Author Topic: RFCs IDs : why/when?  (Read 428 times)

danfoo

  • Jr. Member
  • **
  • Thank You
  • -Given: 0
  • -Receive: 2
  • Posts: 93
    • View Profile
RFCs IDs : why/when?
« on: September 21, 2017, 03:09:30 AM »
Short question: The examples, documentation and some comments on the forum indicates that RFCs can be sent and implemented with or without a custom ID.
In one post I think Aaron wrote something along the lines of "You don't have to specify an ID, but you can do that if you want to optimise bandwidth".

So, when would you specify an ID and what effects does it have? From a coding POV, skipping IDs are certainly easier as you don't have to keep track of which you have used. So any clarification would be very helpful.
Thanks!

devomage

  • Sr. Member
  • ****
  • Thank You
  • -Given: 7
  • -Receive: 67
  • Posts: 250
    • View Profile
Re: RFCs IDs : why/when?
« Reply #1 on: September 21, 2017, 03:38:37 AM »
an Enum byte is another option to gain readability

ArenMook

  • Administrator
  • Hero Member
  • *****
  • Thank You
  • -Given: 337
  • -Receive: 1171
  • Posts: 22,149
  • Toronto, Canada
    • View Profile
Re: RFCs IDs : why/when?
« Reply #2 on: September 24, 2017, 01:22:23 AM »
The only time I consider using IDs for RFCs are when the RFC can be sent very frequently (>10 times per second). Usually paired with the UDP-based SendQuickly for best effect. Personally I tend to stick to named RFCs. The difference is negligible either way, but named ones are obviously easier to find by name.