Re: TRAMP Debian package...
On Mon, Mar 05, 2001 at 11:23:46AM +1100, Daniel Pittman wrote: > I notice that there is now a TRAMP Debian package available, packaged > and submitted by Ola Lundqvist. > > I didn't notice an announcement on the TRAMP mailing list about the > availability of this package, so it came as a bit of a surprise to me > when I scanned the new packages list. ;) Well. I thought it was a nice program, so I tested it, packaged it and said to myself, why not upload it too. Checked wnpp (nothing there) so instead of filing a bug and close it right away I uploaded it... Maybe it was wrong, nut that is what I did... :) > Anyway, I was wondering, Ola, if you had a particular plan for managing > the bug reports against the upstream sources? Well I'll forward them as I use to. :) One thing already, it does not work on the gtk version of xemacs, but I'll get back to you later about that. I do not have time right now... > I believe the correct plan here is to forward bug reports to the mailing > list as well as to yourself, but Kai and others here may have their own > point-of-view on the matter. Ahh I was not aware of that mailinglist :) If I was I should have announced it, of course. > Also, it would be nice to know when a new release of TRAMP was being > made available. Yes that would be nice. > Daniel > > -- > If the doors of perception were cleansed every thing > would appear to man as it is, Infinite. > -- William Blake, _The Marriage of Heaven and Hell_ -- - Ola Lundqvist --- / [EMAIL PROTECTED] Björnkärrsgatan 5 A.11 \ | [EMAIL PROTECTED] 584 36 LINKÖPING | | +46 (0)13-17 69 83 +46 (0)70-332 1551 | | http://www.opal.dhs.org UIN/icq: 4912500 | \ gpg/f.p.: 7090 A92B 18FE 7994 0C36 4FE4 18A1 B1CF 0FE5 3DD9 / ---
check in by daniel: 'tramp/lisp tramp2-enc.el,2.1 tramp2-ops.el,2.3 tramp2.el,2.5'
Update of /services/emacs-rcp/cvsroot/tramp/lisp In directory bonny:/tmp/cvs-serv21527/lisp Modified Files: tramp2-ops.el tramp2.el Added Files: tramp2-enc.el Log Message: Protocol definitions now work. `ssh' and `shell' are now defined and used for testing. Remote detection for base64 encoder/decoder implemented along with framework for such. uuencode stubs and base64 transfer stubs exist but not implemented. TRAMP2 now operates on the remote machine successfully. Basic infrastructure for connection and encoding is in place. Only actual data transfer remains before the infrastructure is complete and the file operation handlers can be ported.
transfer-mode optimization, was: loads but won't save?
On 04 Mar 2001, Kai Großjohann wrote: >> And this reminds me of the todo item which was sent to me by >> Michael Kifer: use out of band transfer for large files and inline >> transfer for small files. Daniel Pittman <[EMAIL PROTECTED]> 05 Mar 2001 10:13:16 +1100 > Worth thinking about... which I will do. I don't know how easy or > practical it will be, though. Is this really going to be /that/ much > more efficient? I suppose it would be quite a bit more efficient ... once you partitioned "small" and "large" :-) For each pair (client, server) ?-) There might be a useful default value, but ... allow me to make an alternate suggestion. To determine that value would require use of a suitable profiler, which would need to be portable enough to run on all the various client platforms on which you'd need to test to get an empirically-reliable result. Instead of doing the (large amount of) required testing, why not just release the profiler, and let folks determine the appropriate value for their local conditions? FWIW, [EMAIL PROTECTED]
password handling, was: Bemühen erfolgreich!
[EMAIL PROTECTED] 04 Mar 2001 22:47:46 +0100 > Did you say you don't like having to put your passwd into the > method? Try omitting it, maybe Tramp auto-groks your passwd prompt. I already tried that. Remember, currently 'plink' won't even take a password itself in a *shell* buffer. That's the problem plink-ssh solved. (Perhaps you should file under "workarounds for defective clients" :-) > But if it doesn't work, then it's probably not worth doing too much > about it, since Daniel has already progressed quite nicely with the > new Tramp. OK, I will look forward to that. If I get some time I may try hacking a wrapper to tramp-connection-function "as an exercise." On Sun, 04 Mar 2001, Tom Roche wrote: >> * While I am delighted to finally have a working tramp, I must add: >> I still believe that tramp should provide an "ange-ftp >> compatibility mode," i.e. use scp, prompting for password. >> If this had been available, I could have been using tramp six >> months ago. [EMAIL PROTECTED] 04 Mar 2001 23:04:07 +0100 > Is it sufficient that Tramp asks you for a password, or does it have > to be scp? Either is fine, as long as scp (or, in my case, pscp) gets happy. TIA, [EMAIL PROTECTED]
TRAMP Debian package...
