On Mon, 9 Dec 2019 at 11:55, mario futire <mariofutire@gmail.com> wrote:
> On Mon, 9 Dec 2019 at 10:55, mario futire <mariofutire(a)gmail.com&gt; wrote:
>
>
> Microprocessors are really cheap.  Devices like mice should support
> customization
> by individual device, which would greatly simplify the device support in
> the OS.
>
> Developer time is a limiting resource for open source software.   There
> have
> been several documents explaining why the effort to maintain X11 has become
> unsupportable and has been consuming resources needed elsewhere.   I expect
> future devices like mice and keyboards will become much smarter and fully
> configurable, allowing simplifications of OS-level support.   Configuration
> will
> be done by running an app that updates the configuration stored in each
> device.   Today you can buy a mouse with limited programmable options
> <https://www.pcmag.com/review/370502/logitech-mx-master-3>
> but the app runs on Windows.

I am not sure what you are advocating

1) using windows

Use of Windows should be avoided, but may be needed for things like device configuration using vendor software that isn't available on other platforms.
More linux users should maintain access to Windows so they are qualified to advise others on the suitability of Windows for a particular purpose and to analyze problems when files need to "work" on both platforms.
 
2) using X11

Fine as long as you don't need anything new, but expect it to decay over time.
 
3) use Wayland and stop complaining because this is open source based on people's free time

I totally agree with 3), but it is a very hard selling point.

Before complaining that wayland doesn't do something that could be done under X11, consider that many
things done under X11 were constrained by decades old technology, and may not be appropriate going
forward.   If you have some free time, start using Wayland now to see where there are limitation affecting
your use cases.

It may be helpful asking vendors like Logitech to provide linux (wayland) versions of tools currently
only available on Windows.
 

I develop software too and I know very well that the legacy system wont go away till the tail 1% is sorted out.
We probably disagree on what is the tail 1% here.

A good part of my career (I'm now retired) was spend sorting out those 1% issues.  In many cases, the legacy
system provided tools that had become available in widely used applications, so it was more a matter
convincing management that we should not devote resources to implementing capabilities that had become
available in general purpose packages being maintained by others, and showing users how to use the
general tools.   Management is generally happy if other people provide software, and users find more
uses for the general purpose software.

Linux is going through a number of big changes: systemd, Wayland, UEFI, and multi-core hardware that are
causing pain for longtime users.   Much of the actual work is funded by big data and AI projects, so may not
address needs of other linux use cases, and adds complexities that aren't relevant to small systems.  

The tail 10% will be things not needed by big data and AI, and some of that will 10% will be  hard to
implement in the newer environments.   If some of those can offloaded to hardware vendors, e.g., by
moving to smarter peripherals, effort can be focused on a fraction of the 10%.   Many of us don't pay for
linux, so we should be willing to pay a bit more for hardware.

--
George N. White III