 |
|
| It is not so hard to compiling muscle by yourself |
|
|
By lelldorin - Posted on May 14, 2008 - 23:12:53 (#22796)
Current version when comment was posted: 4.25 |
|
|
We wrtie a tutorial in order to explain the compiling of muscle and we explaint how to use BeShare in LAN.
http://www.besly.de/menu/search/archiv/net/beshare_LAN_eng.html
|
|
| people need compiled version |
|
|
By lukesky - Posted on March 27, 2008 - 01:25:55 (#22637)
Current version when comment was posted: 4.23 |
|
|
just a few are programmers......so first people
need compiled version......all the best for
beos programmers.....p.s.....beos need flash !....
|
|
| HashFunctor<QString> |
|
|
By Monni - Posted on June 18, 2007 - 12:26:51 (#21739)
Current version when comment was posted: 3.33 |
|
|
Well... I fixed it by grabbing the qhash.cpp from Qt 4.x and pasted it directly in QMuscleSupport.h so it doesn't need whole header at all ;)
|
|
| QMuscleSupport.h |
|
|
By jaf - Posted on June 18, 2007 - 11:22:42 (#21738)
Current version when comment was posted: 3.33 |
|
|
Hi Monni,
You're right, that was an oversight. It will be fixed in the next MUSCLE release. If you want the fix before then, email me and I will send you the updated QMuscleSupport.h file.
-Jeremy
|
|
| Doesn't compile with older Qt versions |
|
|
By Monni - Posted on June 18, 2007 - 10:37:59 (#21737)
Current version when comment was posted: 3.33 |
|
|
For some reasons in QMuscleSupport.h "#include <qhash.h>" isn't surrounded by #if's for QT 4.x, it isn't available on older versions. This causes compilation to fail on older versions.
|
|
| BeOS |
|
|
By Monni - Posted on May 5, 2007 - 02:26:49 (#21635)
Current version when comment was posted: 3.32 |
|
|
BeOS siam 5.0 1000009 BePC unknown
|
|
| Re: BeOS gmtime_r() |
|
|
By jaf - Posted on May 4, 2007 - 16:29:12 (#21634)
Current version when comment was posted: 3.32 |
|
|
And what version of BeOS are you running?
|
|
| BeOS gmtime_r |
|
|
By Monni - Posted on May 4, 2007 - 16:01:47 (#21633)
Current version when comment was posted: 3.32 |
|
|
In my BeOS box return value is actually pointer to struct tm if call is successful, otherwise it's 0 (NULL).
|
|
| Haiku tweak |
|
|
By jaf - Posted on May 3, 2007 - 17:52:09 (#21632)
Current version when comment was posted: 3.32 |
|
|
Okay, done. Thanks for the info.
-Jeremy
|
|
| Re: BeOS tweak |
|
|
By jonas.kirilla - Posted on May 3, 2007 - 17:17:07 (#21631)
Current version when comment was posted: 3.32 |
|
|
As both __BEOS__ and __HAIKU__ are likely to be defined in future Haiku, I think you would have to do something like this:
#if defined(__BEOS__) && !defined(HAIKU)
struct tm * tm = gmtime(&now);
#else
struct tm * tm = gmtime_r(&now, &gmtm);
#endif
BTW, thank you for keeping us up to date with your changes! :)
|
|
| gmtime_r() return semantics |
|
|
By jaf - Posted on May 3, 2007 - 15:47:14 (#21630)
Current version when comment was posted: 3.32 |
|
|
Sure, but what does the integer return value of gmtime_r() mean? Is it "zero on success, non-zero on failure", or is it a time value of some sort?
I don't want to use BeOS's gmtime_r() unless I understand how it works, and I don't know that it is documented anywhere.
|
|
| gmtime_r |
|
|
By Monni - Posted on May 3, 2007 - 14:47:48 (#21629)
Current version when comment was posted: 3.32 |
|
|
I've seen a few BeOS apps use gmtime_r with no problems. It really doesn't need "struct tm *" return value as it is only copy of second parameter. When return value is pointer, it is expected to be NULL on error, but equal to second parameter on success.
|
|
| BeOS tweak |
|
|
By jaf - Posted on May 3, 2007 - 10:32:44 (#21628)
Current version when comment was posted: 3.32 |
|
|
Hi Jonas,
Thanks for the notification. I've changed the code in SetupSystem.cpp to this:
# ifdef __BEOS__
struct tm * tm = gmtime(&now);
# else
struct tm * tm = gmtime_r(&now, &gmtm);
# endif
I'd put in an #if to have it use gmtime_r() under Haiku, but I don't know what preprocessor defines to expect under Haiku...
