« first day (374 days earlier)      last day (3187 days later) » 
00:00 - 19:0019:00 - 00:00

12:00 AM
RELOAD!
[Cardshifter/Cardshifter] 2 opened issues.
[Cardshifter/HTML-Client] 41 commits. 5 opened issues. 2 closed issues. 9 issue comments. 1489 additions. 1214 deletions.
[Hosch250/VSDiagnostics] 11 commits. 2279 additions. 203 deletions.
 
I feel alone when @Duga start babling about issue and repo, but no one else is talking :P
 
[retailcoder/Rubberduck] 1 commit. 22 additions. 38 deletions.
[rubberduck-vba/Rubberduck] 3 issue comments.
[skiwi2/Fragments] 6 commits. 36 additions. 11 deletions.
[Vannevelj/VSDiagnostics] 14 commits. 9 issue comments. 1957 additions. 226 deletions.
[Vannevelj/VSDiagnostics] 19 commits. 1 opened issue. 3 closed issues. 2 issue comments. 6285 additions. 378 deletions.
 
@Marc-Andre I'm here!
 
Great how is the html client going?
 
I think it's going well, but I just got back to my computer after being gone for 2 hours.
 
12:04 AM
And on a personal note are you in school or a you finished with and working ?
 
@Marc-Andre Sorry, I don't feel comfortable sharing that information.
 
No problem at all!
 
Thanks for understanding :-)
 
Hey I'm still a random guy on the internet :P it's easy to understand! And even if it would be in person I understand that people don't want to share this!
 
Thanks. I just do my best to keep my personal information off the internet.
 
12:08 AM
So on another note, what did you learn by doing the html client ?
 
I got to learn about AngularJS.
 
Do you like the framework ?
 
I think it definitely makes interaction with the DOM a lot easier.
 
Great I think I'll have to look at the code, I've never worked with angular yet
 
12:30 AM
Darn. I just finished writing an answer and after I posted it, I just realized that someone else had already posted an answer 10 minutes ago that already had 3 votes and an accept.
 
Never wasted still :)
 
12:52 AM
you are both reviewing the same problem but in different ways
 
1:05 AM
[Cardshifter/HTML-Client] Phrancis pushed commit 2930b787 to sections-layout: updates to cardshifter.html
[Cardshifter/HTML-Client] Phrancis pushed commit ed0a1b05 to sections-layout: fix merge conflicts merging master into sections-layout
 
@SirPython around? I'm a bit confused now
 
Sure! What can I help you with?
 
^ See highlighted files
 
Ah, I see. You made a few mistakes with updating your branch
All the scene files should now be under src/
You have all of master and all of your branch, making your new branch half the old branch and half master.
 
Ah, guess that explains the unexplained merge conflicts
 
1:13 AM
It's a branch baby!
 
More like branchenstein...
Do I understand I could (and should) delete above files?
(and move the card_model directory into src?)
 
I think you should be good doing that
 
Alright
 
Although, I don't know what happened to the files in src/
You might have some filesteins.
 
I've resolved code-level conflicts, just didn't realize some files were moved
 
1:18 AM
I mean, "filesteens"
 
Now, let's say I want to run the site, would I open www/index.html and just run that on the IDEA server?
 
Here is what you would do:
1. Navigate to the root of the repo
2. Run `npm i`
3. Run `npm run develop`

Step 2 needs to only be done the first time.
Obviously, you will need Node and NPM.
(Installing Node will automatically include npm)
Oh yeah: Then, you navigate to localhost:8080/webpack-dev-server in your web browser.
 
oh wow
Alright, I'll try that
 
Good luck :-)
 
I hope there's not a "You'll need it!" implied ;)
 
1:23 AM
You shouldn't need it; there weren't any problems for me.
Well, for some value of problem.
 
Node.js right?
 
Yes.
npm i can take a while to run.
That command installs all the dependencies.
Wow: after reading @DanPantry's answer, I now realize how much AngularJS I don't know.
He said not to use $scope which is probably the most used AngularJS service in my code.
 
