Переводить не буду))
Hello and welcome to that topic dedicated to... aCis !
Roadmap
Link to the past
During those last years of developement, shitload of things have been fixed, refactored, reworked, cleaned up... The amount of work, dedicated time and fixes is just too much monstruous to be named. aCis in numbers, it's :
- 341 revisions, each fixing numerous issues.
- more than 400k edited lines, only counting GS changes, totalizing 202k added and 209k removed lines.
- 4 years of non stop developement.
- 10+ Inner Circle members, dedicated to make the pack better.
- 20+ major sharers, 40+ counting minor ones.
- 147 Customers since the beginning of freemium concept.
- 10 banned evil leakers.
Back to the Future
I want to speak about aCis future, because yup... aCis won't stop, and even better -- it got a roadmap, and I will explain the reason behind my choices.
Priority on quests, hold on AIs
Scripts have begun to be refactored around 2011, with the addition of Java scripts engine. The risk was big ; I decided to drop all jython scripts, and rework them one after one. While I got balls to rework scripts (atm, 364 scripts fully rewritten by the community and tested by me), it's a time-consuming work. The good thing is, that time will soon be used for another goal. Quests are almost all shared under private workbench section, thanks to the dedicated work of few ICs members (notably fernandopm, roko91, junior). Quests is an important part of L2, and a decent pack can't really be stable without those. That's why my current goal is to end it as fast as possible.
AIs are holded on for two main reasons : the current script engine fucks the loading priority, melting AI behaviors. The second reason is, and it will be the next point : many GBs AIs use heavy amount of minions spawns. Spawns which will be handled soon by... The new spawn manager.
Spawn Manager : the next big refactor
Yup, spawns are currently handled under a basic way : loaded x/y/z then spawn on this location. But that's not as L2OFF. aCis is preparing in private workbench a L2OFF spawn manager, with territories system and zones spawn. It means you could spawn and despawn entire territories with a single command. Can you imagine how easy would be events system, or GB minions spawns with such system ?
Unfortunately, that system means HUGE parts of code have to be reworked, notably : castles mercenaries system, 4 sepulchers, 7 signs, day/night spawns, and npc walkers system. It will take time to make it 100%, but the result would be simply a L2OFF behavior, nothing more, nothing less.
SpawnManager is the next big rework, after Geoengine, made by Hasha. Geoengine was needed for incoming SpawnManager. As you can see, we already follow a roadmap...
Scripts : the light at the end of the tunnel
With the newly SpawnManager, we can process AIs such as Frintezza, Zaken, or Ice Fairy without problem. One line can spawn or despawn 10, 20, 40 minions. We won't need boring hardcoded Maps holding values.
Meanwhile, the script engine will be reworked between this period and current period to handle priority system. Quests will be ported into gameserver, to avoid external compilation times, as they will be considered as fully functional. Meaning the drop of 2 other libraries, and sky-rocket server loading time.
Skills data && system : no skill fixes before this milestone.
Skills is an important part of any server, as L2 is a PvP game. So, why don't I fix issues reported on Bug report section ? The main reason is, the current skills XMLs are mixed Interlude and GF L2J data. When both L2J IL or GF are kinda failed on numerous aspects, the mixed result is still better than regular L2J IL. Numerous values are wrong by nature. The point is to re-parse skilldata.txt from L2OFF, and import directly all the new data. It will use, then, 100% L2OFF data. As you can see, it's pointless to manually edit XMLs, which anyway will be replaced by L2OFF parsed data.
Another reason is, the way skills are processed is simply... wrong. A skill is supposed to be a list of effects. L2J gives to a skill its own entity, saying a skill is that or that. It's wrong. The core will be edited to reflect L2OFF system better.
AI system
An emphasis to simplify the way the AI system works (intention system) is under thought.
And otherwise ? Misc stuff !
Like always, between those big milestones will be committed a huge number of fixes, cleanups, reworks. I believe it's thank to details than a pack is successful or not. Fixes can aim anything, anytime. That's what makes the changesets beautiful to read, isn't it ? It goes to packet completion, behavior fix, performance boost, code simplification, new commands,... It's infinite !
aCis meeting
The idea comes from Hasha and is still under the form of an idea ; an aCis meeting will eventually come in next summer (maybe April ? Didn't really checked yet), probably in Prague, CZ. It will be financed with aCis donations. I'm not aware about the different costs, but I probably could participate for a decent amount to host people. We count for the moment Hasha, me, and SweeTs.
I repeat, it's only an idea but it can be funny to meet people. And CZ chicks are.
End word
I hope you appreciate those shared thoughts. The following roadmap will probably take one year, maybe more following IRL issues or sharers dedicated time.
I refresh my thanks to all people who contributed to make aCis what it is it currently, aka (one of) the best and most dynamic Interlude pack. It is cool to see than Interlude community is still alive.
aCis forever !