rudel-devel — Discussion of Rudel development

[Rudel-devel] Development docs/guides/tips/help?

[Noam P. <npo...@us...>]

Hi, I'm trying to work out how the rudel code is put together, though
I'm finding the CLOS based stuff a bit hard to follow.

Any other tips for understanding/debugging the code?


cross-posted: https://github.com/scymtym/rudel/issues/4

[Rudel-devel] can not connect

[Lio N. <lio...@ra...>]

 

Hello. 

Me and my friend have been struggling for a few days now to start the
collaborative environment with gobby and rudel. We did not menage to
connect to gobby with rudel-join-session and we tried every possibility.
We use emacs 25.1.1, gobby 0.5. Rudel version is the one from elpa
0.3.1. Is there a problem with compatibility with versions? 

Usually it hangs at joining (new) [if we use obby/tcp ] 

if we use [obby/start-tls] it hangs at joining (they-finalized) 

We menaged to join together just by gobby. We also menaged to connect
with rudel-publish-buffer (but not via spacemacs - only in plain emacs
with simple init file). 

Any ideas? 

Best regards 

Lio Novelli 
 

Re: [Rudel-devel] eieio calling rudel-backend-factory

[Arne B. <arn...@we...>]

There should already be a fix for that. See

- https://lists.gnu.org/archive/html/emacs-devel/2016-07/msg00575.html
- https://lists.gnu.org/archive/html/bug-gnu-emacs/2016-07/msg00868.html

Best wishes,
Arne

Sam George <sam...@co...> writes:

> Hello.  I'm trying out rudel as I'd really like to have collaborative
> editing abilities in emacs.  My problem is that after loading the rudel
> definitions, the commands:
>
> rudel-host-session
> rudel-join-session
> rudel-backend-dump
>
> all respond with:
>
> slot-value: eieio-oref called on a class: rudel-backend-factory
>
> eieio-version C-x C-e   -->   1.4
> emacs-version C-x C-e   -->   GNU Emacs 25.1.1 (x86_64-unknown-cygwin)
> of 2016-09-17
>
> Where do I start to fix it?
>
> -Sam
>
> ------------------------------------------------------------------------------
> Announcing the Oxford Dictionaries API! The API offers world-renowned
> dictionary content that is easy and intuitive to access. Sign up for an
> account today to start using our lexical data to power your apps and
> projects. Get started today and enter our developer competition.
> http://sdm.link/oxford
> _______________________________________________
> Rudel-devel mailing list
> Rud...@li...
> https://lists.sourceforge.net/lists/listinfo/rudel-devel


-- 
Unpolitisch sein
heißt politisch sein
ohne es zu merken

[Rudel-devel] eieio calling rudel-backend-factory

[Sam G. <sam...@co...>]

Hello.  I'm trying out rudel as I'd really like to have collaborative
editing abilities in emacs.  My problem is that after loading the rudel
definitions, the commands:

rudel-host-session
rudel-join-session
rudel-backend-dump

all respond with:

slot-value: eieio-oref called on a class: rudel-backend-factory

eieio-version C-x C-e   -->   1.4
emacs-version C-x C-e   -->   GNU Emacs 25.1.1 (x86_64-unknown-cygwin)
of 2016-09-17

Where do I start to fix it?

-Sam

Re: [Rudel-devel] rude + el-get + emacs 25

[Arne B. <arn...@we...>]

Jan Moringen writes:

>> i try to get this running. but i run into several problems.
>> is this one of the supported installation modes?
>
> Rudel has been dormant for a couple of years now. I wouldn't generally
> expect it to work in modern Emacs (on any platform).

I just setup a rudel server again using the following snippet:
https://bpaste.net/show/6e5e0fb65bd6

It is run via emacs --daemon -l ~/path/to/file, using a dedicated user
account.

I do not know whether the snippet is the most efficient way to run a
rudel server, but it seems to do its job.

Best wishes,
Arne
-- 
Unpolitisch sein
heißt politisch sein
ohne es zu merken

Re: [Rudel-devel] rude + el-get + emacs 25

[Arne B. <arn...@we...>]

Hi,

Jan Moringen writes:

>> i try to get this running. but i run into several problems.
>> is this one of the supported installation modes?
>
> Rudel has been dormant for a couple of years now. I wouldn't generally
> expect it to work in modern Emacs (on any platform).

At least in this Emacs (GNU Emacs 24.5.1) Rudel works pretty well.

Try connecting via protocol obby with backend tcp to server
d6.gnutella2.info using port 6522 without encryption.