> Before you can install Node, you’ll need to install two other applications. Fortunately, once you’ve got these on your machine, installing Node takes just a few minutes.

XCode. Apple’s XCode development software is used to build Mac and iOS apps, but it also includes the tools you need to compile software for use on your Mac. XCode is free and you can find it in the Apple App Store.
Homebrew. Homebrew is a package manager for the Mac — it makes installing most open source sofware (like Node) as simple as writing brew install node. You can learn more about Homebrew at the Homebrew website
-.-
 
LOL
 
I'm just going to head over to node's page and download it instead, how about that Mr. Fancy Pants
That's efficient, just install XCode so you can download and install the 16 MB node pkg. Sure man.
➜  HTML-Client git:(sections-layout) ✗ cd /Users/francisveilleux-gaboury/IdeaProjects/HTML-Client
➜  HTML-Client git:(sections-layout) ✗ npm i
zsh: command not found: npm
➜  HTML-Client git:(sections-layout) ✗  node -v
v0.12.7
➜  HTML-Client git:(sections-layout) ✗ npm -v
2.11.3
Simple enough ;)
Running npm i now
 
1:38 AM
Hm. I just realized that I think I've cast a close vote on each of my meta questions.
 
Cool, this works now
 
Awesome!
 
Now to make CSS work for me... ;D
 
Good luck (you'll need it this time)
 
TTGB later all!
 
1:41 AM
See you @Marc-Andre!
 
You know, I've been thinking about cimmanon's suggestion to not use tables for the layout (on that lobby question)... and the more I think about it... The more I tell myself: It looks and behaves exactly as intended... Do we really need to go back to the drawing board?
 
@Phrancis Your most is now the most up-voted AngularJS question on Code Review.
 
Later @Marc
 
@Phrancis "First make it work, then make it better"
That is what I am telling myself about the AngularJS code
 
@Phrancis Yes you must change it! Don't even think to not changing it!
<table> for layout are now an illegal thing to do in web :P
 
1:42 AM
Well, should I change it now or should I change it later?
 
I'd say later.
Once we get a working client going, we can go back and fix everything.
 
Not right now, but the sooner the better :P
 
Alright
 
And Bootsrap is giving you a grid to work with
 
True. Still feels I've barely scratch the surface of Bootstrap
 
1:44 AM
Sure ! But it's not a reason to not do it correctly!
 
I was wondering the same exact thing after reading DanPantry's answer: according to him, I've basically messed up on every aspect of AngularJS in my code. Am I going to change that now? Or course not: that's a complete waste of time. However, once a full playable client is up and running and I no longer have to worry about trying to match the Java client, I can then go back and begin to correct the AngularJS.
DanPantry himself may even help to correct it.
 
Now I really need to go!
Cya!
 
@SirPython Sounds reasonable :)
@SirPython at your convenience could you add a message similar to the "No player / Mod" selected, but for the "Deck Builer" button in Lobby that only checks for a mod (and informs accordingly)?
 
@Phrancis There is already one there.
 
I didn't get one, has it been committed to master?
 
1:57 AM
        if($scope.selected_mod) {
            currentUser.game.mod = $scope.selected_mod;

            var getCards = new CardshifterServerAPI.messageTypes.ServerQueryMessage("DECK_BUILDER", currentUser.game.mod);
            CardshifterServerAPI.sendMessage(getCards);
            $location.path("/deck_builder");
        } else {
            console.log("pick a mod pl0x");
        }
 
LOL
 
Oh right. Forgot to change the message :-)
 
I got it, I'll have it post to chat also like the other one
 
> In various scenes in the code, there are places where there a button-like figures with a radio button circle inside. One example of this is the mod section of the lobby.

Often times I find myself trying to click the button-like figure around the radio button, but instead I have to aim my mouse at this tiny circle on the button-like figure. It would be much easier if I could just click the button-like figure itself.
 
