The Mana World Forums

Looks like Elemental Resistance or Elemental Vulnerability aren't working.
Tested #flar on Yetis on Caliban vs Live.
Live 100-150
Caliban 300+

I son't know what all this means. What I do know is that before I could kill a white slime whith one hit of flar. Today its takeing 2-3 while running since they assist.

t3st3r wrote:

• Looks like #flar spell has been broken . Earlier it took for me exactly 2 shots in 100% cases to kill yellow slimes, also usually 2 shots for sea slimes. Now it rather takes 3 shots of #flar 100% of time to kill yellow slimes and 3-4 shots to kill sea slimes. So measurements indicate #flar has become seriously weaker without any warnings. Needless to say it makes gameplay for me 1.5-2 times harder in terms of hunting small stuff, where mages already sucked at this really hard compared to other classes. I guess some bonus does not apply or some monster stats changed or something else that worked before has been "fixed".

Fixed. It was actually only elemental bonus/malus that didn't apply; against "most" mobs it still had full effect, but against water mobs it was weaker.

t3st3r wrote:

• Looks like at least map server has crashed and unexpectedly terminated at night today, disconnecting everyone and taking noticeable time to restart. It terminated at about 3:25 server time or so (time could be incorrect).

Well, I was *hoping* to do better ... coredumps were disabled last restart, but I've enabled them now so I can investigate.

With the new server at least the crashes should be more predictable ...

something is wrong with drops, i just checked it with another player, we both hit the jacko i got 23k he got 19k however he won the drop.
(i couldn't pick the drop it said it belongs to another player).
the jacko we killed before, i made only 9k from jacko, so he probably made more and i won the drop... (he didn't remember how much exp he got from this jacko)
Edit: i chatted with some players in game it seems like the final blow to the monster is what determines now who gets the drop.
Edit 2: i compared it to caliban which still runs the old code, in caliban the one that makes most dmg gets the drop while on main server its the one that makes the last hit.

confirmed. Behaviour of drops has changed. This should be fixed.

checkweight always returns 0

tux9th wrote:checkweight always returns 0

This one is a script bug (in the particular script tested) that is merely exposed by the new code, which no longer corrupts the stack.

Edit: I have discovered the cause of the drop issue.

checkweight issue has been resolved in data.
drop issue is fixed in code.
CDN issue is fixed in nginx config.

Need to restart sometime to bring in these changes ...

Hi All

We have at least 2 people at the moment trying to complete the rossy quest, and getting
riddled with bugs.

First report was that after clearing some caves of monsters it would take a few minutes before they were able to leave the caves.


Now they are being told by the Npc that they have not cleared any caves yet, when indeed they have.

The players names are Chung and 4 elements.

Prsm

verified. Defeating a room doesn't remove barrier or mark quest complete. Working on a fix.

Fixed rossy bug and one crash.

Still have one annoying crash during shutdown ...

Edit: shutdown crash fixed. But since it's during shutdown we can wait until the next otherwise-scheduled restart to fix it. I was planning on totally rewriting that code anyway ...

Please post any outstanding bugs or have we gotten them all?

Remember the only good bug is a dead bug, don't believe me, ask Doogie Howser or Gary Busey's Son and some random b-actress with an ok rack but the military uniform and eagerness makes up for it.
http://www.youtube.com/watch?v=faFuaYA-daw

But I digress.

Looks like testing servers went obsolete.

Regards.

Someone requested help logging in to an account, where the password contains an accented letter.

I tried to change the password using the ladmin tool. With the new source code, ladmin rejects accented letters.

Ladmin-> password test Çenemarchepas
Cowardly refusing to execute a command that includes control or non-ascii characters

The old code accepted this, and obviously people use the character space.

Frost wrote:Someone requested help logging in to an account, where the password contains an accented letter.

This might be a bug; the server is *supposed* to still accept all the same passwords.

But using accented characters is a really bad idea, since it can lead to weak passwords. The most likely cause of the bug is overzealous error-checking, I *thought* I changed the code to give an error if non-ascii characters were used, but perhaps not.

The most likely workaround is to replace the accented characters with a series of underscores, usually 2.

Frost wrote:I tried to change the password using the ladmin tool. With the new source code, ladmin rejects accented letters.

This is not a bug. The old code would silently replace whatever the user typed with something else, which could lead to very weak passwords.

Instead of silently doing something else, the new code brings the error farther forward.

The only automated user of ladmin, the php/sql/cron registration script, only allows alnum characters, so it is not affected.

o11c wrote:using accented characters is a really bad idea, since it can lead to weak passwords.

No, reducing the character space does not result in stronger passwords.

If this is merely a change to the ladmin tool, then players can continue to use accented characters in passwords?