Please preload the update (fixed)

[Entropy]

This is the FIXED version.

Please make a new torrent.

 

http://www.rivanwarriors.com/radu/el_update_131.exe

 

Please preload it, and if you wish you can install it as well, this client is relatively compatible with the current one (some small map discrepancies might be observed until the server is restarted).

As you can see, there are two clients, one is el.exe which is the 586 version, and a SSE2 client for those with newer CPUs, which should be a little bit faster.

 

The full client can be found here:

http://www.rivanwarriors.com/radu/el_131_install.exe

 

If anyone wishes to create a BitTorrent stream and seed it, please give me the URL.

 

Once we have the Mac and Linux client, we'll update.

[ttlanhil]

is it just the execs that changed, or data files as well?

[the_voice]

bittorrent version of the FIXED update

 

Eternal Lands 1.3.1 update

Eternsl Lands 1.3.1 Full Install

[Roja]

2 maps were fixed as well.

[dustpuppy]

Alternate download site for full and update

[ttlanhil]

2 maps were fixed as well.

ahh, okay, thanks. I've re-uploaded the data zip (for *nix users... or anyone who uses CVS, really)

the failed CVS ups in lang\en\encyc are gone, but \lang\en\strings\options.xml~ is there. I've removed that and the executables from the data .zip; but everything else in the .zip is the same as in the update installer

 

the updated zip is still here

[Learner]

2 maps were fixed as well.

ahh, okay, thanks. I've re-uploaded the data zip (for *nix users... or anyone who uses CVS, really)

the failed CVS ups in lang\en\encyc are gone, but \lang\en\strings\options.xml~ is there. I've removed that and the executables from the data .zip; but everything else in the .zip is the same as in the update installer

 

the updated zip is still here

The official Linux .zip's with all the junk and windows files still in there (static & dynamic versions) are available at http://www.other-life.com/el/elc-1.31.zip and http://www.other-life.com/el/elc-1.31_update.zip

[ttlanhil]

from the previous 131 page:

Both exe's crash on my computer when I close the game, anyone else have this issue?

I recon that's a double-free in the text messages... it's fixed in CVS, but that was after this thread started *ducks*

[Jinxx]

I'm was wondering if we could get an explanation as to the different clients included. Also, the SSE2 and the regular one are crashing on #exit, and I have no audio. Thanks

[ttlanhil]

I'm was wondering if we could get an explanation as to the different clients included.

SSE is a feature of newer CPUs, that allows them to do some things a bit faster. the SSE2 client can use some of these operations, so it'll be quicker. on the other hand, SSE2 is only available in newer CPUs, meaning if yours doesn't, it won't work well. that's what the other client is for.

in short: try the SSE one. if it don't work, use the other

Also, the SSE2 and the regular one are crashing on #exit

that'd be the post above

and I have no audio. Thanks

go into config->audio and see if the boxes are empty. if so, try clicking on them

ed:s/sound/audio/

Edited July 30, 2006 by ttlanhil

[Jinxx]

I'm running a Newcastle Core AMD 64 3000 cpu, SSE2 client should be groovy for me. Also I have the audio enabled. I'm getting that Error Initializing Sound : Invalid Operation error like on the last update.

Edited July 30, 2006 by Jinxx

[ttlanhil]

I'm getting that Error Initializing Sound : Invalid Operation error like on the last update.

phooey, that was supposed to have been fixed. I'm talking with jinxx in-game, to try to fix, which will be another fix in CVS after the thread was started. *sigh*

ed: anyone else who has a bug with this version, speak up quick so it can be fixed before ent has to rebuild another

Edited July 30, 2006 by ttlanhil

[Entropy]

Grr..

So does that patch work?

[ttlanhil]

Grr..

So does that patch work?

the audio stuff? I sent jinxx a client with some extra logging in it, to try to find when and how the sound didn't work, but the sound worked... and no, it shouldn't have

last change in sound.c in CVS is days ago, so it's not a recent change there. can't think of any other changes in CVS recently that should be able to affect sound init either

I can only wonder if you got an old version from CVS (it has been mucking about recently. most people didn't get the sound init errors, so it fits) or are linking to libs with some funky optimisation (which would have shown up before, so it shouldn't be that)

dunno what else may cause it

 

ed: okay, now music track changes are causing jinxx's sound to break. more to look into

 

the crash on exit, should be fixed in CVS (assuming it's where I suspect, but what I fix can be expected to crash, and I don't think other changes caused it, so pretty sure)

Edited July 30, 2006 by ttlanhil

[Draym]

I keep experiencing that same problem: crash on #exit and music breaks in track changes

[ttlanhil]

I keep experiencing that same problem: crash on #exit

fixed

and music breaks in track changes

does it stop playing entirely, and knock out sound effects (like rain) as well? or does it just take a few seconds (like, 1-20+) to start the next song? the delay seems to be the problem jinxx is getting with the client I made (which fixed that unknown no-sound-at-all bug)

only looking for new problems

[Draym]

well, the problem I have kill the sound completely, both music and sfx... I noticed it started happening since I got that "too many sounds" bug entering pl storage

[ttlanhil]

the above is now fixed in CVS (at least for jinxx... considering that the problem shouldn't have happened, and the fix shouldn't really work, well...)

