Gentoo Archives: gentoo-user

From: Eric Crossman <edge1035@×××××××××.net>
To: gentoo-user@l.g.o
Subject: Re: [gentoo-user] Document management solution [possibly a bit off-topic...]
Date: Fri, 30 Sep 2005 01:47:35
Message-Id: 1128044586.5145.29.camel@localhost
In Reply to: Re: [gentoo-user] Document management solution [possibly a bit off-topic...] by Nick Rout
1 On Fri, 2005-09-30 at 10:36 +1200, Nick Rout wrote:
2 > On Thu, 29 Sep 2005 16:52:54 -0400 (EDT)
3 > A. Khattri wrote:
4 >
5 > > On Thu, 29 Sep 2005, Steve [Gentoo] wrote:
6 > >
7 > > > Alfresco is what I'd have called a content management system - as
8 > > > opposed to a document management system. I'm interested in managing
9 > > > archives of documents I have received from other people (in dead-tree
10 > > > format)...
11 > >
12 > > If there was something that scanned the document, performed OCR on it,
13 > > checked the OCR output and then built an electronic repository for you I'd
14 > > recommend it. Until then, Alfresco is the closest thing Ive seen that is
15 > > open source. If you're willing to do your own scanning and OCR'ing then it
16 > > will do the rest.
17 > >
18 > > BTW, I would call things like Mambo or Xaraya, content-management tools -
19 > > Alfresco is a slightly different kettle of fish.
20 >
21 > Yes I know what Steve is after, and I'd love to find a way. I was put
22 > off by Alfresco being called "Content Management" because all of the
23 > content management systems I have seen end up bioding something that
24 > resembles [name your favourite news website]
25 >
26 > A closer look at alfresco reveals that it does look more like what Steve (and I ) are after.
27 >
28 > I am a lawyer and I handle hundreds of documents every week, from email
29 > through pdf (both made from an electronic source and therefore has all
30 > the text available, and scanned) openoffice (one enlightened client!),
31 > word, excel, html, faxes, letters (on paper, ya know!) you name it
32 > someone will send me something in it!
33 >
34 > It'd be great to have a metadata system where I could give everything
35 > some keywords:
36 >
37 > client name, file number, matter number, subjects, useful as a
38 > precedent, useful case etc etc etc so that in future I can :
39 >
40 > pull up every document on my computer, my secretary's computer, my mail
41 > server (including attachments), my file server, my palm pilot, relating
42 > to a particular client
43 >
44 > pull up every document about company debentures
45 >
46 > find the case i downloaded and stored somewhere about liability of
47 > guarantors in a consumer credit loan
48 >
49 > find the seminar book for the seminar i went to on asome new area of
50 > law.
51 >
52 > find a letter written by Joe Bloggs sometime in 2003.
53 >
54 >
55 > >
56 > >
57 > > --
58 > >
59 > > --
60 > > gentoo-user@g.o mailing list
61 >
62 > --
63 > Nick Rout <nick@×××××××.nz>
64 >
65
66 I'm not sure if what you're describing exists right now in the open
67 source world, but I can tell you that it certainly does in the
68 commercial world. I used to work in the "metadata" department for a
69 startup here in upstate NY, USA that built a web based application
70 targeting lawyers such as yourself. It was written in PHP/MySQL but the
71 database was being migrated to Oracle due to the rapid growth in the
72 database tables.
73
74 Unfortunately though, in the migration to Oracle, they elected to create
75 a "dynamic" scheme to support adding custom metadata fields as requested
76 per client. It was great for flexibility but the performance was
77 horrible even on quad 3 ghz xeon boxes with maxed out memory. For us
78 programmers, it also made the easy queries difficult and the hard
79 queries near impossible.
80
81 Eric
82
83
84 --
85 gentoo-user@g.o mailing list