Re: beta1

[Guy Harris <guy () alum mit edu>]

On Nov 13, 2003, at 1:43 PM, Neil Spring wrote:

> 1) any chance of removing unused rcsid variables?  grep suggests they 
> are always defined but never used.  Alternately, move it below the 
> include "interface.h" and edit it to say:
> static const char rcsid[] _U_ =

The latter sounds good; that way, you still get the RCS IDs in the 
image, but you don't get the whinging from the compiler.

> 2)
> print-pppoe.c: In function `pppoe_print':
> print-pppoe.c:199: warning: `0' flag ignored with precision and `%X' 
> printf format

I *think* they wanted "%02X" - they print a leading "0x", and then each 
octet, so they presumably want each octet printed as 2 hex digits.

> 3) listening on en0, link-type 1 (EN10MB), capture size 96 bytes
>  seems likely to cause some faqs, if it isn't one already -- I have a 
> 100Mbit link and it says 10MB?

That dates back from when 10 megabits was the *faster* version of 
Ethernet; "EN10MB" distinguished it from the old experimental 3MB 
Ethernet, which I think had a different link-layer header (8-bit, 
rather than 48-bit, addresses).

We could add DLT_ETHERNET as an alias, and use that instead; however, 
applications probably shouldn't start using DLT_ETHERNET, as, if they 
did so, they'd only build with versions of libpcap that have it.

Similar problems exist with DLT_IEEE802, which I suspect was introduced 
because the DLT_ values were originally ARP hardware types, and for 
some reason there was an ARP type for "IEEE 802 networks" - which 
doesn't refer to a particular type of network, so it's somewhat 
meaningless as a DLT_ value, as those values have to specify the 
link-layer header type.  It was subsequently used to mean 802.5 Token 
Ring.

Perhaps we should change the "dlt_choices[]" table to have two names, 
one of which is the current name (for compatibility) and one of which 
is a more descriptive name.  "pcap_datalink_val_to_name()" should 
return the descriptive name; "pcap_datalink_name_to_val()" should 
accept either of the names.

-
This is the TCPDUMP workers list. It is archived at
http://www.tcpdump.org/lists/workers/index.html
To unsubscribe use mailto:tcpdump-workers-request () tcpdump org?body=unsubscribe