[Ant-l] (no subject)

[Libor Forst]

Hi all,

I would like to inform you about a new ANT release with a new approach
to reading chips having issues (cleared with bad race, having punches
from misconfigured units etc.). I decided to do it after my experience
from several last races where these problems occured.
The marshal in the finish is often not experienced enough and so I made
just several crucial checks what is wrong with a chip and what is the
marshal's decision how to continue and the app remembers the decisions.
And - warns the marshal to keep incompletely read chips separately to
be processed later.

The checks and decisions that are saved utnil the marshal leaves the
chip reading module in ANT are:
- When the first chip with incorrect race number is read, the marshal is
  asked whether their ANT is configured for the proper race.
- When the first chipset with incorrect race number is read, the marshal
  is asked whether incorrect chipsets should be read or skipped. It the
  decision is SKIP, ANT writes a warning about keeping the chipset aside
  for further later processing. This is probably the recommended answer
  for inexperienced marshals. However, if this warning is issued for too
  many chipsets, the marshal should contact someone who could check the
  situation. For instance - if the CLEAR unit was accidentaly configured
  for a bad race, all chipsets will cause the alarm. In such case, the
  desicion should be ACCEPT and all chipsets should be read despite the
  race number problem.
- If incorrectly cleared chips are set to be processed, the marshal is
  asked whether only records tagged by punching units as suspicious
  (due to bad clearing) which are probably exactly the ones from the
  current race should be read, and the untagged (probably old) should
  be skipped. The recommended answer is again SKIP. The answer ACCEPT
  may be the good answer for an experienced user in the EC.
- When a chipset contains only some chips with a bad race number (what
  usually means that some chips were not cleared), the marshal can make
  a decision whether these chips should be skipped or read. Probably,
  the right answer is again SKIP.
  
All skipped chips should be then read out later in the EC with an
access to the result system so that they could be imported to the
system and immediately fixed if needed.

If the problem was caused by a misconfigured ToePuch unit, the next
step depends on the result system. In TOP, I made a few new features
to help admin to solve these issues.

BR L