I've been evaluating CVW on Linux recently. There are some problems with the Linux client. Everything used in the system is also in need of an update to get the best security fixes.
It looks like this code drop is over 2 years old now. I don't see activity on the forums for the year 2003. It doesn't look like this is an active project.
Perhaps the users of CVW have moved on?
If there is another application that is like CVW, can someone please post a follow up to this with a URL or name?
I haven't seen much activity within CVW either but have a client who is very interested in using it.
One of the biggest problems I found was the stability on newer versions of Linux. Installation under Red Hat 6.2 seemed to work Ok but on 7.x machines it didn't work (I haven't tried 8.0).
I would be very interested in using it if it didn't seem un-supported. The two posts today are the most activity I've seen in a long time!
Unfortunately I haven't seen any projects that are quite the same as this. While I don't need all the features I definitely need a good place for people to collaborate and store files ... there are some file management systems that have message forums but this provides the messages and files in the same room ...
It's sad it's not more active.
<i>One of the biggest problems I found was the stability on newer versions of Linux. Installation under Red Hat 6.2 seemed to work Ok but on 7.x machines it didn't work (I haven't tried 8.0).</i>
What kind of stability problems did you meet? Was it on the client or the server components?
Maybe if we gather an active community of potential users we can find some developers... what about the code, is it clean and readable (haven't yet looked at it)
I am interested on knowing whether there are any people actively using CVW around here, and about their experiences...
Just to answer a few questions from this and your other post:
- There are no huge issues that I'm aware of. CVW has been tested in a lot of different environments (including low-bandwidth networks) and generally does fairly well. As I mentioned in my other message, neither the server or client have been rebuilt in quite a while - this may account for some of the stability issues noted
- The system could be secured, but it would require a little effort. The client would be rather straightforward, as well as the document server. However, the CVW server would be a bit more difficult. No impossible, but not as straightforward.
- The code is not as clean as it could be. The biggest challenge is the client, which evolved over 3-4 years and could use some cleaning up.
It's true that development on this system is pretty inactive right now. Most of us have been moved on to other projects. We had hoped that someone would have been interested in taking over CVW, but we saw very little contributions back and most people were interested in just setting it up and running their own system.
The good part is that CVW is a pretty stable system. However, it was last built with Java 1.2. And the last client I ran on Linux was I believe was a RedHat 6.1. I think that recompiling the system for newer versions of os's and java vm's would help a lot. If anyone wanted to try and build the servers and/or clients, I would be willing to post them back on SF.
Of course the best forum to gather users together would be if someone had access to an external machine that could host a CVW server. This could probably do a lot in generating interest for other users. We would have some assets here in providing limited help in getting things working, but this would need to be a largely community-driven activity.
What does everyone think?
I have a co-located server where I would be more than glad to run CVW, it runs RH 7.2 has 1 GB RAM and a 60 GB IDE HDD.
If CVW doesn't take a huge bandwidth and resources I could open it to the community
I am willing to recompile the stuff, I certainly would need some help (I am a MD with some knowledge of linux, not a programmer). I can provide SSH shell access to the developers.
As for the project, well, I am currently spending most of my time running after funding and it will be like that in the foreseeable months. Of course if I had some working experimental CVW stuff to show to the potential donators it could help.
The final aim is to be able to offer tele-interaction service to rural consultations in a southern country. As far as I have searched, CVW looks like an excellent tool for it and satisfies the need for an open source, nonproprietary tool that can be further developed.
My name is Max Klohn, I work at Geneva (Switzerland) Institute of Public health, and my direct e-mail is:
amk (at) trauco.net
I am MD of a start up business set up as a virtual organisation. I need a CVW type system and can find nothing as versatile as CVW. There are many comments in the forums over the past year or two from a wide range of people. With a little effort I am sure we could start an active group to develop and use CVW. I am basing my business in financial services, entirely on Linux and open source systems, using Apache, MySQL, PHP as far as possible (exclusively?). Who's prepared to help?
(lots of MDs around I see;-) )
Here's an attempt at a draft "To do":
1) Getting CVW up to date with the current software.