I notice that there is now a TRAMP Debian package available, packaged and submitted by Ola Lundqvist. I didn't notice an announcement on the TRAMP mailing list about the availability of this package, so it came as a bit of a surprise to me when I scanned the new packages list. ;) Anyway, I was wondering, Ola, if you had a particular plan for managing the bug reports against the upstream sources? I believe the correct plan here is to forward bug reports to the mailing list as well as to yourself, but Kai and others here may have their own point-of-view on the matter. Also, it would be nice to know when a new release of TRAMP was being made available. Daniel -- If the doors of perception were cleansed every thing would appear to man as it is, Infinite. -- William Blake, _The Marriage of Heaven and Hell_
Re: loads but won't save? was: plink and tramp-2.0.2.1: more progress
On 04 Mar 2001, Kai Großjohann wrote: > On 04 Mar 2001, Daniel Pittman wrote: > >> By default, I want this to have as much automatic configuration as >> possible. So, my plan is that, by default, the method for encoding >> and decoding files on the remote machine will be auto-detected. > > That's a good idea. But I think you might not be able to > automatically detect whether you should do inline or out of band > transfer. Hrm. Good point. I will have to think a little more on that... > Well, with multi-hop methods you automatically do inline. Which makes the auto-detection a lot more useful, I guess, especially since one of the main points of the new design is to make connecting multiple hops easy - so that ssh and su connections work. :) > And this reminds me of the todo item which was sent to me by Michael > Kifer: use out of band transfer for large files and inline transfer > for small files. Worth thinking about... which I will do. I don't know how easy or practical it will be, though. Is this really going to be /that/ much more efficient? Daniel -- I hold on to the blissful notion that the journey is worthwhile, and I tip-toe across the chasm with a song and a smile. -- Harry Chapin
Re: tramp questions and suggestions
On Sun, 04 Mar 2001, Tom Roche wrote: > * Time to update tramp at the the Ohio State Emacs Lisp Archive? Must do that. > * What, if any, documentation of my travails should be done? Perhaps > something like a HOWTO? Yes, please. There is already a FAQ section. Just submit your entry, and it will be added. > * How to whack the tempfile problem? as previously mentioned: > > *Backtrace* > > Signaling: > > (error "Not a tramp file name: > > d:/temp/_ar@[EMAIL PROTECTED]") Oops. Whee. I completely missed that one. Can you submit a bug report with the M-x tramp-bug RET command? Then I can see more of your setup. > * While I am delighted to finally have a working tramp, I must add: > I still believe that tramp should provide an "ange-ftp > compatibility mode," i.e. use scp, prompting for password. If this > had been available, I could have been using tramp six months ago. Is it sufficient that Tramp asks you for a password, or does it have to be scp? > * Speaking of passwords: how might I hack tramp-connection-function > (or whatever more appropriate fn) to get a password and pass it to > plink? À la plink-ssh: > >> (let ((name (concat "plink-" host)) >> (password (comint-read-noecho "password: " t))) >> (pop-to-buffer >> (make-comint name >> plink-program nil ssh-string host "-pw" password)) > > Are there connection-function-hooks? The passwd stuff should work automatically. If I remember your passwd prompt correctly, it should be part of the normal prompt pattern. We will see... kai -- Be indiscrete. Do it continuously.
Re: Bemühen erfolgreich!
