6:10 p.m.
What was really meant by stating
description "Emacs MIME support. Not needed for gnus >= 5.8.0"
for tm in `package-index.LATEST.pgp'? It is unlikely that gnus
itself, without using tm, is able to create full featured (multipart,
with specified encodings and so on) MIME messages. It looks to me
that currently gnus can only create single-part text/plain messages
and create encoded words in message headers for some fixed set of
charsets.
6:28 p.m.
>>>> "Golubev" == Golubev I N
<gin(a)mo.msk.ru> writes:
Golubev> What was really meant by stating
Golubev> description "Emacs MIME support. Not needed for gnus
Golubev> >= 5.8.0"
Golubev> for tm in `package-index.LATEST.pgp'? It is unlikely
Golubev> that gnus itself, without using tm, is able to create
Golubev> full featured (multipart, with specified encodings and so
Golubev> on) MIME messages. It looks to me that currently gnus
I am using Gnus without TM.
gnus-version "Gnus v5.8.8"
It works fine for me.
Golubev> can only create single-part text/plain messages and
I have created attachments with content-type application/octet-stream
with it.
I haven't missed anything since I started using Gnus >= 5.8.0 without
TM.
If you let me know what feature you're looking for I could try it out
for you in my installation.
Best regards,
Adrian
Golubev> create encoded words in message headers for some fixed
Golubev> set of charsets.
--
Adrian Aichner
mailto:adrian＠xemacs.org
http://www.xemacs.org/
6:51 p.m.
"Golubev I. N." <gin(a)mo.msk.ru> writes:
What was really meant by stating
description "Emacs MIME support. Not needed for gnus >= 5.8.0"
for tm in `package-index.LATEST.pgp'?
As of 5.8, Gnus gained full native MIME comprehension. TM conflicts
with Gnus 5.8 and later.
It is unlikely that gnus itself, without using tm, is able to create
full featured (multipart, with specified encodings and so on) MIME
messages. It looks to me that currently gnus can only create
single-part text/plain messages and create encoded words in message
headers for some fixed set of charsets.
No.
Gnus can create and interpret pretty much any MIME packaging you wish.
It uses an internal system known as "mml" ("MIME markup language", I
believe), providing a reasonable way of introducing additional and
alternative parts, including multiparts.
During composition in a *Messge* buffer, see the Attachments menubar
pulldown, and read the Emacs-MIME info area (*Info* -> Emacs-MIME or,
in very recent alpha Gnus ["Oort"], info://Emacs-MIME/ [that quasi-URL
can be clicked directly in recent Oort Gnus]). Note also the keymap
under `C-c C-m'.
11:33 p.m.
>>>> "Golubev" == Golubev I N
<gin(a)mo.msk.ru> writes:
Golubev> What was really meant by stating
Golubev> description "Emacs MIME support. Not needed for gnus >=
5.8.0"
"tm is bad news, bad mail, bad MIME, bad Lisp, deprecated, please
avoid it." Restated politely. It's actually pretty stable and usable
now, but it does insinuate itself into Emacs in an irreversible way
and is definitely orphaned by its author. It is bug-prone when the
Emacs environment changes.
Golubev> It is unlikely that gnus itself, without using tm, is
Golubev> able to create full featured (multipart, with specified
Golubev> encodings and so on) MIME messages.
Sadly, this is false. None of the add-on MIME packages ever got out
of beta, really. Before SEMI, the author of tm and later SEMI
regularly sent out broken MIME using tm, and his colleagues at the
Mule lab didn't use it. He then abandoned tm to develop SEMI, with
the same historical pattern.
So most of the major Emacs MUAs (Gnus and VM, at least, I think Mew as
well) do their own MIME now, and have done so for years. Gnus 5.8.8
does better than VM 6.9x in many areas (for example, Gnus MIME-encodes
headers, VM doesn't), but both do all of the necessary decoding and
provide hooks for external viewers, and will encode bodies and create
attachments.
Gnus does at least attempt to distribute MML as a separate package
within a package, but I doubt that either Kyle or the Japanese MUA
authors will pick it up very soon.
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Don't ask how you can "do" free software business;
ask what your business can "do for" free software.
7:16 p.m.
That historical discourse about emacs MUAs is very interesting, but
how do I MIME-encode headers and create attachments in that same gnus?
Its own '.texi' does not say much about it. All I have is MML menus.
What if I am on tty? How do I control charset of encoded words in
headers?
