Discussion Acis Stability

[AccessDenied]

i find that the bug with stutter stepping still occurs with geodata but not as often as before

that sound like update problem no geo

[AccessDenied]

So the pack has problem with skills? imma check them

 

Tryskell if you want add me on skype, i can work on quests if u want help..

Edited March 2, 2016 by AccessDenied

[SweeTs]

Quests are basically done, so.. :P

[Tryskell]

So the pack has problem with skills? imma check them

 

Tryskell if you want add me on skype, i can work on quests if u want help..

 

Skills engine is a mix between IL and Freya, but all values and formulas have been reverted back to IL. L2J fix skills one by one (like, unfortunately, most of packs out there), and unlike them I think it's a deep waste of time to do that. So I got errors from L2JFreya era.

 

About skills formulas, problem is inherent to L2J, since L2J exists. And that's the biggest issue (even if, and I repeat it, those formulas are the same from L2J to L2JFree, unchanged for years) . If you want to work on something, skills formulas (Formulas.java) would be a good thing.

 

About quests : http://acis.i-live.eu/index.php?topic=970.0 only the most annoying quests aren't yet done (20+ cond, 100+ htms, complex ones or many drop rates, require numerous checks or waiting SpawnManager - or a mix of all that).

Edited March 3, 2016 by Tryskell

[Nik]

Skills engine is a mix between IL and Freya, but all values and formulas have been reverted back to IL. L2J fix skills one by one (like, unfortunately, most of packs out there), and unlike them I think it's a deep waste of time to do that. So I got errors from L2JFreya era.

Have fun with skills.

 

If you do it properly that's targeting, effects, conditions, cast and the whole stats system to be done.

Basically skill engine is pretty fine if you ask me. The amount of work needed to make it retail-like (better than all current packs out there) is reworking half of the pack to change something players already are used to.

 

Honestly, I've worked on aCis and it is the most comfortable project to work on... it has the least "interlude l2j" feeling to it, and by that I mean the amount of poor code interlude l2j has... back in the days it didn't felt poor because it was the way people codded, but now years later, java and its community gotten better and codding ways are much different. Not to mention that aCis supports Java 8 and I really, really hate projects that doesn't support it.

For me interlude is already way too outdated, and its good that aCis is rather closer to current l2j than old interlude l2j. But if I was hosting interlude-style server, it would be based on L2Classic and not interlude crappy client. Maybe one day we would see aCis for classic :P

[Sdw]

Well that's something we can do ez

[AccessDenied]

So skill form are messed up a bit? well good that you used them from freya since its one of the most 

stable ch i saw so far.. really beautiful and easy in development even tho skills require a big rework 

i havent test aCis skills, it look quite beautiful. Tryskell again i'll say the rework you did is like you changed the whole structure of it..

but again it look nice. right now checking packets wt ollydbg and then ill check skills.

So far i see nice job from you even if again i change your pack structure somehow mostly cause i were used to l2j .. freya..

 

btw aCis standars gonna be higher than any OFF server in few months. i like the design of it, beside i love how 

people say off has no bugs.. is like telling me linux get no viruses.. bitch plz..

Edited March 3, 2016 by AccessDenied

[Tryskell]

Skills engine by itself is fine, there is nothing wrong about it. But the way effects are handled is wrong. By extension, as we gonna simply use parsed L2OFF data, tags will be refactored (added, removed or edited).

 

Skill formula issue comes from the fact L2J stacks chance on a accumulative way (notably for resistances). 2 skills with 30 resist fire = 60% resist fire. So you can easily reach 100+%. Another issue is the matk modifier, which is plain wrong notably for enchantment cases.

 

Most (90%) skills work as they should, the problem is mostly missing enchanting routes, and few other invalid values (power, duration, etc). The real problem is 2 or maybe 3 formulas.

 

I doubt I would work on another chronicle than IL, mostly because I hate datapack related stuff, and L2OFF is kind enough to provide it.