I'm happy that it finally works for you. Did you say you don't like having to put your passwd into the method? Try omitting it, maybe Tramp auto-groks your passwd prompt. But if it doesn't work, then it's probably not worth doing too much about it, since Daniel has already progressed quite nicely with the new Tramp. kai -- Be indiscrete. Do it continuously.
tramp questions and suggestions
* Time to update tramp at the the Ohio State Emacs Lisp Archive? I note > http://neutral.verbum.org/search?q=tramp&archive=incoming > Search Request: tramp > Matches 0-1 in the Incoming archive: > tramp-1.393.tar.gz which is a link. * What, if any, documentation of my travails should be done? Perhaps something like a HOWTO? Perhaps tramp could have a documentary facility allowing folks to add their accounts of getting tramp setup on whatever platform, with whatever toolset they were able to use. Thus mine would be (how to setup tramp on) "GNU 20.7 on win2k with plink": others might do (e.g.) "XEmacs with cygwin on NT" or "GNU 21 on NetBSD with OpenSSH" or "XEmacs on RedHat 7 via RPMs" or whatever. Dunno how to implement ... something like a FAQ-o-matic? Is there an 'info-matic'? * How to whack the tempfile problem? as previously mentioned: *Backtrace* > Signaling: > (error "Not a tramp file name: > d:/temp/_ar@[EMAIL PROTECTED]") * While I am delighted to finally have a working tramp, I must add: I still believe that tramp should provide an "ange-ftp compatibility mode," i.e. use scp, prompting for password. If this had been available, I could have been using tramp six months ago. * Speaking of passwords: how might I hack tramp-connection-function (or whatever more appropriate fn) to get a password and pass it to plink? À la plink-ssh: > (let ((name (concat "plink-" host)) > (password (comint-read-noecho "password: " t))) > (pop-to-buffer > (make-comint name > plink-program nil ssh-string host "-pw" password)) Are there connection-function-hooks? FWIW, [EMAIL PROTECTED]
Bemühen erfolgreich!
[EMAIL PROTECTED] 04 Mar 2001 12:34:20 +0100 > Tramp is quite anal about having all kinds of method parameters. > Don't let any of them default to anything. Specify them all. So > just add (tramp-rcp-args nil) into the method spec, and you should > be all set. Look at the other methods so you're sure you're not > missing anything that should be set to nil. I couldn't see anything else: resulting code copied to http://www4.ncsu.edu/~tlroche/tramp/_emacs.txt Started a new emacs, and C-x C-f /r@t:[EMAIL PROTECTED]:.emacs ... unfortunately tramp died, before displaying the .emacs buffer, with *Backtrace* > Signaling: > (error "Not a tramp file name: > d:/temp/_ar@[EMAIL PROTECTED]") (Full contents of *Backtrace* and *debug ...* copied to http://www4.ncsu.edu/~tlroche/tramp/backtrace_buffer.txt http://www4.ncsu.edu/~tlroche/tramp/debug_buffer.txt respectively.) Note that I have (setq tramp-auto-save-directory "d:/temp") and it has (at top, sorted time descending, broken for mail) > d:/temp: > total 24638 > drwxrwxrwx 0 Mar 4 14:47 . > -rw-rw-rw- 3618 Mar 3 23:21 air-sched.htm > -rw-rw-rw- 0 Mar 3 23:02 cdd18.tmp > -rw-rw-rw- 0 Mar 3 22:54 cdd16.tmp > -rw-rw-rw- 17613 Mar 3 21:53 > why-cooperation-with-rms-is-impossible.mp3 > drwxrwxrwx 0 Mar 3 21:43 soffice.tmp > -rw-rw-rw- 14937 Mar 3 20:15 > _ar@[EMAIL PROTECTED] left over from last night's trial. I whacked that file, started a new emacs, and C-x C-f /r@t:[EMAIL PROTECTED]:.emacs ... got the .emacs buffer, made some changes, C-x C-s > Wrote /r@t:[EMAIL PROTECTED]:/ncsu/tlroche/.emacs Success, at last! (I plink'ed in and less-ed the file, just to be sure :-) Thanks to everyone for their assistance! I have some questions and suggestions, but I'll do them separately. Vielen dank, [EMAIL PROTECTED]
Re: loads but won't save? was: plink and tramp-2.0.2.1: more progress
On 04 Mar 2001, Daniel Pittman wrote: > By default, I want this to have as much automatic configuration as > possible. So, my plan is that, by default, the method for encoding > and decoding files on the remote machine will be auto-detected. That's a good idea. But I think you might not be able to automatically detect whether you should do inline or out of band transfer. Well, with multi-hop methods you automatically do inline. And this reminds me of the todo item which was sent to me by Michael Kifer: use out of band transfer for large files and inline transfer for small files. kai -- Be indiscrete. Do it continuously.
