Re: [O] How to get a paper scanner into org-mode workflow

[Karl Voit]

* Torsten Wagner <address@hidden> wrote:
> Hi,

Hi Torsten!

> I plan to buy a document scanner with ADF and duplex function to scan all
> incoming/intermediate/outgoing papers, convert them into PDF and link those
> into my org-files.

I definitely recommend you to but Fujitsu ScanScap S1500. I do have
the Mac version.

> I was wondering if someone did something like this already? 

I am in the process of digitizing all my physical life so far to
complete my lifelogging-habbit. I also removed the spine of many
books in order to be able to scan them for full-text search and
archiving purposes!

> I use Linux and hence I am looking for a Linux friendly solution.

Although I prefer GNU/Linux by myself, at least the Mac software
bundle of the Fujitsu is really worth the money. Its scanning and
OCR features are just like you need, almost no hazzle with anything
(except the awful bad idea to use JPEG as the only scanning format).

Before the Fujitsu, I bought a HP OfficeJet Pro 8500A Plus which I
am using as a (pretty decent) printer. Because the scanning hardware
and software is *really* complicated/bad compared to the Fujitsu
solution: reliability, usability, featureset, speed - BIG
difference.

I played around with Linux scanning/OCR solutions as well but I
could not get any reasonable fast, reliable, and suitable OCR
results at all :-(

Note: The scan workflow on my Mac mini at home is more or less the
only reason why I did not install GNU/Linux on this thing yet. (Most
of the time I am using my Debian Wheezy notebook to get things done
efficiently.)

> I want to make this finally as smooth, easy and painless as possible to
> make sure I do not pile up papers because I am to lazy to scan them on a
> day by day basis.

Sure. BTDT.

> Ideally, I want to fill the ADF unit, press a button, receive a single PDF
> with all pages scanned duplex in a preset folder, and keep a path to that
> file in org-modes killring to place it as a link at the right place. OCR
> would be nice too to get a full searchable system.

OCR ist a must, if you invest time and money in such a workflow.

Except the killring thing, I do have a similar workflow: paper in
scanner, press single button on the scanner, PDF (with OCR) appears
on the Mac, I enter the filename (usually starting with
YYYY-MM-DD... or YYYY-MM-DDThh.mm...), and it's done. Most things I
keep in ~/archive/YYYY/ and some things I file in
~/institutions/COMPANYNAME/

Because of the date/time-stamp file name part, I can index all those
files using my Memacs[1] file-name module very easily. Using a
yasnippet/custom-Org-link combination, I can easily link to those
files rather quickly. So much for the Org-mode related stuff :-)

> CC. Any recommendation on a well working scanner under Linux. At the moment
> my favorite is the Fujitsu ScanSnap S1500. Seems to work out of the box and
> comes with a good performance.

Oh, you already did good research on this topic :-)

If you try out the Mac SW bundle (I guess Windows is similar), you
are not going to use Linux for this I am afraid :-)

But if you do get different results and thus a workflow with
GNU/Linux which is as smooth as the Mac one, I am *very* curious
about your experience, notes, and lessons learned!

PS: Because of a cross country skiing vacation I am looking forward
to, I will probably not be able to read this ML until next weekend
or so.

  1. https://github.com/novoid/Memacs
-- 
Karl Voit