@Duga @SirPython I'm hoping to get rid of the radio button itself at some point in favor of something that looks more like a game UI, rather than a HTML4 form :)
(e.g., just highlighting the value/thing with different CSS colors and such)
 
2:02 AM
[Cardshifter/HTML-Client] SirPython deleted branch game-view
 
^^ Too lazy to fix merge conflicts.
 
2:13 AM
[Cardshifter/HTML-Client] Phrancis pushed commit 68ec79d3 to sections-layout: delete unused files; add error for deck builder from lobby if game type is not selected
[Cardshifter/HTML-Client] Phrancis pushed commit 326cd927 to master: update local master with remote changes
 
@Phrancis I'm getting an error from the server whenever I try to send the picked cards from the deck builder. Do you think you could help me out if I gave you a server log?
 
[Cardshifter/HTML-Client] Phrancis pushed commit a1c6bb66 to master: merge sections-layout into master
 
@SirPython Yeah I can certainly try
 
Alright, I'll go get the error again and create a paste bin.
 
OK
Try to log the time it happened as closely as possible
Is it expected that the file is named src/deck_builder/template.html?
(as opposed to deck_builder.html)
 
2:18 AM
I don't know. Not all sub-directories of src/ named the html file template.html
^^ That is the server log.
I logged in as Test as first, but then logged out to change my name to a more readable one.
 
Look at lines 168 & 169
java.lang.RuntimeException: Could not read JSON expected message from 7: SIRPYTHON @ /127.0.0.1:62127, message was:
 
Oh.
 
        at com.cardshifter.server.clients.ClientWebSocket.handleMessage(ClientWebSocket.java:107)
        at com.cardshifter.server.model.ServerWeb$InnerServer.onMessage(ServerWeb.java:77)
        at org.java_websocket.server.WebSocketServer.onWebsocketMessage(WebSocketServer.java:469)
        at org.java_websocket.WebSocketImpl.decodeFrames(WebSocketImpl.java:368)
        at org.java_websocket.WebSocketImpl.decode(WebSocketImpl.java:157)
        at org.java_websocket.server.WebSocketServer$WebSocketWorker.run(WebSocketServer.java:657)
 
I'll check the API code to make sure there aren't any typos.
 
^^ 170 after you scroll past the card set JSON
lol ok
brb
 
2:42 AM
@Phrancis I'm going to turn in for the night. I'll work on finishing the fix tomorrow.
See you.
 
started for removing .max properties from cards.

Having a .max property on the cards will cause the server to crash. I don't know if this fix works and I don't have time to know right now.
[Cardshifter/HTML-Client] Phrancis pushed commit a03de0dc to sections-layout: change ng-* to data-ng-* in html files
 
@SirPython OK good night :)
 
3:01 AM
hey @Phrancis, i just found this game for iOS called Loot & Legends. its free to play so you should check it out when you get a chance. its kind of hard to explain, but it has a collectible card mechanic that is very unique
 
3:13 AM
> Looks like this will remain in-page for the foreseeable future.
 
3:27 AM
@bazola OK is it fun?
 
[Cardshifter/HTML-Client] Phrancis pushed commit 52a39632 to sections-layout: Add documentation & TODOs, rearrange sections a bit - #27
 
        <!-- LIST OF ALL CARDS - Displays one full row below for every card -->
        <table>
            <tr data-ng-repeat="card in cards">

                <!-- Creature Type -->
                <td>
                    <!-- TODO: Talk to @SirPython about why this {{card.properties.type}} is not showing anything inside the table cell. -->
                    {{card.properties.type}}
                </td>
                <!-- Card name and link to see card info. -->
                <td>
                    <a data-ng-click="showDetails(card)">{{card.properties.name}}</a>
^^ @SirPython See the empty cells to the left of the card name re: above TODO
 
not sure how fun you will find the game. the combat is based on tabletop D & D and is somewhat repetitive. what is interesting about the game is the way that you collect gear for your characters, and how each piece of gear adds cards to the deck of cards you use for that character during battle
 