(sorry, no security, so don’t use it for anything more sensitive than an
etherpad)

>> is rodel even running on OS X?
>
> It think it has been in the past (never be me, though). But as said,
> changes in Emacs may have broken it in general.
>
> This is probably a disappointing reply, but I don't have time to work
> on Rudel anymore.

While that might be somewhat disappointing, rudel does work well for me,
and if the changes to get it to work in Emacs 25 aren’t too extensive,
I’d like to try my hand at them. In my opinions, rudel is one of the
severly under-reported parts of Emacs.

Best wishes,
Arne
-- 
Unpolitisch sein
heißt politisch sein
ohne es zu merken

Re: [Rudel-devel] rude + el-get + emacs 25

[Jan M. <sc...@gm...>]

Hi.

On Fri, 2016-07-01 at 19:40 +0200, Christian Köstlin wrote:

> i try to get this running. but i run into several problems.
> is this one of the supported installation modes?

Rudel has been dormant for a couple of years now. I wouldn't generally
expect it to work in modern Emacs (on any platform).

> is rodel even running on OS X?

It think it has been in the past (never be me, though). But as said,
changes in Emacs may have broken it in general.

This is probably a disappointing reply, but I don't have time to work
on Rudel anymore.

Kind regards,
Jan

[Rudel-devel] rude + el-get + emacs 25

[C. K. <chr...@es...>]

hi,

i try to get this running. but i run into several problems.
is this one of the supported installation modes?
is rodel even running on OS X?

best regards,
christian

Re: [Rudel-devel] "Could not join session"

[Jan M. <sc...@gm...>]

Hi.

> > To fix the problem, make sure that the start-tls transport backend is
> > available (Can be verified via M-x rudel-backend-dump RET). Then do
> > everything like you did before but select the start-tls transport
> > backend when asked for a transport backend.
> 
> Perfect! With Rudel working, I no longer have to use Google Docs to
> collaborate on documents (not a good solution if one is using LaTeX). 
> Many, many thanks.

Nice, have fun.

Btw., I committed a change that should make the problem you encountered
easier to spot.

> > Sorry for the getting the user interface so wrong. Has not been my
> > focus.
> 
> Some day, I hope to pitch in with my 0.00002 <currency of your choice>. 
> (I think I spotted a couple of glitches in the menu. Will look more
> carefully and report.)

Thanks.

Kind regards,
Jan

Re: [Rudel-devel] "Could not join session"

[Mandar M. <man...@gm...>]

> To fix the problem, make sure that the start-tls transport backend is
> available (Can be verified via M-x rudel-backend-dump RET). Then do
> everything like you did before but select the start-tls transport
> backend when asked for a transport backend.

Perfect! With Rudel working, I no longer have to use Google Docs to
collaborate on documents (not a good solution if one is using LaTeX). 
Many, many thanks.


> Sorry for the getting the user interface so wrong. Has not been my
> focus.

Some day, I hope to pitch in with my 0.00002 <currency of your choice>. 
(I think I spotted a couple of glitches in the menu. Will look more
carefully and report.)

Thanks again,
Mandar.

Re: [Rudel-devel] "Could not join session"

[Jan M. <sc...@gm...>]

Hi.

> Thanks for your feedback.
> 
> I obtained the bzr version today, and tried this again, with a similar
> result.

Thanks for trying all the stuff. Actually, I could have figured the
problem out from your first message. Sorry.

I overlooked the line in your transcript that says:

transport backend: tcp

Using the tcp transport does not allow upgrading the connection to TLS
encryption later. That's why the gobby server terminates the connection
when the Rudel client fails to respond to the TLS handshake properly.

To fix the problem, make sure that the start-tls transport backend is
available (Can be verified via M-x rudel-backend-dump RET). Then do
everything like you did before but select the start-tls transport
backend when asked for a transport backend.

Sorry for the getting the user interface so wrong. Has not been my
focus.

Good luck,
Jan

Re: [Rudel-devel] "Could not join session"

[Mandar M. <man...@gm...>]

Hi,

Thanks for your feedback.

I obtained the bzr version today, and tried this again, with a similar
result.

> No obvious errors so far. I suppose, you chose to use encryption as the
> prompt indicated?

Yes, I did.


> I don't have any experience with sobby, but if it works like a gobby
> server, there should not be any problems due to that.

Right. To verify this, I ran the gobby client in Ubuntu, and was able to
successfully connect.

 
> Also, encrypted connections (as required for communication with gobby
> and probably sobby) currently rely on the external gnutls-cli program.
> The program can be installed via the gnutls-bin Debian package.

I have this too (/usr/bin/gnutls-cli).