IMHO we could have a new release of the server components in less than two weeks.
a) running with the current release (1.3.x) of the java applet server, Tomcat
Priority: high, since there are vulnerabilities and fixes.
Status: up and running, no modifications needed.
b) using the current release of lambdamoo (to be found at sourceforge)
Priority: low, since the moo sems very stable, but will be testing it.
2) Taking some steps in the way of internationalization
a) Some people have got their MOOs to spit out UTF encoded chars and the like. No such ambition for now, for me if CVW spoke iso-8859-1 ant it didn't ate accentuated characters for breakfast it would be a huge step.
Status: compiled in the following patch by Sindre Soerensen, will be testing it soon.
*** MOO-1.8.0r5-orig/server.c Wed Jan 27 03:44:38 1999
--- MOO-1.8.0r5/server.c Wed Jan 27 03:46:28 1999
*** 48,53 ****
--- 48,55 ----
+ #include <locale.h>
static pid_t parent_pid;
int in_child = 0;
*** 1175,1180 ****
--- 1177,1184 ----
Making international characters work in LambdaMOO:
(this file may also be available at:
LambdaMOO is filtering all input from network-connections through the libc
function isgraph. When the locale environment variables are not set, all
non-American characters are regarded as graphical, i.e. they are not allowed
into the server.
Modern c libraries have support for locales. To make LambdaMOO use locales, two
lines must be inserted into the file server.c (see patch).
for example; when a West European language is chosen, the character set will
be ISO-8859-1 (Latin-1). That leads to 255 characters being available. (or that
is 254, it seems that the server is using the 255th character (in Latin-1: /
y with umlaut) for something else, but I haven't researched this.
See http://www.uni-passau.de/~ramsch/iso8859-1.html for a table of characters
These steps are required:
1. grab the patch for your MOO-server version
2. cd <directory holding the LambdaMOO source files>
3. patch -p1 -c < MOO-version-international.patch
if you are using csh, put this somewhere in the MOO's startup-scripts / rc-files:
setenv LC_CTYPE de_DE
or if you are using bash etc.:
modify the source-code to say
(where de_DE is substituted with your locale.)
"de_DE" is one of the german-language locales. on a unix-system, the
will give you a list of available locales. the command
will confirm which locale categories are correctly set. note that on a
Solaris system "de" may well work, while on GNU/Linux you may have better
luck with "de_DE".
Refer also to man (1) locale or man (7) locale.
5. compile the server as usual
if you think the above seemed too complicated, you can instead make the
modifications manually; like this:
1. open server.c from LambdaMOOs source code.
2. add the line
e.g. below the line
3. add the line (if using german locale)
e.g. before the line
This method has been tested with Korean, French, German and Norwegian locales.
Use at your own risk, it has not been thoroughly tested at all, yet! I am not
responsible for any damage to database-files etc.
You should also be aware that a lot of MOO / Mud-clients have no support for
Sindre Srensen, Sun Jan 24 23:58:10 CET 1999
b) much more remains to be done, we should have some way of setting the locale and auto translation for the main commands, but that is futuristic talk.
3) Finding a way to integrate mrouted functionalities and its configuration into the end app
Especially taking into account the fact that many networks are actually at the hands of the hostile dumb network administrator from hell (TM) that wouldn't allow a linux box or a patch to server without a mountain of forms, pressures from the top, delay and red tape.
4) security fixes, security fixes and more security evolution.
It is obvious that the original developers ran this stuff behind strong and capable firewalls ;-) not everyone has this chance, so the system should be hardened to cope with attacks from the ouside and from the inside. Any easy way to tunnel the communication stuff through SSL for example would be great!
5) Misc suggestions:
* Drop out the packaging of mysql with the server, since working mysql distributions are to be found in most servers and distributions today, and the applet uses the standard port to communicate with them.
* Go for some nice icons! May I suggest Pixelmagick: http://jimmac.musichall.cz/index.php3
Some more possible to-do's ;-)
One of the biggest problems we consistently have run into was the audio/video tools in CVW. When we put the client together, we used the dated vic/vat tools. Now a few years later, these tools are *very* old. These tools often won't work with newer hardware. Although there aren't a lot of options available, the University College of London that are a little more up-to-date. Their vic tool can replace the current LBL vic as is (if I remember correctly). Their 3.x rat tool needs to be modified slightly to work (if I can find the patch I wrote to do this, I'll make it available). Of course, what would be *really* nice would be to tie in the 4.x rat tool into CVW, possibly making it a detachable audio tool from the client.
