Discussion: MEOS help

I remember that pi said that GVOC were using MEOS. Maybe there is someone who can help me out.

We're going to use MEOS this weekend for our state champs, but I have a question about "zero time". Please read through the following if you have some MEOS experience - the question is at the end.

I believe that zero-time is intended to provide a sanity check so that when reading the 12-hour timestamp from an SI5 card MEOS knew whether it should add on 12 hours (ie any timestamp before the zero-time must have 12 hours added on). But if I change the zero-time (to say 09:00), all the start times in the reports for the event get 9 hours added on. This isn't helpful in the start-lists.

If I set the zero-time to 00:00, then any SI5 users who finishes after noon will have junk times, but the reports look good

Current plan (after some painful testing):
Set time-stamp to 00:00 until day of race (Sunday) in case I need to create new reports.
Set time-stamp to 08:00 on day of race (first start is 09:00).

Question to MEOS users: Does this work, or is there a horrible gotcha that I haven't figured out yet ?

(We will be running Or as backup. I'm doing a head-to-head comparison).

Running latest 2.5.217 version of MEOS. Once the source for 3.* is out, I'll look at a fix for this if its not already done.

Cheers,
Andy

I don't quite understand - aren't you just supposed to set zero-time to one hour before the actual first start time? (I seem to recall some such instruction in meOS). And then assign start times. You seem to be saying that you already have the runners' start times set and now want to change the zero time?

(We will be using meOS for Barebones next weekend)

With all event software I have used so far, you have to set the zero time to some safe margin before the first start. Otherwise you'll face this 12 hour clock issue with any early starters, pre-runners etc. Usually I set it really early, like 3 hours prior to first start. But I set the zero time BEFORE drawing the start list.

Once you have drawn the start list, changing the zero time will shift all start times. It seems like the implementation in these software packages is always that all start times are relative to the zero time. For example, in OE2003 I was never able to find a way to change the zero time without shifting the entire start list. Whether you can do this in MeOS, I don't know... sorry.

Thanks, looks like the problem is the user, not the software.

The start times were assigned externally and imported. To make them "right" I had to set a zero-time of "00:00". I now want to change this zero-time to make sure that SI5 punching works.

I'm happy that shifting the zero time will work as long as I don't print any start lists.

The 12 hour clock issue can be addressed without a zero-time - the logic in SIDResults does a pretty good job. It processes punches in punch order and if it looks like the runner has time travelled backwards, adds on 12 hours.

Andy

If the zero time concept was removed from all o-event software I'll be one of the first to cork up the champagne!

If changing the zero time from 0:00 to 8:00 adds 8 hours to all start times, won't this mess up the results as well as the start lists?

I haven't yet used meOS, but I suspect that if you did what you're describing, meOS would calculate a split time of about -7.9 hours to the first control for each runner.

Have anyone actually tried this?

Well, we tried it today and it worked fine - the time offset derived from the zero time is applied to every single punch, so the results are all good.

I think that most runners enjoyed some of the new functionality - the split analysis for example.

Funnily enough, even after all this worrying about it in MeOS, I left our backup system (running Or) with zero-time set to 00:00, so that ended up borked. IMHO its a horrible concept, if you can avoid it with a smart algorithm.

If the zero time concept was removed from all o-event software I'll be one of the first to cork up the champagne!

Somehow they have been brainwashed into declaring it necessary. I don't know how to explain it to them. They don't appear to want to listen.

You can fix it yourself in MeOS, its open source C++.

I had to make a couple of small changes already (one to show course on the by minute start list, one so that I could import splits to AP).
(source is at https://www.assembla.com/code/meos-oz/subversion/n... - I don't think I committed the splits XML change yet).

I think all that you need to do is hide the existing zero-time window (so that its always 00:00:00), then alter the SI5 processing algorithm so that it uses the logic from SIDResults mentioned above. That would banish zero-time for users, without touching the source code too much.