EMIT bug?

[Duncan]

I was at Interland at the w/e and saw some interesting EMIT behavior. When downloading it was being reported that some of the controls some people had visited had codes that were a power of two different to what they had actually visited. E.g. the control's real code is 67 but on downloading the EMIT card says they visited 195. Only happened for a few people, and for them only for one or two controls. In all cases it was obviously a bug and they were reinstated because the control they had allegedly visited either wasn't out in the forest or was miles away, but at other events that wouldn't always be the case. It is symptomatic of getting one binary bit wrong when transmitting or storing or reading the control code.

Anyone ever seen this behavior before?

[Spookster]

Anyone ever seen this behavior before?

Yes, though with SI not Emit. In fact it's happened a few times over many years, but I've only recently found out why.

Sue Birkinshaw finished the Beeley event a month ago with a mis-punch. Apparently she had punched control 45, when she should have been to control 91. Control 45 wasn't even in the forest, and nor were there any other controls near control 91, so it was clearly a punching issue. I re-instated her, because I believed she had punched at the right control, and this wasn't a championship race.

A few days later she worked out what had happened.
91 = 1011011 in binary
45 = 101101 in binary, which is very like 91(1011011), but with one binary digit missing.

She reckoned her e-card is old, and had failed to properly record the punch.

SI includes some kind of verification process when you punch (it doesn't bleep and flash until the data is verified as being recorded on the card). Maybe in this case the right data was written to the card, but by the time she downloaded it had "lost" a bit.

[graeme]

and this wasn't a championship race.

...what would you have done if it was a championship race?

[Red Adder]

The earlier epxlanation dooesn't work as neatly on this

67 is 0100 0011 in binary

195 is 1011 1111

Isn't technology wonderful.

[Spookster]

and this wasn't a championship race.

...what would you have done if it was a championship race?

Applied rule 2.6.5 of Appedix I differently.

The Organiser and Controller may consider disqualification to be harsh in cases where the competitor clearly believes they have visited the control, particularly at low key events or for junior competitors. For consistency it is preferable to enforce disqualification at all events. It shall certainly be adopted for Level 1 events.

It being a Level 2 non-championship event, I "preferred" not to enforce disqualification.

[Snail]

Red Adder wrote:

67 is 0100 0011 in binary
195 is 1011 1111

You must work in a different version of binary to the rest of us!

195 is 1100 0011 so only one digit different to 67, as Duncan surmised

[Angry Haggis]

I ran into this a few years ago with SI - main difference was the last bit disappeared rather than the first.

[Nottinghamshire outlaw]

I was the 'victim' of strange emit behaviour at the Interland match (and strange route choice behaviour too but can't blame that on emit!!) - my two first splits were way too long and short respectively (but total for first 2 was OK) and my run in was 2 mins 29 instead of 21 secs!! The controller managed to print it out again with the correct splits and I gained 2 .08 - even with that it was still a rubbish time (need more technical practice!) The Controller said that their computer seemed to have problems reading the newer brikkes (the scrolling variety).

[distracted]

Was it with the newer cards with the screens? They are far more unreliable than the old style cards - be glad it didn't reset itself halfway round your course...

[NeilC]

Card failure would present the organiser/controller/jury with a difficult decision. If a competitor lost or smashed their card against a rock during a race, and as a result could not prove that they had completed the course properly, then it seems reasonable that the competitor is at fault. If however "bit loss" is seen then is it really fair to blame the competitor? However the organisers are also not to blame (unless they provided the card). I would like to think that a jury would consider a "bit loss" case sympathetically - so perhaps BOF need to publish a handy list of the binary codes for all the commonly used control codes.

[Spookster]

If however "bit loss" is seen then is it really fair to blame the competitor?

There doesn't have to be blame. Sport includes an element of luck, and getting disqualified because your e-card didn't function correctly would just be bad luck!

I think most results computing people could work out the binary equivalent of decimal 31 to 255 if they needed to.

[AndyC]

Trouble could be that planners often end up with similar control codes physically near each other and a reading of a zero rather than a one in the last place (not quite what is being reported) or vice-versa could move you from say 121 to 120. The rules do suggest that you shouldn't put easily confused codes on similar features in the same area but this would make the logistics impossible if you have to read binary as well.
Sensible analysis by officials in each case is required it seems

[seabird]

Not knowledgeable enough about the issues being discussed here to be sure this is relevant - but using sI at the Clydach Middle Distance Race in 2009 I managed to run an entire course of non-existent controls, according to download.

We never did untangle how this had come about.

[RJ]

SPORTident continuously update their firmware for the control boxes, with new versions being made available from time to time. Little problems like that would soon be sorted out with SI. For the last 100+ events that we have run we have had no problems with ecards or control boxes..... just the one that had a failing battery that prompted a more careful check on battery life in all the control boxes!

The current 'thing' to watch out for is the ecard with a number >1,000,000!!!!!!

[NeilC]

I updated the firmware in all our BSF7/8 units last year just so that I could use my v9 (>1000000) SI Card at our summer sprint events rather than my old and slow v5.