@Marc-Andre ah, well, for that part I think there's no way we could show unofficial mods on the website. So it would only show for official mods, I guess. (or do you have some other proposal?)
@SimonAndréForsberg I don't which is why I should use some "static" data to construct the page. The thing is I see no safe way to grab some data from the clients and host it on a website without going over some technical limit that could potentially be too hard. Let's start with showing the cards we have the moment and worry when people would start modding and hosting servers.
@Marc-Andre agreed. And for the current cards, I'd go with either reading exported XML, or reading data from Java memory. I don't think reading the "JSON data" will be a good solution.
@SimonAndréForsberg Doing this for every request to the actual page, or doing this once at the start of the server and caching it seems a bit over the top to me. It's not simple, I need to manage a dependency, the web site will be more tied to the game logic, and I'm "wasting" time. I think I would prefer reading a static source of data. Would be easier to maintain I think
@Marc-Andre Agreed. So exporting data, then? Can be fairly easy done with XML, as I have a XML export feature already ready (can easily change it to JSON if you prefer that more) (using the Jackson library)
@Phrancis a simple invocation of the /export command, me thinks.
In the far future I think this perhaps this could be done via a DB? Maybe not, but just a thought. For right now, the scale is very small so I think KISS is probably best
Lately, I've been an enemy of anything that is related to Jackson and JAXB so, yeah I know it should be easy, but I've learned the hard way that most of the time it's not
if the mods were written with the same XML format, the makers of a mod could package it along with the images and submit it to us, and we could feed that to the java server that is parsing our mods to create a website for it
Well I must say that I do mind the graphics, but the thing I didn't like at the time was the UX.
@bazola That is one clean way to do it. I think that at one point, we could offer to "host" custom site for the mods, like we did for cyborg-chronicles.
I don't have the interest to perfect my skills on back end. I work all day on doing back end things so... What I really want to do is some web site/ web application
Ehehe well, my sister has already one kid and is going to have a second in september. So it's hard to not think about it. A lot of responsibility, but they are a true wonder.
I don't know if I will ever be ready to have kids until I have one, but I sure do know I want one and this make things easier for me. I don't know what you think about this (and it's personal so I don't want to ask) but the it sure is a big decision. I see a lot of people who have made that decision too lightly
I agree it is a really big decision. I have a few friends who have had kids, and their life certainly is changed as a consequence.
I do think it's considered too lightly in especially younger adults/teenagers, many of which find themselves insanely busy and broke because of it, though happy nonetheless
I have to say, being over 30 myself now (wife 28), it feels like a "now or never" thing a bit. Haven't decided how to feel about that yet.
Well it sure looks like it, but I do know a collegue and one of my uncle who had a kids at 40, so it's not too late. I find 30 it's the closing period on the best time for having kids based on age. The real best time to have kids is when you're ready
> I think the latest commits provide "good enough" documentation for the time being, although new documentations should be written as new things are implemented.
> I'm not a big fan of the "PhrancisThings" in the current code base, just because... Well, it's not my game, it's the Cardshifter game. It was good as a naming pattern for a good while, but I think a PhrancisThing should be refactored to a CardshifterThing throughout.
@Phrancis I have found some incorrect statements in your documentation. Should I inform you in chat, add comments on your commit, make some edits myself, or create a github issue?
@SimonAndréForsberg I think I prefer comments on the commits, generally. Allows me to learn the logic better as I go along. But, if you prefer to edit directly, that's also OK
> "an array of whilePresent must be used for each resource that is modified" is not correct. a MODIFIER object is required for each resource that is modified.
> it's not really "sent to the server's resource map". And I'm not sure it's worth explaining the inner workings of things here. It should either be a single EFFECT object or an array of EFFECT. Not MODIFIER (modifier is for the whilePresent stuff only)
> again, I don't think the "sent to the server's resource map" is relevant here (and is also not a correct description of how it works behind the screens)
> This is NOT equal to "self". Consider for example the following MODIFIER: { res: ATTACK, change: 1, priority: 1, filter: { owner: "active", creature: true } } This means that the active player has +1 attack on each creature, no matter who 'controls' this modifier. If it is your turn, all your creatures have +1 attack. Compare this to: { res: ATTACK, change: 1, priority: 1, filter: { owner: "owner", creature: true } } This means that if you play the card (and you therefore 'control' the...
modifier), your creatures will have +1 attack, no matter who's turn it is.
> "All available cards" makes it sound like it affects all cards in Battlefield, all cards in Hand, all cards in Deck, etc... (there's missing a "Deck" description here btw). The "Cards" zone is where the model of all cards is stored, when you summon a card for example, the card is looked up from this zone and copied. So changing cards in this zone can change the newly summoned cards.
> Might be good to add an example for how to use priority. I think I had an example in chat that could get included here. Using for example the modifiers "All your creatures have attack +3" and "All your creatures have attack set to 0" will end up with all your creatures have either 0 or 3 attack, depending on the modifier priorities.
private void createMapImageForRegions(Map<MapPoint, String>regions) {
int regionWidth = 60;
int numRegions = 60; //need to know this ahead of time
BufferedImage img = new BufferedImage(numRegions*regionWidth, numRegions*regionWidth, BufferedImage.TYPE_INT_RGB);
for (MapPoint point : regions.keySet()) {
int i = point.x * regionWidth;
int j = point.y * regionWidth;
int color = this.getColorForBiomeType(this.getBiomeTypeForRegion(regions.get(point)));
for (int x = 0; x < regionWidth; x++) {
@bazola img.setRGB is a slow way of drawing. You should also be able to avoid flipVertical by getting the mathematics right while drawing in the first place.
