19 May
2005
19 May
'05
7:02 p.m.
> Doesn't this sort of imply that PDF is the
wrong choice of format for
> jobs like these? PDF is terrible way to package the documents.
It's just better than any other practical method
;-)
So many talk about ASCII being the only "right
way", as Al can attest to
time and accuracy makes image oriented PDF's the way to go.
I'm pragmatic. I have hundreds of thousands (probably millions now..) of
pages of paper. I wanted easy access to all of it. Eric also had a lot of
stuff that I didn't have. The machines I use are Macs and Linux boxes.
There are PDF viewers for both of those platforms that had page indexing.
The level of PDF used is the minimum to support wrapping collections of
scans together with very simple metadata (the page number). That's all
of PDF that I use. The overhead is minimal, it can be read across most
computers currently in use, and if it's working with the right sort of
browsers will only transfer the page being viewed instead of the whole
document.
If something different comes around, the PDF spec is public, and by using
such a small subset it should be simple to translate.
I've never found much use for the fancier scripting stuff in tumble. It
takes long enough to just do the minimal post-processing of the scans that
I do now to think about writing a script for each document to do any sort
of fancier indexing.
19 May
19 May
7:37 p.m.
On Thu, 2005-05-19 at 16:02 -0700, Al Kossow wrote:
Sure... I'm not sure what it is about PDF that bugs me (I'm not that
much of a fan of it even for textual stuff TBH!). In the case of doc
scans it just feels a bit odd downloading a bunch of images wrapped in a
proprietary format I suppose, when there are far nicer tools for dealing
with images than there are PDF viewers for handling PDF content (and
because the PDF files are just full of images, there's no other
metadata).
I tend to 'explode' any PDF files of scans (from whatever source) here
once downloaded into their own directory; I just find it easier to
manipulate via whatever image tool is most suitable for whatever I'm
doing at the time, rather than being stuck with a PDF viewer. I suppose
if I wanted to add metadata to that, I'd include an ASCII text file in
the directory full of images with the relevant info in (I've done that
with ROM and Disk images many a time; not needed to do it with Doc scans
yet*)
* For single page magazine ad scans I've tended to enlarge the image
canvas a little at the top or bottom with a blank area and included
there details about what magazine it was from, date etc. - there are
likely much better ways (and that obviously doesn't scale to thousands
of docs!) but it's been better than nothing up until now.
Maybe I'm atypical in usage :-) I'll rarely want to download scans and
*not* keep a copy on local storage just in case, so I've never used the
"view a PDF file in a web browser" side of things.
>
Doesn't this sort of imply that PDF is the wrong choice of format for
> jobs like these? PDF is terrible way to package the documents.
It's just better than any other practical
method ;-)
So many talk about ASCII being the only
"right way", as Al can attest to
time and accuracy makes image oriented PDF's the way to go.
I'm pragmatic. I have hundreds of thousands (probably millions now..) of
pages of paper. I wanted easy access to all of it. Eric also had a lot of
stuff that I didn't have. The machines I use are Macs and Linux boxes.
There are PDF viewers for both of those platforms that had page indexing. If something different comes around, the PDF spec is
public, and by using
such a small subset it should be simple to translate.
Yep true... plenty of tools already exist to pull PDF files apart. Well,
you'll be converting all your bitsavers content to futurekeep format
soon :-)
cheers
Jules
7:51 p.m.
On Thu, 2005-05-19 at 23:37 +0000, Jules Richardson wrote:
On Thu, 2005-05-19 at 16:02 -0700, Al Kossow wrote:
Standard disclaimer applies that bitsavers is no less great for that
statement, by the way :-)
cheers
J.
I'm pragmatic. I have hundreds of thousands
(probably millions now..) of
pages of paper. I wanted easy access to all of it. Eric also had a lot of
stuff that I didn't have. The machines I use are Macs and Linux boxes.
There are PDF viewers for both of those platforms that had page indexing.
Sure... I'm not sure what it is about PDF that bugs me (I'm not that
much of a fan of it even for textual stuff TBH!). 
20 May
20 May
3:46 a.m.
On Thu, 2005-05-19 23:37:18 +0000, Jules Richardson <julesrichardsonuk at
yahoo.co.uk> wrote:
Pointers for tools? Even while I'm out of time, I'd like to learn more:)
MfG, JBG
--
Jan-Benedict Glaw jbglaw at lug-owl.de . +49-172-7608481 _ O _
"Eine Freie Meinung in einem Freien Kopf | Gegen Zensur | Gegen Krieg _ _ O
fuer einen Freien Staat voll Freier B?rger" | im Internet! | im Irak! O O O
ret = do_actions((curr | FREE_SPEECH) & ~(NEW_COPYRIGHT_LAW | DRM | TCPA));
On Thu, 2005-05-19 at 16:02 -0700, Al Kossow wrote:
I tend to 'explode' any PDF files of scans (from whatever source) here
once downloaded into their own directory; I just find it easier to
manipulate via whatever image tool is most suitable for whatever I'm
doing at the time, rather than being stuck with a PDF viewer. I suppose
if I wanted to add metadata to that, I'd include an ASCII text file in
the directory full of images with the relevant info in (I've done that
with ROM and Disk images many a time; not needed to do it with Doc scans
yet*)
That's basically what I did for the TeX+Images -> PDF script, one .TXT
file per image. This way, you can easily distribute work if needed.