-Jeremy
|
|
| server build fails on gmtime_r() |
|
|
By jonas.kirilla - Posted on May 3, 2007 - 04:20:52 (#21627)
Current version when comment was posted: 3.32 |
|
|
gmtime_r() returns int on BeOS.
"
SetupSystem.cpp:311: initialization to `tm *' from `int' lacks a cast
make: *** [SetupSystem.o] Error 1
"
FWIW, in Haiku gmtime_r() returns the expected struct tm *.
|
|
| oopsie |
|
|
By jaf - Posted on March 15, 2007 - 12:14:40 (#21433)
Current version when comment was posted: 3.31 |
|
|
... ever since I mis-typed the file size. It should be fixed now :^)
|
|
| wrong file size |
|
|
By AlienSoldier - Posted on March 15, 2007 - 10:32:18 (#21432)
Current version when comment was posted: 3.31 |
|
|
since when MUSCLE is 15M? :P
|
|
| CueStation |
|
|
By kancept - Posted on July 6, 2006 - 09:14:43 (#20101)
Current version when comment was posted: 3.22 |
|
|
CueStation runs fine here in R5 and in ZETA. Now to get off my butt and buy an old LD44.
|
|
| re: MUSCLE use for LCS |
|
|
By jaf - Posted on March 30, 2006 - 10:35:24 (#19334)
Current version when comment was posted: 3.20 |
|
|
Hi Nutela,
We use MUSCLE as a networking layer for CueStation (the control application for our client/server audio mixing system) and CueConsole (our modular control surface). As a result, our systems are fairly unique in our industry in that they can be co-operatively controlled and monitored by multiple people on multiple PCs (and/or Macs) simultaneously. See here for details:
http://www.lcsaudio.com/Products.html
As far as porting things to Haiku... we could probably port the server side now (assuming Haiku's networking is robust), but to port the GUI/client side we would need to have Qt 3.3.x ported to Haiku. So let me know when that happens :^)
Jeremy
|
|
| Oh and.... |
|
|
By nutela - Posted on March 30, 2006 - 08:58:43 (#19333)
Current version when comment was posted: 3.20 |
|
|
when are you going to port those super-uber-ultra-cool apps from LCS to Haiku? Drool : ))
|
|
| Musle use for LCS? |
|
|
By nutela - Posted on March 30, 2006 - 08:57:54 (#19332)
Current version when comment was posted: 3.20 |
|
|
Hi jaf you already told me but I forgot most of it and it would be interesting for others to retell, what was the purpose of developing Muscle(d) (for LCS)? : )
|
|
| I love you and.... |
|
|
By ablyss - Posted on August 14, 2004 - 05:11:18 (#13858)
Current version when comment was posted: 2.52 |
|
|
...and all the other BeOS programmers who keep the BeOS spirit alive.
Cya!
|
|
| UDP & Monni |
|
|
By Monni - Posted on January 29, 2004 - 02:09:29 (#10816)
Current version when comment was posted: 2.46 |
|
|
Sure... I do have a camera that should work with both BeOS and Winblows... Then I could investigate Windows Imaging API and make Unizone support it too... But first you need to make the framework in BeShare.
Actually I asked because one Finnish BeOS User Group member noticed this and he is almost as bad Zeta basher as me and wanted to see Bernd's face when we talk all the crap about his toys.
|
|
| Re: UDP |
|
|
By jaf - Posted on January 28, 2004 - 12:20:24 (#10798)
Current version when comment was posted: 2.46 |
|
|
Only if I can get you to be the on-camera entertainment, Monni... ;^) (actually, UDP support was added because we needed UDP support in some of our LCS software -- it isn't used by the 'vanilla' muscled or BeShare)
|
|
| UDP? |
|
|
By Monni - Posted on January 28, 2004 - 11:54:47 (#10797)
Current version when comment was posted: 2.46 |
|
|
Does this mean you are preparing for pr0n cam support for BeShare... That would be like Wh-oooooo-a or something ;)
|
|
| String::Arg() |
|
|
By Monni - Posted on November 11, 2003 - 02:46:55 (#9588)
Current version when comment was posted: 2.45 |
|
|
I think this method is a good idea, but it is kinda ugly and doing all things exactly QString way isn't always best way to do things. Also... QString has direct methods to convert single numerical value to a string representation without using %1...%9 and they are a lot safer than String::Arg(). For example storing URL's in QString doesn't work with [aA]rg()-style methods (Because % is reserved character in URLs).
|
|
| Thanks for the good work |
|
|
By jefro - Posted on July 28, 2003 - 13:39:40 (#8339)
Current version when comment was posted: 2.42 |
|
|
And thanks for the reply. Sorry I don't communicate well.
Great work here. Thanks for another great beos app./tool.
|
|
| re: Thanks |
|
|
By jaf - Posted on July 24, 2003 - 19:20:17 (#8309)
Current version when comment was posted: 2.42 |
|
|
You're right, it could work for those types of data, if someone were to implement the appropriate client software to support it. I must have misunderstood your question -- I thought you were asking if it could do those things without any additional programming.
|
|
| Thanks |
|
|
By jefro - Posted on July 24, 2003 - 18:59:54 (#8308)
Current version when comment was posted: 2.42 |
|
|
I agree that it is a building block for other apps. When I read the documentation I seemed that there could be a lot of things that could be done with it. Since it has shown itself to be stable and portable that is a very big start. It has already proven itself as a method to transport files but for some reason I could imagine a replacement for x on beos or remote desktop. Why would that type of data not work?
|
|
| re: Sometimes... |
|
|
By jaf - Posted on July 20, 2003 - 23:31:42 (#8239)
Current version when comment was posted: 2.42 |
|
|
Hi Jefro,
I think perhaps you misunderstand what a muscle server does. It is not designed to export a shared filesystem (the way CIFS or BeServed do). Instead, it implements a custom messaging protocol that various muscle-aware apps (e.g. BeShare or Ozone) can use to send messages from one computer to another. These clients can use this message-passing facility for any purpose they want -- for example, BeShare and Ozone use it to implement Napster-style file-sharing and chat, while Battleship and FoxRabbitCarrot use it to implement multiplayer games, and another program might use muscle to implement a distributed computation net or "render farm". But the muscle server itself only knows how to pass messages from place to place, and so it wouldn't be useful by itself for NFS/CIFS type applications. Consider it more of a building block for use in larger systems, then something that is useful entirely by itself.
|
|
| Sometimes.. |
|
|
By jefro - Posted on July 20, 2003 - 22:12:55 (#8238)
Current version when comment was posted: 2.42 |
|
|
Sometimes I think and then say better not ask a stupid question but here is maybe on or two.
Could muscle be used for a beos remote desktop similar to windoz's?
Secondly how hard would it be to set up a windows muscle server and use it to transfer files like CIFS/SMB?
Or did I not understand what is going on?
|
|
| oops (wish i could edit my comment) |
|
|
By sikosis - Posted on June 28, 2003 - 03:13:18 (#7936)
Current version when comment was posted: 2.41 |
|
|
thats ... ReflectServer.cpp
|
|
| found it |
|
|
By sikosis - Posted on June 28, 2003 - 03:12:15 (#7935)
Current version when comment was posted: 2.41 |
|
|
in these two files:-
server/admin.cpp
reflector/Reflector.cpp
I added #include <sys/select.h>
and then it compiled. ive got bone 7a btw ...
|
|
| muscle vs. bone |
|
|
By jaf - Posted on June 28, 2003 - 03:03:21 (#7934)
Current version when comment was posted: 2.41 |
|
|
Hm, I just downloaded the archive to my BONE machine and compiled it and it worked fine, so my guess is that there is something wrong with the BONE headers installed on your machine. Perhaps you have an older version of Bone installed? (The one you need to have is bone 7a). Also make sure that the only bone headers folder you have is the one at /boot/develop/headers/be/bone. Earlier bone versions had that folder elsewhere, and if you had an earlier bone version installed, that "other" bone folder might still be around and confusing things.
|
|
| got bone |
|
|
By sikosis - Posted on June 28, 2003 - 02:49:24 (#7933)
Current version when comment was posted: 2.41 |
|
|
ive got bone installed and working with muscle 2.4 currently - however, 2.41 doesnt build. is there some include file missing perhaps ?
