2001.10.03 19:05 "OJPEG & Wang Images (long)", by Scott Marovich

2001.12.02 16:28 "Re: OJPEG & Wang Images (long)", by Jon Saxton

Since this is a response to a fairly old message, I have quoted the original in its entirety.

I agree with you that the things you mention would justify a 7.0 specification but you are talking about major changes to the structure of TIFF files.

Meanwhile we are faced with commercial vendors who continually generate TIFF files encapsulating JPEG images which are difficult to handle in a standard fashion. At least some of these vendors refuse to comply with draft TTN #2 because it is not part of a published standard.

It has been more than five years since TTN#2 appeared and I feel that it is about time that it were ratified. So instead of jumping to a major new specification, can we not at least publish a minor revision and perhaps call it 6.1 or something like that?

The annoying thing about the refusal of commercial vendors to support compression 7 is that in every case that I have seen, they already have the code to handle it. My sample is admittedly small, but all those vendors support compresssion 6 with tag 513 non-zero to mean exactly the same as compression 7. For them to endorse the newer specification would entail the most insignificant of code changes. Meanwhile I am forced to generate TIFF files in that format just to satisfy third-party applications.

I find myself in the ridiculous situation that in generating TIFF files I have to be sensitive to the application environment. For JPEG-in-TIFF, some viewers only support compression 6, others only support compression 7.

Until we get a published standard, I don't see any end to this. If that means a minor revision then so be it.

[Incidentally, Tim, I have a draft specification for random access TIFF files. I am currently implementing that for a special project which deals with very large TIFFs containing thousands of images where retrieving the 6900th image in a "reasonable time" is just not practical via a linear traversal of the IFDs. Support for random access is via a simple extension to the TIFF format which is completely transparent to extant readers.]

Jon Saxton <js@triton.vg> OS/2 software developer
U.S. agent for Triton Technologies International Ltd
http://www.triton.vg/