I set debug-on-error to t and reran, and got the appended message. I'd be
much obliged if you could take a look at your convenience to see if this
tells you what I'm doing wrong.

Thanks again,
Mandar.

---

Debugger entered--Lisp error: (rudel-join-error)
  signal(rudel-join-error nil)
  byte-code("A\211\203

Re: [Rudel-devel] "Could not join session"

[Jan M. <sc...@gm...>]

Hi.

> I couldn't find a mailing list for users, so I'm posting my query here. 
> Sorry if this is inappropriate.

It's OK. For bug reports there is also the issue tracker at
http://sourceforge.net/tracker/?group_id=249139

> I'm running sobby on my machine. When I run Gobby (version 0.4), it asks
> for the host (localhost), port (6522), and a username. I can then connect
> and start working.
> 
> When I run rudel-join-session, it asks for:
> 
> protocol backend: obby
> transport backend: tcp
> server: localhost
> port: 6522
> encryption: y
> username: some name
> color: some colour
> global password: I leave this blank. What should I provide here?
> user password: I leave this blank. What should I provide here?

Looks okay - Password are not supported currently, but using empty
strings should be fine.

> In the *Messages* buffer, I get the following:
> 
> Discovering Sessions ...
> Initializing Zeroconf ...
> Discovering Sessions ...
> Use encryption (Required by Gobby server, not supported by Rudel server)? (y or n) 
> Received Obby welcome message (version 8)
> Joining done [2 times]
> if: Could not join session
> 
> Could you please give me some hint about what I might be doing wrong?

No obvious errors so far. I suppose, you chose to use encryption as the
prompt indicated?

> Versions, etc.:
> 
> 1. sobby 0.4.5-1ubuntu2
> 2. GNU Emacs 23.2.1 (i486-pc-linux-gnu, GTK+ Version 2.20.1) of 2010-09-02
>    on rutherfordium, modified by Debian
> 3. rudel from subversion (checked out on March 7)

I don't have any experience with sobby, but if it works like a gobby
server, there should not be any problems due to that.

The svn version of Rudel is rather outdated. Maybe you could try the bzr
version? Sorry.

Also, encrypted connections (as required for communication with gobby
and probably sobby) currently rely on the external gnutls-cli program.
The program can be installed via the gnutls-bin Debian package.

Good luck,
Jan

[Rudel-devel] "Could not join session"

[Mandar M. <man...@gm...>]

Hello,

I couldn't find a mailing list for users, so I'm posting my query here. 
Sorry if this is inappropriate.

I'm running sobby on my machine. When I run Gobby (version 0.4), it asks
for the host (localhost), port (6522), and a username. I can then connect
and start working.

When I run rudel-join-session, it asks for:

protocol backend: obby
transport backend: tcp
server: localhost
port: 6522
encryption: y
username: some name
color: some colour
global password: I leave this blank. What should I provide here?
user password: I leave this blank. What should I provide here?

In the *Messages* buffer, I get the following:

Discovering Sessions ...
Initializing Zeroconf ...
Discovering Sessions ...
Use encryption (Required by Gobby server, not supported by Rudel server)? (y or n) 
Received Obby welcome message (version 8)
Joining done [2 times]
if: Could not join session

Could you please give me some hint about what I might be doing wrong?


Versions, etc.:

1. sobby 0.4.5-1ubuntu2
2. GNU Emacs 23.2.1 (i486-pc-linux-gnu, GTK+ Version 2.20.1) of 2010-09-02
   on rutherfordium, modified by Debian
3. rudel from subversion (checked out on March 7)


Mandar.

ps: Thanks for what promises to be an extremely useful Emacs addon.

Re: [Rudel-devel] Rudel without dbus

[Jan M. <sc...@gm...>]

Hi.

> Is there a way to use rudel without dbus? I don't have dbus support
> compiled in my emacs and do not need to use Avahi however starting or
> join a session in rudel fails when it cannot find dbus functions which
> it expects to exist.

Could you try changing the lines at the end of
zeroconf/rudel-zeroconf.el from

;;;###autoload
(rudel-add-backend (rudel-backend-get-factory 'session-initiation)
		   'zeroconf 'rudel-zeroconf-backend)

to

;;;###autoload
(when (featurep 'dbusbind)
  (rudel-add-backend (rudel-backend-get-factory 'session-initiation)
		     'zeroconf 'rudel-zeroconf-backend))

and regenerate the autoloads in rudel-loaddefs.el. This should prevent
the zeroconf backend from being loaded. Hopefully, this in turn prevents
the problem you observed.

