« first day (381 days earlier)      last day (3181 days later) » 

12:00 AM
RELOAD!
[Cardshifter/HTML-Client] 7 commits. 1 opened issue. 84 additions. 22 deletions.
[Hosch250/Rubberduck] 5 commits. 3362 additions.
[Hosch250/VSDiagnostics] 1 commit. 38 additions. 1 deletion.
[Vannevelj/VSDiagnostics] 1 opened issue.
[Vannevelj/VSDiagnostics] 5 commits. 3 opened issues. 2 closed issues. 160 additions. 13 deletions.
 
 
1 hour later…
1:06 AM
@Phrancis Do you still need help with this?
It's about a year old.
 
added handling for UpdateMessages, added findUser(by id)

The findUser by ID function is a function that will either return playerInfos.user or playerInfos.opponent based on the ID that was passed in. In case it is useful later, the function will return null if the ID does not pertain to either user.

Now, upon receiving an UpdateMessage, the code does this simply: finds the user with UpdateMessage.id using the new findUser function, and set's it's UpdateMessage.key property to UpdateMessage.v
 
@SirPython So much for that, thanks Graphics.SE
 
I have an answer if you still need it.
 
I know HTML supports SVG fine. I don't think either JavaFX nor libGDX do though.
Probably would be good to KISS and just stick to PNGs & JPGs
 
1:21 AM
But with libGDX it seems a little difficult.
 
Yeah. I'm pretty much ready to forget SVG, it'd be cool but YAGNI.
 
I thought it would be helpful for different size screens, because no resolution is lost.
 
1:46 AM
Hmm. Looks like libGDX requires you to sign a CLA to be allowed to contribute.
 
2:08 AM
@SirPython I think all said and done, the images will be small enough to where it probably won't matter much
 
I wasn't really referring to size, but more to distortion.
 
Ah right
This is about real-life size on the iPad
 
@Phrancis Is that HearthStone?
 
Yeah
 
I should download that to see what it's like.
 
2:17 AM
I think our images will have a different ratio, probably 16:9 or 12:9, to look more like a MtG / YGO physical card
MtG looks about 12:9 to me
YGO looks like a plain square. I don't like the layout as much though, and the print at the bottom is SO TINY
 
Yeah, we'll definitely need something bigger than that.
 
Oh, I like Pokemon's card layout the best I think
That looks like 16:9 ratio
They really went all-out with their newer, nicer looking cards
 
I agree. They seem to follow KISS: there isn't too much information on each card, but it's enough.
 
Yeah, I like that
 
Also, the lower one really has a "pop" to it.
 
2:28 AM
In our case, probably would be best to save the full description and flavor for a "zoomed in" or "detail" view of the card (say when you click it)
 
Almost as if the creatures is leaving the card itself.
 
Yeah it's gorgeous, wish we had that kind of budget :)
 
@Phrancis Something like that. Although, a right-click would be easier so it doesn't interfere with the other left-clicking functionalities.
 
yeah
Does JS support right-click and other keys fairly easiiy?
 
I think so.
I'll smoke the docs.
Hmm. MDN doesn't say anything about right-clicks.
See this for some cool things about menus: hacks.mozilla.org/2011/11/…
 
3:43 AM
Hey man I started feeling like crap a bit ago, think I'm about ready to pass out in bed
 
See you tomorrow, then. Hope you feel better tomorrow!
 
Wife cooked some fresh shrimp from the ocean she got on her trip to SC this weekend, can't stand the smells it's given me a headache and nausea
Didn't sleep well last night either. Today was kind of rough, physically
See you tomorrow
 
Ooh, I know what you are talking about. Well, a good night's sleep should help you.
 
Store bought shrimp is bad enough, but this ocean fresh stuff is the worst
gone...
 
 
5 hours later…
8:40 AM
hey
 
 
2 hours later…
10:22 AM
Hi. Just wanted to let you know that I'm back. I am a bit sick though so I will lay quite low for a few more days.
 
Welcome back @SimonAndréForsberg!
Hope you get well soon
 
> There's a lot of changes that has been done here, while some of them could be useful, others are not something I want to merge into this.

There's no way I am going to include that "If you would like to donate, please send to..." for example.
> Where is game board even getting `currentUser` from? I looked over the source and its not being injected via the constructor nor is it supplied via the template.

Either way, you're right, this smells. That controller is way too big and does too much.
 
@Duga I'd say that something went wrong there... That pull request makes no sense, maybe it should've been a single commit or something
 
@skiwi s/something/Malachi/
 
@Phrancis Downsides with AngularJS:
It's not isomorphic (javascript disabled = no website for you)
And it forces you to use a service like prerender.io (not free) if you want to have it indexed by search engines
AngularJS should be used for web applications, not web sites.
I would not recommend using it for the main project site of the TCG.
If you want to use a framework for a website, React is probably the one you want (or Backbone, but that's more complex); i.e, an isormophic framework
That's not a purists approach, btw - Another big metric for websites is the time it takes to load the page. AngularJS takes quite a while to load because you have to fetch the web page and then wait for Angular to parse itself. There are studies that show the longer the page takes to load, the more viewer retention significantly drops off. I would assume this is important if you want to 'show off' the game so to speak
 
11:06 AM
hello
Simon!
 
Hello @all
 
Hey @jacwah
 
@Phrancis!
 
hey @Phrancis and @jacwah and @DanPantry
 
11:12 AM
@DanPantry Those are all extremely valid points, thanks so much for sharing... Perhaps it would be wise for me to open an issue on that for discussions?
@SimonAndréForsberg I'm about to leave for work, but I'd love to hear about your trip whenever you get to feeling better
 
ttgtw, should be back in a little while
 
11:57 AM
Hey everyone! Welcome back @SimonAndréForsberg !
 
@Phrancis I think so
 
12:22 PM
Hey
 
hey @Phrancis! :D
 
Hey @Phrancis :-)
 
12:52 PM
> I believe this is caused by you not having chosen a mod to play with. Related to #299.
> I thought there was a destroy() effect already... if not, that should be added.
> Technically, the resources don't need to be with ALL_CAPS, that's just what I used originally (they used to be Java enum values once upon a time, where the naming convention says to do it like that). They can absolutely be shown lowercase. But all this is irrelevant to the actual issue here.

Specifying the full mod-dir every time might be a bit excessive, `imagePath 'chinese/jade-emperor.jpg'` perhaps?
 
I see Simon is catching up !
 
> Thanks for taking the time to investigating this. If you have any clue why those aren't included in the dirty classpath, I'd love to hear about it.
> Related to another issue with some clients showing the lobby even if the server is down (won't look for it at the moment, or I have just been dreaming this...). Clients should not switch to the lobby screen until they receive the 200 status. If they receive another status, they should just show the provided message to the user.
> I think this is entirely server-related. The issue is probably (without having looked at the code recently) that when the client has disconnected, the server sends this information to all clients (including the client itself), but when sending the info to the client it cannot do that because the socket is closed, so it tries to disconnect the client, and then informs all clients that the client has disconnected (including the client itself), but it cannot send the message to the client becau
> > This happens only when not using (the latest snapshot)

That, is exactly why there is snapshots.

Use an updated version of both the client and the server and everything is fine. I don't see any issue here.
> Don't allow more than one consecutive space, such as some name --> some name
> I think ZoneChangeMessage helps reduce the network traffic used by the game. I don't really see a reason for GetZoneMessage. From what I'm hearing, it would be only be used early on in the game because.... well... because why exactly? What is the problem about deciphering ZoneChangeMessages? You need to do that at some point anyway. Or do you intend on sending GetZoneMessage requests after each turn and determine what cards have moved where?
> As far as I know, CardInfoMessages are only sent for cards in known zones. Please clarify.
> First of all, my intention is that the ids should be seemingly more random in the future. (At the moment, if you know the deck list it is theoretically possible to know what cards you are going to draw next by looking at the card ids)

Secondly, I think "Put this card into your deck and shuffle your deck" is a better example of problematic card. Graveyards tend to be public, so you would know which card is sent to the hand.

I feel however that there are some benefits of sending the card i
 
1:34 PM
@Phrancis What's the status of the HTML-Client? How much progress has there been during my absence?
 
Welcome back, @SimonAndréForsberg!
 
hi and thanks
 
So far, the client only needs to be able to do one more thing: be able to correctly handle sending action messages, and receiving usable action messages.
@Duga I'll try to find the point in the server log that I am referring to.
@SirPython And, it has only been tested on Mythos.
 
> My bad. (1), I meant to say ZoneMessages, and (2) I now know that the ZoneMessages entities properties only shows the ID of the card, no actual information.
> @Zomis "Or do you intend on sending GetZoneMessage requests after each turn and determine what cards have moved where?" Yes, that was the intention: simply update all the zones after every... never mind. Thinking about it again, it seems too overkill.
 
2:11 PM
good to see ya @SimonAndréForsberg
 
hi @bazola
 
hope your trip was good and sorry you are not feeling well
 
3:15 PM
Aug 6 at 22:54, by jacwah
That's what I get too, my question was why it doesn't look in extra-resources/mods
@jacwah In your IDE, setup the run config to use extra-resources as the working-directory. then it will be fine
 
monking
 
3:40 PM
hey @DanPantry
It feels so good at work to actually know a bit about most technologies ;)
WebSockets, ORM, MVC, etc.
 
@skiwi it does, doesn't it! :P
I tend to learn all the new tech so I'm ahead of the curve
 
Ahead of the curve! ahaha at my place there is still a lot of COBOL and some Java 4, so being ahead of the curve here is not a big problem!
 
Was looking at Sails ORM and it does really resemble Groovy's GORM a lot
@Marc-Andre Ouch... lol
Also installed a JSON formatting extension on my browser today...
Why hadn't I done that earlier?!
The thought of such thing existing never even occured to me
 
> I think that on top of the changes to the client's handling of status code, the server should close the connection if the login attempt fails. It should at least not start to send lobby data as you can see in the logs above.
 
@Marc-Andre That code is probably older than me.
@skiwi What, like, prettyprint?
 
3:50 PM
@DanPantry Just something that formats it and gives you the option to hide/show things
 
so yes ;D
 
> > Use an updated version of both the client and the server

Running the server off the `develop` branch fails. This is not an issue with an outdated server.
> Would it be worth having a version check handshake to prevent this BTW? Notwithstanding @Jacwah's comment
 
4:37 PM
@DanPantry There is code from the 80's so yeah.
 
unity is really not too bad! ive already got a prototype of an open world game with all of my models spawning, and i have already made powerups that you can pick up that make your character move faster and jump higher, and i already have them randomly spawning above the buildings when the world is created :)
 
Unity is awesome! :D
 
> there were some things that I didn't want to merge with your code Zomis. I apologize, I don't have a good grasp on using github just yet. I agree with your conclusion. I am guessing that github automatically grabbed some changes from after I submitted the merge request but before you guys checked it out?????
 
@Lokkij MonoDevelop, shivers...
 
You can use Visual Studio with it just fine
 
4:54 PM
> Correct, when you open a Pull Request, then you are requesting your branch to be merged in target branch. So all commits you make after opening the PR are "added".

FYI, in this image you can see when you synchronized your PR, that is, when you added new commits on an open PR. Unfortunately the actual opening of the PR is not in the list, probably Duga was down.

![schermafdruk 2015-08-18 18 53 55](https://cloud.githubusercontent.com/assets/2961606/9336574/96ec0f52-45da-11e5-9030-2d68cf892
 
5:13 PM
> Agreed. At the moment, it sends the lobby data because of the querymessage that it gets. Closing the connection will solve that.
 
Some interesting stuff up here
Wish I had time right now to look at that JavaFX for mobile
 
> I'm not sure what happened [here](http://chat.stackexchange.com/transcript/message/23501459#23501459) but the markup was suddenly messed up when creating a link:

> [Vannevelj/VSDiagnostics] `[build]()` for commit 5f4c34cc on ???: AppVeyor was unable to build non-mergeable pull request
> For example: http://chat.stackexchange.com/transcript/message/23480071#23480071

You'll notice however that other build messages do contain a correct link.
> > I have come to the conclusion that it was never possible for the HTML-Client to connect to any server other than a local one.

I believe that is because the other servers are outdated.

> Running the server off the develop branch fails. This is not an issue with an outdated server.

It is, actually. The changes to fix this issue seems to have been done on the `rules` branch, which has not been merged into `develop` [as far as I can see](https://github.com/Cardshifter/Cardshifter/netw
> Could you provide the JSON request that was sent to Duga for this? Should be available in the github webhook settings (you'll just need to find the right ones)
> Could you provide the JSON request that was sent to Duga for this? Should be available in the github webhook settings (you'll just need to find the right ones)
> Request Headers:

Request URL: http://stats.zomis.net/GithubHookSEChatService/hook?roomId=26639
Request method: POST
content-type: application/json
Expect:
User-Agent: GitHub-Hookshot/2283109
X-GitHub-Delivery: 7e3adb00-44f9-11e5-9594-7fd57f909fe3
X-GitHub-Event: pull_request

Request Payload

{
"action": "opened",
"number": 170,
"pull_request": {
"url": "https://api.github.com/repos/Vannevelj/VSDiagnostics/pulls/170",
"id": 42604618,
"html_url": "https://gi
> Request Headers:

Request URL: http://stats.zomis.net/GithubHookSEChatService/hook?roomId=26639
Request method: POST
content-type: application/json
Expect:
User-Agent: GitHub-Hookshot/4963429
X-GitHub-Delivery: 01753300-45c5-11e5-8dfb-2a2b9f7ec5a9
X-GitHub-Event: pull_request

Request Payload:

{
"action": "opened",
"number": 177,
"pull_request": {
"url": "https://api.github.com/repos/Vannevelj/VSDiagnostics/pulls/177",
"id": 42718413,
"html_url": "https://g
 
5:36 PM
Hello
 
> The chat message is about a build event, the provided JSON request is about a pull-request event. Please provide the JSON request that caused that particular chat message.
> The chat message is about a build event, the provided JSON request is about a pull-request event. Please provide the JSON request that caused that particular chat message.

(And please add a new issue for that Internal Server Error message for that particular pull request JSON data there...)
 
5:55 PM
@skiwi im gonna try to give it a watch
 
> - use `CreatureTypeComponent.has` method and don't use the `.getCreatureTypes()`
- use a Set instead of a List for the creature types. Order does not matter and they should be unique
 
6:31 PM
@SimonAndréForsberg I solved it since then and added it to the build/run guide :)
 
@jacwah good, thanks.
haven't completely read the chat transcript yet
 
> @Zomis So the snapshots are not based off the develop branch? If that's the case, it's the reason for this confusion.
 
@SimonAndréForsberg That would be quite a feat!
 
> The snapshots are based on whatever commit I run gradlew snapshot from.
> Request Header:

Request URL: http://stats.zomis.net/GithubHookSEChatService/hook?roomId=26639
Request method: POST
content-type: application/json
Expect:
User-Agent: GitHub-Hookshot/2283109
X-GitHub-Delivery: 8135cb80-44f9-11e5-8986-5862787ec725
X-GitHub-Event: status

Request Payload:

{
"id": 286776118,
"sha": "08c44809526ed2d81ecf70433a688bf9278b2966",
"name": "Vannevelj/VSDiagnostics",
"target_url": "https://ci.appveyor.com/project/Vannevelj/vsdiagnostics/buil
 
@Duga Lol well obviously, but usually projects build snapshots/nightly builds from their development branch @SimonAndréForsberg
 
6:41 PM
> Request Headers:

Request URL: http://stats.zomis.net/GithubHookSEChatService/hook?roomId=26639
Request method: POST
content-type: application/json
Expect:
User-Agent: GitHub-Hookshot/2283109
X-GitHub-Delivery: 260cd480-450b-11e5-91f6-913427f7afa7
X-GitHub-Event: status

Request Payload:

{
"id": 286884334,
"sha": "08c44809526ed2d81ecf70433a688bf9278b2966",
"name": "Vannevelj/VSDiagnostics",
"target_url": "https://ci.appveyor.com/project/Vannevelj/vsdiagnostics/bui
 
@jacwah true that. but this is not anything that is done automatic. you have a point though that it would be preferred to make snapshots from develop
 
I think that snapshots should at least include the commit hash their made from somewhere
Aug 11 at 20:11, by jacwah
The commit hash should be distributed with snapshots...
 
@jacwah I have considered that, yes. I'm planning on adding that.
 
That would help if a similar issue crops up in the future :)
 
7:02 PM
s/if/when/
could you make it a github issue just so that we/I don't forget it?
 
> I'm not sure what happened [here](http://chat.stackexchange.com/transcript/message/23501459#23501459) but the markup was suddenly messed up when creating a link:

> [Vannevelj/VSDiagnostics] `[build]()` for commit 5f4c34cc on ???: AppVeyor was unable to build non-mergeable pull request

Request Header:

Request URL: http://stats.zomis.net/GithubHookSEChatService/hook?roomId=26639
Request method: POST
content-type: application/json
Expect:
User-Agent: GitHub-Hookshot/4963429
X
 
7:15 PM
Maslow's hierarchy of needs, updated for 2015. http://t.co/b267vu6xyp
 
> I guess the only reasonable course of action is to check if `"target_url"` is empty or not and if it is, just show this:

[Vannevelj/VSDiagnostics] build for commit 5f4c34cc on ???: AppVeyor was unable to build non-mergeable pull request
 
@SimonAndréForsberg Sure
 
> That makes sense.
> When a PR builds on AppVeyor, the message shown in chat will contain a link to

1. AppVeyor
2. Commit
3. Branch

Would it be possible to also link to the exact PR?
> Could you show an example of the JSON data that is sent to Duga? If the JSON contains the required information, the answer is yes. Otherwise no.
 
Phew busy day in here today!
 
You know it's really funny that here we are talking about lists, sets, etc.
And in Qt/QML all I have is a dumb Javascript array :|
 
Just one of them? ^ :)
 
Howyamean?
 
Like... you have only one, so you keep having to ... jokes aren't as funny after you try to explain them :|
 
7:46 PM
> Hmm. I don't immediately see anything that points towards it which is strange because the UI in AppVeyor itself indicates that it knows it's working on a PR.

{
"id": 287881646,
"sha": "783d44335558f8d0eff587281bdd60aa9632d2a8",
"name": "Vannevelj/VSDiagnostics",
"target_url": "https://ci.appveyor.com/project/Vannevelj/vsdiagnostics/build/1.8.134",
"context": "continuous-integration/appveyor",
"description": "AppVeyor build failed",
"state": "failure",
"commit": {
"sha":
 
@Phrancis I don't get it
 
Don't worry, I don't either
~goes back to his tables...~
 
4 mins ago, by skiwi
And in Qt/QML all I have is a dumb Javascript array :|
a = singular. you have only one array.
And in Qt/QML all I have is dumb Javascript arrays :|
 
Oh...
Ooooooooooooooooooooooooooooooohh
Well, QML is not that bad ;)
 
> seems like the only thing that references the PR is the commit message,

> "message": "Merge pull request #179 from Vannevelj/issue174\n\nCloses #174",

Don't think that is a good option to use though.
 
8:04 PM
> It's always possible to know the exact code a release was built from because of tags, but the only identification of a snapshot is the version number preceding it and the date it was built. This is not enough, as shown by the recent confusion in #353.

I think the easiest solution would be to create a text file in the root of the package archive containing the commit hash and optionally the branch name. It could be called `version` or similar. An easy was to get this information is parsing t
> There are gradle plugins that can be used to interact with git. Should be possible to use that to find out what the current branch and commit SHA is.
 
 
1 hour later…
9:28 PM
 
9:53 PM
@SirPython Hey!
 
Hey, @jacwah!
How are you?
 
I'm good! Fixing everything @Simon has commented on :)
How are you?
 
I'm good as well! Fixing the code before SimonAndréForsberg comments on it :)
 
10:11 PM
the game just got a lot more fun :)
0
Q: Do you want to be a super (voxel) hero?

bazolaI have implemented an upgrades system in my Unity3d game, and I am pretty sure that I am not doing things in an optimal way. Any advice regarding best practices would be much appreciated. The idea is that when the player collects a powerup, their run speed and jump speed increases. Collect enou...

 
10:23 PM
 
@SimonAndréForsberg Thanks for the comment about parameterized tests! I didn't know about them and they're really handy. Updated code above ^
 
[Cardshifter/Cardshifter] build for commit cde2bb64 on ???: The Travis CI build passed
 
@jacwah I was just having some problems connecting to an SE site, too.
 
It's working now, huh
 
10:39 PM
> @danpantry I completely agree about the controllers doing too much. Now looking back on it, it's like reading C code where everything is in the main method (I plan on starting to fix this after the client is fully playable).

The `currentUser` comes from [`src/login/controller.js`](https://github.com/Cardshifter/HTML-Client/blob/master/src/login/controller.js#L18).
 
 
1 hour later…
11:45 PM
Loki Astari vs. Simon André Forsberg: 2985 diff. Year: -2791. Quarter: -1195. Month: +105. Week: +30. Day: +20.
200_success vs. rolfl: 3449 diff. Year: +2702. Quarter: +3036. Month: +457. Week: +240. Day: +110.
Mat's Mug vs. Simon André Forsberg: 1281 diff. Year: +2171. Quarter: +667. Month: +945. Week: +218. Day: +135.
200_success vs. janos: 15528 diff. Year: -4072. Quarter: -1151. Month: -337. Week: +146. Day: -29.
 

« first day (381 days earlier)      last day (3181 days later) »