In the "would be nice" department: Being able to drag-and-drop from a user's desktop into CVW. DnD works within CVW (I was just learning about it at the time) but never made it available from outside the application. Wouldn't think it would be a huge undertaking.
i'm also looking at a crm package such as TUTOS (www.tutos.org) and would like to tie everything together as a seamlessish offering. no point in re-inventing wheels. any suggestions on the best way forward. which packages. should be standard MySQL, PHP, etc.
See my posting under Is CVW an Inactive Project.
From all the people using CVW and who have written comments to the forums over the past two years there must be enough interest and skill to re-activate development of CVW. Comment/ e-mails please.
I am very interested in using and deploying CVW into an international design community. I find that it works well. I have the following issues, however.
1. I am meeting resistance from clients becasue they have to download and install a client. Would it be possible to make the client just a normal web browser?
2. The issue of having to explicitly import files is a little painful. Why can't CVW use a standard file system on the server, accessed via Samba or NFS? The building/room/folder/... metaphor is exactly duplicated by the root/directory/subdir... structure of a file server. Further, the server ould provide the authentication required for access based on requests from CVW.
The model that I envision is of a web-browser client making requests to a CVW server that connects to a standard file structure on the back end.
3. There is a real need for a chat-like interface that allows the chat in the room to be logged as in a forum. Saving scrollback doesn't do it because it creates a file on the user's computer (client-side). We need the chat history to be saved in the room, as a document, with persistence. Further, I need to know when peolple come and go, but I do not need to know about how one person got there.
It seems the first issue around is the acute shortage of developers.
I see lots of ppl interested in features, but there's currently no active development team for CVW.
If ppl are really interersted in the project (I am!) we need to start discussing ways to pool resources to attract developers.
And try to document the source code while the ppl who wrote it are still reachable...
I am a novice of CVW, After going through the overview of CVW , we just tried it and test, compare its performances with other software(server in linux, clients in windows).In the aspect of audio and video, Maybe it exists some quality defects compared with microsoft netmeeting, just like latency and articulation aspect. but the best thing is CVW support Multicast, but i didnt try it, if someone have experienced in this, please give me some advices. In addition, we just prepare to analyse the source code of CVW. is there some softwares to debug and compile MOO code in windows environment? which part of code are the communication and management of CVW ?
By the way, i am very interesting in this project and very glad to see there are many voices after long "hibernation" peroid.
hope your advices please contact with me :email@example.com; my name is andy
IMHO the moo is a fairy stable piece of code, being around since aeons in computing history.
Programming moos seems to be a whole world in itself ;-) yet doesn't seem out of reach. The most likely environment for developping mos in windows seems to be telnet ;-)))
There's also a couple of "graphical" interfaces around for developing, seems still alpha.
If you want to run the moo server under windows maybe you could try with cygwin...
Most stuff that ppl would be really interested in editing and evolving actually reside in the java servlet and the java client.
Speaking of this, and for general reference, here's some hints from Steve that might come useful if you plan to go in code inspection:
A few classes that might help you to start looking at things:
MCPCommand - The representation of the MCP (MOO Communication Protocol) command being sent to/from the server. I'm not sure you would have to change anything in here, but might give you some info on the information
CVWServerComm - A good place to look at incoming messages. Try not to get too nauseas from the endless "if" statements.
RoomScrollback - The object that displays text in the client. Might have to set a locale in here to get text to display correctly (??)
CVWCoordinator - Everything goes through here and everything depends on it. If there has ever been a larger class written in java, I've never seen it. Hopefully, you won't have to spend too much time in it.
As far as your other issues:
It doesn't surprise me that the tcl client wouldn't work. It would
definitely have to be recompiled, and probably a bit of work done to convert the older tcl commands to the 8.3 syntax. Also, the tcl client was never updated to handle the new doc server, so all its document commands would have to be completely rewritten (if you wanted it to do
All the java clients had a shared library to do some stuff Java couldn't do at the time, such as setting file permissions and reading registry entries. See "WindowsDocServerUtils.java" for more details.
