On Oct 2, 2019, at 6:40 PM, Mario Rugiero <mrugi...@gmail.com> wrote:

> El mié., 2 oct. 2019 a las 19:48, Mario Rugiero (<mrugi...@gmail.com>) 
> escribió:
>> I used '1' because that's what Linux does when advertising newer
>> versions of syscalls.
>> '_ext' does look better, I think I'd go with that.
> On the other hand, numeric versioning is more future-proof.
> 'pcap_set_buffer_size_ext_ext_ext wouldn't look good.

Any future extension would probably add another argument - I doubt there'd be a 
reason for a size argument bigger than a size_t - so presumably it's be setting 
something other than just the buffer size, and it'd be 
pcap_set_buffer_size_and_suitcase_size or something such as that, unless you 
could just have separate pcap_set_buffer_size_ext and pcap_set_suitcase_size 
calls.

Either that, or there'd be options or a mode or something such as that, e.g. 
flags such as "if the size is too big, just clamp it at the maximum" or 
something such as that, in which case it'd be pcap_set_buffer_size_with_options.

So I don't expect to have to go with pcap_set_buffer_size_ext_ext_ext or with 
pcap_set_buffer_size2.
_______________________________________________
tcpdump-workers mailing list
tcpdump-workers@lists.tcpdump.org
https://lists.sandelman.ca/mailman/listinfo/tcpdump-workers

Reply via email to