Discussion:
[9fans] hardware device line -> what about the "file resource hardware devices" that are supported
(too old to reply)
Christoph Paschke
2012-03-13 16:21:17 UTC
Permalink
I'm a newbe at Plan 9 and I got Inferno and Octopus just running. I writing first limbo programs and I like it.

+ Made in plane C
+ resource is a file and not complicated proprietary drivers
+ it can be run such easy and small with emu at each system (I have Mac)
+ interconnect of resources

And when in one example it is explained, that you just send command "snap" to a cam input stream and you catch the file later by /image folder I thought:

Ok, let's google if there is any description for a CAM or a USB-Scanner what can be addressed such way with Plan 9. I found nothing.

So, my question: Doesn't it makes sense that some hardware like CAMs or USB-Scanners are working with Plan 9. I mean: There is a good system but as I understand there is absolutely no device what I can buy and connect to it.
There not need be 1000 CAMs supported, I would be happy if already only 1 CAM, 1 Scanner, 1 Printer would be supported and usable with Plan 9.

Example: I want write some Limbo programs and I would also support for example some printing functions (one laser printer, one label printer) and a barcode scanner for example.
And if I mean printing, I will say that it need work and not absolutely catastrophic results.

So think about Apple: There is also not 1000 of hardware what you buy, you stay in the Apple Line. Ok, today it is growing fast and more and more companies offer the matching Mac drivers now.
But I remember that before it was always VERY LESS that was supported. Most was only a small Apple supported line of hardware, in most cases high value.

So, for me, who wants start programing at Plan 9 base would be very interesting to find a good "Plan 9 Line of Hardware" that work with the file resource concept.
+ One Laser Printer, maybe one Inkjet
+ One Label Printer
+ One Barcode Reader, maybe one RFID reader
+ One digital camera
+ maybe One Meshing device like Zigbee / Digimesh
+ One USB-Scanner
+ One Hardware for a fax system

So, only one supported hardware line would be already enough and I already could start with some nice user programs.

Is this a stupid question? Sure I could now spend endless times in trying to reach such aim myself, but I believe there are so much tec freaks out there, that can do such things much better and faster and I could concentrate more at starting user programs?

- Chris
Nemo
2012-03-13 17:24:29 UTC
Permalink
For the octopus we wrap underlying MacOS X or Linux devices, and then use them.
For native plan 9 I don't know if there's a working usb camera.
I think there was some, long ago, but I'm not even sure.
erik quanstrom
2012-03-13 17:27:32 UTC
Permalink
Post by Nemo
For the octopus we wrap underlying MacOS X or Linux devices, and then use them.
For native plan 9 I don't know if there's a working usb camera.
I think there was some, long ago, but I'm not even sure.
wasn't that a tv tuner?

- erik
Tristan
2012-03-13 21:54:22 UTC
Permalink
Post by Christoph Paschke
So, my question: Doesn't it makes sense that some hardware like CAMs or
USB-Scanners are working with Plan 9. I mean: There is a good system but as I
understand there is absolutely no device what I can buy and connect to it.
Is this a stupid question? Sure I could now spend endless times in trying to
reach such aim myself, but I believe there are so much tec freaks out there,
that can do such things much better and faster and I could concentrate more
at starting user programs?
folks don't write code until they want it. until now, nobody has cared
enough about cameras or scanners or printers (though i thought ethernet
connected printers work more or less (but i've never printed anything))
to do anything. now you care, maybe you'll write them. the beauty of plan
9 is that it's easy to write drivers if you have decent documentation.
which you will for at least some of the devices you listed.

or maybe some common user-oriented device support like this would make a
good Google Summer of Code project?
--
All original matter is hereby placed immediately under the public domain.
erik quanstrom
2012-03-13 22:09:11 UTC
Permalink
Post by Tristan
folks don't write code until they want it. until now, nobody has cared
enough about cameras or scanners or printers (though i thought ethernet
connected printers work more or less (but i've never printed anything))
to do anything. now you care, maybe you'll write them. the beauty of plan
9 is that it's easy to write drivers if you have decent documentation.
which you will for at least some of the devices you listed.
or maybe some common user-oriented device support like this would make a
good Google Summer of Code project?
you'd do much better in gsoc if you limited your scope.