It appears that current version of gnus lacks documentation badly.
7:42 p.m.
On Mon, 5 Nov 2001, Golubev I. N. wrote:
That historical discourse about emacs MUAs is very interesting, but
how do I MIME-encode headers and create attachments in that same gnus?
Headers are automatically encoded when necessary, attachments are made
with C-c C-a in the message buffer or via the menu.
Its own '.texi' does not say much about it. All I have is
MML menus.
Composing mail is described in the Message manual.
What if I am on tty?
Gnus doesn't notice the difference,
How do I control charset of encoded words in headers?
If you type a Latin-2 character, the latin-2 MIME type will be used. Etc,
you do not normally need to indicate to Gnus anything about the charsets,
but for details see the Emacs MIME manual.
It appears that current version of gnus lacks documentation badly.
Did you see the Message and Emacs-MIME manuals? I think most of what you
ask are covered by them.
4:05 p.m.
Your message is really helpful wrt to basic keybindings and how to
start. But I think that such a startup should be (easily) available
to everybody through conventional emacs techniques. One should not
have to post questions or read faq.
Composing mail is described in the Message manual.
All that end user `sees' is gnus. So gnus documentation should be a
point to start and figure out everything one needs. When it describes
`gnus-group-mail' and friends, it should refer to `message' manual.
Currently `gnus.info' does not mention `gnus-msg-mail' at all, and
functions' documentation does not state explicitly where to read about
mail composition.
> What if I am on tty?
Gnus doesn't notice the difference,
I am sure that gnus doesn't, but user does. With menubar it is easier
to understand what MML does. MML minor mode description (which is the
only thing universally available on any device) is quite terse and
does not state explicitly that it is stuff for composing MIME messages
described in this-and-this `.info'.
Etc,
you do not normally need to indicate to Gnus anything about the charsets,
but for details see the Emacs MIME manual.
Hope that it describes what to do without MULE or when same text may
be sent in different charsets, and required charset may change.
> It appears that current version of gnus lacks documentation
badly.
I think that the statement still holds. There are apparently not
enough cross-references, and learning it is not as easy as it could
be.
Even now I observe some surprises in message mode. It appears to use
`sendmail' stuff, and does that in a strange way. One has to change
default `sendmail-program' value, not buffer-local one, to have it
affect sending in `message' mode. So one cannot (easily) reuse his
`sendmail-program' stuff.
And there is no way back, that is, you cannot stick to `tm' either,
since it is also considerably hosed. So mail editing in xemacs
becomes quite uncomfortable.
5:04 p.m.
On Wed, 7 Nov 2001, Golubev I. N. wrote:
Your message is really helpful wrt to basic keybindings and how to
start. But I think that such a startup should be (easily) available
to everybody through conventional emacs techniques. One should not
have to post questions or read faq.
The key binding is mentioned if you do C-h m, it is also available in the
menubar.
In Oort Gnus the somewhat confusing menubar name "MML" was changed into
"Attachment" to make things easier to locate for users that doesn't know
what MML is. Please suggest other specific things that can be improved as
well, and I'm sure they will be added.
> Composing mail is described in the Message manual.
All that end user `sees' is gnus. So gnus documentation should be a
point to start and figure out everything one needs.
Right, but this seems to be covered -- in the top-level node "Composing
Messages" in the first paragraph, there is a reference to the message
manual.
When it describes `gnus-group-mail' and friends, it should refer
to
`message' manual.
Referring to the "Composing Messages" node is perhaps better, I'll add
this.
Currently `gnus.info' does not mention `gnus-msg-mail' at
all, and
functions' documentation does not state explicitly where to read about
mail composition.
gnus-msg-mail is a XEmacs thing, it should be documented in the XEmacs
manual. Gnus does not use the function as far as I know.
> > What if I am on tty?
>
> Gnus doesn't notice the difference,
I am sure that gnus doesn't, but user does. With menubar it is easier
to understand what MML does. MML minor mode description (which is the
only thing universally available on any device) is quite terse and
does not state explicitly that it is stuff for composing MIME messages
described in this-and-this `.info'.
Right, I'll submit a patch that adds a few sentences to the MML minor
mode with more information.
> Etc,
> you do not normally need to indicate to Gnus anything about the charsets,
> but for details see the Emacs MIME manual.
Hope that it describes what to do without MULE or when same text may
be sent in different charsets, and required charset may change.
