On Wed, 23 Oct 2002 10:32:01 -0700, sane-de...@www.mostang.com wrote: > Send Sane-devel mailing list submissions to > sane-de...@www.mostang.com > > To subscribe or unsubscribe via the World Wide Web, visit > http://www.mostang.com/mailman/listinfo/sane-devel > or, via email, send a message with subject or body 'help' to > sane-devel-requ...@www.mostang.com > > You can reach the person managing the list at > sane-devel-ad...@www.mostang.com > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Sane-devel digest..." > > > Today's Topics: > > 1. Re: No Sane Device Found (Gene Heskett) > 2. Epson Perfection 2450 Photo (Karl Bellve) > 3. Re: Epson Perfection 2450 Photo (Henning Meier-Geinitz) > 4. Re: Fwd: I killed my Epson perfection 1260 photo. Twice. (Kees > Lemmens) > 5. Re: Artec USB scanner information (Jaeger, Gerhard) > 6. xsane-0.89 source and win32-version released (Oliver Rauch) > 7. USB kernel scanner driver (Henning Meier-Geinitz) > 8. Re: USB kernel scanner driver (Major A) > 9. Re: USB kernel scanner driver (Sergey Vlasov) > 10. Re: USB kernel scanner driver (Major A) > 11. Re: USB kernel scanner driver (Jonathan Buzzard) > 12. Re: USB kernel scanner driver (Henning Meier-Geinitz) > > --__--__-- > > Message: 1 > From: Gene Heskett <gene_hesk...@iolinc.net> > Organization: None that appears to be detectable by casual observers > To: Henning Meier-Geinitz <henn...@meier-geinitz.de>, > sane-de...@www.mostang.com > Subject: Re: [sane-devel] No Sane Device Found > Date: Wed, 23 Oct 2002 10:19:36 -0400 > > On Wednesday 23 October 2002 04:11, Henning Meier-Geinitz wrote: > >Hi, > > > >On Wed, Oct 23, 2002 at 02:43:47PM +0800, ' ' wrote: > >> I tried the information, and PNM is already enabled, the shared > >> and dynamic library is in /usr/local/lib/sane. But still the > >> reply is no deveice found. I'm using Red Hat 7.3 wherein Sane > >> 1.0.7 is installed in the system. I uninstall it with make > >> uninstall then install 1.0.9. What's the possible problem this > >> time? > > > >Is pnm also enabled in dll.conf?
> > > >Bye, > > Henning > > Henning; A reminder about RH and changing from tars to rpms > > I don't mean to be 'out of line' here, but please don't forget that > redhat does not use /usr/local for the default path, they use '/', > so he probably now has 2 'etc/sane.d' directories, making it at > least twice as hard to convert to using tars. I tried the > --relocate-whatever on a recent rpm install, but I still wound up > with a new /etc/sane.d directory and a non-functioning install. I > guess that only works if one is building from the source rpm. > pnm is not enabled in dll.conf (#pnm). I also checked my system, it has two /etc/sane.d, one from the root '/' and the other is from /usr/local/etc/sane.d. I think the previous installation is in root directory and maybe in uninstalling the previous copy was unsuccessful. Any suggestion on how will i uninstall the previous installation? I tried to uninstall it I tried using it but pnm is not also enabled that's why i tried re installing another version. > -- > Cheers, Gene > AMD K6-III@500mhz 320M > Athlon1600XP@1400mhz 512M > 99.18% setiathome rank, not too shabby for a WV hillbilly > > --__--__-- > > Message: 2 > Date: Wed, 23 Oct 2002 10:39:11 -0400 > From: Karl Bellve <karl.bel...@umassmed.edu> > Organization: Biomedical Imaging Group > To: sane-de...@www.mostang.com > Subject: [sane-devel] Epson Perfection 2450 Photo > > > I just bought this scanner to work with our Linix workstations but I > am > having trouble getting XSane to see it. This is on a Redhat 7.3 > system > with the default RPM installs. > > sane-find-scanner returns that it finds a scanner on > /dev/usb/scanner0 > with vendor=0x4b8 and product=0x0112. > > I modifed /etc/sand.d/epson.conf to reflect /dev/usb/scanner0 > > I modifed /etc/modules.conf with an options scanner vendor=0x4b8 > product=0x0112. > > scanner module is loaded. > > Scanner is detected on the USB bus, shown via dmesg. > > I get error messages in dmesg with the following: > > scanner.c: openscanner(x): Unable to access minor data > > X is a number from 1 to 15. > > > -- > Cheers, > > > > Karl Bellve, Ph.D. ICQ # 13956200 > Biomedical Imaging Group TLCA# 7938 > University of Massachusetts > Email: karl.bel...@umassmed.edu > Phone: (508) 856-6514 > Fax: (508) 856-1840 > PGP Public key: finger k...