- erik
John Floren
2012-03-13 22:27:01 UTC
Permalink
Post by erik quanstrom
Post by Tristan
folks don't write code until they want it. until now, nobody has cared
enough about cameras or scanners or printers (though i thought ethernet
connected printers work more or less (but i've never printed anything))
to do anything. now you care, maybe you'll write them. the beauty of plan
9 is that it's easy to write drivers if you have decent documentation.
which you will for at least some of the devices you listed.
or maybe some common user-oriented device support like this would make a
good Google Summer of Code project?
you'd do much better in gsoc if you limited your scope.
- erik
I'd also like to recommend that projects take place outside the kernel
when possible. It's a lot easier to do things in userland on a single
Plan 9 box (or VM), while I've found that kernel work is best done
with at least a CPU server, a "victim" PC with serial output so we can
catch crash messages, and then another box where you can sit to
actually write code--a much more complex thing for a student with no
hardware budget to set up!

John
Tristan
2012-03-13 22:55:15 UTC
Permalink
Post by John Floren
Post by erik quanstrom
Post by Tristan
or maybe some common user-oriented device support like this would
make a good Google Summer of Code project?
you'd do much better in gsoc if you limited your scope.
i would imagine so, but it wouldn't be up for me to specify which one.
i'm not qualified for gsoc. (read some as 'one of x' not 'a few x')
Post by John Floren
I'd also like to recommend that projects take place outside the kernel
when possible.
i believe pretty much everything he mentioned would be in usb space,
which is (thank you everyone responsible) not kernel.
Post by John Floren
It's a lot easier to do things in userland on a single Plan 9 box (or VM),
while I've found that kernel work is best done with at least a CPU server, a
"victim" PC with serial output so we can catch crash messages, and then
another box where you can sit to actually write code--a much more complex
thing for a student with no hardware budget to set up!
i've tended to use the cpu server as the sitdown computer, bad for
reading pdfs but other that... but agreed!

enjoy,
tristan

(appologies to John, who'll likely get this twice, except this bit)
--
All original matter is hereby placed immediately under the public domain.
Jeff Sickel
2012-03-14 01:38:13 UTC
Permalink
Post by Tristan
Post by Christoph Paschke
So, my question: Doesn't it makes sense that some hardware like CAMs or
USB-Scanners are working with Plan 9. I mean: There is a good system but as I
understand there is absolutely no device what I can buy and connect to it.
Is this a stupid question? Sure I could now spend endless times in trying to
reach such aim myself, but I believe there are so much tec freaks out there,
that can do such things much better and faster and I could concentrate more
at starting user programs?
folks don't write code until they want it. until now, nobody has cared
enough about cameras or scanners or printers (though i thought ethernet
connected printers work more or less (but i've never printed anything))
to do anything. now you care, maybe you'll write them. the beauty of plan
9 is that it's easy to write drivers if you have decent documentation.
which you will for at least some of the devices you listed.
I care about cameras. In fact, I care about >96 cameras capturing
frames at the same time and staying in sync. Can I leverage Plan 9
ideas to do this? Maybe, but only time will tell.
Post by Tristan
or maybe some common user-oriented device support like this would make a
good Google Summer of Code project?
Whether or not something like this is a good summer project or not
would be up to the consortium.

-jas
erik quanstrom
2012-03-14 13:02:29 UTC
Permalink
call me today. 815 715 9180.

- erik
erik quanstrom
2012-03-14 13:13:45 UTC
Permalink
Post by erik quanstrom
call me today.
the guys in dark suits are here to take away my
email card ...

sorry for the spam.

- erik

Loading...