[sane-devel] Questions and suggestions

[m. allan noah]

2009/1/23 Mattias Kregert <mattias at kregert.se>:
>
> Good morning, and thank you all for this wonderful piece of software!
>
> I have been using sane for six years now, and recently got some new hardware
> and will be using saned and the net backend in the near future.
> Unfortunately i have some moderately annoying problems with net/saned which
> i need help with, and a couple of feature requests in case you need more to
> do :)
>
> - The network should be fast enough to keep up with the scanning speed, but
> still the scanner pauses a little between each page - as if it is waiting
> for data to be transferred over the network before it decides to
> start scanning the next page. When watching network activity, i see that
> data is sent in bursts with a not-so-small delay between each burst.
> My completely novice assumption is that saned scans one page, sends it to
> the net backend, scans next page, sends next page, and so on. If this is the
> case; can saned instead be configured to immediately start scanning next
> page and queue it for transmission to the client, not waiting for the first
> page to complete its journey over the network?

no, the protocol is serialized. saned has no way of knowing ahead of
time that you actually want to scan the next page. you might just stop
with one scan. but- see comment below.

> Or is is some other problem, like saned compressing data before transmission
> or something? And if that is the case, can i turn it off?

it does not compress. i wish it did :)

> - I have some sort of permission problem with saned. I have a usb scanner
> which can not be accessed through saned, even if i
> chown /proc/bus/usb/NNN/NNN to saned:saned. The only solution i found so far
> is to run saned as root.
> When running saned -d128 nothing useful is reported, so i have no clue about
> what is causing this. Any ideas?

most modern linuxen dont use files in /proc/bus/usb. i bet the real
files are elsewhere.

> - Feature request: When looking for scanners (with scanadf -L) saned only
> reports scanners not currently in use. It would be nice to be able to see
> all attached scanners, perhaps with some indication if they are being used
> right now.

most backends need to talk to the scanner to make sure it is really
supported. if the device is already open, the kernel will prevent us
from doing that.

>
> - Feature request: It would be nice to be able to ask a particular host
> about its scanners, like "scanadf --L=myscannerhost.mydomain", not having to
> add it to net.conf first.

interesting idea, but scanadf does not know anything about networking.
It just asks the sane lib for all the known scanners. perhaps an
external tool is a better approach.

> - And last (and actually least): I use the fujitsu backend, but with a
> FI-6130 scanner i get weird colors (RGB-BGR problem). I use version 1.0.19,
> but perhaps this has been fixed in cvs?

The cvs version of sane-fujitsu is MUCH improved over that in 1.0.19.
That bug is fixed, and several new features were added. One that might
interest you is the --buffermode option, which tells the scanner to
read ahead with the same parameters. Not sure if the 6130 supports
that...

allan

>
>
> Best, cheerfullest, looking-forward-to-next-relase-est regards
>
> /Mattias
>
> --
> sane-devel mailing list: sane-devel at lists.alioth.debian.org
> http://lists.alioth.debian.org/mailman/listinfo/sane-devel
> Unsubscribe: Send mail with subject "unsubscribe your_password"
>             to sane-devel-request at lists.alioth.debian.org
>



-- 
"The truth is an offense, but not a sin"