At 18:41 2001-02-08, you wrote:
> >...and non-binary WML without WML-script (sorry for that
> >sentence...).
> >The unbinaryzation, or whatever, is done by a special
> >proxy/gateway.
> >/gustaf
>
>Unbinarization is not wise to do from a customer perspective as it will
>increase data
>traffic to 300% in some cases. Even on GPRS, with initial average speeds of
>about
>28.8kbit/s the user will still be cost sensitive. Handset manufacturers,
>please
>don't go there! (although it will benefit the network operator on the short
>term).
I do not know where you have found these figures, but they differ anyway a
lot from others.
We have also made practical tests in DoCoMos network and KDDI's (WAP).
The figures we got was that text data was 20% more efficient in binary
mode. But since most sites today consist of text and to a large extent GIF
images (which are not affected by the binary format), the effect is
considerably lower, in the order of 10% as an average.
But this is only part of the picture in terms of the performance. In
practice it has been shown in most operators networks (and KDDI especially)
that the WAP gateway is one of the largest problems for large subscriber bases.
/g
>About binarization (for those that don't know this): at the WAP gateway the
>plain
>text file will be converted to a smaller binary code to decrease the amount
>of traffic
>on the network. Practically, tags of several characters are converted to a 1
>byte
>code.
>
>Best regards,
>Paul
>
>[ Did you check the archives? http://www.appelsiini.net/keitai-l/ ]
--------------------------------------------------------------------------
http://xped.io | gustaf_at_xpedio.com | t +46 8 674 50 47 | m +46 70 916 4747
[ Did you check the archives? http://www.appelsiini.net/keitai-l/ ]
Received on Fri Feb 9 13:03:56 2001