AntLog - double elimination result logging tool

Place discussions about upcoming events here in this thread.

Moderators: petec, administrator, BeligerAnt

User avatar
Remote-Controlled Dave
Posts: 3716
Joined: Sun Apr 03, 2005 5:30 pm
Location: Antrim, Northern Ireland
Contact:

Re: AntLog - double elimination result logging tool

Postby Remote-Controlled Dave » Sun Mar 29, 2015 6:24 pm

Talking to people at the AWS, I don't think there would be much objection to reducing entries from 4 per team to 3 as long as it was absolutely necessary, and everyone had to do it. Its not as fun but most I mentioned it to said it would be an ok solution if it came to it.
Die Gracefully Robotics
Winner - AWS 39

User avatar
EpicentrE
Posts: 1075
Joined: Mon Jun 09, 2003 12:00 am
Location: Coventry
Contact:

Re: AntLog - double elimination result logging tool

Postby EpicentrE » Sun Mar 29, 2015 6:48 pm

I was discussing this with Andy on the way back; I too wouldn't object to reducing entries to 3, as long as we didn't make the third have to be a walker or a cluster. I think walkers and clusters have their own advantages nowadays so that they don't need to be incentivised.

In fact, for all events, I'd be happier having 3 antweights per team, but an arena which produced longer fights, rather than having more robots having very short fights.
Scott Fyfe-Jamieson, Captain of Epic Robotics. Champion of AWS38/41/42.
http://www.epicrobotics.co.uk

User avatar
BeligerAnt
Posts: 2431
Joined: Wed May 15, 2002 12:00 am
Location: Brighton
Contact:

Re: AntLog - double elimination result logging tool

Postby BeligerAnt » Sun Mar 29, 2015 7:29 pm

Without wanting this thread to drift further into a "how to run the AWS" debate...

Alex, we tried running two arenas several times in the past and it has never worked sucessfully.

The original 4 robot limit was introduced in the days when walkers and clusters were harder to build and consequently rare so few people actually entered what is now considered a "full" team. I don't see a problem with reducing the limit to 3 on a case-by-case basis.

I suggest we now get this thread back to a discussion of AntLog itself.
How should it be improved?
What are its failings that really need to be addressed?
What enhancements would be nice but not essential?
Does anyone want to help with AntLog3?

My plans for Antlog3 include coding it in PHP and hosting the project on GitHub to make collaboration simpler.
It will run on a webserver with a MySQL database to store the data. It should be able to provide both an admin back-end to set up the event and log results and a user front-end for competitors to monitor their own progress.
This is a big project with no clear timetable attached yet...
Gary, Team BeligerAnt

User avatar
joey_picus
Posts: 1132
Joined: Tue Jan 13, 2009 1:51 pm
Location: Lancaster, Lancashire
Contact:

Re: AntLog - double elimination result logging tool

Postby joey_picus » Sun Mar 29, 2015 8:15 pm

My coding skills are near-nonexistent but I do own OpenOffice (.org) on at least two computers so I hereby volunteer for testing duties :) and if there's anything else I can do - I have some slight experience with server maintenance - I'm more than willing to try. One thing I would like in that future incarnation, if not planned already, would be some way of searching the upcoming fights and/or groups for your own robots - I have no idea how large a job that is though...

I don't know if there's any major failings of the current setup of AntLog, but my experiences yesterday did highlight a potential problem with how hard it is to fix an incorrect result entry - or even to spot it - which is something that will inevitably happen as humans are involved. (no offence to the humans who were involved who put everything right, it happens!) I had a couple of words about that with Gary already and it's a hard thing to really deal with, short of having two separate people entering the results which is a little unwieldy to say the least, so does anyone more intelligent than me have any ideas?
Joey McConnell-Farber - Team Picus Telerobotics - http://picus.org.uk/ - @joey_picus
"These dreams go on when I close my eyes...every second of the night, I live another life"

User avatar
Remote-Controlled Dave
Posts: 3716
Joined: Sun Apr 03, 2005 5:30 pm
Location: Antrim, Northern Ireland
Contact:

Re: AntLog - double elimination result logging tool

Postby Remote-Controlled Dave » Sun Mar 29, 2015 9:48 pm

We did encounter bugs with AntLog that me and Mark had to put right but he'd be much better at explaining them.

