October 28th and 29th Battles

Moderator: Community Team

Post Reply
User avatar
Zachrin
Posts: 7529
Joined: Fri Jun 12, 2009 7:52 pm

October 28th and 29th Battles

Post by Zachrin » Wed Oct 30, 2019 11:35 pm

After reviewing some bug reports regarding a battle that occurred in Widu, there are some notes that we'd like to announce to help clarify any potential confusion. If you die while traveling to another node, you will be returned back to the last town that you had been in and not a town you were traveling to/through. This is regardless of whether you are in a group or an army. If you are in a group/army, the members will follow the orders given even if the leader does die.

We did discover a bug that occurred with the order of events processed resulting in the Army being incorrectly moved to Cork. The code ran as follows:

1) Leader Died and was returned to Cork
2) Army was updated to location of Leader (Cork)
3) Dead members were removed from Army
4) Lt was promoted as Captain was dead

It should have ran:

1) Leader Died and was returned to Cork
2) Dead members were removed from Army
3) Lt was promoted as Captain was dead
4) Army location should then be updated to the location of the current Captain

In this case, the army should have stayed in Widu and not Cork. Since the army leaders are no longer in Widu, we will not be adjusting it's location. The 3 who were killed on October 28th were dropped from the army. However, since the army was incorrectly listed in Cork, this allowed the army to readd members who were not in the correct town to request admission to rejoin the army. This led to them battling when they should not have been able to. As such, the result of the battle would not have changed in this matter. However, those who were not in the correct town would not have been able to participate and that portion would need to be adjusted. The one player in that situation who died will be returned their stats for that portion as they should not have been able to battle in the first place.

A concern was brought forth that First Strike occurred multiple times in a particular battle. After reviewing the logs, it shows that the First Strike correctly ran at only the start of the battle. There were 3 battles on October 28th: 1020307, 1020309 and 1020318. These 3 battles were all sparked by Attack Lists with one being from one army and 2 from the other army. As such, each battle is entitled to a First Strike.

A concern was brought forth regarding being healed after having fled/died in a battle. There was a bug that allowed them to be healed after they have fled/died.

A concern was brought up regarding fleeing and dying. Fleeing currently is coded as follows:

A) Defenders must survive long enough after starting to flee after the rounds attackers from their current attacker.
B) Attackers must survive long enough after starting to flee after the rounds of all defenders battles with them during that round.

If you are part of an aggressive army, statement B applies and you must survive the entire round with your attacks and the defenders being able to get a lick in. If you are part of a defensive army, statement A applies and you just have to survive the attacker who is giving you a lick. As such, the code as it is currently intended to run worked properly. At some point, I'd like to revise Statement B as I'm not fond of the results of what happens with larger battles. However, what happened wasn't a bug. (Issue 1557 was left open for thinking on if or how to improve it for the future. However, that's not part of the here and now)

There was also a concern regarding disarm removing from the incorrect individual. I have verified that disarm does work correctly. With this said, if a person is disarmed, they do have a chance to grab their backup weapon. This was the case in the first instance of the other side being disarmed. Future battle events showed that the original person who did the disarming did use their weapon. In addition, Weapon Proficiency improvements do show before the log for that round as it happens before the logging of the round, but is all part of that same subsection of battle.

Both bugs found from the October 28th battle have been patched for the reset running on October 31st, 2019. On a side note, Battle numbers are recycled every day. So, if a battle number is the same as the previous day, this can be expected. The battle numbers help discern differences in battles on a day to day basis.

Given that people were participating from the Cork town in Widu that shouldn't be, the battles on October 29th is not official canon. Any player who participated in the battles on the 29th may request their Health Points adjusted or their Golems repaired if they were affected by those battles in Widu.
Image
This world is but just a canvas for our imagination - Henry David Thoreau

Post Reply