Ethereal-dev: Re: [Ethereal-dev] RE: [OID] Field Information

Note: This archive is from the project's previous web site, ethereal.com. This list is no longer active.

From: Ulf Lamping <ulf.lamping@xxxxxx>
Date: Thu, 11 May 2006 01:28:25 +0200
Graeme Lunt wrote:
I have just checked the change in to provide the "Field Information" context
menu item (18125). I've added the function to oid_resolv.[ch] rather than packet-ber.[ch] as
you suggested.
However this means the URL template is not (currently) configurable as I
didn't know where "OID" preferences should appear. Any suggestions?
For now the URL template is fixed to the www.alvestrand.no site (which
contain links to the elibel site).
Hmmm, I'm a bit disappointed by your current solution as it is very specialized. To quote my previous mail:

"I would think to have a clean mechanism to set a specific help URL corresponding to a protocol tree item is the more important part for now."

Well, your implementation is the opposite:

- "Field Information" suggests a general purpose mechanism, but it's (currently?) limited in effect to the FT_OID fields so for most Ethereal users pretty useless - No one else can add URLs in the way you've implemented it, as it would require to add a new FT_... value (which would be a very bad idea only for this reason) - the "Field" in "Field Information" is just duplicated information -> better: "Related Information"?
I have another change (which I haven't checked in) that introduces a new
field type for URIs (FT_URI) . This allows the new "Field Information" menu
item to then bring up the indicated location. My particular interest is the
URI form of a GeneralName which is used in a number of X.509 certificate
extensions.
But is a URI field type a sensible addition?
I don't see a need for a FT_URI. This would make it impossible to attach a URL to an existing proto_item (e.g. a FT_UINT8). A better idea might be to use a FI_URI, working much like the FI_GENERATED flags. Or am I wrong here and this should really be dealt like the FT_NUM?

So the way you've implemented it, the whole thing is pretty much incomplete.

Regards, ULFL