And without meaning to drift off topic again, with all due respect Gary, I can only think of 3 AWSs where we used two arenas and it always worked completely fine.
Die Gracefully Robotics
Winner - AWS 39

User avatar
BeligerAnt
Posts: 2431
Joined: Wed May 15, 2002 12:00 am
Location: Brighton
Contact:

Re: AntLog - double elimination result logging tool

Postby BeligerAnt » Fri Apr 17, 2015 11:07 pm

Well, after a few weeks of brushing up my PHP and SQL skills and getting to grips with the Yii framework, I have a working "first prototype" of AntLog 3.0 :)

This is much sooner than I expected as using the Yii framework made for much less coding than I had anticipated.

screenshot.jpg
AntLog 3.0
screenshot.jpg (57.52 KiB) Viewed 1881 times


It still needs quite a bit of work to get to where I want it but so far you can:
[*]Create a team
[*]Create robots (must belong to a team and weight class)
[*]Create events for a particular weight class
[*]Add entrants to an event (8 to 128 in 2, 4 or 8 groups)
[*]Run all the fights of an event, automatically handling byes

The "fights" view shows all the possible upcoming fights (not just 4). In the final stages you still lose visibility of upcoming fights because of the dependencies. I might look at ways to improve this but I think there are better things to look at.

The "entrants" view shows each entrant's robot name, team name and number of "lives" left - starts at 2 then goes down to 1 then 0 as you lose fights. The table can be sorted by robot and number of lives, and filtered by team and event so it's easy to see the status of all the robots in a team.

The "event" view shows all the robots signed up for an event. The "robots" view also shows which robots are signed up for the current event. Most of the views allow filtering and sorting, but some of it requires more work.

The biggest improvement over AntLog 2 is in the visibility of fights. Since AntLog 3.0 is a web application running on a web server it can support multiple clients viewing the database at a time. If the venue has a wireless network available anyone can view the data on a laptop, tablet or smartphone. I need to work on simply setting up a private network for venues that don't provide wi-fi.
Although AntLog 3.0 is a web application (makes use of web technologies) it does not require an internet connection. It runs happily on a laptop, and I believe a Raspberry Pi would be able to run it as long as it wasn't doing too much else!

There is still much to do.
Figure out the login and access restriction features in Yii to restrict the entry of fight results to the event organiser.
Now that the core logic is working, improve the user interface as it's very clunky in places.
Much more testing as I've only run one small event through it so far!
Detect the last fight in an event and actually announce that there is a winner. [DONE :) ]
Add a lot more error checking, input validation etc...
Add a facility to upload teams and robots from the AWS signup form.
Refactor some of the code where it is badly hacked together :oops:
Sort out a GitHub repository to share the code.
... and probably lots more...
Gary, Team BeligerAnt

User avatar
peterwaller
Posts: 3476
Joined: Fri Feb 15, 2002 12:00 am
Location: Aylesbury Bucks
Contact:

Re: AntLog - double elimination result logging tool

Postby peterwaller » Sat Apr 18, 2015 10:39 am

Sounds and looks rather impressive I must say. :o

User avatar
joey_picus
Posts: 1132
Joined: Tue Jan 13, 2009 1:51 pm
Location: Lancaster, Lancashire
Contact:

Re: AntLog - double elimination result logging tool

Postby joey_picus » Sat Apr 18, 2015 12:21 pm

Looks very nicely done, good work :) I presume accessing the results is reasonably simple from a browser on a Wi-Fi enabled device, to ask the blindingly obvious? (maybe I'll have to finally upgrade my phone...)
Joey McConnell-Farber - Team Picus Telerobotics - http://picus.org.uk/ - @joey_picus
"These dreams go on when I close my eyes...every second of the night, I live another life"

User avatar
BeligerAnt
Posts: 2431
Joined: Wed May 15, 2002 12:00 am
Location: Brighton
Contact:

Re: AntLog - double elimination result logging tool

Postby BeligerAnt » Sat Apr 18, 2015 3:13 pm

Viewing the results should be easy (at the event). Viewing the results after the event relies on the event organiser uploading them to a website somewhere...

(Something else to work on ;) )

Currently, from the home page, select Events, view the completed event you are interested in and click the Results button:
results.jpg
Results view
results.jpg (50.18 KiB) Viewed 1856 times


This is something that needs streamlining, and positions need extending beyond the first 8 places.
However, limiting access to certain functions by organiser login appears reassuringly simple :)
Gary, Team BeligerAnt

User avatar
Rapidrory
Posts: 1140
Joined: Sun Nov 24, 2013 9:54 pm

Re: AntLog - double elimination result logging tool

Postby Rapidrory » Sun Apr 19, 2015 7:30 am

Excellent!

Is the 128 entrant limit an inherent limit or just a user defined one? It would be hard to run an event with more than 128 entrants, but it'd be even harder if the software doesn't support it :P We worked out as load of contingency plans for the Reading AWS just incase we went over antlog's limit, but it would be nice not to have to worry about it if it's a simple change.
Rory Mangles - Team Mangler

Robots: Too Many...

