Think about L2JServer future

[MELERIX]

the contributions are not ignored.

the problem is that "there is a lot of contributions" (mostly Java ones), and only few active devs checking these things, so obviously is hard check and test all of these manually.

and specially more hard if community doesn't help too -.-

then, why something is not committed yet ? generally is because code doesn't match standars (format, typos, wrong coding, etc) or in some cases code is too custom, and some others reasons.

anyway that doesn't means that the contribution will not be committed, this just means that code should be verified, and not only by active devs, it should be verified and tested by the community users too.

in short words... proper test results and good feedback = fast development.

now, about money... I don't think it will make L2J Dev's work more interesting.

difference here is that some people prefer to work for public and for free (self motivation), and others prefer to work privately and for themselves (selfish attitude).

[marcose]

the contributions are not ignored.

the problem is that "there is a lot of contributions" (mostly Java ones), and only few active devs checking these things, so obviously is hard check and test all of these manually.

and specially more hard if community doesn't help too -.-

then, why something is not committed yet ? generally is because code doesn't match standars (format, typos, wrong coding, etc) or in some cases code is too custom, and some others reasons.

anyway that doesn't means that the contribution will not be committed, this just means that code should be verified, and not only by active devs, it should be verified and tested by the community users too.

in short words... proper test results and good feedback = fast development.

now, about money... I don't think it will make L2J Dev's work more interesting.

difference here is that some people prefer to work for public and for free (self motivation), and others prefer to work privately and for themselves (selfish attitude).

Maybe we need a L2J Tutorial, dedicated to java for l2j.
Like the Wiki, but everyone can put they codding style and a vote system to valorate or read it and add info to the topic. So, we make in l2jserver java for l2j experts.

I can help whit that, I don't have an java licenceature but i know something about the language.

[BiggBoss]

agree with a explanation (mostly visual) about how l2jserver is internally organized, what does each class do and how they work together. About learning java, oracle java tutorial & practice

[jurchiks]

@Zoey - you completely missed the point of my post. You only suggest asking the Adv. Users, but they have the most biased opinion because they're in your team. Others, who are NOT associated with any dev team, would have a completely different opinion.
I don't usually listen to l2j team members because they've proven to be highly stubborn (I'm not saying I'm not stubborn myself, but still) and frequently unable to give exact reasons for their actions or even answer reasonably at all.

@MELERIX - contributions are ignored, a vast majority of them. That's the only reason why so many people have stopped contributing; because their contribs are either ignored or not taken seriously.
And even if they are not ignored, then why is there only a couple of commits per week MAX? You're not working on anything yourselves and not checking the contribs either, it's obvious.
It doesn't take months to develop an SQL-to-XML conversion, it's a work for a couple of days at most, and a large part of all commits are bugfixes and minor stuff - 49 commits in last 30 days, 25 of them are "fix", "formatting" or "cleanup", the last two of which is what Eclipse does either automatically or with 3 keyboard keys, 4 more are either reverting previous changesets or deleting unused/old code and the rest is what you could actually call real development.

I'm not saying you should now reveal a grand plan to rewrite all of l2j core in a couple of weeks (nobody would believe it anyway), but at least don't lie about doing a lot and not having the time to check the contributions, because you have more than enough time, you just don't want to do it (which is understandable - not many would want to do it -, but shouldn't be lied about).

But we're not here to talk about that, are we?
Somebody should clean the topic to get it back on track already.

[BiggBoss]

maybe many contributions are ignored because they are not well done or doesnt worth to commit them, because those wont mean any improvement and will just fuck up many diffs?

[jurchiks]

There aren't that many that are "not worth committing". Obviously your criteria for "worth it" is different; I think if code works (even if with bugs; those can be fixed in the process) and the idea is good then all you need to do is push the contributor a little to improve it (with hints of course) and when it's ready - commit it. Not like "this isn't perfect, ignore it". How many contributions have you seen that are perfect or near-perfect? I certainly haven't seen that many to ignore the less perfect ones.

[K4N4BS]

Lots of L2j Devs says "We commit only working and 'perfect' things, each contribution should be fully working...",
But, one question comes up to my mind... If they commit just 'perfect' things, why the hell they commit "bug/typo fix" every hour/day?

[BiggBoss]

i dont say they must be perfect, but as you said, you working code even with bugs is committeable, i dont think so. If you made your code work, debug it and then post it. Its not the same to make an user software, were the user wont see any bugs, performance decrease, etc.. than a software where hundred of connections send and read info every x milliseconds, executing the code you have in the server.

I find more motivational to see your code running without bugs in your pc than watch how the trac get full of stack traces cause of your stuff

[BiggBoss]

Lots of L2j Devs says "We commit only working and 'perfect' things, each contribution should be fully working...",
But, one question comes up to my mind... If they commit just 'perfect' things, why the hell they commit "bug/typo fix" every hour/day?

Typo fixes: Usually, those fixes are commited for text (like html, sql data, java Strings). Server will work with

Code: Select all

<html><body>Hello!</body></html>

than with

Code: Select all

<html><body>Hollo!</body></html>

You dont see any error and just see the html appear, so you dont take a deeper look on whats on the text.

Bug fixes: As i said, code WITHOUT bugs is commited. That does not mean we pass any of them. We post our fucking stuff in the user contributions, to let other ppl (mostly community users) to test it, and find whats missing or whats bugged.
Usually, noone test the stuff, so all feedbak we have comes from our own test. If we didnt find any bug while we tested it, for us is code WITHOUT bug. Thats why it gets commited and later need hot fix, when the bugs are found

[K4N4BS]

So, i guess devs shouldn't talk about "Fully Working commit"...
And, since both ways (Dev team and community) can make a typo/bug sometimes (Come on, we are humans),
Dev team shouldn't commit community contributions? Even they have some little bugs, or even help the code author to complete and refine it?
L2jServer is a great project with an big community, but as i see, since a time ago, dev team isn't giving a shit to community members, too much people have stopped their contributions because they doesn't see dev team interested in their work, or even dev team ignore them because they haven't years of codding experience...
The point is, i think dev team should really incentive community, in contributions and test,
But now, the team just says "Write and test it, then we commit", and not "We want help you to finish this, to get this finished soon".

That's my point of view, L2jServer as lost it's definition of "Community", since community is Help each other, not Help Dev Team.

P.S: I know there are devs that really want help community, they incentive members, but they are minority.

[BiggBoss]

Read again my post. Of course we post Fully working shit cuz with all the test we have (Mostly our test and few community members) we didnt found any bug or we fixed all whats is found. When we commit it we think is fully working cuz all test says so.
But if we know a contribution its not working, or its working with bugs, it wont get commited, not matter if it comes from dev or user, and even if we want to incentivate users, we wont commit bugged shit just to make them happy

[K4N4BS]

And Unstable branch is used for...?
Seriously, since it isn't used to help develop something, i don't see the point of keep "Unstable" name...
Lets take Debian as example, they have 3 development branches (Stable, Testing and Unstable),
It's user's choice, take something stable, test something or really test something...

[BiggBoss]

You couldnt give me a better argument to say how devs take more care for users than user for l2j
Unstable was created to make the users easy to test all stuff. Instead of having to apply a patch, you only have to update to HEAD.
But even if its name is unstable, no code with known bugs is committed to it.

[K4N4BS]

And what about take Debian development template?
Stable, Testing and Unstable version?
Actual Unstable can be renamed to Testing, since it isn't stable and isn't unstable...
This should speed up development, incentive community members and help dev team.

P.S: I never say "Commit every shit", i say "Commit useful community contributions".

[BiggBoss]

thats not up to me to decide