"too many sounds" bug entering pl storage

next to investigate. I know this is part of the .elm now, but I'll see what else I can dig up

[Learner]

the above is now fixed in CVS (at least for jinxx... considering that the problem shouldn't have happened, and the fix shouldn't really work, well...)

"too many sounds" bug entering pl storage

next to investigate. I know this is part of the .elm now, but I'll see what else I can dig up

Current CVS and previous version can bot be crached by not having any sounds, enable sounds & music, and then disabling them. In my tests it seemed that disabling music made it crash.

[Knuckles]

#cls still makes the Client go funny..

[ttlanhil]

meh, scratch that, jinxx was still getting crashes sometimes. not all the time, not even with the same sequence of events (tp lakeside to wsc), but sometimes.

sometimes it's only music that dies, not all sound, meaning it's in stream_sound, oter times it's sound as well, so update_position. with a call to the openal error check function in every single sound function that calls something in openal, it can still get an "Invalid Value" at the start of stream_music... and according to what I've added for error checking in sound.c it should not be possible to get this bug... if you assume the problem is in elc. if, on the other hand, it's openal... well... that'd explain why I can't isolate it, and why i can't even reproduce it.

I can't reproduce learner's bug either. he's gonna try debugging it tomorrow, because I can't

sorry folks, it's gonna take a bit longer

 

I'd like to thank jinxx again for downloading test clients over and over (at least a dozen times) as I changed debug logging and such, trying to isolate the fault, and then doing a bunch of mapchanges and tp jumps to reproduce the bug

 

meh. chasing bugs I can't reproduce is driving me batty... need to go work on a feature so I can't test it locally and see progress happening

 

as for the portland storage "too many sounds" error... it's in the map file.

too many particle systems with sounds attached are loaded from file (maybe the map-maker thought they were culled-by-distance, so could use a lot? they're only paused at distance, all loaded, so there's a limit on the number you can have map-wide.)

the limit of my mapmaker experience is getting object ids, so someone else will have to fix this one

 

#cls still makes the Client go funny..

arg, that was supposed to be fixed. was done a couple of days back (unless ent only updated the version numbers for this build and didn't cvs-up, or cvs is still acting up, in which case the fix may not have made it in) Edited July 31, 2006 by ttlanhil

[fjornada]

Client works fine for me. Grabbed Learner's release, just got some (minor?) errors during startup, like:

 

languages/en/strings/help.xml:24: parser error : StartTag: invalid element name
<<<<<<< help.xml
^
languages/en/strings/help.xml:24: parser error : StartTag: invalid element name
<<<<<<< help.xml
 ^

 

but the enc. worked without problems. a little FPS improvement in crowded ares.

 

One question, in the encyclopedia:

 

Steel Graves: 5-6 Armor

Tit Graves: 4-5 Armor

 

is it right?

[Entropy]

Why wouldn't it be right?

[Knuckles]

#cls still makes the Client go funny..

arg, that was supposed to be fixed. was done a couple of days back (unless ent only updated the version numbers for this build and didn't cvs-up, or cvs is still acting up, in which case the fix may not have made it in)

 

When ever I type #cls and then go into Consol, all of the Consol goes funny, I can take/show you a screenie if it well help any.

 

#edit - Ent has already said somewhere that Steel Plate is better than Titanium.

Edited July 31, 2006 by Liquid

[ttlanhil]

languages/en/strings/help.xml:24: parser error : StartTag: invalid element name
<<<<<<< help.xml
^
languages/en/strings/help.xml:24: parser error : StartTag: invalid element name
<<<<<<< help.xml
 ^

heh, unfinished CVS merge. it shouldn't do you any harm though

Steel Graves: 5-6 Armor

Tit Graves: 4-5 Armor

is it right?

yeah. tit is better elemental protection, but not as good at physical (and, I think ent said, tit breaks easier. grep the forums for that though)... it's good to have more strategy in your choice of kit!

Why wouldn't it be right?

because people expect tit to be better than steel *shrug*

When ever I type #cls and then go into Consol, all of the Consol goes funny, I can take/show you a screenie if it well help any.

I imagine it's various coloured characters for a few spaces, that's the bug with new #cls I've seen before Edited July 31, 2006 by ttlanhil