[Coco] RE: Rainbow magazines]

Roger Merchberger zmerch at 30below.com
Wed Jun 8 17:31:54 EDT 2005

Rumor has it that Rswoger at aol.com may have mentioned these words:
> >From Bob
>You guys are right. I just tried to OCR a JPG and a GIF. Many mistakes. I
>tried PNG and TIFF as suggested also available as saves from paint.

Lordy, why are you using Paint?

Check out ACDSee - very nice Image viewer/converter; it'll do batch 
convert/rotate/resize/etc. very easily. The earlier versions are quite 
small & fast as well; tho creature feep is moving in on the latest version...

There are many other freeware/shareware/beerware converters out there that 
are batch capable - they might save you quite a bit of time.

>  PNGs were better and TIFF were even better yet.

Did you actually look at the the size of the TIFFs? [[ See below for the 
answer. ]]

>I believe that obtaining program listings using OCR would be a real PLUS.

Assuming whomever does it is willing to actually double-check what the OCR 
program spit out -- these proggies are by no means anywhere close to perfect.

>I for one hate PDF files because of their size and the requirement for a
>reader which is also large and slow.

[[ Regarding above WRT TIFF files... ]]

PDF files are large when they contain only scanned graphics -- if the text 
is OCR'ed first, the files themselves become correspondingly smaller; I use 
OpenOffice as my spreadsheet, and it has a one-click PDF creator... and 
also output's HTML. The HTML is on average 27K; the PDFs average 36K; and 
the PDF's look one heckuva lot better than the corresponding HTML.

I can post examples of each to the web if you wish to see them.

As such, TIFFs are either uncompressed or losslessly compressed, so are 
usually quite large; and a lot of programs can't handle compressed TIFFs at 
all, let alone multipage TIFFs. Yes, they exist, no, most software can't 
deal with 'em very well at all. :-( The compression that PDFs can use is 1) 
on average a little more efficient than what's available with TIFFs, and 2) 
multipage files are not only well-supported, but much easier to navigate.

Oh, and Adobe Reader versions < 5 are actually rather speedy, 5 is so-so, 6 
is really where they start doggin' it down. Needless to say, I don't run 6. 
;-) There are also several freeware readers out there, some of which are 
quite speedy. PDF is a fully documented standard also, so there'll be 
readers available for it for a good long time.

Another great example of PDFs that work great is the release of the book 
Thinking Forth from Leo Brodie, that our very own John Hogerhuis helped 
bring about. Almost 300 pages, and it's less than 4.4 megs. AFAIK, there's 
no other on-screen reader that could come close to the 'ease of use' of 
that project.

>  As you see on my web site, I like something
>small and fast like I give you all when you read the newsletters.

Small is less of an issue as people won't be downloading it -- they'll be 
browsing 'em off of the CD-ROM/DVD-ROM disk. "Fast" is difficult to gauge 
as if it's low enough quality it'll be tough (and slower) to read (and 
subsequently OCR later -- personally, I've have very little luck OCRing 
anything under 300dpi; 600 is better (with the subsequent 4x increase in 
image size)) and so a quick 'opening' time may lead to a slower 'usability' 

TIFF can't easily handle mixed-mode data -- if the data/text/listings were 
OCRed, how would you access it? With PDFs, it's a highlight, copy & paste 
function. With TIFFs, it's not gonna happen. PDFs are also 
hypertext-capable - if indexed, you could click on an index reference, and 
go right to that page.

A lot is going to depend on what tradeoffs will be made between quality, 
speed, ease of use, and of course (and most importantly) the ability to 
reach the widest audience available with the product.

I'm not saying that PDFs are the be all and end all of data viewers -- just 
that 1) whoever does the job needs to weigh the pros & cons of what they 
use carefully and 2) don't exclude a large portion of the target market due 
to the format choice.

Roger "Merch" Merchberger

Roger "Merch" Merchberger  --  SysAdmin, Iceberg Computers
  _±±_                          zmerch at 30below.com
(©||®)  If at first you don't succeed, nuclear warhead
  _)(_   disarmament should *not* be your first career choice.

More information about the Coco mailing list