Re: ange-ftp and passive mode
Harry Putnam <[EMAIL PROTECTED]> writes: > host user (list `passive) "Toggle passive mdoe"))) You are right :-) `passive' command toggles passive mode on NetBSD too. > And removed the colon at `|^Passive mode")' > > I am guessing the `setq' is there to recognize the feed back from the > ftp client .. That is, if passing the toggle command in the first part > produces a message from the ftp client like: > > ftp> passive > Passive mode off. > ftp> > > Then the rexexp needs to recognize the `Passive mode off' output. > > Or is that not how it works? Good change. It will make the code works with other ftp clients. I put colon to the regexp because NetBSD client says as bellow, but I guess it is NetBSD specific. ftp> passive Passive mode: off; fallback to active mode: off. ftp> passive Passive mode: on; fallback to active mode: on. -- Yuji Yamano
Re: loads but won't save? was: plink and tramp-2.0.2.1: more progress
On 04 Mar 2001, Kai Großjohann wrote: [...] > Tramp is quite anal about having all kinds of method parameters. > Don't let any of them default to anything. Specify them all. So just > add (tramp-rcp-args nil) into the method spec, and you should be all > set. Look at the other methods so you're sure you're not missing > anything that should be set to nil. > > I agree that this is not nice, but OTOH you don't add a new method > every day, so maybe it's not too bad. On the topic of connection methods and their definitions, I want to see what people think of one of my plans for the tramp2 code. By default, I want this to have as much automatic configuration as possible. So, my plan is that, by default, the method for encoding and decoding files on the remote machine will be auto-detected. It will be possible to override this in connection definitions but, by default, all the detection will be automatic. This, of course, presumes that I don't find any real problems getting reliable detection of encoding methods remotely working. Daniel -- You call this paradise |Just take my hand girl You call these mountains|Just take my hand girl We call this poverty|I'll take your tears girl We call these hills |Step aside -- Jams Ray and The Performance, _Mountain Voices_
Re: loads but won't save? was: plink and tramp-2.0.2.1: more progress
On Sat, 03 Mar 2001, Tom Roche wrote: > When I try to _save_ the buffer, I get the error > >> Method `t' didn't specify tramp args > > which appears to be caused by tramp's desire to call > tramp-rcp-program in order to save the backup file > > /r@t:[EMAIL PROTECTED]:/ncsu/tlroche/.emacs~ > > Any ideas how to fix this? FWIW I tried setting He he, that's a fun typo. It should have been ``didn't specify rcp args''. When I renamed rcp.el into Tramp, I made one substitution too many... Tramp is quite anal about having all kinds of method parameters. Don't let any of them default to anything. Specify them all. So just add (tramp-rcp-args nil) into the method spec, and you should be all set. Look at the other methods so you're sure you're not missing anything that should be set to nil. I agree that this is not nice, but OTOH you don't add a new method every day, so maybe it's not too bad. kai -- Be indiscrete. Do it continuously.
check in by daniel: 'tramp/lisp tramp2.el,2.4'
Update of /services/emacs-rcp/cvsroot/tramp/lisp In directory bonny:/tmp/cvs-serv16206/lisp Modified Files: tramp2.el Log Message: TRAMP2 actually works, end-to-end, on a local connection. The handler does the right thing and all. If you get real enthusiastic, test it with: M-: (tramp2-handler-for-noop "/!:[/bin/sh -i]localhost::test.file")
check in by grossjoh: 'tramp/lisp ChangeLog,2.6 tramp.el,2.7'
Update of /services/emacs-rcp/cvsroot/tramp/lisp In directory lucy:/export/home/grossjoh/work/kai/tramp-devel/tramp/lisp Modified Files: ChangeLog tramp.el Log Message: * tramp.el (tramp-handle-file-directory-p): Use `test -d' rather than `cd' to check if a file is a directory. Suggested by Stefan Monnier.