Maybe I'm atypical in usage :-) I'll rarely
want to download scans and
*not* keep a copy on local storage just in case, so I've never used the
"view a PDF file in a web browser" side of things.
No, I consider that as normal, fair and expected usage :)
Just a honest question. There is some kind of warez szene
scanning/ORCing/correcting scans of current best-selling books. How do
they do the job? They've basically got to solve the very same problem.
If something
different comes around, the PDF spec is public, and by using
such a small subset it should be simple to translate.
Yep true... plenty of tools already exist to pull PDF files apart. Well,
you'll be converting all your bitsavers content to futurekeep format
soon :-) 
19 May
19 May
10:16 p.m.
I created indexes for the five WesternDigital databook PDF's I downloaded
and sent them to Al.
In a week or so I'll try a google search and see if it shows up.
I suggest anyone downloading PDF's that are not indexed spend just a few
minutes and create an index, it will help the next person.
I created a text file using the exact same file name as the PDF file (with
txt instead of pdf, and it was really 5 files). I created a three line
header and followed with one line per device with 24 characters for device
name, 8 characters for printed page number, 8 characters for PDF page
number.
Where more than one device was mentioned for a page I entered separate lines
with different device names and duplicated page numbers, I only entered the
1st page number not a range.
If we all share the labor it shouldn't take long to deluge Al with index
files ;-)
It is possible that Al will not instantainiously post the index files so for
now if anyone else sends Al an index file it might be handy to say so here
so others will know they don't need to duplicate the work.
Randy
10:29 p.m.
From: "Randy McLaughlin" <cctalk at randy482.com>
Sent: Thursday, May 19, 2005 9:16 PM
I created indexes for the five WesternDigital databook
PDF's I downloaded
and sent them to Al.
In a week or so I'll try a google search and see if it shows up.
I suggest anyone downloading PDF's that are not indexed spend just a few
minutes and create an index, it will help the next person.
I created a text file using the exact same file name as the PDF file (with
txt instead of pdf, and it was really 5 files). I created a three line
header and followed with one line per device with 24 characters for device
name, 8 characters for printed page number, 8 characters for PDF page
number.
Where more than one device was mentioned for a page I entered separate
lines with different device names and duplicated page numbers, I only
entered the 1st page number not a range.
If we all share the labor it shouldn't take long to deluge Al with index
files ;-)
It is possible that Al will not instantainiously post the index files so
for now if anyone else sends Al an index file it might be handy to say so
here so others will know they don't need to duplicate the work.
Randy
After sending the index files to Al I realized what I forgot: More
keywords.
I should have included the following text in the files:
WesternDigital
Western Digital
Datasheet
Data Sheet
That way if someone googles 'western digital wd1100 datasheet' it would show
up higher on the list (more keyword hits).
Randy
20 May
20 May
3:38 a.m.
On Thu, 2005-05-19 16:02:51 -0700, Al Kossow <aek at bitsavers.org> wrote:
I've never found much use for the fancier
scripting stuff in tumble. It
takes long enough to just do the minimal post-processing of the scans that
I do now to think about writing a script for each document to do any sort
of fancier indexing.
Personally, I think the extra-features of tumble aren't not yet
complete. But this is *not* a complaint: I'm a hacker and could probably
add missing features, but (soon switching my employer, so I'm now
coaching my remaining co-workers...) I'm horribly out of time. Maybe
that'll get better in some time...
MfG, JBG
--
Jan-Benedict Glaw jbglaw at lug-owl.de . +49-172-7608481 _ O _
"Eine Freie Meinung in einem Freien Kopf | Gegen Zensur | Gegen Krieg _ _ O
fuer einen Freien Staat voll Freier B?rger" | im Internet! | im Irak! O O O
ret = do_actions((curr | FREE_SPEECH) & ~(NEW_COPYRIGHT_LAW | DRM | TCPA));
7004
days inactive
7005
days old
6 comments
4 participants
participants (4)
-
aek＠bitsavers.org -
cctalk＠randy482.com -
jbglaw＠lug-owl.de -
julesrichardsonuk＠yahoo.co.uk