Kind regards,
Jan

[Rudel-devel] Rudel without dbus

[dclist <dc...@gm...>]

Is there a way to use rudel without dbus? I don't have dbus support
compiled in my emacs and do not need to use Avahi however starting or
join a session in rudel fails when it cannot find dbus functions which
it expects to exist.

Re: [Rudel-devel] Details on EtherPad communication

[Joe C. <hol...@gm...>]

>> I've read your developer documentation on the subject, as well as tried
>> as best I can to poke through your source code, but my team is still
>> stuck on the issue of simple establishing a connection with an existing
>> EtherPad document and posting/polling from it for changes.


That's exactly where we got stuck too :)  But we seem to be missing
only a small bit of info.  I was thinking over the holiday break to
return to this.

Re: [Rudel-devel] Details on EtherPad communication

[Jan M. <sc...@gm...>]

Hi Jimmy.

> I'm involved in a project trying to produce a JavaScript library to
> allow WYSIWYG-based collaborative editing using EtherPad as a backend
> but not the EtherPad web interface.
> 
> I've read your developer documentation on the subject, as well as tried
> as best I can to poke through your source code, but my team is still
> stuck on the issue of simple establishing a connection with an existing
> EtherPad document and posting/polling from it for changes.

First of all: I'm not sure whether the Rudel code is the ideal starting
point for understanding EtherPad's client-server communication. The
EtherPad backend in Rudel is incomplete and also not as tidy or well
documented as we would like it to be. Also, the code hasn't been touched
for some months due to lack of time. However, if your are looking for a
basic example of how to establish a connection and download the initial
document state, there could be some useful things in there.

> Any help you can provide or resources you can point me too would be much
> appreciated.

My suggestion would would be to directly approach the EtherPad
community.

@Joe: I though, maybe you could help or would know who to ask.

Sorry for not being terribly helpful.

Kind regards,
Jan

[Rudel-devel] Details on EtherPad communication

[Jimmy T. <ji...@je...>]

Hello,

I'm involved in a project trying to produce a JavaScript library to
allow WYSIWYG-based collaborative editing using EtherPad as a backend
but not the EtherPad web interface.

I've read your developer documentation on the subject, as well as tried
as best I can to poke through your source code, but my team is still
stuck on the issue of simple establishing a connection with an existing
EtherPad document and posting/polling from it for changes.

Any help you can provide or resources you can point me too would be much
appreciated.

Cheers
-- 
Jimmy Theis

Re: [Rudel-devel] Hi and chat

[Jan M. <sc...@gm...>]

On Thu, 2010-02-11 at 16:17 -0500, Jean Schurger wrote: 
> Hi,
> 
>   I've just discovered rudel, it's a pleasure to be able to use obby,
> and still have all the power of emacs features like org-mode,...

That was the reason for writing it.

> I have a question: is there a way to send 'chat' messages ?

You can do this (assuming you use the obby backend) using the following
function:

(defun rudel-chat-send-message (text)
  (interactive 
   (list (read-string "Text: ")))
  (let ((connection (oref rudel-current-session :connection)))
    (rudel-send connection "obby_message" text)))

Disclaimer: I just hacked it together and only tested it in the trunk
version. I think it should work in all recent versions, though.

You could either bind this function to a key or build a proper user
interface around it. For example, one could call this function when RET
is pressed in the rudel-chat-log buffer. Basically like this:

(defun rudel-chat-send-current-line ()
  (interactive)
  (let ((end   (point))
	(start (save-excursion
		 (beginning-of-line)
		 (point))))
    (rudel-chat-send-message (buffer-substring start end))
    (delete-region start end)))

(defadvice rudel-chat-handle-buffer (around fix-insert (sender text)
activate)
  "Fix inserting SENDER and MESSAGE into chat log buffer."
  (let ((buffer (or (get-buffer rudel-chat-buffer-name)
		    (pop-to-buffer rudel-chat-buffer-name))))
    (with-current-buffer buffer
      (save-excursion
      (goto-char (point-max))
      (insert (format "\n%s: %s"
		      (rudel-chat-format-sender sender)
		      text))))))

;; eval in *rudel-chat-log-buffer*:
(local-set-key (kbd "RET") #'rudel-chat-send-current-line)

Good luck.

Kind regards,
Jan

[Rudel-devel] Hi and chat

[Jean S. <je...@sc...>]

Hi,

  I've just discovered rudel, it's a pleasure to be able to use obby,
and still have all the power of emacs features like org-mode,...

I have a question: is there a way to send 'chat' messages ?

Jean.