Yes, see the "Charset Translation" node. (Located via `s mule RET'.)
> > It appears that current version of gnus lacks documentation
badly.
I think that the statement still holds. There are apparently not
enough cross-references, and learning it is not as easy as it could
be.
Agreed, please submit more places where you think cross references are
appropriate. Diff format prefered, but suggestions are useful as well.
Even now I observe some surprises in message mode. It appears to
use
`sendmail' stuff, and does that in a strange way. One has to change
default `sendmail-program' value, not buffer-local one, to have it
affect sending in `message' mode. So one cannot (easily) reuse his
`sendmail-program' stuff.
This is a bug that was fixed a while ago, but I don't think there has been
a new mail-lib release that includes it yet.
(A new mail-lib release would be a good thing, the mail-lib in CVS solves
alot of problems and add several new features.)
And there is no way back, that is, you cannot stick to `tm'
either,
since it is also considerably hosed. So mail editing in xemacs
becomes quite uncomfortable.
With yours and others help it will become more comfortable. Without the
help, it will stay uncomfortable.
8:46 a.m.
>>>> "Simon" == Simon Josefsson
<jas(a)extundo.com> writes:
> Currently `gnus.info' does not mention `gnus-msg-mail' at
all,
> and functions' documentation does not state explicitly where to
> read about mail composition.
Simon> gnus-msg-mail is a XEmacs thing, it should be documented in
Simon> the XEmacs manual. Gnus does not use the function as far
Simon> as I know.
This is becoming a FAQ. Anyway, are you sure? It comes from
gnus-msg, which is part of the Gnus distribution and authored by larsi
and Umeda-san. The docstring says
Start editing a mail message to be sent.
Like `message-mail', but with Gnus paraphernalia, particularly the
Gcc: header for archiving purposes.
Simon> (A new mail-lib release would be a good thing, the mail-lib
Simon> in CVS solves alot of problems and add several new
Simon> features.)
Sigh. I'll get on this.
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Don't ask how you can "do" free software business;
ask what your business can "do for" free software.
9:19 a.m.
On 8 Nov 2001, Stephen J. Turnbull wrote:
>>>>> "Simon" == Simon Josefsson
<jas(a)extundo.com> writes:
>> Currently `gnus.info' does not mention `gnus-msg-mail' at all,
>> and functions' documentation does not state explicitly where to
>> read about mail composition.
Simon> gnus-msg-mail is a XEmacs thing, it should be documented in
Simon> the XEmacs manual. Gnus does not use the function as far
Simon> as I know.
This is becoming a FAQ. Anyway, are you sure? It comes from
gnus-msg, which is part of the Gnus distribution and authored by larsi
and Umeda-san. The docstring says
Start editing a mail message to be sent.
Like `message-mail', but with Gnus paraphernalia, particularly the
Gcc: header for archiving purposes.
Yes, but I believe the function and the package is there for the
`mail-user-agent' stuff in (X)Emacs. So it is the interface glue that VM,
RMAIL etc also have, and the function should probably be documented where
`mail-user-agent' is described. (And hopefully it already is, but I didn't
check.)
Simon> (A new mail-lib release would be a good thing, the
mail-lib
Simon> in CVS solves alot of problems and add several new
Simon> features.)
Sigh. I'll get on this.
I'm having trouble committing things currently, and I'd like to commit a
one-liner I sent to xemacs-patches several weeks ago first -- I sent a
mail to the cvs maintainer so hopefully it will be fixed soon.
11:15 a.m.
>>>> "Simon" == Simon Josefsson
<jas(a)extundo.com> writes:
Simon> I'm having trouble committing things currently, and I'd
Simon> like to commit a one-liner I sent to xemacs-patches several
Simon> weeks ago first -- I sent a mail to the cvs maintainer so
Simon> hopefully it will be fixed soon.
Er, I'm the XEmacs CVS maintainer now.... OK, I've fixed that alias.
(gomen'nasai, Jareth-san. taihen gomeiwaku wo kaketeshimaimasita (^^;;;)
What's the problem? Are you aware that cvs.xemacs.org is no more?
We're planning to move to SunSITE.dk as soon as we can get some
security issues straightened out. In the meantime, CVS changes are
done using your "jas(a)barbar.josefsson.org" SSH public key, using
CVS_RSH=/usr/bin/ssh
CVSROOT=:ext:xemacs＠tanko.sk.tsukuba.ac.jp:/usr/CVSroot
or whatever is appropriate at your system. Checkouts and updates can
be done anonymously
CVSROOT=:pserver:cvs＠sunsite.dk:/pack/anoncvs
The cvs login password is "cvs". The canonical reference for this
stuff is http://www.xemacs.org/Develop/cvsaccess.html, but it doesn't
describe the temporary server at "tanko". Note that due to my
university's brain-damaged firewall, the only way you can check
service is via ssh -v, which will hang after announcing the connection
(you don't have shell access; it automatically starts the CVS server,
which is waiting for CVS :ext: client-server protocol). No ping,
traceroute, etc, etc, due to firewall.
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Don't ask how you can "do" free software business;
ask what your business can "do for" free software.
12:11 p.m.
On 8 Nov 2001, Stephen J. Turnbull wrote:
What's the problem? Are you aware that cvs.xemacs.org is no
more?
We're planning to move to SunSITE.dk as soon as we can get some
security issues straightened out. In the meantime, CVS changes are
done using your "jas(a)barbar.josefsson.org" SSH public key, using
CVS_RSH=/usr/bin/ssh
CVSROOT=:ext:xemacs＠tanko.sk.tsukuba.ac.jp:/usr/CVSroot
Ah, the cvsaccess.html page says to use "martinb(a)tanko.sk.tsukuba.ac.jp".
Ok, now ssh -v works, but commit doesn't:
$ xemacs-cvs-rw.sh commit ChangeLog smtpmail.el
cvs server: failed to create lock directory for
`/usr/CVSroot/xemacs-packages/xemacs-packages/mail-lib'
(/usr/CVSroot/xemacs-packages/xemacs-packages/mail-lib/#cvs.lock): No such
file or directory
cvs server: lock failed - giving up
cvs [server aborted]: lock failed - giving up
cvs commit: saving log message in /tmp/cvsKSWWkp
$ cat ~/bin/xemacs-cvs-rw.sh
#!/bin/sh
exec env CVS_RSH=ssh cvs -w -z3 -d
:ext:xemacs＠tanko.sk.tsukuba.ac.jp:/usr/CVSroot ${1+"$＠"}
$
The cvs login password is "cvs". The canonical reference
for this
stuff is http://www.xemacs.org/Develop/cvsaccess.html, but it doesn't
describe the temporary server at "tanko".
Tanko is mentioned in several places, but not all (search for
cvs.xemacs.org), same for "martinb" vs "xemacs".
Thanks..
1:28 p.m.
>>>> "Stephen" == Stephen J Turnbull
<stephen(a)xemacs.org> writes:
>>>> "Simon" == Simon Josefsson
<jas(a)extundo.com> writes:
Simon> I'm having trouble committing
things currently, and I'd
Simon> like to commit a one-liner I sent to xemacs-patches several
Simon> weeks ago first -- I sent a mail to the cvs maintainer so
Simon> hopefully it will be fixed soon.
Stephen> Er, I'm the XEmacs CVS maintainer now.... OK, I've fixed
Stephen> that alias. (gomen'nasai, Jareth-san. taihen gomeiwaku
Stephen> wo kaketeshimaimasita (^^;;;)
Stephen> What's the problem? Are you aware that cvs.xemacs.org is
Stephen> no more? We're planning to move to SunSITE.dk as soon as
Stephen> we can get some security issues straightened out. In the
Stephen> meantime, CVS changes are done using your
Stephen> "jas(a)barbar.josefsson.org" SSH public key, using
Stephen> CVS_RSH=/usr/bin/ssh
Stephen> CVSROOT=:ext:xemacs＠tanko.sk.tsukuba.ac.jp:/usr/CVSroot
Stephen> or whatever is appropriate at your system. Checkouts and
Stephen> updates can be done anonymously
Stephen> CVSROOT=:pserver:cvs＠sunsite.dk:/pack/anoncvs
Stephen> The cvs login password is "cvs". The canonical reference
Stephen> for this stuff is
Stephen> http://www.xemacs.org/Develop/cvsaccess.html, but it
Stephen> doesn't describe the temporary server at "tanko". Note
Stephen, I used to commit via
:ext:martinb＠cvs.xemacs.org:/usr/CVSroot
in cvs.xemacs.org times.
Oh, shared commit access has been moved from acount martinb to xemacs.
Somehow my mind failed to register this (registry full, perhaps?).
I'll do a quick update now.