4:06 AM
@bazola Not found on iOS store, weird
 
[Cardshifter/HTML-Client] Phrancis pushed commit a3c7dc5c to sections-layout: chance increment and decrement buttons to glyphicons
 
ttgtb
 
 
3 hours later…
7:28 AM
@SirPython thats webpack, npm run develop runs webpack-dev-server which watches your files for changes and detects syntax errors etc
 
7:56 AM
@SirPython personal preference but I see no use in ever using $scope directly. The only exception is in the link function of directives when you want to use $watch. You can, of course, use $scope if you want, but I think the controllerAs syntax is much cleaner, and you need to use objects in the scope anyway otherwise you run into prototypical inheritance issues inside isolate scope
@Phrancis some packages use native compilation so -shrug-
worth it though
webpack-dev-server is one of those, and karma. sorry. :-(
it could be worse, I had to install visual studio 2013 for it
 
8:31 AM
> it would be useful to have a develop branch on this repository for potential contributors to pull from and point PRs at. That way we don't end up breaking working, released code (master) with development code (develop). All branches would branch off of the develop branch and be merged back into develop until a new release is ready, at which point it would be merged into master and tagged
 
@SirPython It would be useful to have unit tests to help facilitate this. ;-) Don't worry about saying you've 'messed up on angular'. It is better to have things work than have them be pretty.
My answer wasn't to make you feel like you've messed up on angular, but to point out potential improvements. You haven't done anything wrong, you just haven't necessarily done it how I would have. But that's the point of code review :-)
btw, install packages using npm. ng-FitText is on npm. npm i ng-fittext --save and then require('ng-fittext'); angular.module('foo', ['ngFitText']);
@Phrancis this is my fault. I think it would make sense to have the default name be template.html but you can really name it whatever you want as long as the intent is clear. I don't think it makes sense to name a file deck_builder.html inside of the deck_builder folder as its already clear that I'm looking at deck builder
 
9:26 AM
Monking @Dan
 
hey
 
Hey @skiwi
 
9:44 AM
monking @skiwi @jacwah
 
@DanPantry Cool to see you active here aswell ;)
 
@skiwi where else am I going to procrastinate?
 
@DanPantry Right, I'm doing that for a few weeks already
I'm having holidays, but they don't feel like
Every week some stuff is going on, including driving lessons and work
 
 
1 hour later…
10:49 AM
<h1>Monking!</h1>
 
hey @Phrancis
 
@DanPantry It's not so bad at all, except the game server is not liking it
 
I'm just reading through the newest MtG set otu of pure interest ;P magic.wizards.com/en/articles/archive/card-image-gallery/…
 
If I have a game server running as well as npm run develop together, every time I make a change to the client and the app refreshes, the game server crashes with a StackOverflowException
 
Those cards take a minute to understand lol
Man, if MtG is still a thing in 5 years I might play it tabletop lol
 
10:58 AM
@Phrancis That sounds interesting lol
game server is in java right?
 
Yes
I have a server log and want to file an issue to report but I'm not sure if I should file it as a server issue in the main repo, or a client issue in the HTML one
 
Also in Groovy actually
 
Only thing I know for sure is that it's due to trying to close a websocket connection
Probably would make most sense as a server issue I imagine
@skiwi ^ agree?
 
@Phrancis If the server crashes, it's always a server issue
 
ok
 
11:07 AM
It might be that the client sends something it shouldn't be sending, but that shouldn't crash the server
 
OK
Seems like the server gets caught in a infinite back&forth messaging and eventually just quits. I'll provide a full log with the issue then
 
11:24 AM
I should point out here that I have not actually modified the operation of the messaging :p
 
> I've been noticing this issue lately and has been particularly frustrating now that the web client runs with `npm` where every change it the client source code causes an app refresh, which then causes the error and requires restarting a game server to continue.

I've copied a [full server log](http://pastebin.com/psrpZnVH) and below are the lines where the error occurs.

_Note that for this log, my username was `pl0x`.