it does say -DBONE in the compile lines.
|
|
| Bone/net_server confusion? |
|
|
By jaf - Posted on June 27, 2003 - 18:35:18 (#7925)
Current version when comment was posted: 2.41 |
|
|
Usually when I see those errors it means that the Makefile is trying to compile muscled for BONE on a non-BONE system, or vice versa. Look in the compile lines generated by the Makefile... if they have the keyword -DBONE in them, the Makefile thinks your system has BONE installed... if this is not the case (or if you do have BONE installed but the -DBONE keyword is missing) then the BONE autodetect isn't working. The BONE autodetect in the Makefile looks for the folder /boot/develop/headers/be/bone. If that folder exists (specifically, if it finds the file bone_api.h in that folder) it assumes BONE is installed. You may need to delete that file if you had bone installed and then later removed it...
|
|
| build errors |
|
|
By sikosis - Posted on June 27, 2003 - 18:23:36 (#7924)
Current version when comment was posted: 2.41 |
|
|
I get a bunch of build errors when trying to "make" muscle
/boot/home/apps/muscle/reflector/ReflectServer.cpp: In method `status_t muscle::ReflectServer::ServerProcessLoop()':
/boot/home/apps/muscle/reflector/ReflectServer.cpp:243: `fd_set' undeclared (first use this function)
/boot/home/apps/muscle/reflector/ReflectServer.cpp:243: (Each undeclared identifier is reported only once
/boot/home/apps/muscle/reflector/ReflectServer.cpp:243: for each function it appears in.)
/boot/home/apps/muscle/reflector/ReflectServer.cpp:243: parse error before `;'
/boot/home/apps/muscle/reflector/ReflectServer.cpp:248: `readSet' undeclared (first use this function)
/boot/home/apps/muscle/reflector/ReflectServer.cpp:248: implicit declaration of function `int muscle::FD_ZERO(...)'
/boot/home/apps/muscle/reflector/ReflectServer.cpp:249: `writeSet' undeclared (first use this function)
/boot/home/apps/muscle/reflector/ReflectServer.cpp:266: implicit declaration of function `int muscle::FD_SET(...)'
/boot/home/apps/muscle/reflector/ReflectServer.cpp:364: implicit declaration of function `int muscle::select(...)'
/boot/home/apps/muscle/reflector/ReflectServer.cpp:399: implicit declaration of function `int muscle::FD_ISSET(...)'
make: *** [ReflectServer.o] Error 1
|
|
| Duh! |
|
|
By MercuryMan2000 - Posted on April 7, 2003 - 13:04:36 (#6762)
Current version when comment was posted: 2.30 |
|
|
Ok, I realized after I posted my question that it was stupid!
Please ignore!
|
|
| hmm |
|
|
By jaf - Posted on April 7, 2003 - 12:51:45 (#6761)
Current version when comment was posted: 2.30 |
|
|
If ./muscled help is crashing, it sounds like maybe your muscled executable is munged somehow. Try doing a "cd muscle/server; make clean; make" to regenerate a fresh muscled executable from scratch. Assuming you have a correct executable, BeShare should work properly on a muscled server with no arguments supplied -- make sure that the BeShare client(s) are actually sharing some files (if they are, it will say how many files are being shared by each BeShare client in that client's window's title bar), and note that two "I'm Firewalled" BeShare clients won't be able to see or download each other's files (since there is no way for either to connect to the other)
|
|
| ./muscled help crashes |
|
|
By MercuryMan2000 - Posted on April 7, 2003 - 12:16:19 (#6760)
Current version when comment was posted: 2.30 |
|
|
So I have no idea what command line switches I can throw to it. ./muscled by itself runs, another BeOS machine with BeShare can connect, but not find any files. How do I specify what dir structure to share?
Thanks!
|
|
| "BeShare" on Windows |
|
|
By Monni - Posted on February 13, 2003 - 08:06:13 (#5909)
Current version when comment was posted: 2.25 |
|
|
Ozone isn't dead, and despite the fact that VitViper doesn't like Unizone, it is also still alive and under development.
- Monni -
|
|
|
|
|
|
Comment Pages:
<< prev | 1 | 2 | next >> |
|
|
