6 comments

[ 3.9 ms ] story [ 19.8 ms ] thread
File format cannot be open sourced. Tool for working with the format can, or the format can have published specification. Format itself doesn't have source code.
To be clear, what was "open sourced" was a C header file containing struct definitions and parsing functions for that file format.

An implementation could be considered one way of providing a format/protocol specification.

> "open sourced" was a C header file [...]

Indeed. In my opinion too little to call it a day, but I haven't worked on implementing file handling, so I may be wrong here.

> An implementation could be considered one way of providing a format/protocol specification.

Not quite. Look at OpenVPN, which has hilarious situation: code is open, but protocol is effectively closed. You can't practically derive the protocol from OpenVPN's source (yes, I tried that). Unless you can put weeks of effort, of course.

So it's a closed protocol because it's hard to understand? That doesn't seem to make much sense.
No. It's closed because there's no specification or even an alternative implementation and you can't easily derive one.
This brings back some fond memories. One of the first interesting projects I worked on professionally was a post-mortem debugger that ran under OS/2. We used Microsoft's compilers and it was a bit of a challenge to reverse engineer enough of the debugging symbols to make something useful. Definitely could have used this back then :)