Oh no, not this flamewar again...
On the plus side: yes, your development cycle and intermediate release
is easier.
On the min side: You don't have the full API (on Win32 that is).
IMHO when a protocol is complete (stable RFC or whatever standard bodies
publish) and dissector is fully developed (still can have TODO's, but
has significant functionality) and stable (works and doesn't get worked
on extensively), then it should go in.
Then again leaving some plugins in the tree as code samples is fine too.
Thanx,
Jaap
Richard van der Hoff wrote:
Luis EG Ontanon wrote:
Richard,
Once that you are at it why don't you move it to epan/dissectors and
remove the plugin.
I guess I could. The good thing about it being a plugin is that when
you're hacking on it, you don't need to relink the entirety of
libwireshark on each compile cycle.
What's the disadvantage of it being a plugin?