Re: crypto flag?
Johannes Berg <[email protected]> Wed, 07 Dec 2011 22:07:49 +0100
| Newsgroups | org.netbsd.radiotap |
|---|---|
| Message-ID | <[email protected]> |
On Wed, 2011-12-07 at 11:47 -0800, Guy Harris wrote: > On Dec 7, 2011, at 10:50 AM, Johannes Berg wrote: > > > What would you think about a radiotap flag that tells wireshark what > > happened in the packet? > > Other than saying "that tells the program reading the file what > happened in the packet" - this shouldn't be thought of as > Wireshark-specific - a flag of that sort makes sense; if there's some > information that is necessary or every very helpful when processing a > file, and the program that generates the file knows that information, > the ideal is to have that information stored in the file somewhere > that allows programs reading the file to get it. Yes, you're right, it's not wireshark specific. I think that we know this information when generating, and if not we can leave it out. I think it'd have to be something like name: crypto flags bit number: xxx structure: u8 unit: bitmap 0x01: IV present 0x02: frame encrypted ... come to think of it, we could record a bit more info about crypto I guess, at least in the case where we actually have keys (which happens if you record on the same machine that is doing something) I'll think about this a bit more to see if there's something else we can record. johannes