-- 
Jean Schurger
http://schurger.org
GPG: http://schurger.org/jean.asc

Re: [Rudel-devel] Rudel on OS X

[Tim H. <tim...@gm...>]

I suppose the rest of this conversation occurred privately (I hit
reply-to instead of reply-to-all, and away it went).

Jan has released a fix in 0.2-3 using an approach I have tested.

As you were.

Tim

Re: [Rudel-devel] Rudel on OS X

[Jan M. <sc...@gm...>]

On Wed, 2009-11-11 at 17:22 -0700, Tim Harper wrote:
> On Wed, Nov 11, 2009 at 5:16 PM, Jan Moringen <sc...@gm...> wrote:
> > Hi Tim.
> >
> >> I downloaded and installed rudel today on OS X.  Installation went
> >> pretty well, but I got an error saying "dbus-call-method" wasn't
> >> defined.  Anyways.. it boiled down to I had to install dbus (I could
> >> install it with no_x11 and that made it not pull down and compile the
> >> whole x11 universe), then I rebuilt Emacs 23.1 from source (making
> >> sure that during ./configure it picked up dbus), and it all worked.
> >
> > Other people had the same problem and couldn't get Emacs to work with
> > dbus. Interesting that it worked for you.
> >
> > Did you use the 0.2-2 release or check out from one of the repositories?
> > I thought the problem was fixed, in a rather hackish way, in the 0.2-2
> > release (and the 0.2 branch for that matter). However, if you checked
> > out trunk, you possibly did not get the fix, since we work on a long
> > term solution there.
> 
> I was running on the 0.2-2 release downloaded from sourceforge.  I
> tested it with a co-worker on OS-X, it worked fantastically 

In case you still have the dbus and non-dbus versions of Emacs 23
around, could you try evaluating (place point inside the fragment and
hit the OS X equivalent of C-M-x) the following piece of code (the
result will be displayed in the echo area):

(and (require 'dbus nil t)
     (require 'zeroconf nil t)
     (dbus-get-name-owner :system "org.freedesktop.Avahi"))

It was written to prevent loading of the zeroconf stuff on OS X
(ironically).

> (the
> default colors were bad though (white on light blue), and it didn't
> seem to respect the color we specified, but I'm sure that'll take some
> tweaking).

It would be very helpful, if you could open an issue in the issue
tracker on the sourceforge page and provide a description of how to
reproduce this specific problem.

You should be able to change "your" color after joining a session at any
time using C-c c c or the menu entry.

> >> If you'd like me to send a patch for the INSTALL file, I'd be happy to.
> >
> > That would be nice, since I don't have OS X available for testing.
> 
> I talked to technomancy on #emacs, and he was surprised it wasn't
> fixed.  He said it could be fixed by commenting out a few lines.  I
> don't know if it would be better to just fix it, or if it would be
> better to modify the INSTALL file with the work-around instructions.
> What do you think?

The best solution would be to make the code mentioned above work
properly. If that cannot be done, we should add a description of the
workaround to the INSTALL file.

Thanks again for the feedback.

Kind regards,
Jan

Re: [Rudel-devel] Rudel on OS X

[Jan M. <sc...@gm...>]

Hi Tim.

> I downloaded and installed rudel today on OS X.  Installation went
> pretty well, but I got an error saying "dbus-call-method" wasn't
> defined.  Anyways.. it boiled down to I had to install dbus (I could
> install it with no_x11 and that made it not pull down and compile the
> whole x11 universe), then I rebuilt Emacs 23.1 from source (making
> sure that during ./configure it picked up dbus), and it all worked.

Other people had the same problem and couldn't get Emacs to work with
dbus. Interesting that it worked for you.

Did you use the 0.2-2 release or check out from one of the repositories?
I thought the problem was fixed, in a rather hackish way, in the 0.2-2
release (and the 0.2 branch for that matter). However, if you checked
out trunk, you possibly did not get the fix, since we work on a long
term solution there.

> If you'd like me to send a patch for the INSTALL file, I'd be happy to.

That would be nice, since I don't have OS X available for testing.

Thanks for the feedback.

Kind regards,
Jan

[Rudel-devel] Rudel on OS X

[Tim H. <tim...@gm...>]

Hi all,

I downloaded and installed rudel today on OS X.  Installation went
pretty well, but I got an error saying "dbus-call-method" wasn't
defined.  Anyways.. it boiled down to I had to install dbus (I could
install it with no_x11 and that made it not pull down and compile the
whole x11 universe), then I rebuilt Emacs 23.1 from source (making
sure that during ./configure it picked up dbus), and it all worked.

If you'd like me to send a patch for the INSTALL file, I'd be happy to.

Thanks,

Tim