Trigger appears to start at lines 116-117:

[2015-08-11 00:0
 
@DanPantry I've noticed the error before that, just less frequently (it only happened on manual page refreshes)
 
> Note that you can opt-out of the auto refresh by just going to http;//localhost:8080 instead of http://localhost:8080/webpack-dev-server. This doesn't fix the issue obviously.
 
It would appear the issue occurs when establishing a websocket connection
 
> And at that point it looks like it's going into an infinite back & forth of these two messages
I think it looks like the server tries to send a "client disconnected" message before actually removing the client from its internal client array
 
11:35 AM
@Duga Semicolons might not work :D
 
@skiwi I fixed it. :-(
> Send to AI Idiot: AvailableModsMessage [mods=[]]
abusive log @Phrancis
 
It is pretty big indeed
Hence the snippets in the issue itself =;)
 
btw, I'm always happy to write the server-side websocket implementation in javascript :^)
gets shot
 
ttgtw
I'll check chat from mobile from time to time ping me if you need me :)
 
@Phrancis I need you ;-) Can we please not put 'demo' code (./card_model) in the master branch on the repo? D:
On the HTML-Client repo, thati s
 
11:43 AM
> Was `card_model/` meant to be included in this commit? I don't think it should be on the master branch, it looks like demo code. I would personally have expected this to be in a feature branch. If it's on the master branch then any contributors get it when we update.

cc @Phrancis @SirPython
> Also, as for data-* prefixes, we can do this in a post-processor because we are now using a build step. See here for a tool we could use. If we are going to take this approach we may also want to add a build step for validating HTML5?
 
12:10 PM
Ugh
Stuff like that is fun to implement...
 
12:31 PM
Hey everyone!
 
hey @Marc-Andre
 
@Marc-Andre Hello!
 
12:53 PM
o/ @Marc-Andre
 
1:26 PM
I'm trying to run the web client, but the server crashes instantly...
[2015-08-11 15:25:15,689]  INFO  ServerWeb [WebSocketWorker-11] (        ServerWeb.java: 46) - Connection opened: org.java_websocket.WebSocketImpl@6e779c0d
[2015-08-11 15:25:15,689]  INFO     Server [WebSocketWorker-11] (           Server.java:158) - New client: 6:  @ /127.0.0.1:42184
[2015-08-11 15:25:15,693]  INFO  ServerWeb [WebSocketWorker-11] (        ServerWeb.java: 66) - Connection message from: org.java_websocket.WebSocketImpl@6e779c0d: {"username":"jacwah","command":"login"}
[2015-08-11 15:25:15,694]  WARN  ServerWeb [WebSocketWorker-11] (        ServerWeb.java: 84) - Connection er
Anyone recognize this?
OK it doesn't crash, but the thread handling the client dies
 
That's strange.
Are you running the client by running npm run develop?
 
Yeah
 
I'll try to run it too.
 
> Put this on the back burner until I can figure out how to neatly replace assets with externals at build time with webpack. Until then this will only result in performance degradation
 
@Duga @DanPantry I literally had my mouse right over the "merge" button.
@jacwah Are there any errors in your web bbrowser's console?
 
1:31 PM
Firefox can't establish a connection to the server at ws://127.0.0.1:4243/. cardshifter.js:840:0
"Websocket error(error 1)" cardshifter.js:987:14
 
Sorry, stupid question.
Try restarting your server and the client.
Or just the server.
 
Wait I think that message was from before
It doesn't pop up now
 
Restarted both of them
 
[Cardshifter/HTML-Client] danpantry pushed commit 69607ee6 to master: add development instructions to README
[Cardshifter/HTML-Client] danpantry pushed commit 3332422e to master: add npm to scripts
Merge pull request #54 from danpantry/chore/development-instructions

add development instructions to README
 
1:36 PM
This is weird...
 
@SirPython LOL
@SirPython @jacwah FYI my changes didn't alter the server interface so any functionality it had should be intact
@SirPython the problem with the PR I had for the uglification is that I can't find a nice way that works on windows to choose to use 'externals' in production build and the vendor chunk in the dev build
I'll have to wait until I get home for that, but it should be pretty simple using environment variables
@SirPython develop branch where? ;o)
 
Ah, right. Sorry. I'll add that now.
 
