On 27/04/15 22:23, Peter Law wrote:
> I'd also like to add some general comments about where I think things
> went particularly well this year.
Here are a few comments from me, some about where things went well, and
some about where they didn't. I'll get round to ticketifying them when
I'm not on fire with TYP stuff; I'd appreciate it if someone could
remind me if I haven't done it by this time next week, as I might forget.
Because it's much easier to point out where things went wrong than to
make general statements on where things went well, let me first say that
I agree with Peter's sentiments that this was probably the competition
where everything went the smoothest. I also agree with Peter's points,
so I won't repeat them here. Most of these things are just nitpicks,
albeit ones which, if fixed for next year, would still improve things
somewhat.
* The network worked. Admittedly, the WiFi mostly didn't appear to be
able to cope with the loads (though I'm not sure whether or not this was
the primary issue with it — I was getting weird things like duplicate
ping packets (!) when attempting to use the WiFi even with no
competitors). But since all teams had an ethernet port, this didn't
matter nearly as much as it did last year (when the WiFi also didn't
work but teams' only alternative was tethering), and so everything
network-related went reasonably well. I would suggest the main cause of
this is the hiring of kit from an actual hire company, rather than using
SR's mostly clapped out hand-me-down switches and cables some of which
had serious issues like lack of autonegotiation. I would certainly say
we should continue with providing Ethernet next year. If we can get the
WiFi working too, that's a significant advantage. I know Phil from the
wireless society (who was working the live streaming with BATC this
year) suggested to me he'd quite enjoy working with the WiFi himself — I
suspect it would also help to have a dedicated person doing just the
WiFi, to reduce the load on Jeremy (though this year I get the
impression that it wasn't nearly as bad as it has been in previous
years; please correct me if I'm wrong, Jeremy).
* Following on from this, the live stream worked really well. Thanks a
lot to Phil/BATC for doing this. It was higher quality than last year
thanks in part to the use mostly of software rather than hardware video
equipment. The overlay didn't happen, though speaking to Phil in the pub
on Saturday night it did seem like the sort of thing that would still be
potentially feasible with a software (ffmpeg) solution and chroma
keying. We need to remember that we can use the control room on level 2
next year to have the live stream on the venue monitors upstairs, so
that we can plan for an ethernet cable to here when we're doing the
network rather than having to lay it after competitors have already
arrived, and perhaps to set up a dedicated laptop for it on the Friday
so we don't need to rush around on Saturday and Sunday morning with
Blueshirts' personal laptops trying to work out why Windows won't boot.
The page to display the two streams side-by-side worked well on the
second day after a bit of wrangling to make it work in Firefox.
* No spare kit on helpdesk, or anywhere in the venue. This was an issue
when one team apparently had to borrow a Ruggeduino from another team
who weren't using theirs. This would have been more of a problem had it
been a motor or power board that failed!
* "No tactical switch-offs" was a very good rule, but I think there
needs to me some clear definition (even if just internal, as opposed to
something we give to teams) of when switch-offs to prevent robots
damaging themselves is allowed. I know the linesmen (including me) plus
Sam had a few deliberations during the knockout on whether or not to
honour a team's switch-off request. In particular, one of the requests
was only resolved when the robot started smoking, at which point it's
probably too late (fortunately this didn't cause any issues as the team
was knocked out anyway).
* Team photos happened well! Every single team had a team photo taken.
Organising to send up teams after league matches somewhat worked, but
wasn't sufficient (and didn't appear to catch all teams, as some seemed
to have no idea they were going on well after the instruction was given)
— the idea of Tyler's to appoint me as a photo shepherd when I had no
other jobs was a good one, as I was able to liaise with him to get the
list of teams who hadn't yet had a photo, and confront them individually
and agree a time when they would go (for the teams who were reluctant to
do so). This should definitely be repeated in the future.
* Blueshirt group photo didn't happen, which is a bit of a shame. This
should happen in the future.
* Radios (ignoring those for shepherding since I wasn't part of that)
worked perhaps a little worse than in the past. People often seemed to
be away from them, and I still think some of the people who could have
usefully had radios didn't. Perhaps a small briefing on radio protocol
and how the radios should be kept (ie within earshot of at least one
person, and on full volume so they can hear them over things that might
be going on) would be useful, as I believe has happened in the past. The
fact that we had six (!) licensed radio amateurs to pass along messages
using handhelds also helped avoid going up and down stairs.
* Scheduling worked quite well, although it turned out that my schedule
was somehow messed up and I only got given the correct one on the day of
the competition when I turned up in the wrong place! The shifts were
about the right length this year (longer than SR2013 but shorter than
SR2014, unless I'm much mistaken) with enough breaks. It generally felt
like we had about the right number of blueshirts — whether this was an
illusion due to the good scheduling and general lack of major problems
this competition, or whether we actually had more blueshirts than last
year, I have no idea.
* Robots not going up and down the stairs was not strongly enforced this
year. I know there was an accident on the stairs (which was handled
well); I have no idea whether these two facts were related or not. In
either case, next year I would recommend having a person on the ground
and second floors whose sole job it is to redirect people who are
carrying robots away from the stairs. I believe there was a lift
shepherd who somewhat filled this role in 2014, but correct me if I'm
wrong — how was it done then? This was particularly an issue after the
group photo, but I don't really know how this could be sanely resolved
without a teleporter.
* MY OLD SCHOOL WON WOOOOOOOOOOOOO
Murray.