@molmed.umassmed.edu > > --__--__-- > > Message: 3 > Date: Wed, 23 Oct 2002 16:45:11 +0200 > From: Henning Meier-Geinitz <henn...@meier-geinitz.de> > To: sane-de...@www.mostang.com > Subject: Re: [sane-devel] Epson Perfection 2450 Photo > > Hi, > > On Wed, Oct 23, 2002 at 10:39:11AM -0400, Karl Bellve wrote: > > I get error messages in dmesg with the following: > > > > scanner.c: openscanner(x): Unable to access minor data > > > > X is a number from 1 to 15. > > That's not a problem. SANE checks /dev/usb/scanner0, > /dev/usb/scanner1, and so on. As you only have one scanner, it's not > found at the remaining 15 minor device numbers. The scanner driver is > a bit too verbose. > > Bye, > Henning > > --__--__-- > > Message: 4 > From: Kees Lemmens <c.w.j.lemm...@its.tudelft.nl> > Reply-To: c.w.j.lemm...@its.tudelft.nl > To: sane-de...@mostang.com > Subject: Re: [sane-devel] Fwd: I killed my Epson perfection 1260 > photo. Twice. > Date: Wed, 23 Oct 2002 17:07:05 +0200 > > Hi, > > On Tuesday 22 October 2002 20:51, Volker Kuhlmann wrote: > > > (sane-frontends-1.0.9-pre1) everything worked like a charm. Except my > > > first scanner died after a few hours of scanning. In the middle of a > > > > > > replacement epson 1260. But today the same thing happened to my 'new' > > > scanner: in a middle of a scan the epson dies.... I think I have used > > > the new scanner for about 1 hour today, 3 hours yesterday and 3 hours 4 > > > days ago. > > > > Looks like Epson made a dud batch, that happens with cheap consumer > > rubbish^H^H^H^H^H^Happliances. I recently bought a pretty normal > > telephone, and the first 2 died with the same fault within hours. > > > > When you get it replaced, insist on the dealer giving you a new one from > > a different batch. The dealer mightn't have any so you may have to wait. > > > > Volker > > Had a similar problem with an Epson 1240 about 6 months ago: failed > with the > same symptoms after one day of use. It was sent back to Epson to be > repaired > (this took 6 weeks .....) but since then it has worked flawlessly > although it > is used quite often. > > -- > groeten, > Kees Lemmens. > > > --__--__-- > > Message: 5 > From: "Jaeger, Gerhard" <gerh...@gjaeger.de> > To: Henning Meier-Geinitz <henn...@meier-geinitz.de>, > SANE Mailing List <sane-de...@mostang.com> > Subject: Re: [sane-devel] Artec USB scanner information > Date: Wed, 23 Oct 2002 17:40:45 +0200 > > Hi, > > On Tuesday, 22. October 2002 19:42, Henning Meier-Geinitz wrote: > [SNIP] > > Artec 1236 USB > > No information > > This in fact is a LM983x based scanner, but I do not have any Vendor > or > Product ID. Support can be realised, but I never saw this device... > > Cheers, > Gerhard > > --__--__-- > > Message: 6 > Date: Wed, 23 Oct 2002 17:59:20 +0200 > From: Oliver Rauch <oliver.ra...@rauch-domain.de> > To: SANE-MAILING-LIST <sane-de...@mostang.com> > Subject: [sane-devel] xsane-0.89 source and win32-version released > > Hello. > > XSane-0.89 source and win32-version have been released. > > It can be downloaded from: > > http://www.xsane.org > > There have been a lot of improvements: > > - 3 pass scanning: data interleaving is done in buffer, patch by > <oliver.friedr...@hamburg.de> > > - replaced fork() by spawnvp for ocr command in xsane-save.c (patch > by Franz Bakan) > > - xsane_save_jpeg does reduce a 16bit/sample images to 8bit/sample > at first because jpeg > does > not support 16 bit/sample > > - xsane_save_ps does reduce 16bit/sample images to 8bit/sample while > saving > > - When scanning to file (mode SAVE) then the warning about image > depth reduction 16->8 > bit/sample > for jpeg and postscript is not printed any more. > > - Since april 2000 there is defined a pnm (ppm/pgm) format for 16 > bit binary images. > With this format the unofficial RAW16 format that has been used by > xsane until version > 0.88 > becomes senseless. The RAW16 format has been removed, xsane does > create 16 bit binary > pnm files > now (the difference between the RAW16 and the pnm16binary format > is that the > pnm16binary format > always uses MSB first while RAW16 was written in machine byte > order). > In preferences-Setup->Image there is a selection if the pnm16 > image shall be saved in > ascii or > binary format. > When you have any problems with the pnm16binary format then make > sure at first that the > program > you use does support it correct. imagemagic does support it, the > xv version 3.10a does > not > support it. > > - read_pnm_header now does read the gamma, contrast, brightness, > threshold values, so > when these > data is saved in an image then it is correct now. > > - gamma, brightness, contrast and threshold are added in the header > of pnm files now > > - when 16 bit image is scanned and "reduce 16bit image to 8bit" is > selected and the image > is saved as pnm then the file is created correct now > > - batch scan thumbnail: replaced gdk_image by gtk_preview > gdk_image needs pixel colors based on hardware, gtk_preview uses > truecolor pixels. > > - added po file converstion to utf8 in po/Makefile[.in] > > - added documentation for batch scan window > > - changed page size in range from (max-min)*1e-40 to (max-min)*1e-30 > because some systems like solaris had problems (one long bar that > could not be moved) > > - xsane-preview.c: in case of an error the error error text is > created before > preview_scan_done is called when parameters->* are needed for > creating the > error message > > - made html pages compatible to html 4.01 > > - changed header of postscript files, may be this solves the > printing problem on some > systems > > - added "%%EOF" to end of postscript files > > - updated translations: cs, de, fr, ja, nl, ru, sv (update for > 0.88), tr > - new translations: pt, pt_BR > > - add no_device icon in error message: no devices available > > - added help_no_devices dialog > > - added device icon in device_scanning_dialog > > - added a frame in the decision dialog > > - when preview_autoselect_scanarea does not find an image then > the complete image is used > > - added accelerator key ESC to decision dialog: > when a reject button (right) is available then ESC=reject, > when no reject button is available then ESC=OK. > > - set block_update_maximum_output_size_clipping in > preview_rotation_callback, > now rotation does work like expected when in copy mode > > - added paper_orientation menu in copy mode > > > Best regards > Oliver > > -- > Homepage: http://www.rauch-domain.de > sane-umax: http://www.rauch-domain.de/sane-umax > xsane: http://www.xsane.org > E-Mail: mailto:oliver.ra...@rauch-domain.de > > --__--__-- > > Message: 7 > Date: Wed, 23 Oct 2002 18:29:52 +0200 > From: Henning Meier-Geinitz <henn...@meier-geinitz.de> > To: SANE Mailing List <sane-de...@mostang.com> > Subject: [sane-devel] USB kernel scanner driver > > Hi, > > While sending some questions and patches to the linux-usb-devel > mailing list, I was asked by Greg KH, the Linux USB maintainer: > > | So the userspace SANE tools don't need the scanner driver anymore? > If > | so, I'd be glad to remove the driver, like the others that we > dropped a > | while ago for this same reason. > > I answered: > > | SANE 1.0.9 will have support for both the scanner driver and > libusb. > | However, some backends (drivers) still have hardcoded links to > | /dev/usb/scanner? so it will take at least one more version to get > it > | right. So basically, no, the scanner driver isn't needed in future. > If > | the removal is only in Linux 2.5, I think that's ok. For 2.4, I > | wouldn't do that because people tend to update their kernels more > | often than their SANE package and shouldn't lose their ability to > | access USB scanners. > | > | I don't know vuescan very well, so maybe it depends on the kernel > | scanner driver. I have no idea about other packages. > | > | But that's just my personal opinion. > > The background is, that there isn't a really good reason to have a > generic driver for USB scanners in the kernel. There is no USB class > for scanners, so the scanner driver must be constantly updated with > new vendor and product ids. Furthermore, the scanner driver doesn't > provide any functionality that can't be accessed through the generic > /proc/bus/usb interface, i.e. libusb. > > The reason why this came up is the ongoing trouble concerning the > maintainance of the kernel scanner driver. Adding new device ids is > quite slow, and some long-standing patches haven't been integrated > for > months (years?) and generally, the interface is too inflexible. Also, > it just doesn't make sense to keep the vendor and product ids in two > different places (kernel and SANE). > > Furthermore, libusb supports more platforms then the Linux-alike > scanner modules. > > Since SANE 1.0.9, sanei_usb can also use libusb. If a backend just > uses sanei_usb_attach_matching_devices() with "usb vendor product" > lines or sanei_usb_find_devices() directly, there is no need to > change > the backend at all. I's only necessary to keep the vendor and product > ids somewhere in the backend, either in the code itsself or (better) > in the config files. > > So the basic question is: Is there anything that the scanner driver > can do that libusb can't and that's needed by any backend? Are there > any other reasons to not remove the USB scanner driver in Linux some > day in the future? > > If there are no objections, the plan will be to change the backends > to > be able to use both libusb and the kernel scanner module. > > The following backends seem to support USB scanners: > > canon630u looks like libusb-ready > coolscan2 looks like libusb-ready > epson already has a check for vendor/device ids but may need a > way > to search for all epson scanners automatically > mustek_usb is libusb-ready > plustek will need some work to autodetect all supported scanners > (currently the approach is: check ids of /dev/usb/scanner > for > everything we now) > snapscan looks like libusb-ready > umax looks like libusb-ready > umax1220u looks like libusb-ready > > I've only looked at the config files and sometimes source code, so I > may be wrong. > > Bye, > Henning > > --__--__-- > > Message: 8 > Date: Wed, 23 Oct 2002 17:47:18 +0100 > From: Major A <and...@users.sourceforge.net> > To: SANE devel <sane-de...@mostang.com> > Subject: Re: [sane-devel] USB kernel scanner driver > > > > The background is, that there isn't a really good reason to have a > > generic driver for USB scanners in the kernel. There is no USB class > > for scanners, so the scanner driver must be constantly updated with > > new vendor and product ids. Furthermore, the scanner driver doesn't > > provide any functionality that can't be accessed through the generic > > /proc/bus/usb interface, i.e. libusb. > > How does libusb handle permissions on linux then? /proc/bus/usb can > only be accessed by root. With the USB scanner module, one had a way > of setting permissions via the device file. > > How would one restrict access so that although a user can scan, they > can't fiddle with the USB hard drive or CD-ROM or automatic > nuclear-weapon detonator? > > saned is great at this, but it would now have as root (I tend to run > it as a dedicated "saned" user), and I don't think we should lay the > burden of setting saned up on the normal user who just wants to scan. > > > So the basic question is: Is there anything that the scanner driver > > can do that libusb can't and that's needed by any backend? Are there > > any other reasons to not remove the USB scanner driver in Linux some > > day in the future? > > If it happens in linux 2.5.x, then that's fine by me, provided 2.6 or > 3.0 is sufficiently far away. > > > If there are no objections, the plan will be to change the backends to > > be able to use both libusb and the kernel scanner module. > > Isn't that done transparently by the sanei_usb interface? > > BTW, VueScan uses the USB scanner module. > > Andras > > ===================================================================== > ====== > Major Andras > e-mail: and...@users.sourceforge.net > www: http://andras.webhop.org/ > ===================================================================== > ====== > > --__--__-- > > Message: 9 > Date: Wed, 23 Oct 2002 21:01:58 +0400 > From: Sergey Vlasov <v...@altlinux.ru> > To: sane-de...@www.mostang.com > Subject: Re: [sane-devel] USB kernel scanner driver > > On Wed, 23 Oct 2002 17:47:18 +0100 > Major A <and...@users.sourceforge.net> wrote: > > > How does libusb handle permissions on linux then? /proc/bus/usb can > > only be accessed by root. With the USB scanner module, one had a way > > of setting permissions via the device file. > > Giving access to all USB devices is relatively simple: usbdevfs has > several mount options to control it: devuid, devgid, devmode; bus*, > list*. > > > How would one restrict access so that although a user can scan, they > > can't fiddle with the USB hard drive or CD-ROM or automatic > > nuclear-weapon detonator? > > This is trickier. Looks like the hotplug scripts are able to do this, > however, this needs some tweaks... > > > saned is great at this, but it would now have as root (I tend to run > > it as a dedicated "saned" user), and I don't think we should lay the > > burden of setting saned up on the normal user who just wants to scan. > > You can give the "saned" user the right to access all USB devices. > > --__--__-- > > Message: 10 > Date: Wed, 23 Oct 2002 18:11:16 +0100 > From: Major A <and...