Jump to content

Intrepid

Legendary Member
  • Posts

    5,521
  • Credits

  • Joined

  • Last visited

  • Feedback

    0%

Everything posted by Intrepid

  1. Question was because l2j used to schedule such updates to releases of eclipse versions that supports it the best, and i dont know when this release will happen. :D
  2. Just checked out acis, i see its still java7 when will you change it to java8 with which eclipse update? Or when l2j does?
  3. If you go open source it doesnt encourage people to commit simply because they just want everything ready thats how people work. The few who does commit and give ideas are going to do it no matter what kind of project model you are following. From my past experience i can tell that if you are open source people expect you to deliver everything to them on a silver plate with nothing to get in return, but if you are private people actually seem to have more interest in what you do. Tho many people atleast those i spoke with, dont consider acis as an alternative for their needs because of interlude.
  4. I give it to you good work needs time, yet heres another view on that subject. Generally a layout of an svn project(could be another thing to highlight that SVN is old as anything bitbucket is way better in every sense) is trunk,branches and tags. So an ideal development process goes like this: I want to update h5 to god for example so i release to branches the latest stable h5, than update the trunk which always has to be an unstable use at your own risk version with something big like packets, when thats done create a snapshot into tags as a restore point if something goes wrong. After that update the trunk with again something big like skills and create a snapshot into tags, after that lets say update quests and again create a snapshot into tags. I know thats way too ideal to actually follow and it rarely happens but it could be like this also. From that view i agree with xdem also because if you follow the basics of the svn structure than it could be done many ways not jut with a rare release when a bunch of stuff is done.
  5. And see this is my issue with this, people always think like okay this works alredy so dont even look into it. Heres an example, thread management in l2j is really old it works yes but theres so much room for improvement since ages that its not even funny considering how big of an impact it has. Ofcourse l2j did start some things that make perfect sense such as exporting stuff into the datapack but the way its executed is something that again have tons of room for improvement but nobody does anything about it because it works. The thing is that even tho sometimes there are great improvement for example i really like the extension of the skills and items on dp side in l2j, many times what happens is they either need ages to do something that has been done since ages(fbiagent shared the geoabstraction years ago and it took them years to implement an alredy fine code) or the other thing is they know about an issue and instead of fixing it they just increase memory heap size(it was a "temp fix" for their lack of care about performance which eventually become a fix for it for the time beeing). Also but this is something that i can understand from a user friendlyness point of view, the only reason why they cant organize their projects properly is they dont intend to use maven or any other advanced project management tool, which even tho harder to use from a user standpoint because it needs a bit more background setup gives a huge boost in productivity.
  6. See im not a superior user at all, i just prefer the thought process and workaround of l2jfree over l2j like you prefer l2j over l2jfree or emu whatever pack, matter of taste really. Indeed its private its not used for any server for the simple reason that they dont care about creating a server or making money with it, all they care about is to code for their own satisfaction because they like it. Lordrex in that sense is difficult to understand that i will give to you. The reason to add back stuff or to modify things that "work alredy just fine" is something that my obsession sort of. I like it when things operate the way i want them to operate over the way someone decided to in some cases. Also the reason that java got better in some cases doesnt mean that javolution should be dropped entirely since there are still cases when javolution would be a better choice overall. And just a little addition, L2J long ago had an obsession as far as trove went, they were so obsessed with it they actually overused them instead of properly selecting what fits best for the exact code part to be better. That is what i always struggle to understand, Theres no overhead created by having an additional library in my opinion for some cases where its better. The reason i've heard many people say is that a coding pattern should be followed like a scheme to give a base to a certain code which is a nice idea but shouldnt be that strict in my opinion. See even tho i dont intend to do anything other than just playing around in eclipse for my own liking that atleast for me wouldnt be enough if it would consist of adding a bunch of event like things. I like to do things in a similar mindset like lordrex does now because it has a good point, you dont need to shout out to the world what you do, if you really like it just do it but for yourself and yourself only.
  7. As i said before a lot of times and as everyone know who actually follows a little bit of everything as far as packs go that l2jfree and l2emu is the same now and its private, if you would know what i know about their progress and how they work you would hide in a corner and cry. :) You seem to be blind for some reason to my first post i alredy i said i just miss coding a bit and i want to play around for the sake of nostalgia but keep hating if that makes you happy. :) Now if you cant do anything else than talk down people liek your something superior maybe its better to staying silent or understand the 1st post. :)
  8. Well than maybe you could link some fresh ones as i've seen a few comparing the newest stable javolution against java 1.7 but maybe not enough of them?
  9. Well based on the performance tests i've seen since the release HashMap is the only one that is better in reading(and reading only) than javolution. Others improved but not as much.
