackballs, trackpads, joystick-like devices, etc. to control "some indexed buttons with a 2d location + scrolling indications". This is because there's no particularly better word for that input concept than "mouse". I believe this is the case for "gamepad" too.
The need to emulate mouse is forced by web compatibility. It's hardly a success story for coherent API design - beyond lowest common denominator functionality, you have to use completely separate APIs for touch, for example. Besides, it's surprising to me that gamepad can be seen as a representative name for wheels or 3d mice. - WBR, Alexey Proskuryakov
participants (2)
-
Alexey Proskuryakov
-
Scott Graham