That's really cool for you. Congrats.
And thanks, btw, for dodging my question about an authoritative compatibility matrix from 3DConnexion for current 6DOF devices vs. the gamepad API you recommended.
I have worked professionally in software quality assurance for years. Proving one (1) good case doesn't proove, that there don't exist multiple failing combinations. That's why I was asking for a dependable compatibility matrix.
Obviously you're not interested in the failing cases. Which is a valuable information in itself. Because now I know, what _NOT_ to recommend to MY customers. You also don't realize that you're deconstructing your very authority as the head of SDK development, by trying to proove your customers wrong. Instead of listening to them in a quest to improve. What a sad experience this has become.
On a completely related note:
SpaceNavigator USB for sale. Excellent condition: Minimum price asked: must cover postage for shipping.
Please direct your offers to:
chris@vielzutun.ch
Simple(r) example application?
Moderator: Moderators
Re: Simple(r) example application?
We cannot possibly provide that assurance. How could we? In fact, I don't know of anyone that can.
One of the motivations for us to develop 3DconnexionJS (our solution for JavaScript programs) was precisely to address the issue with inconsistencies between browser implementations of the Gamepad API.
Re: Simple(r) example application?
Well, that's quite a statement
. You recommend a solution which you know already that it doesn't hold its (and your!) promise. I propose we forget about the previous exchange and start again from scratch.
...
...
Without knowing the details of the navlib's implementation I bet, all it would take is a maximum of five lines of code to pass raw sensor data right through to whomever wants to consume it. And of course to expose that part of the API in your documentation. All it would take would be some good will on the part of 3DConnexion.
![Shocked :shock:](./images/smilies/icon_eek.gif)
...
...
Without knowing the details of the navlib's implementation I bet, all it would take is a maximum of five lines of code to pass raw sensor data right through to whomever wants to consume it. And of course to expose that part of the API in your documentation. All it would take would be some good will on the part of 3DConnexion.
Re: Simple(r) example application?
We cannot recommend the Gamepad API, although we wish we could. As mentioned, there are too many inconsistencies (and limitations) in this API to offer the level of integration needed to fully support a SpaceMouse. You will find other statements to such effect from us elsewhere in this Forum.
I only stated that the Gamepad API can be used to retrieve the device raw device data but, again, that's not what we recommend.
For web solutions, we recommend the use of 3DconnexionJS, a JavaScript solution that matches SDK version 4. As surely you know, SDK 4 takes a different approach (camera control paradigm) than simply a transport for the device raw data (which requires developers to reimplement the entirety of navigation features customers have come to expect).
Since you stated 3DconnexionJS was not what you were looking for, I shared what I knew that might perhaps work for you.
Re: Simple(r) example application?
You appear to have a short memory, because that's exactly what you did, in your post (in this very thread) of Thu Mar 31, 2022 8:43 am. It's true that you recommended Gamepad AP in the context of raw data retrieval. But even this limited claim of yours doesn't hold any water.
Let me fix that for you:ngomes wrote: ... the entirety of navigation features customers have come to expect.
There are so many levels on which your claim deserves to be challenged. I don't have the time to cover all of those, so I'll have to pick the most relevant ones:... the entirety of functions which many (maybe even most) current(!) customers have come to expect.
There are currently >100.000 websites fully or partially based on Three.js. With likely a several orders of magnitude higher number of website visitors, all of which are potential 3DConnexion customers. The very people who could be firsthand tapping into this huge reservoir of potential new 3DConnexion customers are the software developers which you are patronizing in a really obnoxious way.
If developers don't get the support they need for JavaScript integration, all those cool applications (which might potentially draw in a significant portion of the above mentioned website vistitors) are not going to be written in the first place. Personally, I fand this refusal of 3DConnexion, represented by the bottleneck which you constitute, truely staggering.
Yes, I'm painfully aware of the first part. Thanks for reminding me.As surely you know, SDK 4 takes a different approach (camera control paradigm) than simply a transport for the device raw data (which requires developers to reimplement the entirety of navigation features.
As for the second part: the little bit of trigonometry involved is stuff for 9th or 10th graders. Not really rocket science.
And chiefly:
This is the Software Developers forum. People who come here can be trusted to know how to implement rotations and translations in 3D space. Those who feel insecure may possibly embrace what you provided for them. The ones who do know how to implement interfaces which are obviously beyond your imagination find their devices crippled, due to a wanton decision of yours personally. Because you can.
Over and out.
Re: Simple(r) example application?
Ladies, gents,
I didn't experience a change of mind. Rather, I'm aware that 3dconnexion has a bigger reach in terms of SEO than my own blog. And I understand, that folks in search of a solution may eventually find this thread before they find my blog post(s) with full source code (70 SLOC for the bare minimum, approx. 300 SLOC for an intuitive driver of type "camera control", based on the current/recent version v140 of Three.js. No external libraries and no background processes are required, no strings attached. You DO need a browser which supports WebHID, though.
That said:
I did solve the thread title for myself, without, and as it appears: against the expressed will of the 3dconnexion software clerks. And I'm willing to share my findings with you.![Cool 8-)](./images/smilies/icon_cool.gif)
The story starts here:
https://discourse.threejs.org/t/spacemo ... last/37146
I didn't experience a change of mind. Rather, I'm aware that 3dconnexion has a bigger reach in terms of SEO than my own blog. And I understand, that folks in search of a solution may eventually find this thread before they find my blog post(s) with full source code (70 SLOC for the bare minimum, approx. 300 SLOC for an intuitive driver of type "camera control", based on the current/recent version v140 of Three.js. No external libraries and no background processes are required, no strings attached. You DO need a browser which supports WebHID, though.
That said:
I did solve the thread title for myself, without, and as it appears: against the expressed will of the 3dconnexion software clerks. And I'm willing to share my findings with you.
![Cool 8-)](./images/smilies/icon_cool.gif)
The story starts here:
https://discourse.threejs.org/t/spacemo ... last/37146