@users.sourceforge.net> > To: SANE devel <sane-de...@mostang.com> > Subject: Re: [sane-devel] USB kernel scanner driver > > > > Giving access to all USB devices is relatively simple: usbdevfs has > > several mount options to control it: devuid, devgid, devmode; bus*, > > list*. > > Oh, that's good to know. So far, I've only tried chmod, with no > success... > > > You can give the "saned" user the right to access all USB devices. > > Great, it's simple once you have the possibility of setting > individual > permissions. I was assuming you could only set permissions for the > entire usbdevfs tree, and I haven't actually used hotplug for USB > yet. > > Andras > > ===================================================================== > ====== > Major Andras > e-mail: and...@users.sourceforge.net > www: http://andras.webhop.org/ > ===================================================================== > ====== > > --__--__-- > > Message: 11 > To: SANE Mailing List <sane-de...@mostang.com> > Subject: Re: [sane-devel] USB kernel scanner driver > Date: Wed, 23 Oct 2002 18:21:50 +0100 > From: Jonathan Buzzard <jonat...@buzzard.org.uk> > > > henn...@meier-geinitz.de said: > > The following backends seem to support USB scanners: > > canon630u looks like libusb-ready coolscan2 looks like libusb-ready > > epson already has a check for vendor/device ids but may need a way > > to search for all epson scanners automaticall > > mustek_usb is libusb-ready > > plustek will need some work to autodetect all supported scanners > > (currently the approach is: check ids of /dev/usb/scanner for > > everything we now) > > snapscan looks like libusb-ready > > umax looks like libusb-ready > > umax1220u looks like libusb-ready > > The other backends that spring to mind as using the scanner module > are > the hp backend, which uses the scanner backend, and the avision > backend > which uses the hpusbscsi module. > > JAB. > > -- > Jonathan A. Buzzard Email: jonat...@buzzard.org.uk > Northumberland, United Kingdom. Tel: +44(0)1661-832195 > > > > --__--__-- > > Message: 12 > Date: Wed, 23 Oct 2002 19:30:26 +0200 > From: Henning Meier-Geinitz <henn...@meier-geinitz.de> > To: Major A <and...@users.sourceforge.net> > Subject: Re: [sane-devel] USB kernel scanner driver > > Hi, > > On Wed, Oct 23, 2002 at 05:47:18PM +0100, Major A wrote: > > How does libusb handle permissions on linux then? > > See man sane-usb :-) > > > /proc/bus/usb can only be accessed by root. > > On a one-user system: just mount it with read/write permissions for > everyone. On multiuser systems, use hot-plug utilities to set up > permissions. > > > With the USB scanner module, one had a way of setting permissions via > > the device file. > > But only for all or no scanner. If you have a machine with more then > one scanner, you can't grant permissions for only a specific scanner > (because you never know, which one will be /dev/usb/scanner0). > > > How would one restrict access so that although a user can scan, they > > can't fiddle with the USB hard drive or CD-ROM or automatic > > nuclear-weapon detonator? > > Use a chmod in a hot-plug script that starts when that specific > vendor/product id is plugged in. > > > > If there are no objections, the plan will be to change the backends to > > > be able to use both libusb and the kernel scanner module. > > > > Isn't that done transparently by the sanei_usb interface? > > Basically, yes. But if you just use a hardcoded > > sanei_usb_open ("/dev/usb/scanner", &fd) > > sanei_usb will just open that file. > > If you use sanei_usb_find_devices or > sanei_usb_attach_matching_devices > and call sanei_usb_open with the "file name" you got from those > functions, then it will also work with libusb. > > Basically, backends that rely on having a device file name like > /dev/usb/scanner0 > in their config files need to be changed. > > > BTW, VueScan uses the USB scanner module. > > Ok. > > Bye, > Henning > > > --__--__-- > > _______________________________________________ > Sane-devel mailing list > sane-de...@www.mostang.com > http://www.mostang.com/mailman/listinfo/sane-devel > > > End of Sane-devel Digest > -- Sent by Pinoymail, http://www.pinoymail.com/
