mono guys cant blind-fix
The fact that mono can have bugs/incomplet impelmentation is possible like in any software. But how do you except the mono developer to fix a bug they dont even know it exist ? Well, you are all programmers here, u know what i mean. If you want a bug to be fixed, the best way could be to report it (assuming it's a bug, reading previous post, it seems to be a problem in mono-internal functions?). At last it could work a lot better than just just sit there and praying "let's hope they will fix the secret bug we didnt told them about"
A general way in debuging a program with multiple componnents/library is :
- isolate the problem : determining which fuction call cause the problem, with what parameters, reducing the code to 5..10 (or any number as small as plossible) lines that expose the problem.
- communicate with mono staff : they have a complete bug management system :
http://www.go-mono.com/bugs.html
they also have a mailing list that you can find on the contact page
http://www.go-mono.com/contact.html
- then wait
now you are allowed to claim "it's not our fault it's the evil mono thing"
But jumping right to step 3 might not be the best solution. Note that i have a very limited .net framework knowlegde and of course i know nothing about the runuo source code (and i cant), but at last i find the above not completely stupid.
It's not a flam post or any thing like that, and i find rather positive that the #1 uo server emulator has some interest in a real server operating system. It's a lot better than sphere for example, they make linux fan waiting for 6 month for a version update and then they nearly annouce "sorry we drop linux support, we use a super new windows only architecture." there is about 15 threads about it in their forums :
http://www.sphereserver.net/board/showthread.php?t=7103
It would be a good timing if it could work in the following months : it would allow a large amount of linux users to upgrade.
and i didnt talk about any "use wine" solution, while runing a client program with wine is sometime acceptable when no other choice are possible, using wine for an internet-wide SERVER is realy something that hurt. It has some strong dependencys on the XFree graphic server software, that should not be installed on linux server (mine just has no screen
. It require indirectly administror privileges to run, that bring the server security to something comparable to a windows-level security (any failure on wine/x graphic server/runuo could cause full system root-level compromision, not just to an unprivilegied account). Plus it add in all case about 200 mb of useless wine/graphic server on the hard drive, and also represent an burden of at last 80 mb of ram, maybe more if wine must emulate multi thread/process (it take 28mb of ram to run only notepad.exe with wine, an Xserver easily take the rest). The speed penality vary a lot, things get worse for a program with a lot of I/O (emulating every winsocks call has a price, just imagin the number of external dll calls the server make to send a network packet, multiply that by the number of users on the server and by the number of packets/s/user)
The support of this or that software is very changing between minor versions of wine (they cant check for regression by testing it with every windows program), meaning that upgrading wine could cause runuo to stop working even if it could work with current wine version, and upgrading is not allways a matter of changing the version number, but could also be a security concern.
So maybe you can try it yourself with wine on your 2 pc local network durring 20 minutes and say "ooOOOooOO it works", but it will never be the 500+ users & infinite uptime with the wine way.
let me add : good jobs making a completely usable server, it's not an easy task i know. And that's about all i have to say. (you can notice i didnt have talked about opensource advantages, i'm a very tolerant linux addict