On 6/14/07, Myrna van Lunteren <[EMAIL PROTECTED]> wrote:
If I dig through the comments in DERBY-1377, it appears Andrew actually handled the ij/.sql scripts. It doesn't explain why the *subsql, *sql1 and sql2 were skipped. Maybe because they don't work by themselves? Or oversight?
I don't think the tool complained about filetypes that it didn't know about. I see that situation has improved. This goes into the oversight bin. It just wasn't on the radar. Some of those .subsql definitely qualify as creative works, like tests/store/updatesetlocks,subsql, others don't (in my eye), like tests/lang/createTestProcedures.subsql
Also, there are explanatoryfiles for instance, README.txt - does that need a header?
Yes, these qualify. We gave all the ones that existed at the time of last release the header. If there are new ones that don't have it, they need it.
How about all the .properties files in the old test harness directories, was it decided those were not 'source' files in their own right?
In most cases, they don't qualify. There are some that do qualify, though, like metadata.properties. Comments are safely ignored in properties files, so the header should be added to all the properties files to be safe.
How about .policy?
Same as with properties, most won't qualify. Some, like derby_tests.policy will. Probably best to update them all and be covered. HTH, andrew
