Best Distro for document management

Sundeniro · 11-11-2008, 03:33 AM

Hi

I have a PC that I wish to retire as my windows machine (I have to have one for work) I thought it would be really useful to me to make this PC a dedicated document management PC i.e. for scanning, optically recognising the documents and then turning them int searchable PDFs.

This is something I do now on WinXP I have been messing with a few live-CDs (completely new to linux) to see whihc ones I like the feel of.

I have tried the following so far

OpenSUSE - runs very slow
Ubuntu - Seems quite nice
Kubuntu - Failed to load
PuppyLinux - Too Cluttered and confusing

I will give others a go but to reduce the time I spend trying to get a system working fully I thought it would be a good idea to post my Spec and ask for suggestions on both best distro for my hardware and software to accomplish my document management needs.

Pentium 4 2.8GHZ
Ram 1.5GP
Graphics AGPx4 512MB dedicated memory
Internet connection currently via LAN but may switch to WiFi but that shouldn't be an issue as I will do that post set up and specifically seek compatible hardware.

If you need any other info please let me know

And thanks very much for your help

Scanner is a Plustek PS256 Smartoffice ADF Scanner.

I also use a Network Printer.

billymayday · 11-11-2008, 04:26 AM

OpenSuse shouldn't be too slow, and I'd be wary about measuring speed based on a live CD.

I'd be looking at OpenSuse, CentOS, Debian and maybe gentoo and slackware

Sundeniro · 11-11-2008, 07:53 AM

why those particular distributions?

I'd really appreciate it if you could expand on your suggestions?

I have been favouring UBUNTU up until now but am happy to admit that I am not knowledgeable on this subject hence my request for advice.

pixellany · 11-11-2008, 08:03 AM

Anything in the top ten on the "hit list" at Distrowatch is going to be fine.

The long tent pole may be that scanner--I would do some searches

Try here: http://www.sane-project.org/sane-mfgs.html

Found this by accident:
http://www.gjaeger.de/scanner/plustek.html

Sundeniro · 11-11-2008, 08:34 AM

Yeah the scanner does seem to be a sticking point, also finding there is little in the way of decent document management and OCR software available on LINUX which is another major issue for me.

I'm a one man business and was interested in setting up a "mirror" PC to do everything I do for my work on my windows box now to see if there would be a feasible point to switch over, but it is looking very unlikely to happen any time in the next few years I think.

farslayer · 11-11-2008, 09:22 AM

I would use Debian and install knowledge tree Open Source document management system on it.

Debian can easily be installed in a minimal configuration without a lot of extra junk you don't need on a server, and it runs well on slightly older spec hardware since it lacks a lot of the Bloat found in some distros.

Sundeniro · 11-11-2008, 09:42 AM

Thanks for the reply but I do not want to run this PC as a Server but as a desktop with a scanner attached, to do it's own OCR and PDF conversion

farslayer · 11-11-2008, 10:29 AM

Ah so it's a document scanning station and not a Document Management System.. gotcha..

How to scan and OCR like a pro with open source tools (Linux.com
OCR for Linux: Teaching Linux to Read (Linux Magazine
Tesseract OCR - Highest rated FOSS OCR for Linux

gscan2pdf gscan2pdf - attaches metadata to PDF upon creation to allow indexing for searches.

Quote:

OCR (Optical Character Recognition)

The gocr or tesseract utilities are used to produce text from an image.

There is an OCR output buffer for each page and is embedded both as an annotation (pop-up note) and as plain text behind the scanned image in the PDF produced. This way, Beagle can index (i.e. search) the plain text, and the contents of the annotations can be viewed in Acrobat Reader.

In DjVu files, the OCR output buffer is embedded in the hidden text layer. Thus these can also be indexed by Beagle.

There is an interesting review of OCR software at http://groundstate.ca/ocr. An important conclusion was that 400dpi is necessary for decent results.

Let us know how you put this all together and what software you choose.

Thanks !!

billymayday · 11-12-2008, 12:53 AM

Quote:

Originally Posted by Sundeniro

why those particular distributions?

I'd really appreciate it if you could expand on your suggestions?

I have been favouring UBUNTU up until now but am happy to admit that I am not knowledgeable on this subject hence my request for advice.

Stable, well supported, a wide variety of software available (brother, for example provides drivers in binary form for CentOS, Suse and Debian at least).

It's a very personal choice as to what distro you like.

As others have said, the main issue will be the scanner end, but you probably have the best chance of finding a ready package on a major distribution (and I'm not a big fan of the *buntus).

Sundeniro · 11-12-2008, 02:10 AM

Hi all

Thanks for the advice it is very much appreciated.

I think my best course of action is to build a much better windows box as previously planned.

Tune this machine in to a LInux box in the corner of the room still not sure which distro and whether a GNOME or KDE environment is better for my needs.

Keep doing my scanning etc on the windows box while I get a good book on LINUX and learn as much about my distro as possible and take a trial and error approach to getting the machine to do what I want. (Might take months but i like learning new things)

Not sure if I should as this on another thread but, is it true i only need a firewall on a linux machine and no virus software or is this just an old wives tale?

Many thanks for all your kind advice.