1999.06.19 08:06 "TIFF: The Next Generation", by Niles Ritter

1999.06.19 08:06 "TIFF: The Next Generation", by Niles Ritter

Dear TIFF community,

I have just now finished catching up on the last six months of TIFF mailing list following a long (unfunded) absence, and I see that there is a "vergence in the force", as they say in Star Wars I. There are many issues which have clearly become of interest to many people and organizations at the same time, including:

  1. The need for an open, non proprietary, robust interchange format for imagery beyond the 32-bit file size limitation. For more than one organization and imaging application this is becoming a Big Deal with no robust alternatives.
  2. Clarification of existings TIFF standards and "best practices" recommendations, including the main contents of the circa '96 TIFF 7.0 draft spec which still languishes somewhere in the internals of Adobe, Inc.
  3. Expansion of the imaging model beyond offset printing applications, including multispectral remote sensing, medical imaging, digital cartography, astronomy and other scientific applications and interchange of raster data.
  4. Simultaneously, simplification of certain features which have in the current model invited bad implementations via cutting of corners.
  5. Incorporation of new compression technologies
  6. Web-capable (stream) oriented format
  7. Hierarchical multi-scale image database representation, via "supertiling" metadata or similar concept, related to issues in item (1).
  8. The need for an open-source implementation of the new/extended standard, promoting its wider popularization in the imaging software world.
  9. The need for the standard to be public, and absolutely freely available.
  10. Establishment of benchmark test image validation suite for the new standard.

The time is drawing near when the TIFF community shall need to take it upon itself to organize an effort to produce the next generation "TIFF", whether it be called TIFF64, TIF2000, NexTIFF, OpenTIF, or Thelma.

Adobe, Inc, the copyright owner of the TIFF name, tag registry and format specification, has over the years kept the format on a maintenance-only level of support, with little effort to promote or refine the standard. This is not surprising, given TIFF's history as the neglected step-child left over from the absorbtion of Aldus by Adobe, and its collision with Adobe's own widely promoted PDF document format.

The discussions that are ongoing are very important and should continue, with the goal of the formulation of a draft proposal for the next generation image format. By extending TIFF if possible, breaking the standard only if the long term benefits far outweighs the cost of backward compatibility.

If we wish to continue with the TIFF legacy, we shall need to make use of the Adobe tag registry, as any attempt to bypass Adobe in defining new tags would invite mutually incompatible standards, and tolling the death of TIFF as a viable interchange standard.

It is my belief that we should actively invite Adobe.com to participate in the discussions and give them the benefit of the doubt. As a caveat, though, their current note on TIFF tag registry support:

    http://www.adobe.com/supportservice/custsupport/NOTES/2596.htm

directs all inquiries to ada@adobe.com, which does not yield favorable results. If anyone has ties to Adobe internal support, an inquiry into this situation alone would be useful. But I digress...

The GeoTIFF committee did not need Adobe to bring GeoTIFF 1.0 extensions to TIFF to fruition, however, so if push comes to shove we won't need Adobe to get the next generation TIFF going either.

-- Niles