| 1 |
Dr. Canek Peláez Valdés: |
| 2 |
... |
| 3 |
> Where do you get that impression from? The OP needs handling keyboard and |
| 4 |
> mouse (as per his first email), and to do that in Linux these days, you |
| 5 |
> basically need udev, because xf86-input-mouse and xf86-input-keyboard are |
| 6 |
> going the way of the dodo. |
| 7 |
|
| 8 |
It is inconvenient that thoose two goes away. |
| 9 |
Regarding udev, it has never supported serial mice, so it doesn't help |
| 10 |
me. |
| 11 |
|
| 12 |
... |
| 13 |
> My point is that it's not his call; it's the call of the developers of the |
| 14 |
> software that he decided to use. |
| 15 |
|
| 16 |
Poeple write whatever software they want to or are paid to do. |
| 17 |
It is my call if I want to use that software or not. |
| 18 |
|
| 19 |
> > Yes I take your point, but bloat is bloat, and bloat is a liability. |
| 20 |
> > |
| 21 |
> |
| 22 |
> There is no bloat; the developers *need* to handle the dynamic hardware |
| 23 |
> case *and* the static hardware case. With udev, they handle both; otherwise |
| 24 |
> there would be two code routes: one for static and another for dynamic |
| 25 |
> hardware. |
| 26 |
... |
| 27 |
|
| 28 |
As I wrote before, udev does not handle serial mice, so udev does not |
| 29 |
solve anything for me nor does it help me in any way to run my systems. |
| 30 |
Udev is just something pushed on me for no gain except possible to |
| 31 |
satisfy some dependancy touted to be beneficial. So in this very |
| 32 |
specific case it can be considered "bloat" if you wish to use that |
| 33 |
kind of words. |
| 34 |
|
| 35 |
My guess is that it is more useful on laptop than on a desktop box |
| 36 |
or an industrial computer. |
| 37 |
|
| 38 |
/// |
| 39 |
|
| 40 |
As a side note, from what I understand, udev today is mostly about |
| 41 |
usb-devices because that is where the dynamic hardware comes from |
| 42 |
today (at least when we are not talking about hotplugging cpus, |
| 43 |
memory cards, io-cards and such (but that is more of a enterprise |
| 44 |
problem than a small system problem. |
| 45 |
|
| 46 |
Serial ports are darn easy to implement in hardware and |
| 47 |
softwere. |
| 48 |
|
| 49 |
E.g. if I have a program connecting to a device using a serial |
| 50 |
and it is disconnected, I can just reconnect it and nothing |
| 51 |
special happens, noting to be done in software except logging. |
| 52 |
The same device via usb, the dis-/reconnect will close the |
| 53 |
port and make it vanish forcing med to find out find out where the |
| 54 |
new /dev file is and reopen and reinitialize it. |
| 55 |
In hardware, mcu's without usb are cheap and their serial port |
| 56 |
are simpe to program and the serial port "stack" is vanishingly small. |
| 57 |
Just look at the tty_* files in |
| 58 |
http://aspodata.se/git/openhw/libarm/ |
| 59 |
http://aspodata.se/git/openhw/libarm/stm32/ |
| 60 |
For usb support, I need an usb stack (which is larger), e.g. |
| 61 |
https://github.com/libopencm3/libopencm3/tree/master/lib/usb |
| 62 |
I need to understand the usb protocol and all thoose structs to fill |
| 63 |
in, and in the end I get a system that is harder to program on the |
| 64 |
host side for no gain other than that +5V is provided by usb. |
| 65 |
|
| 66 |
Regards, |
| 67 |
/Karl Hammar |
Archives