Hi Joerg and others,
Ronnies mail reminded me of something: Shouldn't the dissectors be
moved into epan/packet/... and the smb family of dissectors into
epan/packet/smb/...? I know (well, actually I've read about it in
several mailinglists) that moving files around in CVS isn't fun but
how about starting this at least with new dissectors or those that
are as good as new?
I agree. Soon we will have every dissector tvbuffified and all old
traces of old-style dissectors removed.
Perhaps we should decide on what to do next, moving all dissectors
one by one into epan/packet sounds good to me.
I would propose to audit and move the dissectors one by one into
epan/packet/... but then only if the dissector is
* fully tvbuffified
* reasonably cleanly written and easily maintainable.
* ALL fields for the protocol has display filters defined
and that all display filters have the blurb field explaining the
use of this data properly defined.
* (something I would like) the dissector does not do anything funny
or nonportable as using memcpy() to a struct instead of properly
reading all data with tvb_get_x() accessors.
* All preference options which controls behaviour of the dissector is
documented in the manual and man-page.
* (why not?) If the dissector is described by an RFC or similar open
document make sure that a pointer to that document is specified in
a comment in the beginning of the dissector.
Make sure we also specify pointers to (if any) relevant ietf
workinggroups and iana documents.
Perhaps also, if possible, supply the specification in a separate
directory.
Perhaps someone can produce a list of what we would want of the
audited/fixed dissectors so we know what and how to update them?
best regards
ronnie s
_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp
