« first day (29 days earlier)      last day (3532 days later) » 

1:17 AM
@SimonAndréForsberg I think this might be more a matter of which properties should be sent to the client. The only thing the client needs to work with the card is an ID, and then it needs some properties to display. I was imagining this working like this...
Each card would have a function named something like getDisplayProperties, and this would return a table with just the stuff the UI needs to display the card. This table gets sent over the wire, and the UI drops the properties of the table into an SVG template... for example if the tables have "name," "description," "health," and "defense" properties, the SVG files would have matching tokens like {name} and {health}, and the data would be inserted with simple replacement
The SVG files themselves would be part of the mod, sent to the client at the beginning of the game (there would probably only be one or two "templates" for an entire game), so there is no problem with the results of getDisplayProperties (or whatever) matching the placeholder values in the SVG
 
 
10 hours later…
11:08 AM
Monking-king
 
Monking
I was thinking... is there anything preventing us from making both a Java API and a Lua API?
Just a thought I had
Anyways....
 
> Added API module
> Renamed package name for server.messages -> server.incoming
> Added ids to cards and zones
 
@SimonAndréForsberg Nothing at all, if people want to code in Java, I don't see why they shouldn't be able to do so
 
> Added (unimplemented) Command-Line Options for playing network
> Extracted LuaTools
 
@skiwi My thought exactly :)
 
11:18 AM
> Added getIndex method to Player and size
> Added getname method to player (why wasn't that there before?)
> Added an API module
> Incoming and Outgoing Message classes
 
I'm really curious as to why you do not push every change you make directly... It's like how everyone works ;-)
(Would also integrate better with Travis CI)
 
> ClientIO setName method
> Refactoring and addition of message handlers
 
If your commit is a finished commit (that is, it actually works), I think it's best to push it as well
 
@skiwi Because at the moment I need to enter my username and password every time (I should really fix so that I don't need to do that...)
 
then go fix :D
 
11:21 AM
@skiwi Even these commits are not totally 'finished'. There is a lot of things that needs to be added. It compiles, but it still isn't totally possible to play a network game
 
@SimonAndréForsberg That doesn't matter, that is exactly why it is not on the master branch
 
Correct :)
 
Can you also code the user/pass in the .git file?
I need to also enter it when I do command line work
 
you need to use SSH keys when working on the command line
 
okay, I should set that up as well
 
So then you can also set it up for egit?
 
I don't use egit for the pushing, I always push from the command line
even though I probably could use egit for that as well
 
Ah understood
With Netbeans most basic functionality works from there, for advanced stuff I use command line
 
11:43 AM
I prefer to use the command line for git, I get a better understanding of how it works then
I have setup my SSH key now but I still need to enter the passphrase for it everytime I push, but perhaps that's the plan
 
Building this github bot made me understand git a whole lot better
 
And just in case you run into the same problem I had:
204
Q: Could not open a connection to your authentication agent

Danny Dai SmithI am running into this error of: $ git push heroku master Warning: Permanently added the RSA host key for IP address '50.19.85.132' to the list of known hosts. ! Your key with fingerprint b7:fd:15:25:02:8e:5f:06:4f:1c:af:f3:f0:c3:c2:65 is not authorized to access bitstarter. I tried to add t...

 
12:09 PM
Ah, silly me. I could choose to not have a passphrase on my ssh key
 
12:56 PM
@SimonAndréForsberg Smart
 
1:10 PM
@SimonAndréForsberg What is the preferred way to ignore a property in JSON?
Or even possible to publish it as a kind of raw object?
 
@skiwi @JsonIgnore ?
what do you mean publish as raw object?
 
Raw object might be more friendly
Say a JSONNode, in which you can query the content if it's present
 
perhaps simply declare it as a JsonNode ?
 
Perhaps it'll work
The Travis CI API is a bit weird
 
1:22 PM
and there's no debug func I think
maybe in the build log
> Webhooks are delivered with a application/x-www-form-urlencoded content type using HTTP POST, with the body including a payload parameter that contains the JSON webhook payload in a URL-encoded format.
@SimonAndréForsberg github.com/skiwi2/TestRepo is this also offline for you?
 
>
13:16 UTC We're performing a short maintenance procedure which will temporarily impact access to a small number of repositories
Great
First time I've seen it
 
Cardshifter is offline as well
 
I think the actual repo is still online
Ugly
skiwi2@github.com/skiwi2/TestRepo.git: 503 Service Unavailable
very ugly
@SimonAndréForsberg It's online again
 
1:59 PM
@SimonAndréForsberg Could you help me with a very minimal maven file?
It only has
public class Main {
    public static void main(String[] args) {
        System.out.println("1");
    }
}
as source
 
@skiwi sure, for what?
 
and that's it
Netbean's Ant's build.xml would drive me crazy
 
Umm... maven file? That's a java file?
 
Yes, that's my only main
In main.Main
 
So, you want a pom.xml for that file?
 
2:00 PM
yes
 
Ping: Avoid administrative distraction.
[Zomis/TestRepo] Zomis pushed commit 639c2893 to master
> added pom.xml for skiwi
 
@skiwi There you have one ^^
 
Thanks @SimonAndréForsberg
Can you PR it?
:)
or well
you already have a PR going I see
> We can’t automatically merge this pull request.

to resolve conflicts before continuing.
Interesting
 
Yeah, there's some merge-conflicts that would happen because we have both made changes to our own repository, without merging them
you can setup my repository as a remote though
 
I'll just paste it
Let's see if it works
It works @SimonAndréForsberg
Now I've got to figure the webhook out
Hmm
Caused by: java.lang.IllegalStateException: Cannot convert value of type [java.lang.String] to required type [com.skiwi.githubhooksechatservice.travis.events.BuildEvent]: no matching editors or conversion strategy found
@SimonAndréForsberg
	@RequestMapping(value = "/payload", method = RequestMethod.POST)
	@ResponseBody
	public void build(final @RequestParam("payload") BuildEvent buildEvent) {
		chatBot.postMessage(MessageFormat.format("\\[[**{0}**]({1})\\] [**build #{2}**]({3}) for commit [**{4}**]({5} on branch [**{6}**]({7}) {8}",
			buildEvent.getRepository().getOwnerName() + "/" + buildEvent.getRepository().getName(),
			buildEvent.getRepository().getUrl(),
			buildEvent.getNumber(),
			buildEvent.getBuildUrl(),
			buildEvent.getCommit().substring(0, 8),
It doesn't seem to understand that @RequestParam("payload") has JSON in it
It is currently treating it like a raw String
 
2:15 PM
Make it a @RequestParam("payload") String string and log the contents of it
And show me the contents of the request
 
@SimonAndréForsberg docs.travis-ci.com/user/notifications/#Webhooks-Delivery-Format that you can see here
 
Ping: Avoid administrative distraction.
 
Ah, on github there's an option
Travis CI uses the 2nd version
 
correct
 
But your bot only uses the 1st
 
2:19 PM
So I need to grab the string and do manual conversion?
4
A: JSON parameter in spring MVC controller

jhadesdevThis could be done with a custom editor, that converts the JSON into a UserProfile object: public class UserProfileEditor extends PropertyEditorSupport { @Override public void setAsText(String text) throws IllegalArgumentException { ObjectMapper mapper = new ObjectMapper(); ...

or this
 
3
Q: Spring @MVC and the @RequestBody annotation with x-www-form-urlencoded data?

jsumnersI am trying to figure out why I can't receive a request from a jQuery.ajax call when then Spring @Controller handler method includes a @RequestBody annotation. Consider the following: HTML/JavaScript: <form id="foo" action="/baz"> <input name="bar"> </form> <script> $(function() { var ...

 
2:30 PM
> Added more Javadoc
> Using interface instead of implementation in test
 
Sigh... How hard can it be to properly document your API?
 
Quite hard actually
 
If they over at Github and Travis CI would actually test their own webhook API, they would notice there's suddenly new properties
 
Guess why I just did this:
2 mins ago, by Duga
> Added more Javadoc
 
:)
Aaand a new ondocumented feld
Now it should be working, I swear
It's working!
 
2:39 PM
good
 
Will push and finish it later today as I need to check the documentation, etc.
 
2:56 PM
> Added more comments and did some restructuring
 
Hmm, if I would post that question now it would be around 13K chars
And then I need to write a description of it as well...
 
> Hiding implementation details
 
3:13 PM
^^ That was easier than expected
 
i made pong
with the Lua ECS thing
the ECS can hold its own but still having a hard time visualizing how it will be used with a TCG exactly. Feel free to use it though.
 
4:02 PM
Ping: Encourage flow.
 
4:26 PM
And back
@SimonAndréForsberg you know anything about the "random" pings? :)
Keep connecting repos?
 
Okay
time to write up the readme for the travis part
 
> Added travis hook support.
 
