Looks like .47 fixed a lot of outstanding bugs/issues. The polling interval change (in conjunction with the Ship Delay=None change a few revisions back) will cause balance problems however.

In about revision .40 Ship Delay = None went from truly 0 delay to 250ms.

This divded ppl into two classes:
SLOW: ping > 250ms
FAST: ping < 250ms

It was not a major problem at the time, as the polling interval was 1s anyway. In .47 the polling interval was reduced to 250ms as well; this enforces the divide in connection speeds.

We now have the situation that a FAST connection can photon block any area of space (bubble or dock) with a photon script. A SLOW connection cannot.

For example, stardock is surrounded by defensive figs. Suppose stardock is in 1627, and 3245 is next door.

Under versions before about .40 (ship delay=none) , I could macro "3245^MA100^M1627^M", and get to stardock/ their bubble without wearing a photon, regardless of who was running their photon blocker script.

Under .46 I could as well as the polling delay meant my enemy wouldn't know I'd hit his fig until I'd finished my move delay.

Under .47 with the faster message polling the FAST photon scripter may find I've entered the sector and photon me before I've finished waiting the move delay. (Yes, I did just test this under .47)

The .47 behaviour per se is not the problem- this was always the case for ship delay=half/normal games anyway. The problem is that someone who has a FAST connection can photon anyone inside the move delay, and block off arbitary parts of space this way. Someone with a SLOW connection can never get someone in the move delay.

This polling time setting should be entitled "Can modem/international users kill Toll figs without dying Yes/No", No being the default.

250ms is about the worst possible choice for these settings. If it was 50ms, then everyone would be "SLOW" so the playing field would be level. If it was 500ms then everyone would be "FAST" so the playing field would be level. 250ms is right smack in the middle of ppls latencies, and therefore a bad choice.

Whilst sysops can change the poll interval back to 1s, the default is 250ms, _and_ users cannot see what the setting is by pressing *.

Dr. Bad