*real*savormix

Meh, with the fact that it keeps running into obstacles 24/7 (and thus essentially wasting more time on this than on farming), it's a poor man's bot.

You are just lucky you don't know their language. It's as cancerous and "i'll rather not share" community as any.

Aside from the fact that publishing a botting scenario gives all the information for a server owner to ban precisely those characters that are using it (regardless if actions are pseudorandomized), botting seems to be all about RWT nowadays (face it, be it WoW, L2 or whatever, hardly anyone bots just because they want to play endgame). And thus, successful botters are highly despised by the RWT community. Not only botters do not require RWT services, they are also stealing THEIR (RWT-er) monsters and wasting their time. That's why in general, botters are hostile against botters for the time being.

Since those versions are typically referred to by the names of people who originally shared them, CN is 'meifu' and NA is 'sina'. TW is lost (this is a factual statement and not a name) and CN + NA html/fstring (+ injected AI) works wonders, so it doesn't really matter.

Well, I didn't check if Iss Hiero/DC/Domi is reusing Iss Ench skills, but I'd bet it isn't (despite the fact that the skills do the same). Iss SD got their awakened dances, Iss SM got their awakened songs, Wynn AL got their cats, Wynn SM got their dark servitors, Wynn EM got their unicorns, … You get the picture. While pretty much all 'divine' skills (before awakening) remain the same as in HF (and these still comprise the majority of player's skills), GDL introduced a lot (relatively) of new 'giant' skills. Oh, and did I mention the not-so-small skill enchant related changes in IO? "Skill" is no longer an ID+level; now it's ID + level + sublevel. I understand, IO does look pretty much like GoD. But then again, I could say IO looks pretty same as GF, because Ruin Imp Elders (isolated example, the correct term would be "all monster ai superclasses") act precisely THE SAME WAY as they did in GF. The free teleport bug is still there in IO as it was in GF. And so on… My main point was that the script pack is 3 years old currently. It will only get more obsolete with time, so with the current price it may become more of a decoration than a useful (even if only for l2j) data pack before it finds a buyer. Therefore, crowdfunding. P.S. Please consider the fact that I am an interested party and thus the point of view used in my messages may not objectively reflect the way general market perceives the value of the files for sale.

Typically for KR and JP, you simply use -t parameter in l2encdec.

Yeah, l2-2004 "classic" is the real thing :D Newer clients typically report misleading information, e.g. incorrect enchanted item stats, wrong skill descriptions, incorrect SS/SpS counts on weapons, etc. Still, C1 doesn't have proper windowed mode support as it is (even after modifying some engine config files), so it's a much bigger pain in the ass to multibox, thus I can understand that people would like a modern one. ON TOPIC: I have mentioned a few times that I used to believe this is Lindvior npcdata/pc_parameter. Sorry if I mislead anyone, this is, in fact, Glory Days/GD and not GDL. Thread title is 100% accurate.

Those skills are purely cosmetic, both in l2off and l2j.

Special textures for non-custom messages: "Finish", "Start", "Game over", "1", "2", "3", "4", "5"Otherwise, use Message = 0 and your own message (just like via L2OFF eventmatch control panel): (on IL & earlier) FE 04 00 00 61 00 62 00 63 00 00 00Will send a message 'abc'. Everything can easily be found via the links I have posted earlier.

Countdown is dead simple, just send packet with message '5', then schedule a task (after 1 sec = 1000 ms) to [send message '4' and schedule another for '3'], which will in turn schedule for 2, … until you finally send the packet with 'START'. What exactly prevents you from implementing this direct approach? P.S. Obviously, you can keep the state inside the task and just keep rescheduling it until you send START.

Opcode | Packet definition | Possible messages In IL, FE 04 00 02 00 00Would show you that 'START' message. EDIT: missed a byte inside the packet transcript.

