[sane-devel] Install problem with fi-4120C2 under FreeBSD and OS X 10.4.2

[Darrell Styner]

On Nov 2, 2005, at 9:27 AM, m. allan noah wrote:

> On Wed, 2 Nov 2005, Darrell Styner wrote:
>
>
>>
>> On Nov 2, 2005, at 6:48 AM, m. allan noah wrote:
>>
>>
>>>>> 1. no thats not normal, but it may be caused by lamp warm-up or  
>>>>> some such. need more debug log info of what is going on during  
>>>>> that pause.
>>>>>
>>>> I don't think it's warm-up. The scanner's always ready to go  
>>>> right away under Windows and the delay is the same when you run  
>>>> one scanimage right after another. I'll send you a debug log soon.
>>>>
>>> judging from the logs you sent, it is something about the DATA0/1  
>>> usb toggle again. since almost all these transactions involve a  
>>> single write and 2 reads, i am surprised that it acts stupid on  
>>> the second read, but not the first. the fi-4120c just locks up at  
>>> this point. the 4120c2 self-resets the toggle after 30 seconds, i  
>>> think (its not documented).
>>> i think a retry on the second read might fix that, i will code it  
>>> up and send you a patch.
>>>
>>
>> Thanks, I look forward to the patch.
>>
>> I've got this scanner on eval right now and can send it back and  
>> try another if there's something similar that's known to play  
>> nicer with sane. I'm going to be integrating this into a  
>> commercial product and need it to be rock solid. This intermittent  
>> read problem is shaking my confidence a bit. Do you know of  
>> another USB, ADF scanner in the 1K price range that might be a  
>> better choice in terms of sane compatibility? I've looked  
>> carefully at the "supported devices" list but don't see many mid- 
>> level, ADF scanners that are currently sold and supported.
>>
>
> honestly, these scanners are great, and i know a great deal about  
> how to program them (my company owns ~250 4120C/4120C2/5110C  
> scanners, running a proprietary libusb app on headless linux boxes)  
> so the difficulties with sane are pretty much entirely my fault for  
> not having time to sit down and really fix it.
>
> that said, the vertical paper path of these units does leave a bit  
> to be desired, and i have been thinking of looking at some of the  
> panasonic usb units that have the laser-jet like paper path. i  
> cannot comment on their sane support.
>
> allan

Thanks for the feedback. I've also had some concerns about the paper  
path of the 4120. My first attempt to scan a batch of 15 pages with  
this scanner led to 6 straight paper jams until I moved the paper  
guide about 2mm, which seemed to make it happy. I'm worried that my  
users won't have as much patience. Still, it has a small footprint,  
makes good, fast duplex scans, and the price is reasonable for our  
market. Please let me know if you have any luck with the panasonics  
or something else a little more industrial strength.

>
>
>>
>>
>>>>>> Thanks for the help,
>>>>>>
>>>>> its not a good long-term fix, so be prepared to test something  
>>>>> more permanent later...
>>>>>
>>>> I'll be happy to test anything as long as it's in the next 29  
>>>> days. After that I may lose access to the scanner.
>>>> Am I losing any features by having that line commented out?
>>>>
>>> later model fujitsu scanners can do monochrome scans using any  
>>> one of the three color heads. default is green. this is useful  
>>> for dropout of color printed forms or losing editors marks, etc.  
>>> older scanners you have to change to a colored lamp. having that  
>>> line commented is no worse than it not working, like it was  
>>> before :) if you did want to specify the color, you could not do  
>>> it til this code is fixed.
>>> allan
>>>
>>>> Thanks again.
>>>>
>>>>> allan
>>>>>
>>>>>> On Nov 1, 2005, at 10:53 AM, m. allan noah wrote:
>>>>>>
>>>>>>> as a test, could you try modifying backend/fujitsu.c,  
>>>>>>> recompile and re-install sane backends? i want to see if this  
>>>>>>> fixes the problem...
>>>>>>> around line 3673, you should see:
>>>>>>> if (s->has_dropout_color)
>>>>>>> change that to
>>>>>>> if(0)
>>>>>>> allan
>>>>>>> On Tue, 1 Nov 2005, Darrell Styner wrote:
>>>>>>>
>>>>>>>> Here it is. Note that the end of the file shows an interrupt  
>>>>>>>> where I ^C'd after waiting a couple minutes for it to  
>>>>>>>> continue. If I let it go on at this point the log will grow  
>>>>>>>> rapidly as it fails trying to write the same 31 bytes to the  
>>>>>>>> scanner over and over again.
>>>>>>>> Thanks for looking into this.
>>>>>>>>
>>>>>>> -- 
>>>>>>> "so don't tell us it can't be done, putting down what you  
>>>>>>> don't know.
>>>>>>> money isn't our god, integrity will free our souls" - Max  
>>>>>>> Cavalera
>>>>>>>
>>>>>> Darrell Styner
>>>>>> darrell at avanta.com
>>>>>>
>>>>> -- 
>>>>> "so don't tell us it can't be done, putting down what you don't  
>>>>> know.
>>>>> money isn't our god, integrity will free our souls" - Max Cavalera
>>>>> -- 
>>>>> 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
>>>>>
>>>> Darrell Styner
>>>> darrell at avanta.com
>>>>
>>> -- 
>>> "so don't tell us it can't be done, putting down what you don't  
>>> know.
>>> money isn't our god, integrity will free our souls" - Max Cavalera
>>>
>>
>> Darrell Styner
>> darrell at avanta.com
>>
>>
>>
>>
>
> -- 
> "so don't tell us it can't be done, putting down what you don't know.
> money isn't our god, integrity will free our souls" - Max Cavalera
>
> -- 
> 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
>
>

Darrell Styner
darrell at avanta.com