There are some Java MUD servers out there (currently inactive). The latest/greatest seems to be either Xmud or Momoko.
MoMoKo (might be better, since it self-distributes/federates)
Also ScryMUD, metacosm, jaMUD, JMUD.
Advantages still seem to go to MUD/MOO over other, more complete Java libs/APIs and some of the devs seem to be moving (albeit VERY slowly) into the world of XML, especially in the medical and educational/distance learning arenas.
Also, the Java clients I see out there MAY solve the I18N problems from the server thru the client.
Wow. I am really happy to see that so many people are getting more interested in this program! Kudos to the folks at Mitre that did the right thing and moved this system into Open Source.
As an overview of myself, I have used CVW when it was a Mitre project (Steve remember EFX?) and when it was Open Source, and while many people see the product that is out there, a lot of dollars were spent to get this program operational.
Add me to the list to help out with this project, and while I am not a software developer, I am a systems architect, and I can provide some history as to the use of this product.
Steve, glad to see you are still around!
I would like to say something, actually this can be part of sourceforge developer tools (Oh yes) so that Developer can actually use this.
I can't help any way... (Java is totally beyond my understanding) But I can help using it ;D
We just fired up our first CVW server and are smitten: we love it!, and think a lot of nice work went into.
We would also like to thank Mitre for making this part of the 'commons'.
I have been looking for a free (preferrably) or open source project to contribute to, and I think this may be the one - I have about 2 years of Java experience (20+ in development in general, C, Python, etc.), and have been using Linux for 5 or 6 years now.
Anybody have any suggestions as to what it takes to 're-activate' a project like this?
It seems we have enough interest (I am especially interested in the medical project mentioned above - my father was a medical doctor (as opposed to managing director)/PC tinkerer and would have been fascinated by the project).
I've never been a 'project co-ordinator/leader' on a 'virtual' team, but I certainly have been in the 'bricks and mortar' world - I'll (gulp!) volunteer to at least get this thing going if people are still interested - we'd like to extend it for a project of ours anyway.
Just saw this thread and would like to offer an additional set of kudos to the team at Mitre and to all the past developers...I've actually deployed this product in collaborative environments and distance learning situations pre-09/11/01 (total count:17 site installs). I've also seen the offshoot (Windows-based) that US DoD adopted instead...CVW IS better now and could be MUCH better in the future. It really surprised me when the DoD went with the competition, rather than CVW.
I would think that, given the realities in the world today, this would be an EXTREMELY popular project. The rest of the world has caught up, with Java SSH tunnels and controllers, Java H.323 clients and servers and the rise (comparatively) of broadband and other enabling technologies would make this MCH easier...heck, we can even deploy to more than one handheld and to CELL PHONES!!!
I currently am the manager of the HSQLDB Project and the maintainer of the (inactive for now) Vienna Project here at SF.net...I would be willing to volunteer to add some bandwith here in whatever capacity necessary.
Some of master students are developing different projects of CVW as their thesis in my university. I am included, just started though. Hopefully I can share experience with you guys here.
Another MD wieghs in! Lots of us virtual MDs around I see. From what I see on this list we have enough interested parties to make CVW breathe again.
I'm willing to devote time and resources. If none of the other team members are willing or interested in project managing this, I'd be glad to, since that seems to be the missing ingredient here.
What I suggest is the following:
1. All those actively interested in building CVW further, send me an e-mail (click on my name and follow link to send e-mail).
2. Once we have a core team of interested parties, we can work as a group to resurect CVW.
I look forward to hearing from you.
I am in talks with the Welsh Development Agency and the University of Wales to get support for Open Source development. I have proposed a number of LAMP projects and CVW is high on this list. The feedback is very positive from the above and I hope to have fairly concrete support in two to three weeks time.
I see CVW as an integral part of running my business and would be keen to do whatever I can to help. I have little current technical skills, but know a man that has.
CVW is too good not to make it work.
Log in to post a comment.
Sign up for the SourceForge newsletter:
You seem to have CSS turned off.
Please don't fill out this field.