\o/ They should post in this thread so they can be +1ed. :-)

Note that I'm not planning to personally work on the forum - ultimately, we'll probably replace it with a fully featured modern forum. But, I'm *absolutely* happy to accept code modifications that improve the current forum.

y2k tells me he's working on the double-post bug for games next. And then might look at pulling pre-game chat from vDip. I'm excited.

Really exciting to see these issues being addressed.

Out of curiosity, rather than implementing a 3rd party Forum, what are the thoughts on duplicating this forum into several categories?

NO DOUBLE POST YES Y2K IS A HERO!!!

Abge: This forum hasn't scaled well with the site growth, and was the reason behind all the DB issues we had earlier in the year. Supporting our own improved forum would more than duplicate the amount of coding work webDip requires, and we already have considerably more work than we have developer capacity.

I understand that, but I was wondering if having, say, 4 small forums resolves the db issues of the 1 large forum we have now.

@abgemacht That would require replicating the user table somehow into 4 databases (aiee, the horror, the horror) - otherwise, all 4 "small" forums would still be in the same actual physical database. Probably not going to solve the "too much stuff is lying about in the DB" problem.

Yup excellent point. That isn't what we want.

You can split by table instead of by databasr... Or you can duplicate the number of databases, considering the usernames have a limited number of touchs (accoUnt creation, and maybe banning?) it shouldn't be too hard to do... Though still not KISS enough...

Thank you very much, Doktorkeks42 and y2kjbk. Little things like this are just as important as major changes to keep improving our site and making it the best Diplomacy site there is. I encourage you and anyone else out there with the skill set to chip in and do any little thing you can think of and have the time for. Not only do those little improvements make all our lives that little bit better, but they also mean ATC has more time to work on big things like the scoring changes. Win/Win for all of us.

This site couldn't run without people volunteering their time and effort. :)

i'm really gonna try to get some more stuff done in the coming weeks, might not be until January

THANK YOU Y2K FOR GETTING RID OF DOUBLE POST

THANK YOU Y2K FOR GETTING RID OF DOUBLE PO$T

Orath: Yes, we can improve the current forum to make it work with the site. But, in terms of developer effort to site improvement, integrating an existing forum is less effort overall, allowing us to spend time on improving the Diplomacy section of the site.

On a personal note, I don't want to help code a forum. I want to help improve a Diplomacy site.

"On a personal note, I don't want to help code a forum. I want to help improve a Diplomacy site."

A very understandable position. Of course, the Forum is quite important to the life of the community.

Do stand-alone forums scale significantly better than ours does? Obviously there is a large implementation cost, but once that's done, do you expect the new forum would scale much better?

thanks for the double post fix :)

The forum is EXTREMELY important to the community. We're definitely keeping it, and it will definitely improve.

"do you expect the new forum would scale much better?"

Yes, without question.

OK, that is certainly good to know.

@A Tin Can,

I would argue that forums are not thr right piece of technology. But your point is well made. Taking a third party software and integrating it is a great idea.

Alternative have been invented - the reddit style upvote messaging system. The stackoverflow style question&answer system.

All have different strenghts and weaknesses. A wiki style front page and news site would be very empowering for the community. But i'm not sure it would he tue right piece of technology to integrate...

Looking at the current 'copy&paste' this list system which we see on the forum, having a editable list which anyone can add to would be cool.

Will you be discussing with the community what they think is best? I honestly don't know, haven't fallen for any single piece of tech... (Though stackoverflow style q&a would be great for new players...) tweet-style messaging service?

I wonder if you could combine the different forum styles, allowing users to choose what style they prefer, getting what they want from the same database..

Not for webdip, just in general.

Supporting extra skins sounds like even more work, so I doubt that.

Yeah like I said, not for webdip.

Steephie, entirely possible (if technically difficult). But it is like adding auto google translate for foreign languages and then expecting people to understand each other. The more different the presentation, the less sense people will be able to make of what they are saying - context is a huge part of communication and you take some context away while giving the illusion that everyone is looking at the same thing (if you can overcome the technical hurdles)

Conclusion, i dislike it, would also love to see it trialed somewhere :)

@steephie, yes you can do something like that, you store the matrix representing how each user wants to see a site in the user table and create it on render. Obviously the problems come into play around load time, and matrix complexity taking up more space. Theme related changes only aren't too hard if you're simply changing css related changes though. So if you create say 3 themes that display data in different ways and simply store a theme choice in a user table then the overhead isn't as bad. Which is something I've done for professional sites before in the automotive industry. It's a bit of a hack, but not load time or DB intensive at all.

Hmm...
I wonder if it's less work if you just created three subdomains with different styles tapping from the same database. The players can simply choose their preferred style by going to the right subdomain. Might not be as sexy as something that remembers but still...
You could also open three tabs with three styles that way.

