New HoF brain storm

[WraithLord]

Hi, As you may know Tyrant hasn't logged in quite a long time and the HoF is in disrepair.
I volunteered to take this hot potato to hand

I'd like to use this opportunity to hear any ideas you may have for improving the format.

In any case I will not be doing manual updates of the HoF statistics. I plan to dump all the data to a DB and write a program to generate the report. I will then post the updated reports.

I can do all this myself but it will slow me down. So help, even from one brave soul, would be greatly appreciated. No programming skills are required.

Bring it on. Now's the time to talk

[Belac]

If it's in a DB, you can set it up to sort by types of game. So it could be broken out by era, size of game, wacky rules, # of players, etc.

[WraithLord]

I don't want full blown DB.
I want a human readable, none XML, text file. I'll implement data access and logic myself.

[Gandalf Parker]

I appreciate that. Text files can be used by other programs more easily. Or one of the old "db" standards such as a text file with commas between each item. Many programs can import and export CDF (comma delimited file"

[WraithLord]

That's the format, ex:
[
Winner: Jazzepi
Game: Epiphany
Game Link: http://www.shrapnelcommunity.com/thr...b=5&o=&fpart=1
Players: 11
Nation: LA T'ien Ch'i
Mods: Well guarded neutral Emor citadels as VPs
Era: Late
]

[
Winner: Calmon
Game: the Quick and the Dead
Game Link: unknown
Players: unknown
Nation: LA Ermor
Mods: unknown
Era: unknown
]

The parser is almost ready. It doesn't care much for white spaces and it's flexible so I can adapt it quickly to changes.
I'm using Python btw, Perl would've got the job done faster but I wanted a tidy OO program and not a quick and dirty perl voodoo.

[Psycho]

I tried writing a parser in Ruby some time ago to parse the thread and get the HoF statistics, but it proved to be not so easy. The problem is that you don't have a standard format for posts. Also some player names are written in two or three different variants - with and without capital letters, with and without spaces, etc. You'll probably have to enter a lot of data manually.

And if you're doing it, well then XML is probably the best way. You can easily write an XSLT to transform it to a human readable html file for example.

[WraithLord]

Some old posts are missing almost all information so it's mandatory to extract that info from the game thread.

However you see it there's just no way to avoid manually building the DB. If I get help in this it will cut the time to completion by half. Otherwise my plan is to complete the report gen. engine and then get to manually feeding the DB. It's not that bad, mind you, most of the posts are compatible with my parser (just add [ and ] lines with simple awk/sed script or by hand), so it's down to copy/paste.

[Hadrian_II]

Originally Posted by WraithLord

I don't want full blown DB.
I want a human readable, none XML, text file. I'll implement data access and logic myself.

whats wrong with XML? i think there are free XML libs out for every programming language. If you use XML you dont have to write a custom parser for a custom format (what in my opinion just sounds like unnecessary work).

Originally Posted by WraithLord

However you see it there's just no way to avoid manually building the DB. If I get help in this it will cut the time to completion by half. Otherwise my plan is to complete the report gen. engine and then get to manually feeding the DB. It's not that bad, mind you, most of the posts are compatible with my parser (just add [ and ] lines with simple awk/sed script or by hand), so it's down to copy/paste.

Why not create a simple web form, and fill the db with the input. I suppose it is not harder for players to report victory to a webpage, than to a forum post.

[Trumanator]

I think making it a non XML is a good idea, simply because it makes it easier to access quickly.

[WraithLord]

KISS.

I don't want XML. I personally don't care much for it. It has it's uses of course but IMO it's useful for higher caliber projects, like defining protocols or modeling etc. For simple game report DB it's a bit more than what's required.

Of course both your ideas can work. They'll require a bit more work than what I had in mind that's all. If anyone wants to bring to the table an end product that does the DB/report generation magic then I'm all for that. Go for it

For now I'll stick with my plan. It's simple to fill (no change to current report). Simple to copy to DB. Simple to export. yata yata.

I appreciate and respect the advice given but I'm not convinced. It's a very small use case and I don't see any need to go overboard.

I want to bring to discussion another aspect. Not technical.
I don't think the current ranking system makes much sense in the way it factors wins against magnitude.
Ex:
player A won 3 games of 6 players each.
player B won 2 games of 20 players each.
The current system would put A above B. I don't think it makes sense. Player B has won much bigger games, defeated more opponents and invested much more time. Am I missing something here?