@SimonAndréForsberg Would you mind restarting @Duga? Then I can add the travis hooks in a bit
 
4:33 PM
TTQW
Monking!
 
Monking @Duga
Oh I've got to eat right now, will hook up the bot project first in a bit
 
 
1 hour later…
5:44 PM
@SimonAndréForsberg Should we test in chat here with the build started messages? Or leave them out from the start
 
> Added travis support.
> Added travis build web hook.
[skiwi2/JavaLuaInteropTest] skiwi2 deleted tag latest-master
[skiwi2/JavaLuaInteropTest] build #33 for commit 2b5f8e39 on branch master passed
> Added travis support.
[Cardshifter/Cardshifter] build #1 for commit 8311526c on branch master passed
2
 
6:12 PM
Nice nice
 
 
2 hours later…
8:11 PM
@Malachi Cannot seem to reach you, read your question two days ago... There should be a fair explanation on the github site itself if you are familiar with Tomcat, if not catch me or @SimonAndréForsberg at some point
 
Does Malachi read this room?
try ping him in the second monitor
 
Doesn't he get the ping either way?
@SimonAndréForsberg Is this extra blank line an Eclipse setting?
public class CardAction extends UsableAction {

	private final Card card;
 
@skiwi I don't think so. Not sure.
@skiwi nope, just my personal preference I guess
 
8:31 PM
Huh... I've got a purge branch
@SimonAndréForsberg So I better leave them like that when I work on your code?
We've still got to make at least some rules about coding standards
Do you know a way to check the status of all local branches? So I could delete unnecessary ones
 
The only coding standard I follow (I think) is: Don't change someone else's coding standard
@skiwi git branch purge should remove the unused ones I think
 
@SimonAndréForsberg Yeah, trying to keep up with that one as well
@SimonAndréForsberg Might explain why I have a branch called purge
 
I think it's prune?
 
I believe there are much more important things to focus on than stylistic coding standards
probably yes :)
 
8:34 PM
Nope, now I got a prune branch as well
 
Ah, it's git remote prune
to remove removed remote branches
 
That's for pruning the local remote branches though
I think other one is manual then?
 
What is it that you want to do with your local branches? Which ones do you want to delete?
 
Fixed it now, but it was the core-game and prune and purge branches
 
That's step 1
Uh... I broke something
And why didn't @Duga say so?1
Might be that we need your error log here @SimonAndréForsberg
 
@skiwi what did you break?
You created a branch, have you done anything else?
 
No, that part is fine
but @Duga is supposed to say that Travis build failed
 
[Cardshifter/Cardshifter] skiwi2 deleted branch core-game-improvements
 
@skiwi when/how did Travis build failed?
 
Doesn't Travis only do builds for the master branch at the moment?
 
Nope, it does on all
Generally it only builds releases from master branch though
 
Oh well
But if the other branches don't build, that's alright, right?
 
master is the important one?
 
8:44 PM
All branches should build at all times (and pass tests)
Else other people cannot work on it properly
@Duga didn't notify of a failed build
There will an update for @Duga incoming for that bug
 
@skiwi I'm not sure I agree with that. What if you want someone else to finish it?
In my opinion, the reason for using branches is so that the master branch always branch successfully
 
As far as I've learned it, master means feature-complete and other branches mean that they should still build and pass tests properly
 
If you have the requirement of all branches always needs to pass building, then the whole point of using branches in the first place is reduced
That sounds unreasonable to me
 
How so?
 
What if one developer only wants to finish half the feature, and let someone else take care of the rest? Then it does not necessarily build successfully.
20
Q: Is it good idea to require to commit only working code?

AstronavigatorSometimes I hear people saying something like "All committed code must be working". In some articles people even write descriptions how to create svn or git hooks that compile and test code before commit. In my company we usually create one branch for a feature, and one programmer usually works ...

 
8:49 PM
@SimonAndréForsberg That's something I don't think should happen in the first place for half-working code, I would guess
@SimonAndréForsberg Good question, I'll read that
Firstly I'm wondering how this builds
public class Main {
    public static void main(String[] args) {
        System.out.printl("1");
    }
}
In that example pom you send there @SimonAndréForsberg, did you include the source directory?
 
No, it uses Maven default source directory
I would never make big not staible commit. I prefer to make always small commits so i can always find out in history what happend. And so on Jeff, ask yourself what would happend if you commit hundreds of modified files with hundreds new and changed lines of code in one commit? No one except you would not understand what happend. Maybe it is better to make commits while developing? — Astronavigator Nov 14 '11 at 11:20
^^ That comment
 
@SimonAndréForsberg What one is that then? Easier to modify that one
 
@skiwi src/main/java/com/package/subpackage
I prefer making small commits, and sometimes that means that something might be broken.
 
