1999.01.05 17:10 "Re: Opening to page "x" in a multi-page TIFF file?", by Mark Knibbs

it stood for proprietary. The main problem with using PDF as a file format for scanned documents is that users who do not use a platform for which Adobe provide Acrobat Reader are unable to view the document.

I am not affiliated with adobe, but I do not agree.

How are users *who do not use a platform for which Adobe provide Acrobat Reader* supposed to view PDF scanned documents then? (Given that Ghostscript does not seem to like some PDF files.)

Platforms currently supported by adobe:

[list of platforms including various Windows and UNIX versions]

There is an unsupported DOS version too.

Adobe Acrobat ran on my 386 dx2 with 4MB but of course not with huge pdf files - same for huge tiff files with 'normal' tiff viewer.

The DOS version only works with very old PDF files (i.e. not with most currently available ones). Do Adobe provide binaries for non-x86 Linux versions? What about non-x86 ports of NetBSD, FreeBSD etc.?

PDF is neither very large not complex when the pages are just stored directly with the image data in the original format.

I was talking about the PDF specification itself, not necessarily individual PDF files. By which I mean, since the PDF specification supports many features (i.e. it is complex), PDF viewers need to support these features which makes them large.

Sure, PDF files can contain (more or less) just compressed image data with file size comparable to the equivalent TIFF G4 file. But given that they may also contain structured text etc., PDF viewers need to support all these features.

For example, there is no Acrobat Reader for the OS I use. Ghostscript (free PostScript/PDF interpreter) does exist for it. I recently downloaded a large scanned document in PDF format from a web page. Sadly Ghostscript chokes on this file, and so I cannot view it.

Are you using Amiga or so?

That's one of the computers which I use, but the point is equally applicable to any other OS for which there is no Acrobat Reader.

For Windows 3.x/95/98/NT (Acrobat Reader may not run on low-end machines):

And neither will any of the viewers if the tiff file is big.

I have used Infothek DocView to view multi-page TIFF G4 files (300dpi, 2320x3408 pixels per page) on a 4MB 386SX machine under Windows 3.1. Of course moving from page to page is quite slow, but it does work. For very low-memory machines, it would be possible to write a TIFF-printing program which only requires a few K for the image data, by decompressing and printing a few scanlines at a time.

For UNIX/NetBSD etc. (Acrobat Reader probably not available):

Not true

Not true for a few UNIX-like OSes, but what is Adobe's non-x86 support like? I'm sure that the number of UNIX Acrobat Reader binaries is less than half the number of UNIX OSes in existence.

All plugins run on same platform as acrobat but acobat is available (free) for more platforms.

Yes. I mentioned plugins to illustrate that e.g. publishing TIFF scans on web pages is no less accessible to people running mainstream OSes (for which Acrobat Reader does exist) than PDF. It is of course much *more* accessible for those who are not able to run Acrobat Reader. TIFF viewer software is available for more platforms (almost every platform) than Acrobat Reader.

platforms makes acrobat a far better choice as a plugin.

Also with optimising, pdf can be streamed off the web.

For scanned documents specifically this would also be possible with TIFF. Whether anyone has actually tried this is doubtful though.

A good browser solution would in my opinion be a java based viewer downloaded when needed - too bad that the one available that CAN print only prints in Netscape 4.06+ (due to needed java 1.1 support)

That would probably be a good idea, for people with fast modern computers.

[list of browser plugins]

These are all windows only plugins, though...

I just did a quick web search to get some links. I have no idea whether there are also any free Mac TIFF browser plugins. (I would be interested in the URLs if there are.) There may be TIFF plugins available for a similar range of UNIXes as Acrobat Reader, but have I not looked into this.

Just to recap: I do not necessarily think that storing scanned images in PDF format is the way to go. CCITT G4 tiff is an ok format and from there it can be converted to whatever format deemed necessary.

For viewing, however, I

I guess so. What annoys me is that some web sites which "publish" scanned documents (e.g. the EPO Espacenet site) do so in PDF format. Users of the web site obviously only have access to these and not the original non-PDF files. (And apparently, if you do use an Acrobat Reader browser plugin, it may not allow you do save out the PDF files to disk!)

find the available tiff viewers often lacking in G4 support and 300 dpi images are often decompressed to a silly 3K*2K pixels screen waster. A gif needs to be decompressed and resized to 72 dpi to be managable. So far I have

There seem to be two "classes" of image viewer. There are programs designed for displaying pictures, which might typically cater also for GIF, JPEG etc.

Then there are viewers which are designed with document imaging applications in mind; these are usually cleverer when it comes to scaling the image. One program in the latter category is Infothek DocView, for Windows 3.1/95/NT. This is available from http://www.informatik.com/.

If you use Windows, you should check this out. It antialiases the image when viewing at reduced sizes. For example, even on my poverty-stricken Windows 3.1 machine (16MHz 386SX, 640x480 16-colour desktop), text from 300dpi scans of US patents is legible at 25% scale (effectively 75dpi) -- full page width is visible. It's even quite fast when scrolling up and down a page.

In fact, I have found Acrobat Reader to be *far worse* than Infothek DocView for viewing scanned documents. It's much slower and doesn't antialias scaled- down images.

-- Mark