AWARE SYSTEMS
TIFF and LibTiff Mail List Archive

Thread

2004.12.01 14:55 "[Tiff] Version 3.7.1 Availability", by Bill Bither
2004.12.01 15:26 "Re: [Tiff] Version 3.7.1 Availability", by Joris
2004.12.01 16:04 "Re: [Tiff] Version 3.7.1 Availability", by Bob Friesenhahn
2004.12.01 19:06 "Re: [Tiff] Version 3.7.1 Availability", by Jeff Breidenbach
2004.12.20 20:16 "RE: [Tiff] multithreading issues", by Bill Bither
2004.12.01 19:06 "Re: [Tiff] Version 3.7.1 Availability", by Andrey Kiselev
[...]

2004.12.01 16:04 "Re: [Tiff] Version 3.7.1 Availability", by Bob Friesenhahn

Is there a timeframe for the release of version 3.7.1? There are some

I don't know this answer. I agree that the 3.7.0 release does include some significant flaws. It was the first autoconf-based release and there were many other changes (particularly related to tags), so the existence of flaws was not unexpected.

This leads me to another question. What is the policy for testing LibTIFF? Are there unit tests that are run frequently? Is testing in this manner been a philosophy for a long time, or is it relatively new? I do recall discussion a while ago about a test suite using checksum validation when encoding and decoding images. That's a procedure we also use in-house for our unit testing and is certainly quite valuable to us.

Libtiff is starting to develop a test suite, but developing good test suites is very hard work. Up to now, it seems that most formal testing has been done by other packages which depend on libtiff. For example, GraphicsMagick includes many TIFF tests in its test suite, but it is worth noting that a large amount of the code tested is actually GraphicsMagick code rather than libtiff code.

Bob
======================================
Bob Friesenhahn
bfriesen@simple.dallas.tx.us
http://www.simplesystems.org/users/bfriesen