@bazola From a BufferedImage you can also get the underlying byte buffer that's in a specific format (defined by the BufferedImage) and directly work upon that
the printouts for the colors are valid and yet it draws white
Color color = new Color(this.getColorForBiomeType(this.getBiomeTypeForRegion(regions.get(point))));
//System.out.println("blue = " + color.getBlue());
if (this.topOwnersAndRanks.containsKey(ownerName)) {
color = new Color(this.getColorForRank(this.topOwnersAndRanks.get(ownerName)));
//System.out.println("player color = " + color.getBlue());
}
img.getGraphics().setColor(color);
img.getGraphics().drawRect(i, j, regionWidth, regionWidth);
img.getGraphics().fillRect(i, j, regionWidth, regionWidth);
Each time you call getGraphics() on a BufferedImage you get a new Graphics object, so setting the color on one, doesn't set it on the next one. So cache the graphics object.
BufferedImage bi = new BufferedImage(800, 600, BufferedImage.TYPE_INT_RGB);
Graphics g = bi.getGraphics();
g.setColor(Colo...
and i solved the flipping problem using math also by remembering simple algebra
for (MapPoint point : regions.keySet()) {
int i = point.x * regionWidth;
int j = (60 - point.y) * regionWidth;
Simple arithmetic:
return 7 - input + 5;
(which can be simplified as return 12 - input;)
Let's say the input is 7:
return 7 - 7 + 5 --> return 5
Or if the input is 5:
return 7 - 5 + 5 --> return 7
We got our development environment pretty much set up now, tomorrow meeting with customer to discuss and get approval of the URD, then it's design&implementation time
For the backend we use Django which gives us Models (that are backed by a database) and a REST API
For the frontend we are using React together with Bootstrap, and also using Webpack (which means nodejs and npm)
In our development environment we run the Django server (which also hosts our frontend) and an extra Webpack server that allows us to do hot reloading on Javascript (actually JSX) files
Webpack also takes all JSX files and bundles them into a big bundle.js file
For production/staging we pregenerate the bundle.js file and the Webpack (dev) server is no longer needed
And as tools we are using GitLab and Jenkins, and we made our own "tools" that run on top of an Apache server:
1) On every commit to develop of the docs repository we compile all LaTeX files and generate pdfs, put those online
2) On every commit to develop of the code repository, we update the code on our staging environment that's also reachable online
And then there's still that on every commit we can run tests (Continiuous Integration), and we made a system where the notifications get pushed to a Pushbullet channel
Currently we're still using Django's default SQlite database, but in the coming days we'll be migrating that to PostgreSQL at least on the staging environment and we're also looking into using Flux for React, I believe Flux reduces tries to store state of the REST API whenever possible, such that the backend doesn't get hit that much and faster loading times
Our client had as requirement that everything should run in one Django application, that made it a bit more complicated
I'm thinking if I'm to set this up to my wishes: Would then an Apache server at www.domain.com to host the frontend and another Apache server at api.domain.com for the backend be the way to go?
As Django is an MVC application, we are currently using a single Template index.html that loads the React code, etc.
Well unless you really got a lot of hit, I would run only one apache that would distribute between server. But don't trust me too much on that, my job is not to setup the network infrstucture ;)
@skiwi Django, Js framework (not the one you're talking but they do share some ressemblance between all of them) node.js, Working with REST API Boostrap, never heard of webpack but I understand what his job is
@skiwi Yeah I know quite something it's just not my day job :P I'm a back end guy on some entreprise services nothing web application related
@Marc-Andre Like here with React.createClass() you create a "class" with methods, it can even have mixins, etc. While in reality it doesn't exist, because you are just passing a JS object around
Same for the require, JS doesn't supports imports, yet people managed to get it working
How come certain random strings produce various colors when entered as background colors in HTML? For example:
<body bgcolor="chucknorris"> test </body>
...produces a document with a red background across all browsers and platforms.
Interestingly, while chucknorri produces a red back...
@Phrancis data->>playerNames cannot be both Phrancis and Zomis at the same time though. And I'm not sure how to do such lookups in JSON Arrays in PSQL.
either way, there's a lot that might need to be re-vamped one day.
Would be pretty cool if it did though, then we could just have the Java game server open a connection after a game is done and insert the replay as a new record in the DB. Then just write a few DB functions to crunch data in useful ways.
> There's a part of the JavaFX project that's not being used as well, that was refactored away a while ago (when the JavaFX code started to use the same view for both server-based and local-based games)
Simple arithmetic:
return 7 - input + 5;
(which can be simplified as return 12 - input;)
Let's say the input is 7:
return 7 - 7 + 5 --> return 5
Or if the input is 5:
return 7 - 5 + 5 --> return 7
How come certain random strings produce various colors when entered as background colors in HTML? For example:
<body bgcolor="chucknorris"> test </body>
...produces a document with a red background across all browsers and platforms.
Interestingly, while chucknorri produces a red back...