I could agree with restricting travis-ci builds, but I still disagree personally with the statements there
 
I consider our non-master branches to, at least at this stage, be feature branches.
 
8:54 PM
I have never intentionally committed non-compilable code here, I cannot really imagine how you would get something like that
If I refactor an interface for example, then the interface itself and all refactored files (with minor refactorings) would be a single commit
@SimonAndréForsberg The com is in this case also package name?
 
Of course you should try to always commit compilable, working, code. But sometimes that's just unreasonable to expect
@skiwi yes
 
@SimonAndréForsberg There's a [skip ci] tag always available in commits to not do any CI on that... would that suffice?
It's usually used for updating readmes, etc.
Then you are still thinking along the lines that ideally every commit compiles, but you can specify when that's not the case
 
I would not like that. I think for some branches it is just perfectly fine to not compile
 
How would we go about it then
First you are saying that you should try to make them compile, but that doesn't seem to agree with that it would be perfectly fine if they didn't
I don't mind as long as we get a concensus
and shoudl figure a way to opt-in for travis tests in a branch
 
trying to make them compile does not mean that they always have to compile, which leads to the conclusion that it is ok if they don't compile
 
9:03 PM
I'd want to have it enabled on my branch for example
 
sure, feel free
 
I cannot replicate the bug with @Duga here locally, can you check the error logs to see what's written there?
 
I don't really expect us to work on 'each others branches' much, but if I do, I will keep your preference in mind
 
> When a travis build fails, the failure event does not get posted by the chatbot.
 
I've got to go now, university starts tomorrow, can you add the logs there?
 
9:07 PM
it says nothing
I do not see any log information about it whatsoever
 
That's... not good
It just worked here
Cannot do much to fix it though
 
> Each branch should have the option to opt-in for travis support, most likely via a small configuration file.
This should be by default enabled on the master branch.
 
9:36 PM
[Zomis/TestRepo] Zomis pushed commits 528df4a6 and 639c2893 to test42
> Merge pull request #6 from skiwi2/pull-request-test

pull request test
> Added travis configuration file.
> Merge origin/master
> Added language Java to travis configuration file.
> Moved travis configuration file from source to the correct root.
> Added example Java main and netbeans configuration.
> Lowered compatability to Java 7.
> Switched from Netbeans Ant to Maven.
 
> Fixed on start property in travis configuration file.
> Forced build failure.
> Changed project structure to default Maven structure.
> Fixed compilable code, added failing test.
[Zomis/TestRepo] Zomis pushed commit 289b094d to test42
> merge
[Zomis/TestRepo] Zomis pushed commit d6e113c4 to test42
> merged with skiwi
[Zomis/TestRepo] Zomis created pull request #1: Test42 to merge test42 into master
[Zomis/TestRepo] skiwi2 pushed commits 9d05bbe6 and 28c29ba2 to master
[Zomis/TestRepo] skiwi2 pushed commits 7cda954c and 73eca0c2 to master
 
Oh, shut up @Duga
 
[Zomis/TestRepo] skiwi2 pushed commits 85f9d315 and 3958a14a to master
[Zomis/TestRepo] skiwi2 pushed commits 9f0566df and dd9276f6 to master
[Zomis/TestRepo] skiwi2 pushed commits 697aa06a and 24f9ed54 to master
[Zomis/TestRepo] skiwi2 pushed commits cc0d79aa and 45649b67 to master
[Zomis/TestRepo] Zomis pushed commits fd23b641 and 6cc49980 to master
[Zomis/TestRepo] Zomis pushed commits d1541319 and 289b094d to master
[Zomis/TestRepo] Zomis pushed commit d6e113c4 to master
[Zomis/TestRepo] Zomis pushed commit 69c1003c to master
> Merge pull request #1 from Zomis/test42

Test42
[Zomis/TestRepo] Zomis merged pull request #1: Test42 from test42 into master
[Zomis/TestRepo] Zomis deleted branch test42
 
10:05 PM
> Added some new cards
> Updated readme, made some StackAction methods final
> Showing full Zone name in Swing view
> Improved and added several components and systems
> updated gitignore
 
In some of my projects, I really need to commit and push to github more often...
Those changes are ooooold
 

« first day (29 days earlier)      last day (3532 days later) »