[Freeciv-Dev] Re: (PR#2316) Reduce bandwith take 3

["Raimar Falke via RT" <rt@xxxxxxxxxxxxxx>]

On Fri, Nov 15, 2002 at 10:10:59AM -0800, Per I. Mathisen via RT wrote:
> 
> On Fri, 15 Nov 2002, Raimar Falke via RT wrote:
> > Final overview:
> 
> Ok, so discarding packets does yield a good deal of network savings, after
> all.
> 
> > Share of the chat messages at the total traffic:
> >
> > 2:     24   1675       69 PACKET_CHAT_MSG(13)
> 
> In other words, if we (finally, IMHO) went for a binary protocol for these
> messages, as we have discussed repeatedly over the years, we could shave
> another big chunk off the network traffic. And also get rid of the related
> i18n problems. I also think that it is the right thing to do from a design
> viewpoint - creating a user-readable response from an event is the
> responsibility of the client.

Yes. But. You have to provide a key for the client to retrieve the
translated format string. The last patch which addressed the problem
took the untranslated English format string as a key. This has the
advantages that if the client doesn't know about the string (and so no
translated string is available) it can print the English message. From
the bandwidth point of view however a simple 32bit key would be
nice. But this would create extra problems (assign keys to format
strings, if the client doesn't know the key it can't print andthing,
... ) and so it isn't worth the effort. You can solve this by same
extra code (client can retrieve the English format string for a given
key if the client doesn't know the key and so on ...) but this is just
overkill. It is easier to just compress the data. Even if we only do
compress chat packets.

        Raimar

-- 
 email: rf13@xxxxxxxxxxxxxxxxx
  "With a PC, I always felt limited by the software available.
   On Unix, I am limited by my knowledge."
    -- Peter J. Schoenster <pschon@xxxxxxxxxxxxxxxxx>