  10. Well that depends on how "deep" you go into the pack, i like to change a crapload of stuff like i would instantly add back javolution and use it whenever possible which would alredy make a convertion difficult to newbies.
  11. Ohh well tbh i wouldnt sync much but i like to run over the changesets to pick up whats usefull, and i would turn the pack upside down in many ways to fix a lot of things they messed up while following l2j, Its still interlude right?
  12. But acis have no constantly updated source code to sync and to be honest i am way too lazy to sync a million lines. :D L2J is messed up altho an older version like gracia final would be okay since i dislike their newer chronicles a lot(mainly because of their interface implementations which is way too messed up for my liking).
  13. I miss coding a lot so for time killings sake and for fun i want to do some stuff not for any server or public project or anything like that, im over these things as i dont even play the game anymore. So my question is which project is atm stable enough for testing purposes so it can be used for development? The chronicle doesnt matter im interested in everything atm.
  14. The amount of static data doesnt matter in realtime performance only in startup time where they use an utterly useless and messed up wrapper for data loading. Tho that conversation would get pretty heavy considering 99% of the mxc community is an l2j guy coming from their hate against maven and understanding any other project.
  15. What uses a lot of resources in l2j at higher chronicles is truly the bad coding practices for example their mmocore is by far the worst out of anything that has been used by any java based game emulator, their threadpool is outdated and messed up, their database handling is again outdated and messed up, almost all of their attempts to create a more versatile and easier to work with code failed(just compare their bunch of interfaces and holders with the global restriction of l2jfree). And thats just the tip of the iceberg, the biggest reason why l2j is as bad as it is, is their huge ego(always have been an issue). If something is good and tested for months they dont implement it just because they would need to acknowledge someone else did a good job from another team, so they rather try to implement their own version which rarely turns out to be better or even good than what they could have synced. Thats the reason why russians are ahead this much. :) Ohh and memory stats doesnt matter because of l2j and therefore your pack too because you follow them on sync, tends to cheat and hide their bad practices for example the idea of deleting javolution instead of using a newer version and work around a faster and more versatile collection of classes just because it would require them to learn new things. :)
  16. Actually its more and more the case, guess why they had to increase the minimum memory required to 1g when the pack couldnt even load properly with less and not because of so much stuff beeing in it, just bad coding in general, but thats the case with it since ages so nobody cares especially when l2j is dead. :)
  17. You do realize that trickster is 1 class and your config screws around with every single class that uses a bow(because of how l2j implement the weapon types especially frozen which is the worst pack in history) right? You proove with this comment that any server you make or involved with is crap if thats usefull for you instead of properly balancing the damage via class templates and skills. Shares like these should be junked even deleted instantly for the sake of less crap l2j servers.
  18. The point is codes like that make little to no difference because these affect the server in a bad way. Heres an example, you recude the incoming damage from blunts to light armor, the result will be lower auto attack damage but exact same skill damage so the end result is if you have lets say an arcana mace which is a blunt this patch does nothing in order to reduce the damage of mages so its pointless. If you reduce lets say bow damage against robes than because of the archers rely on kiting and auto attack damage the archers will get destroyed with any mage while its considered a skill matchup in reality. As for the forks, anyone who knows anything about the teams who work still or worked in the past on l2 emulators know that the only ones that should be considered good are l2j for its stability, support and constant updates, l2jfree even tho its private because its code quality and management and ofc some of the russian teams who dont rely on l2j based code but rather doing their own stuff from scratch.
  19. My point was this modification in general have nothing to do with balance, thats mostly a tool for the most unskilled server owners to nerf mages and dagger classes. :)
  20. The title is confusing, this kind of "balance" has been available since ages and have been shared a million times even tho its inside many packs. If you want to share something truly balance related than go for skill related changes(powers,enchant routes,chances,debuff timers) or change the base stats(best result is doing both). Balance cannot be done and never will be done with simply core modifications. :)
  21. By stability i mean clean, fast code that maintains a good structure overtime which is everything thats not l2j or its forks. In a way the only stable pack thats truly is stable and publicly available is the gracia final l2jfree branch. Anything other than that is messed up so yes.
  22. Oracle dropped support for windows xp so heres the link to what you should try out. http://stackoverflow.com/questions/20965564/installing-jdk8-on-windows-xp-advapi32-dll-error If it doesnt work than i guess upgrade if you want jdk 8.
  23. You said interlude okay heres a task that nobody accomplished ever since the existance of l2j. Stable server. :D Jokes aside if you want to get creative implement a new skill engine similar to what neverwinter has. ;)
  24. You have to know the action that executes what you want for example if you want a simple message to show up when you pick up whatever stuff you put your code into the pickUpMe() or whatever its name or you can put it into the addItem method since its called on inventory update or even add it to the getItem packet. Mostly you need to experiment where its better to put it but thats the wasiest way, try to avoid creating a lot of small classes for tiny bit of things.
×
×
  • Create New...