1:53 PM
@DanPantry Not blaming you! I ran from master
 
2:06 PM
BTW.Meetings :|
 
> card_model is demo/spoof code, indeed. Sorry I should not have pushed to master. You can delete it and I'll keep it in my branch for now
 
@Duga commented on null! Nice @Phrancis!
 
2:42 PM
> git add -p ;-)
 
3:00 PM
[Cardshifter/HTML-Client] SirPython pushed commit a04182a6 to master: Update package.json
 
@Marc-Andre no idea how that jappened lol
Finally done with meetings, I want to do some actual work today
I have to say, databases here are MASSIVE and everything in the business depends on SQL, it's a bit overwhelming
 
hey again
 
3:20 PM
now requires a username to log in

During testing, I accidently attempted to log in without a username and I logged in with success (which is not a good thing). Now, the client will block you from logging in if you have not yet entered a username.
[Cardshifter/HTML-Client] SirPython pushed commit 157dbd79 to game-board: now successfullly removes all .max properties from cards in DeckConfig.configs.Deck.cardData so the server doesn't die
 
@Duga A server issue should maybe be opened
 
@jacwah To prevent users from logging in without a username?
 
@SirPython Yeah
 
Should that be something handled by the server or by a client? (I am undecided)
 
Ideally both. The client should do validation for UX, but ultimately the server has to be safe when run from any client (even a modified or malicious one)
In this case it's not a big security issue, but might still cause problems
 
3:26 PM
That is very true.
 
@jacwah maybe this can be added to the issue about duplicate names? Seems closely related
 
@Phrancis Good idea, I'll do it
 
@Phrancis Waiting till Visual Studio is installed? :P
 
3:44 PM
> A closely related issue is that a client can log in with an empty user name. It's also possible to choose user names like " " (only space) and other problematic ones. We need to lay down some rules on what names are valid. I'll leave a suggestion below.

A user name...
- is unique. Two clients can not connect with the same name simultaneously.
- is at least 1 character and at most 20 characters long.
- contains only ASCII letters, spaces, underscores and decimal numbers.
- does not begin
 
		@Override
		public void onMessage(WebSocket conn, String message) {
			logger.info("Connection message from: " + conn + ": " + message);
			ClientIO io = webClients.get(conn);
			if (io == null) {
				logger.error("Message was recieved from unknown ClientIO");
				return;
			}
            try {
                byte[] bytes = Base64Utils.fromBase64(message);
                logger.info("Connection message from: " + conn + ": " + Arrays.toString(bytes));
                io.sentToServer(transformer.readOnce(new ByteArrayInputStream(bytes)));
^ This code is weird
It tries to b64 decode {"username":"jacwah","command":"login"}
And then catches an error that as far as I can see will never be thrown
 
4:04 PM
VS is downloading @skiwi for right now I'm actually setting up Outlook rules because I get some many (hundreds per day) Sql error emails. Setting up an error dump directory
 
@Phrancis I guess you first priority will be solving those so you don't get spammed? :p
Could imagine it's quite annoying
 
4:16 PM
@skiwi I was able to copy a colleague's rules so I just meed to go through and assign the folder for each
 
4:32 PM
I'm looking at the code for the websocket server. I can't understand how it works for you...
The client is not supposed to base64 encode messages, right?
 
@jacwah I think the libGdx client uses encoded or serialized messages of some sort, that may be why this is there; perhaps @bazola can shed some light?
 
@Phrancis Does it use Websockets?
 
4:57 PM
@jacwah I'm pretty sure it doesn't. I think it uses port 4242 which is the same the JavaFX client uses
 
The code I'm talking about is in the websocket server
 
@Phrancis Ah cool
 
@jacwah Ah. Have no clue then :o
 
Supporting the GWT client might've caused some weird stuff @jacwah
What file is it?
 
