3 ms·

You can read all of the comments by the devs who aren't even considering the accessibility option. Remapping mouse buttons come across to them as someone trying

by moftz 8y ago

You can read all of the comments by the devs who aren't even considering the accessibility option. Remapping mouse buttons come across to them as someone trying to map keyboard shortcuts to the mouse buttons for better productivity in whatever apps they use for work.

pgeorgi 8y ago
> Remapping mouse buttons come across to them as someone trying to map keyboard shortcuts to the mouse buttons for better productivity in whatever apps they use for work. That's because key mapping and button mapping is more or less the same problem, as far as computers are concerned. From the blog: "Evoluent - the manufacturer of my mouse - also provide a handy tool for Windows and Mac so that I can set the mouse buttons up just the way I like them". There's a technical reason for that (actually, two): USB HID allows devices to expose how many buttons they have, plus some characteristics, but it's utterly useless to expose the information that "button 17 is the one in the top right corner when seen from above". The other reason is that USB device vendors (no matter which device type) don't feel the need to implement standards faithfully, so devices need special handling in drivers all the time. So Evoluent provides its own Windows and Mac tool. In the bug linked in the post, the devs are discussing adding such special cases, or - instead, and preferred - providing an API through which third parties (no matter if Evoluent or that chap who can write xinput configs) could develop frontends that could do such configuration. Now, these APIs shouldn't enable third parties to essentially log everything entered by the user. And that's where the discussion is at. I'm not sure if the Evoluent tool reconfigures the device (through some proprietary vendor commands to the mouse) or reconfigures Windows to remap these buttons. Interestingly the former case might already be supportable on Chrome OS, since it supports the WebUSB APIs. Since these USB commands would be non-standard and most likely undocumented by the hardware vendor, it would be up to Evoluent to provide such a frontend to configure the device. The latter case (configuring a remapping) would be similar to that API that is under discussion. It doesn't go as fast as the blog post's author wants, okay. But screaming into the void (or even here) doesn't help with advancing that process.

by moftz 8y ago

pgeorgi 8y ago

> Remapping mouse buttons come across to them as someone trying to map keyboard shortcuts to the mouse buttons for better productivity in whatever apps they use for work. That's because key mapping and button mapping is more or less the same problem, as far as computers are concerned. From the blog: "Evoluent - the manufacturer of my mouse - also provide a handy tool for Windows and Mac so that I can set the mouse buttons up just the way I like them". There's a technical reason for that (actually, two): USB HID allows devices to expose how many buttons they have, plus some characteristics, but it's utterly useless to expose the information that "button 17 is the one in the top right corner when seen from above". The other reason is that USB device vendors (no matter which device type) don't feel the need to implement standards faithfully, so devices need special handling in drivers all the time. So Evoluent provides its own Windows and Mac tool. In the bug linked in the post, the devs are discussing adding such special cases, or - instead, and preferred - providing an API through which third parties (no matter if Evoluent or that chap who can write xinput configs) could develop frontends that could do such configuration. Now, these APIs shouldn't enable third parties to essentially log everything entered by the user. And that's where the discussion is at. I'm not sure if the Evoluent tool reconfigures the device (through some proprietary vendor commands to the mouse) or reconfigures Windows to remap these buttons. Interestingly the former case might already be supportable on Chrome OS, since it supports the WebUSB APIs. Since these USB commands would be non-standard and most likely undocumented by the hardware vendor, it would be up to Evoluent to provide such a frontend to configure the device. The latter case (configuring a remapping) would be similar to that API that is under discussion. It doesn't go as fast as the blog post's author wants, okay. But screaming into the void (or even here) doesn't help with advancing that process.