Best regards,
Adrian
Stephen> that due to my university's brain-damaged firewall, the
Stephen> only way you can check service is via ssh -v, which will
Stephen> hang after announcing the connection (you don't have
Stephen> shell access; it automatically starts the CVS server,
Stephen> which is waiting for CVS :ext: client-server protocol).
Stephen> No ping, traceroute, etc, etc, due to firewall.
Stephen> --
Stephen> Institute of Policy and Planning Sciences
http://turnbull.sk.tsukuba.ac.jp
Stephen> University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573
JAPAN
Stephen> Don't ask how you can "do" free software
business;
Stephen> ask what your business can "do for" free
software.
--
Adrian Aichner
mailto:adrian＠xemacs.org
http://www.xemacs.org/
1:54 p.m.
>>>> "Simon" == Simon Josefsson
<jas(a)extundo.com> writes:
Simon> Ah, the cvsaccess.html page says to use
Simon> "martinb(a)tanko.sk.tsukuba.ac.jp".
Thanks for the report. Fixed in workspace now.
Simon> Ok, now ssh -v works, but commit doesn't:
Simon> cvs server: failed to create lock directory for
Simon> `/usr/CVSroot/xemacs-packages/xemacs-packages/mail-lib'
Your Repository file doesn't jibe with the structure of the repository
now, I guess. I suspect that the contents of your mail-lib/CVS/Root
file is the single line
:ext:martinb＠cvs.xemacs.org:/usr/CVSroot
(which doesn't matter because CVSROOT overrides, but doesn't change
it; you can change to the CVSROOT I gave by hand with no problem,
though and then you won't have to set CVSROOT), and your
mail-lib/CVS/Repository file is the single line
xemacs-packages/xemacs-packages/mail-lib
Change that to XEmacs/packages/xemacs-packages/mail-lib.
It's also possible that you're not committing from the working
directory itself. If not, cd'ing there might help. If this kind of
manipulation fails, you'll have to check out again. If so, I'm sorry
for the inconvenience.
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Don't ask how you can "do" free software business;
ask what your business can "do for" free software.
2:31 p.m.
>>>> "APA" == Adrian Aichner
<adrian(a)xemacs.org> writes:
APA> I'll do a quick update now.
Already done and committed. ;-)
--
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Don't ask how you can "do" free software business;
ask what your business can "do for" free software.
5:28 p.m.
"Stephen J. Turnbull" <stephen(a)xemacs.org> writes:
mail-lib/CVS/Repository file is the single line
xemacs-packages/xemacs-packages/mail-lib
Change that to XEmacs/packages/xemacs-packages/mail-lib.
This was the problem, thanks. But changing this makes anonymous CVS
stop working. The anonymous CVS server doesn't seem to be tightly
synced with tanko either, so I guess I'll only use tanko for now.
6:57 p.m.
>>>> "Simon" == Simon Josefsson
<jas(a)extundo.com> writes:
Simon> "Stephen J. Turnbull" <stephen(a)xemacs.org> writes:
> mail-lib/CVS/Repository file is the single line
>
> xemacs-packages/xemacs-packages/mail-lib
>
> Change that to XEmacs/packages/xemacs-packages/mail-lib.
Simon> This was the problem, thanks. But changing this makes
Simon> anonymous CVS stop working. The anonymous CVS server
Simon> doesn't seem to be tightly synced with tanko either, so I
Hi Simon, please report specific problems you are seeing.
We need this to find the cause.
SunSITE.dk is synced daily.
It might be though, that packages don't get synced correctly. They
don't appear in the rsync log.
I'll have a look.
Best regards,
Adrian
Simon> guess I'll only use tanko for now.
--
Adrian Aichner
mailto:adrian＠xemacs.org
http://www.xemacs.org/
7:21 p.m.
Adrian.Aichner(a)t-online.de (Adrian Aichner) writes:
Hi Simon, please report specific problems you are seeing.
cvs.sunsite.dk and tanko.sk.tsukuba.ac.jp doesn't have the same
"Repository" path. The following illustrates the problem, it is not
enough to change the cvs root with -d, you need to edit CVS/Repository
as well, as Stephen described.
anon cvs ok, but ssh not ok:
[jas＠dhcp128 mail-lib]$ cat CVS/Root
:pserver:cvs＠sunsite.dk:/pack/anoncvs
[jas＠dhcp128 mail-lib]$ cat CVS/Repository
xemacs-packages/xemacs-packages/mail-lib
[jas＠dhcp128 mail-lib]$ cvs upd
? ChangeLog.jas
? status
? todo
cvs server: Updating .