NanoTwo antweight speed controllers. (viewtopic.php?f=1&t=2128)

User avatar
BeligerAnt
Posts: 2431
Joined: Wed May 15, 2002 12:00 am
Location: Brighton
Contact:

Re: AntLog - double elimination result logging tool

Postby BeligerAnt » Sun Apr 19, 2015 5:20 pm

I thought about increasing the number of robots, but to be honest the limitation is not the software but the time available at an event!

Even with the smoothest run event, fitting more than 254 fights into one day would be a challenge. That would be about 50 more fights than at the last AWS! I would rather see the number of robots per team reduced to remain within the 128 robot limit than end up with an event that cannot easily be run in a single day.

As soon as the event spills over to 2 days the running costs double, most people's entry costs increase significantly so the number of entrants drops significantly - making a bigger field and a 2 day event unnecessary...

It is quite possible (although quite a lot of work) to expand the double elimination table to 256 robots, as 8 groups of 32 or as 16 groups of 16. When the codeis in a usable state it will be hosted on GitHub as an open source project so anyone can modify it to support more entries or whatever...
Gary, Team BeligerAnt

User avatar
Remote-Controlled Dave
Posts: 3716
Joined: Sun Apr 03, 2005 5:30 pm
Location: Antrim, Northern Ireland
Contact:

Re: AntLog - double elimination result logging tool

Postby Remote-Controlled Dave » Mon Apr 20, 2015 10:28 am

Could you not have been more creative with your example names? :-p
Die Gracefully Robotics
Winner - AWS 39

User avatar
BeligerAnt
Posts: 2431
Joined: Wed May 15, 2002 12:00 am
Location: Brighton
Contact:

Re: AntLog - double elimination result logging tool

Postby BeligerAnt » Mon Apr 20, 2015 5:51 pm

Ha ha! No I couldn't Dave :)

Team Alpha deliberately has robots called Ant A*, Flea A* etc. That way it's easier to check that the logic is correctly separating robots into groups so that teams don't have to fight themselves.

Names A1, A2 etc are much less typing than calling them Aardvark, Apple etc... And I daresay team Xray may be limited to very few robots that way... ;)
Gary, Team BeligerAnt

User avatar
joey_picus
Posts: 1132
Joined: Tue Jan 13, 2009 1:51 pm
Location: Lancaster, Lancashire
Contact:

Re: AntLog - double elimination result logging tool

Postby joey_picus » Mon Apr 20, 2015 9:32 pm

Xylophone, Xenon, Xenophobia, Xena Warrior Princess (walker), Xanatos (flea), Xylocaine (nano walker)...:P

I suspect the issues with 128+ robots are a large part of why a lot of less friendly competitions have qualifying rounds and seeding (and also that I imagine, from an organisational point of view, 129 robots might even be harder than 256 without lots of group rejigging?).
Joey McConnell-Farber - Team Picus Telerobotics - http://picus.org.uk/ - @joey_picus
"These dreams go on when I close my eyes...every second of the night, I live another life"

User avatar
BeligerAnt
Posts: 2431
Joined: Wed May 15, 2002 12:00 am
Location: Brighton
Contact:

Re: AntLog - double elimination result logging tool

Postby BeligerAnt » Mon Apr 20, 2015 10:33 pm

Someone had to Joey! :D

You may be right about qualifiers, and I hope we don't have to go that way...The logistics of running a one day, all-comers event are enough for most people... Seeding doesn't actually reduce the number of fights it just gives an (unfair?) advantage to the top-ranked entries - something we should avoid like the plague IMO.

Once the tabes have been calculated it's actually just as easy to run any number of entrants as any other. It's just that with 129 (or 65, 33, 17, 9) entrants there are two unlucky souls that get one fewer byes than everyone else...
AntLog2 used separate tables for each group size as it was copied mostly from Peter's original paper charts. However, each group size is actually a 16-way group partially populated and with all the byes removed. The entries have to be fitted into the group in a particular order to ensure that the byes fall out as early as possible, but it means that AntLog 3.0 uses only one huge table and automatically propagates the byes. It does mean that a 16-robot event starts with round 3! (This might get hidden so the rounds always appear to start at 1 regardless of the size of the field!)

The only problems with running more than 128 entries are 1) validating the logic in the table, and 2) the length of time required to run the event.

Just to put it into perspective, 128 robots = 254 fights = 4.25 hours if fights can be run one per minute (every minute)...
Gary, Team BeligerAnt


Return to “Events”

Who is online

Users browsing this forum: No registered users and 4 guests