Whoever runs pa-db.com, THANK YOU and... please update it?

Great site.

I have an updated instance here:
http://www.nanodesu.info/pa-db/
and yeah yarmond build something awesome.

You should also take a look at the PA Matches wiki.

Has more info on the units, the unit history, and a bunch more PA articles.

It's the largest and most accurate PA wiki available.

for units pa-db is more accurate, as it is generated directly automatically from the game files
But yeah your wiki has more shiny stuff in it.

Which is why I said "wiki" since the wiki goes into more info about PA while the unit database is just a unit database. *shrugs*

Let's see. Does the dev team know what they're doing?

http://www.nanodesu.info/padb/unit/tank_laser_adv
http://www.nanodesu.info/padb/unit/fighter_adv
I'll take that as a no.

Okay. Carry on.

Yeah t2 fighters are completely imba vs levellers. And of course we should expect the balance to be 100% finished.

...wait a sec

The game has a lot of balance issues right now. It's practically a half finished rebalance. Of course stuff is not balanced right now

I think his point was that the Peregrine and the Leveller are just strictly better versions of a T1 unit by design, they don't actually have a role to play, and so there doesn't seem to be a role that either of them are designed to play late game.

Same thing with Inferno and Vanguard. I don't see why you would ever build both. It seems to be very strictly one or the other.

Personally I never understood the fear people have over different tech levels that are just stronger versions, so that did not occur to me.
Making the jump at the right time in itself is the interesting part of it.

It's actually super useful pa-db lags behind because you can compare stats with Colin's version.

Colin, any chance for old builds to remain archived?

I was thinking myself. pa-db is written in phyton, which is a language I am not very fond of, so adding some features would be not easy for me. I don't have any development environment setup for it.
Conceptually it should actually not be a change that big, just add some sort of version selector.
Dunno. Another thing I could do, which is kind of a workaround, is to run multiple instances of pa-db that use different data as a base. But that is kind of weird to do.

I'll reply to this with my reasons later.



Can someone please preserve the information somehow? We could start creating an excel spreadsheet of the different information for comparison.

That is the lame way of doing it.

You can save webpages directly to your browser right?
I would on mobile, but I have hardly any data left for the month and am fast running out of storage

The PA Matches wiki has a history for each unit.

Well, Uber specifically stated that they wanted every unit to be viable at all stages of the game...

What about simply doing:
http://www.nanodesu.info/padb/58772/
http://www.nanodesu.info/padb/59549/
etc.

With a redirect from http://www.nanodesu.info/padb/ to the current build?

Edit: A suggestion for both the DB and the wiki - include the units' strategic icons, and use more symbols instead of words, it makes scanning the page for info easier. For example, a green "+" instead of "HP", the metal icon instead of "Metal Cost" etc. Not everything should be symbols, but the key info at least.

Edit 2: I think one of the main things that turns me off the wiki is the layout of the page, especially the massive banner. If I load up a unit, say http://pamatches.com/wiki/units/bots/stinger/, I can only see 2 bits of info without having to scroll down: The unit name and description. For 1050 pixels, this seems excessive.

I didnt really plan to start this t1/t2 discussion again. I am fine with both ways, so please let them do what they want

To preserve the information only a few json files of the game need to be stored. The padb software runs against them.
raevn, that pattern was what I thought about. The issue is more that padb is written in phyton. I really am not good with it. So all I could do would be to start multiple instances of padb and put them on the different path. But running multiple instances might be not a good idea once there are more and more versions.

What about caching the output? E.g., instead of keeping the json files of the units for each build on the server and having the pages generated on demand, write an additional script that calls all the pages and saves them to static html files.

Hmm actually that might be it.

... I just remembered this program:
http://www.httrack.com/

Gonna try. If it still works like I remember it actually should be simple.

Lol I just found that program.

It's a bit of a fight with relative/absolute links but I think thanks to mass replaces with notepad++ I got it working. New url, because I am too lazy to change the setup of the running pa-db server:
http://www.nanodesu.info/pa-db/