AOTY score page goofed up?

[INSAYN]

Is it just my computer, or does yours also show no points for CBY's perch, and for some reason Rawkfish's two fish are not being added together.  Hopefully it's just my computer here at work that is goofy. 

[polepole]

Is it just my computer, or does yours also show no points for CBY's perch, and for some reason Rawkfish's two fish are not being added together.  Hopefully it's just my computer here at work that is goofy. 

I see points for CBY's perch.

And Rawkfish's is indeed messed up. This is a longstanding, known bug in the system that happens when 2 fish scores from the same person are equal.  We've been fixing it by adding 0.01 inch to one of the fish to make them unequal.  I'll do that to one of Rawkfish's fish.

That being said, bsteves, anything thoughts on fixing this bug for real?

-Allen

[INSAYN]

Is it just my computer, or does yours also show no points for CBY's perch, and for some reason Rawkfish's two fish are not being added together.  Hopefully it's just my computer here at work that is goofy. 

I see points for CBY's perch.

And Rawkfish's is indeed messed up. This is a longstanding, known bug in the system that happens when 2 fish scores from the same person are equal.  We've been fixing it by adding 0.01 inch to one of the fish to make them unequal.  I'll do that to one of Rawkfish's fish.

That being said, bsteves, anything thoughts on fixing this bug for real?

-Allen

If that's the case, we should just round all of Rawkfish's catches to a nice even number to give CBY a fighting chance this year.   

Yeah, I tried closing everything on my computer and viewing it again.  CBY's perch doesn't show points for me.  I can see points for it if I click on the image icon and view the entry.   It shows the points there just fine. 

[rawkfish]

And Rawkfish's is indeed messed up. This is a longstanding, known bug in the system that happens when 2 fish scores from the same person are equal.  We've been fixing it by adding 0.01 inch to one of the fish to make them unequal.  I'll do that to one of Rawkfish's fish.

I figured that's what was up.

If that's the case, we should just round all of Rawkfish's catches to a nice even number to give CBY a fighting chance this year.   

No one would want an win like that, would they? 

[bsteves]

I think the answer to the bug is that I need to rely less on pure SQL and more on Ruby's ActiveRecord when calculating and ranking everyone's score.  Currently the following SQL query is used to create the total scores and ranking for the site.  It suffers from a glitch were it doesn't deal with ties well.  If anyone has any suggestions I'm all ears.  Until then the add/subtract a tiny fraction of an inch from your tied fish seems to be the work around.

SELECT id as user_id, users.login, Q1.Score from users left join (SELECT user_id, login, sum(maxspscore) as score FROM (SELECT t1.user_id, t1.login, t1.speciesname, t1.maxspscore, COUNT(*) AS lowerscore FROM (SELECT users.id as user_id, users.login, targets.speciesname, Max(length*targets.points) AS maxspscore FROM users INNER JOIN (targets INNER JOIN catches ON targets.id = catches.target_id) ON users.id = catches.user_id WHERE (catches.date >= users.date_paid) AND (catches.date >= '2012-01-01') and (users.date_paid >= '2012-01-01') GROUP BY users.login, targets.speciesname) AS t1 JOIN (SELECT users.id as user_id, users.login, targets.speciesname, Max(length*targets.points) AS maxspscore FROM users INNER JOIN (targets INNER JOIN catches ON targets.id = catches.target_id) ON users.id = catches.user_id WHERE  (catches.date >= users.date_paid) AND (catches.date >= '2012-01-01') and (users.date_paid >= '2012-01-01') GROUP BY users.login, targets.speciesname) AS t2 ON t1.user_id = t2.user_id AND t1.maxspscore <= t2.maxspscore GROUP BY t1.user_id, t1.maxspscore HAVING lowerscore <=10) AS Table1 GROUP BY user_id order by score DESC) as Q1 ON users.id = Q1.user_id where date_paid >='2012-01-01' AND (users.youthdiv = 'no') order by Q1.score DESC, users.login;

[[WR]]

Brian
Hit up the brain trust on Linux Journal and Linux Magazine.  .coms respectively. There are some resident experts on Ruby and the other open source data base programs at each location. Maybe they can give you a working patch that will solve this problem.