I see. To end this offtopic discussion, I simply thought you've forgotten what prelude (or C1, for that matter) was like and are now dubbing things "hardcore" despite the fact that there was nothing "hardcore" about Lineage II back in the day (key point I was trying to prove). Not to mention that there was the party bonus exp ratio (changed countless times since then) set up in a way to penalize solo leveling and thus there were only a pair of classes who could solo exp somewhat effectively. Unless you exp exclusively with a full party, this might cause you to feel 0.8x instead of 1x. But then again, it is hard to compare real early Chronicle gameplay with this mix&match one. I haven't checked, but perhaps absolute values are, in fact, 0.8x. It doesn't mean the game is exactly that much "harder" (exping-wise).

Care to define "lower than x1"? The general lack of SP and adena was a design feature in Lineage II (unrelated to rates) before they started the whole noobing down trend. I don't even recall what was the last chronicle when you actually had to play as "real" 1x on official servers. (Not counting the adena drop decimation to unseen amounts much later in Lindvior and Epeisodion) Also, I don't recall needing to kill 20+ mobs to get to lvl 2 in KR Classic.

There's nothing "hardcore" about the classic mode. It's a mix-and-match type of L2, primarily designed by sales people to monetize players' nostalgia. Aside from the .unrs (maps) and spawns, I don't really see anything even close to what L2 was in Prelude - C3. And I played on Korea Classic when it first opened.

Don't forget than in fundraising you could even set a larger goal (if you really want to). Still, GD is 3 years old, and seeing what changed in Lindvior, Epeisodion, as well as Ertheia and IO, it looks like that in current L2 there's hardly anything left from GD – not even GD's 4th classes and their skills are in use anymore (according to IDs observed in IO). And unlike HF, which was last of its kind before major & completely unnecessary changes, GD is in the middle of major and unnecessary changes that were later reverted or reworked. The price could be justified if it was an Epeisodion script pack with those Ertheia/IO html/geo bonuses. But GD… it's like C5. Nobody needs C5/GD, since C4/HF and IL/Ertheia are unquestionably going to be extraordinarily more popular. But this is just my opinion. TL;DR: GD isn't worth 300 USD to me, so I prefer the fundraising scenario, like others.

But there is a reason to blame NCSoft (and yes, I mean NCSoft, which is typically referred to as NC Korea, even though NCSoft manages both Korea and Taiwan, leasing prebuilt serverpacks to contractors in other regions). NCSoft made a decision in the very beginning that they are the only ones to decide who lives or dies what is "Lineage II". That's why IAM INTERACTIVE was knocked out by terminating their contract. That's why everything that regional contractors are allowed to do is limited to creating pre-approved premium items. The worst part is that NCSoft only cares about Korea/Taiwan (both of which are THE ONLY regions that require real ID; tencent does not, they merely have firewalled their Tiantang 2 servers away from the rest of the world). GoD is the perfect example for introducing features that cannot work in international servers, e.g. mentorship and features only Korea asked for, e.g. class unification. As bad and illogical those features are, players will want them. Therefore you are forced not to ignore the official servers. Even if you don't care ;)

Check your facts, 2009 was 6 years ago and L2 had already faced the IL vs Kamael and later conflict and the popularity was noticeably declining. Of course, by 2011, when NA got HF, people suddenly started playing on GF servers that they had previously en-masse abandoned for GE and/or Freya (mostly due to what you described), but the trend was still straight downhill. L2 was popular 10 years ago (2005), at C3/C4 transition times, when most of the important features have been finally shaped up, and it had received great visibility from Prelude/C1/C2. These golden times (technically, you can add C1/C2 here as well) lasted roughly until IL/Kamael. After that, it was slow but steady slope downhill (hardly visible in K/K+, a little more apparent in Gracia, but starting to show more and more with GE, Freya, HF. At HF L2 was almost dead in official servers, and people (exceptions only confirm the rule) were playing exclusively on IL or GF private servers. Since GoD, it's no longer possible to ignore the downward slope trend, since L2 is going full speed down the hill, getting worse (= less and less players, and the remaining ones usually only log to do one daily and instantly logoff) with every update. Similar stuff in private servers, made even worse by the fact that almost all of them are years and years behind Lineage II (of course, some people love IL more than anything; HF seems to be trying to become the moral successor of IL).