This one goes on the 'You never know'-list by the way. If someone would love to work on this for real, feel free to let me know :-)

"I wonder if it's less work if you just created three subdomains with different styles tapping from the same database"

WebDiplomacy is not designed to do this. A lot of files are cached locally after they're first generated, which means that there's some surprising/incorrect behaviour if you try this.

I looked in to having a "beta" site where people could view potential changes to the site design/interface before we made them public. Even when disabling all the caching (which was messy), it's still not quite stable enough for me to be confident releasing it to everyone.

I really would like a good solution for having a beta test site, though.

Steephie: If you really do want to learn to work on large code projects - starting small with some simple webdiplomacy changes would be a good first step. We have lots of good changes for newcomers to the code on the github list.

https://github.com/kestasjk/webDiplomacy/issues?utf8=%E2%9C%93&q=is%3Aopen+is%3Aissue+label%3Aeasy-difficulty

I've worked on a little python script that recognize based on the opening post that it is diplomacy related or other, so it can divide the forum into 2 tabs. It worked on a small learning set and limited input.
But my javascript knowledge is rather limited and I can't figure out how to script for this forum.

@gonnor, does it recognise when someone goes off on a tangent?

In general, i prefer a system where recognising that kind of difficult thing is done by human computation. But i do like learning algorithms in principle...

It was a general idea which I figured would be useless for webdip.

@ATC: I agree, but I've got enough things going on at the moment. Some things planned in January:
Wrap up school-part of boardgame
Register company
Lots of tests
Meeting with Admission Board at university
Create website for company
Help first client and get the word out for more clients
Hopefully recruit first employee
Take care of contracts, insurance, financial stuff

I am absolutely positive that I forgot stuff.
I won't do February :-)

My point: I agree in principle but for the time being, I don't want to spend time on doing something to help me learn skills to do something I might want to do while I could be spending that time on any of the above.

@orathaic, it's based on Naive Bayes learning technique. At this moment, i gave it the opening posts of 20 threads manually, and gave it the label diplomacy or other, so it can learn what the common words are.
I wanted to automate my learning database, by scraping the threads based on 2 buttons i added to the forum. But i can't figure out to bind the action to the button. atm it doesn't even alert('button pushed');

This is a cool hobby project, but probably not something we'd add to the site (there's no need for automatic classification where a checkbox on new forum posts would do).

Having said that, if you want help with the JavaScript, shoot me a PM. You want to look into the onclick event. And don't use anything from w3schools- so much of the JavaScript info on that site will lead you astray.

Yeah, what ATC said, i don't know much python, but javascript might help...

You could create a simple site with the webdip forums in an iframe, and a javascript button - but i'm not sure how easily the data is to scrap. iframes from different domains have different rules (for security reasons)... Actually there may be a work-around. And i'm making things much more complicated than they need be...

Pm/email if you have your project on github and i might throw something together. It sounds fun :)

Yeah I figured, ATC, just wanted to throw it out. I've made it couple of months ago, but got stuck and the project sidelined. For my thesis I have to work with machine learning techniques, hence this little project to learn differents algorithms.

What's with w3schools? I had to make a website for school and used mostly w3schools and stackoverflow. Wrote html, javascript (jquery) and php

@orathaic, it won't be for the near future though. Got exams now, and after that i really need to jump in my thesis.

I don't know what's with them, but everyone I know who's learnt Javascript from them has ended up horribly confused. A lot of their explanations are incorrect, but the code examples all work. This is particularly insidious, as it leads people to think things work for the wrong reasons. Also, I don't like how they make their branding intentionally confusing so it seems like they're associated with the W3C.

If you want to learn JS, I recommend Douglas Crockford's excellent book "JavaScript: the Good Parts", followed by the JQuery documentation if you're planning to do anything in the browser.

Douglas Crockford is notable as the author of JSLint, and healso plays diplomacy - here's his map: http://www.crockford.com/wrrrld/diplomacy.html

"I've made it couple of months ago, but got stuck and the project sidelined"

I know that feeling. I have half a stalemate line detector written somewhere. (and dozens of other half projects)

I *think* w3school's CSS/HTML tutorials are alright, but I tend to go to other sources instead because their JS information is so poorly QA-ed, so I don't want to take the risk. I remember a number of the hardcore JS/HTML stackoverflow users used to call them "w3fools".

I think this is a really cool project, by the way! I've also thought about writing something to do truth detection or sentiment analysis on press, though I don't think it would actually be accurate enough to be useful.

I would only use w3schools if you need to know what to do with that one line of code, not as something to study.

CodeAcademy was always the one we were told to use

For learning, for looking stuff up or for both?

@atc: has the double posibg in a game after refreshing still happen?