2002.09.05 11:56 "Re: Decode group4", by Scott Wagner

Jan et. al.,

My two cents - Frank Cringle's codec as implemented in LibTiff is extremely efficient and is well-tested. At this point the codec itself has acquired the mantle of being somewhat of a "reference standard" implementation of G3/G4/Huffman encoding. I think encapsulating it as a C++ class is a great idea, but if you do so (especially if you put it back in the public domain), it would be a good idea to leave the codec engine itself alone and just wrap encapsulation code around it. I have fiddled with this type of codec before - both my own implementations and those of others - and have learned that such implementations are often prone to bizarre and difficult-to-find bugs.

If you use "a C++ class implementation of Frank Cringle's codec," you will be incorporating a lot more than just a few lines of code - the years of testing this codec have gone through are of great value.

/*********************************************************************
* Scott Wagner  (swagner@itek.com)      NOTICE: My personal information,
* Staff Engineer                        including my E-mail address, may
* A.B.Dick Company                      not be sold, distributed, or made
* Rochester, NY USA                     public without my consent.
*********************************************************************/