Edited March 3, 2016 by Tryskell

[minideed]

I should put my Depmax64 IL files back online and let the Acis team have full access to it so they can get missing HTML's and other things. Just have to contact DepMax64 and have my license updated and moved onto a new server. :D

 

I have always liked Acis as an L2J pack since they are focused on retail stuff and not useless custom functions.

[Sdw]

Naaaaaaah, it's way bigger than that.

 

L2J xml sucks by design (e.g: skill start at level 6, you have to make it start at 1 for useless shit)

 

Current aCis xml is based on really L2J one which sucks even more. 

 

Beside that :

• L2J targeting is fucked up, not even close to retail and you'll have tons of issues with it if you keep it that way.
• L2J stats system is fucked up as well, just check critical calculation , or your given exemple on resist and it's mainly stats calculation, not really formula
• The fact is not effects based, most of them sharing the same code base, with shitty add/mul/blabla which makes it easy to makes typo
• L2J made a lot of retail conditions simplification, there is even missing one, and doesn't handle all type of it (target, passive, regular)

and that's just a really quick sum up. You have no idea how many dirts we found when we started this quests with L2J data.

[AccessDenied]

Skills engine by itself is fine, there is nothing wrong about it. But the way effects are handled is wrong. By extension, as we gonna simply use parsed L2OFF data, tags will be refactored (added, removed or edited).

 

Skill formula issue comes from the fact L2J stacks chance on a accumulative way (notably for resistances). 2 skills with 30 resist fire = 60% resist fire. So you can easily reach 100+%. Another issue is the matk modifier, which is plain wrong notably for enchantment cases.

 

Most (90%) skills work as they should, the problem is mostly missing enchanting routes, and few other invalid values (power, duration, etc). The real problem is 2 or maybe 3 formulas.

 

I doubt I would work on another chronicle than IL, mostly because I hate datapack related stuff, and L2OFF is kind enough to provide it.

Yeap i saw this one in my freya project too.. i had to deal with the fact that skills were overpower or fail.. especialy when it came to anchor-silence-resistance e.t.c i had to deal with the enchanced skills too... and player had easily 505005050% resistance to water or 1/1 anchor with +10.

even if formulas are ok still skills need check up and xml fix. you know what imma stop my project for a bit and try 1 by 1 acis skills

Edited March 3, 2016 by AccessDenied

[Tryskell]

Yeap i saw this one in my freya project too.. i had to deal with the fact that skills were overpower or fail.. especialy when it came to anchor-silence-resistance e.t.c i had to deal with the enchanced skills too... and player had easily 505005050% resistance to water or 1/1 anchor with +10.

even if formulas are ok still skills need check up and xml fix. you know what imma stop my project for a bit and try 1 by 1 acis skills

 

If you want to help on skills data - and waiting for a bigger rework - you can eventually see with SweeTs. I won't touch to skills before at least one complete cycle, but I can commit sharers stuff meanwhile. SweeTs was preparing a diff patch gathering fixes.

 

I'm aware about what you say, Sdw :P. What's wrong about targeting ? You mean "fan" form things and alike ?

[Sdw]

Not just that. Well unsure you have effectpoints implemented yet, but what I ask you to have a monster cast in AoE a root affecting only monster (yeah that shit happens) ? Having AURA without target type, affect type, affect scope is just retarded :D

[AccessDenied]

If you want to help on skills data - and waiting for a bigger rework - you can eventually see with SweeTs. I won't touch to skills before at least one complete cycle, but I can commit sharers stuff meanwhile. SweeTs was preparing a diff patch gathering fixes.

 

I'm aware about what you say, Sdw :P. What's wrong about targeting ? You mean "fan" form things and alike ?

Well generally imma sit and check (at least the most used skills) and re-edit their xmls cause im 110% sure they wont work properly.. i bet silence work after 150 uses even if is +12

and without the enemy wearing earrings :D so ill re-edit and if after weeks e.t.c i end up with a nice result- sweetS can join and test with me if he want we can share xmls to you and re-place

cause as far i know from my freya... skills are always FUCK YOU -.- 

 

love ya tryskell <3

[Tryskell]

Skill data isn't the problem in this case, it's formula. I invite you to use the debug config to see exactly how the % is calculated, you could be surprised about the output.