ServerWeb.java:65 (onMessage)
 
5:02 PM
@SirPython I have a question; I saw in deck builder you used {{card.properties.name}} in one place, and {{card.max}} another place. How do you differentiate which properties are part of which set? I'd like to document that as I will need to be able to query these fields in several areas of the client
(in particular for the card models where this will be crucial)
 
@jacwah If you look at the history here github.com/Cardshifter/Cardshifter/commits/…
Then in the very first commit if you look at the class there's nothing weird going on
This commit is the one that changes stuff: github.com/Cardshifter/Cardshifter/commit/…
As I thought, it has to do with supporting GWT
 
Nice detective work @skiwi
 
:P
 
The question is still how this works with the web client
 
Does it work?
 
5:11 PM
Not for me, but apparently for everyone else
Server threads dies on my machine
 
Hmm... how are you running the client?
In a browser?
 
Yeah
Using npm run develop
And then connecting to localhost:8080
 
What browser?
 
Firefox 39
 
That seems recent enough and should have ful support.... weird
 
5:14 PM
[Cardshifter/HTML-Client] Phrancis pushed commit c2c12f94 to sections-layout: Added ngFitText directive to www/index.html as this will be needed in some sections of the client. The code related to this directive is located in `/ngFitText/ng-
 
Working on Cardshifter while on lunch break == win
2
 
@skiwi LOL wow
@skiwi Either way the server shouldn't crash...
 
@jacwah Right, yes
 
What causes it to crash is that the b64 decode has assert length % 4 == 0
But that's garbage anyway because the message it receives is not base64 encoded...
 
5:19 PM
[Cardshifter/HTML-Client] Phrancis pushed commit 3f3a5c53 to sections-layout: Removed ngFitText directive references temporarily while card_model is being restructured for better functionality.
 
BTW.Work
 
@jacwah sorry if its already answered but yes, the GWT version of the libGDX client uses web sockets, and the mobile and desktop versions use regular sockets. And it does also crash for me as well when trying to use the websocket version with the cardshifter server
 
@bazola OK thanks for clearing that out
14 mins ago, by jacwah
The question is still how this works with the web client
^^ This though
 
well if the websocket stuff in the server is not working, then i don't think the web client will be able to interface with it either
 
Then how can the people working on the web client connect to servers? It's connecting through WebSocket AFAIK.
 
5:27 PM
for me, it is possible to connect to the server with the GWT version of cardshifter, and it is possible to start a game, but after a while it causes the server to crash and it stops working
 
Is "the GWT version" and github.com/Cardshifter/HTML-Client the same thing?
 
no
the gwt version is here: pixelpocalypse.com/cardshifter
 
OK I think we're talking about different things then. I'm asking about the HTML client
 
The HTML is the new client that the relatively new guys here are building, the GWT client is the same client as our existing desktop client but then uses a library/framework/tool to be able to host it on a website
 
So if the GWT client sends base64 encoded messages through WebSocket and the HTML client doesn't encode its messages, how can they both work? They don't on my computer, but obviously on other's.
 
5:44 PM
It's all magic to me
 
As much as GWT client is cool tech, I think it might be better to drop it in favor of the native html client once it works good
 
So JavaFx was "dropped" in favor of GWT and now GWT should be replace by an html client ? Am I wrong in the timeline here ?
 
[Cardshifter/Cardshifter] build for commit f82f972d on ???: The Travis CI build passed
[Cardshifter/Cardshifter] build for commit 068eb6c2 on ???: The Travis CI build passed
> The PR above (#348) now has Server.isValidUserName where validation easily can be implemented. I added a check for empty names.
 
6:41 PM
@Phrancis When you have time, can you take a look at my PR on #345?
 
> You could use `catch( InvalidUserNameException | UserNameAlreadyInUseException e) {
client.sendToClient(new WelcomeMessage(0, false, ex.getMessage()));` and just create the message with the exception. It would remove the if construct.
 
00:00 - 19:0019:00 - 00:00

« first day (374 days earlier)      last day (3187 days later) »