Ind

currently this mempool implementation sirius black has added is not used so it wastes the thread it creates on server boot and is broken, for example if(p->num_nodes_free < p->elem_realloc_thresh) racond_signal(l_async_cond); if the free nodes are out and the spawner thread hasnt been run yet and you try to get another node it crashes the main thread. Do we get rid of it? is it there any point in it anymore(since greenbox has overhauled the ERS system)?

login server should display information regarding your issue as long as your client connection settings are correct

type 'bt full' after the segmentation fault and post the output here. thanks

Hello and welcome, this is rAthena's development digest for the month of October, 2012. As I've mentioned in our previous digest, all new digests, including this one, are no longer numbered and account for a specific month. I'm afraid October was our worst month stat-wise to date -- we've never had any month with less than 100 commits, however these low figures are not expected for November for we've received a new developer, and some of our existent staff will be able to be more active as of November. Development Team Lighta joined as Core Developer Development Highlights Updated Renewal Exp/Drop modifiers, also made them easier to be customized with a new db file: level_penalty.txt Documentation updates all over the place. Misc. Stats During the period there were 46 commits. Of these 46 commits 17 were bug-fixes.

Monthly Digest #5 Hello and welcome, this is rAthena's fifth development digest. Below is a small report for August 14th - September 30th period (past 47 days). As you may have noticed this digest has taken some additional time to come out, we delayed it intentionally so that all upcoming digests will refer to its own month -- in November the 1st we'll post a October digest, and so on. Development Team Akkarin joined as Documentation Mananger -- and has been doing great, I must add. Masao is no longer a Script / DB Developer Gepard is no longer a Core Developer Development Highlights Updated msg_athena table (credits to mkbu95! thanks!) Updated Renewal Casting System Implemented Official Monster HP Bar ERS System Redesign - by GreenBoxImproves overall server performance as it is used by our red-black tree memory database which is then used by a handful of features [*]Client MD5 Hash Check Support [*]@commands @reloadquestdb @unloadnpcfile @addperm/@rmvperm (mostly useful for permission debugging) @cart (required since @option doesn't work for new carts) @newmount is now @mount2 [*]Many Documentation updates by Akkarin, Arcenciel and Euphy [*]As usual many great skill behavior fixes by Malufett and team, and I feel I must emphasize this. Skill behavior fixing is often troublesome and takes time, for our team has to get on their official server characters and test reports in order to code them in rAthena. [*]And Euphy's and Joseph's impressive slaughter spree of script file sizes (aka optimizations all over the place). Misc. Stats During the period there were 172 commits. Of these 172 commits 85 were bug-fixes.

I believe that is normal, back on iRO before trans were released my friends monk would kill moonlight flower with a single asura (lex'd of course, but still it was very high damage)

I support this name change well I can't really give a go since, as you said, its not finished. Post the final one here when you're ready, at this point however I can say you've done a wonderful job

can you post the whole diff then?

i failed to think on a efficient way to have this coded, perhaps a multi-dimentional array? e.g. modifier_array[MAX_PLAYER_LV][MAX_MOB_LV], exp modifier = modifier_array[sd->status.base_level][md->level]?

I agree with malufett and brian

no one has shown interest in doing it until now, go ahead.

start storing the char name in the same table, much more efficient.

due to the position of the error i'd say a oninit label is falling thru and reaching the ontouch.

Thank you :3 I'll grab my sword and start ripping stuff out of it.

do any of you know if we're allowed to modify libconfig? (there are a couple things in it that are quite annoying, for example if it finds a duplicate entry in the file it drops the whole file data rather than throwing a warning or replacing.)

bapti-bipti-bump! this would be most helpful when debugging as well

Oki, group setting then. Added in r16751 as discussed

Discussion started here (check to see previous comments): http://rathena.org/board/topic/69681-sql-files-redesign-proposal/ So this topic aims at discussing whether this is viable and secure, and what methods are best suited for us

the updater discussion has delayed this topic, so I'm moving on with the topic's original goal and we can create a new one solely for the updater discussion. Added this in r16750

it appears ultramage removed simply because they werent official, that code however won't work anymore. Since the interest for these features are huge I'd kindly ask anyone whos interested to redesign (theoretically only -- we'll code it later) the commands in any manner they find efficient and present it in a new topic for discussion

why it wouldn't work with @option: newer carts aren't option masks (and it could conflict e.g. as arcenciel pointed out). Sorted in r16749 as per Trojal's suggestion

Oki added here r16749

hum o.o i see no problem with it

fixed, may take a while to appear back online (due to the cache).