MarkZD

Just an observation, it's not because it's only used on char-server, it's because it's used out of mapserver category and it doesn't fit very well into one of the other categories(login, char etc) or is/could be in more than one server, will eventually be in interserver. , true up to now.

Only in database, the struct will be the same for all items, being stackable with a pointer having its unique uid for each item, which will be used for non stackable as well to save a little ram. I'll not give much more info about that now, I'm gonna organize my stuff to travel, see you after 16 days.

I got a little upset when I was just showing a point before saying that and trying to know about your point and you were agressive, so I said it. I know it's not every time true, I accept I was not fully right. Thanks for showing your points anyway, it was just the way you started which was not good. Ridicule is a thing, criticizes is another, and since start your were with the first option. We(Kisuka & me) already talked on Irc and solved the above subject. About the name, nsiuid will be changed later. Renamed variables name from nsiuid to unique_id: r17086

Yep, everytime an user walks, uses item, skill, atk, the idletime var is changed.(maybe it should update on sit down/stand up too?) It'd not add a performance issue to just check it with the current implementation.

I'm planning about it, but it'd be a file where user would specify items, by name and some other thing as chance to get it to be unique and they'd be advised to only put rare items as it'd could take to severe performance issues if apple was defined as unique and people started to buy lot of them on NPC. Lot of proccessing to create the uid's and insert items on database later. Although, it's not yet sure we'll put it.

-- OFF TOPIC START Better the whole log: As it's seem, I just explained the reason of name and I showed my point about why this name was chosen instead of unique_id. He tells about my programming etiquette, but his human etiquette is worse. Talked the guy, which says that is developer, which take hours to realize a mere descriptive name, which is explained for user, in configuration on src/config/core.h as mentioned on changelog. -- OFF TOPIC END uid is a commonly used name, the rest is just a complement, you should know it's a unique_id if you see uid, since you know every programming basic etiquette. http://en.wikipedia.org/wiki/UID http://docs.oracle.c...server/UID.html I'm just here to help, as I said if the majority wants the name change so much I don't see really a problem, we can do it. I agree and I'd like to know from other if it's so hard to know nsiuid is related to unique id. Btw, I think nsiuid a good name, but it could be nsi_uid if it's easier for you to read, I think it's even better than unique_id for this case.

I didn't look yet, I'll just leave it here as a record, maybe someone wants to give a look/say something.

r17082 I didn't get the other error although, it doesn't make much sense as this variable doesn't exist in clif_tradestart and anywhere till now. Maybe the compiler became crazy because of the first error, try to compile with r17082.

Implemented Non Stackable item id at: r17080 I forgot to put on the commit, but credits go also to Xantara and Sirius White, which adapted to rA and created the base code, respectively. I removed some unnecessary queries, renamed somethings, fixed some glitch and add a safer way to get the uid than the first, also put the option to enable disable and tested.

I'm taking over it.

IMO, it should be done with a special parameter or on another function and let the @mobinfo keeps regarding only to the mobs. So remove the atcommand_mobinfo_type, and let the new command enabled by default. #ifdef RENEWAL_DROP &&|| RENEWAL_EXP, shows the referenced infos, otherwise show user a message informing it just work with those renewal settings or just shows the default drop, exp.

http://pastebin.com/r3A2BTGC

Implemented at: 17017

What is your point in using msgstringtable.txt? If you just want to show some msg in commands(@anything), write it in msg_athena.conf so in the atcommand you send it to the user using clif_displaymessage(fd, msg_txt(message_number)); Assure you're using a number which doesn't exist and the number is smaller than 1401. If you want numbers bigger than 1400, you need to change the limit in atcommand.c changing the defined number: #define MAX_MSG 1400 Each text can have up to 1024 characters.

For the parrying, u just need to change your parrying in db/(pre/re)/skill_require_db.txt Find: And replace with:

You just need to change the map/teleport/relog to see the difference after activating/deactivating.

If you don't have any custom, try downloading all again, buildbot says it's ok: http://build.rathena.org/console

You didn't applied the diff correctly. Look at your skill_strip_equip function, there is: if (rnd()%100 >= rate) if (rate > -1 && rnd()%100 >= rate) return 0; But it should be: if (rate > -1 && rnd()%100 >= rate) return 0;

Implemented at: 16992 We opted by replacing the src svn with the old version 16966. We agreed.

I am starting to work on this now. Complete. Here's the diff, if someone want to check: http://www.mediafire.com/download.php?peh69n9fbvadwr1 Patch the revertTo16967 before. I pretended to commit it a little later, but not sure after talking to mkbu95. It'd be good if we could make a meeting with everyone at irc.

I did it in the exact way you requested, you didn't say you want it to be 100% chance, the skill has a base chance so it of course will not remove every time. If you want it to be 100%, remove this piece from the diff: rnd()%100 < i &&

I have already put it, it was a problem when pasting it to uPaste, updated the link. Edit: Implemented the safer way in battle.c r17004

http://rathena.org/b...sconnect-issue/ Sample of bug: Result: No damage, since damages smaller than 1 will return miss. Here's a patch I made which will make formulas a little safer on systems(compilers) which implements int as int32: http://upaste.me/24dc1734408c633f Or we can try using long long, there's other ways to do. I'd like to hear your opnion. EDIT: RIght after posting I though it'd be more elegant to use int64(long long), instead of using the float trick. Sample: If none has any complain, I'll go for it.

I'm marking it as approved as the majority is in agree.