JIT cannot optimize based on something that isn't there :D And since generics only exist at compile time in Java, JIT will keep the CHECKCAST instruction equivalent in the compiled code, even though it may be obvious from actual code logic that the object cannot be of any different type at that point. So two possible approaches to this performance issue (since there is definitely more than one branch involved in CHECKCASTing): a configurable compiled bytecode postprocessor (preferably) or a configurable agent that would remove unnecessary CHECKCAST instructions from bytecode. After all, refs on the stack are untyped. Of course, if unrelated CHECKCASTs are removed, you wont be getting Exceptions at the time of invocation, but (unchecked) Errors instead. Hence it must be configurable to only process parts of the code.

And pretty much all widespread OSes (embedded included) have a native C API. Of course, the driver model can allow C++ code, but interface points will always be C (you can forget about dealing in classes at the hardware level, although you can use them for internal state management inside the driver). C++ also has real generics (though that has its side effects on the compilers), which are easier because you can declare any reasonable type limits, unlike in Java where you can't do this and that and also that other thing because in runtime there's no generics at all and thus nothing of this can be enforced. Two most dreaded features a language can have. An unlimited complexity preprocessing stage and the ability to make operators do things that only make sense to a single person. Of course, Java has those annotations and technically they can be used to induce an unlimited complexity preprocessing stage. But this is somewhat less popular in Java, because it is not used by default (at least not for code generation), while a preprocessing stage always exists in C/C++, just teasing to use more than imports or constant declarations.

Ironically, the reason why I started with java was due to l2j. That's definitely not a good thing, since I started with 0.4.1, then C2 and up, which meant dealing with some horrible code. Eventually (around the time Wooden & KenM released the multiplexing mmocore) the code got better, design patterns were finally introduced, etc. Regardless of the language, I've never found a better way to learn something than just by doing something you need for yourself. Don't get discouraged that at first your code will be bad (or at least highly influenced by other languages you already know) and thus the performance is likely to suffer. As you keep writing more code, you'll get used to the language and its specifics. As C++ is an old language, and was designed to have compilers completely compatible with C code, it has become increasingly complex, Eagle_Eye makes a correct point. One of the main problems with C++ is that it's currently 2015 and you still can't avoid interfacing with C APIs at some point. Of course, the situation is much better, then, say, at 2010. You can call me a purist or whatever, but due to all the legacy compatibility "features", interfacing with C is prone to fatal errors if you forget just for a second that something that holds true in C++ world, may be interpreted differently after it crosses into C. Usually you should expect memory issues if you dared to assume something that doesn't hold true (and if you didn't take the time to learn all the features since K&R, you will, sooner or later). On the other hand, if you strictly stay in the C++ world (it's not too hard now with Boost and similar libs), C++ can be as easy as Java. I think beecrypt tried to make itself an example of this. Though, if performance matters to you, you will be forced to delve into the native side of things, which is the JVM for Java and all available compilers for your platform for C++. Obviously, the JVM is much easier. Changes in optimizing compilers (currently referred simply as 'compilers'), even when you take away security fixes & default option changes, can both boost your application speed significantly without any effort on your part, or, (e.g. with a race condition fix) effectively make your application ultra slow, even though changing a few lines or reordering a few statements will make it as fast as it used to be, etc. You'll also have to learn all about branch prediction, at the very least. To sum it up, it always depends on how much you want to learn. It doesn't take much time to learn about a language if you simply want to modify existing software. However, there is no way to skip the time you will have to learn how the software you are going to modify works.

Nonsense, I didn't use ANY system shared in this thread, but got my own 746, added the Engine.dll and can log in to the game world flawlessly. Check whether your server is using custom protocol or antibot. I'd also suggest using the default dsetup.dll, as IL with dsetup noGG was always causing issues (GPFs are the easiest to identify, but there were others).

Hi, after a long time lurking around I finally decided to create my own account here. I will mostly continue lurking, but if I post, you will now be actually able to track my posting history. I guess you can thank people who kindly allowed me to use their accounts to access bonus content to finally convince me to join. See you all around, _dev_ (and many other names)