* Manually updating the licenses guarantees that it is a appropriate to update the license.
For the directories we are running it on, this appropriateness is determined in advance.
* Due to the wide variety of comment formats, you would have to guess the format based on the extension (which itself isn't always accurate).
No, you can use regular expressions to determine the comment format.
* There are actual comments in many files that cannot be tampered with.
You underestimate the power of regexps.
* In many files, there is a comment before the license that gives settings to the text editor such as how many spaces are in a tab, etc.
You underestimate the power of regexps.
Once I get my CVS tree access, I can update the licenses manually.
There are approximately 32,000 files in the Mozilla tree (find . | grep -v CVS | grep -v ".obj" | grep -v "dist/"). Doing them all by hand would be an extremely long and tedious job.
This has already been done fairly successfully in the past for a large number of files. The current Python relicenser copes with most comment formats, and fails safely when it can't cope. However, the author is somewhat busy and we need something that can be actively maintained and improved.
Gerv