M mail-extr.el
M rmailout.el
[jas＠dhcp128 mail-lib]$ cvs -w -z3 -d :ext:xemacs＠tanko.sk.tsukuba.ac.jp:/usr/CVSroot upd
? ChangeLog.jas
? status
? todo
cvs server: Updating .
cvs server: cannot open directory /usr/CVSroot/xemacs-packages/xemacs-packages/mail-lib:
No such file or directory
cvs server: skipping directory
[jas＠dhcp128 mail-lib]$
ssh cvs ok, but anon cvs not ok:
[jas＠dhcp128 mail-lib]$ cat CVS/Root
:ext:xemacs＠tanko.sk.tsukuba.ac.jp:/usr/CVSroot
[jas＠dhcp128 mail-lib]$ cat CVS/Repository
XEmacs/packages/xemacs-packages/mail-lib
[jas＠dhcp128 mail-lib]$ cvs upd
? ChangeLog.jas
? status
? todo
cvs server: Updating .
M mail-extr.el
M rmailout.el
[jas＠dhcp128 mail-lib]$ cvs -d :pserver:cvs＠sunsite.dk:/pack/anoncvs upd
? ChangeLog.jas
? status
? todo
cvs server: Updating .
cvs server: cannot open directory /pack/anoncvs/XEmacs/packages/xemacs-packages/mail-lib:
No such file or directory
cvs server: skipping directory
[jas＠dhcp128 mail-lib]$
7:38 p.m.
>>>> "Simon" == Simon Josefsson
<jas(a)extundo.com> writes:
Simon> Adrian.Aichner(a)t-online.de (Adrian Aichner) writes:
> Hi Simon, please report specific problems you are seeing.
Simon> cvs.sunsite.dk and tanko.sk.tsukuba.ac.jp doesn't have the
Simon> same "Repository" path. The following illustrates the
That's right. I'll try to get a symlink created on SunSITE.dk to fix this.
But that's not a sync problem.
Can you give an example of what you called "doesn't seem to be tightly
synced with tanko either"?
Simon> problem, it is not enough to change the cvs root with -d,
Simon> you need to edit CVS/Repository as well, as Stephen
Simon> described.
Simon> anon cvs ok, but ssh not ok:
Simon> [jas＠dhcp128 mail-lib]$ cat CVS/Root
Simon> :pserver:cvs＠sunsite.dk:/pack/anoncvs
Simon> [jas＠dhcp128 mail-lib]$ cat CVS/Repository
Simon> xemacs-packages/xemacs-packages/mail-lib
Simon> [jas＠dhcp128 mail-lib]$ cvs upd
Simon> ? ChangeLog.jas
Simon> ? status
Simon> ? todo
Simon> cvs server: Updating .
Simon> M mail-extr.el
Simon> M rmailout.el
Simon> [jas＠dhcp128 mail-lib]$ cvs -w -z3 -d
:ext:xemacs＠tanko.sk.tsukuba.ac.jp:/usr/CVSroot upd
Simon> ? ChangeLog.jas
Simon> ? status
Simon> ? todo
Simon> cvs server: Updating .
Simon> cvs server: cannot open directory
/usr/CVSroot/xemacs-packages/xemacs-packages/mail-lib: No such file or directory
Simon> cvs server: skipping directory
Simon> [jas＠dhcp128 mail-lib]$
Simon> ssh cvs ok, but anon cvs not ok:
Simon> [jas＠dhcp128 mail-lib]$ cat CVS/Root
Simon> :ext:xemacs＠tanko.sk.tsukuba.ac.jp:/usr/CVSroot
Simon> [jas＠dhcp128 mail-lib]$ cat CVS/Repository
Simon> XEmacs/packages/xemacs-packages/mail-lib
Simon> [jas＠dhcp128 mail-lib]$ cvs upd
Simon> ? ChangeLog.jas
Simon> ? status
Simon> ? todo
Simon> cvs server: Updating .
Simon> M mail-extr.el
Simon> M rmailout.el
Simon> [jas＠dhcp128 mail-lib]$ cvs -d :pserver:cvs＠sunsite.dk:/pack/anoncvs upd
Simon> ? ChangeLog.jas
Simon> ? status
Simon> ? todo
Simon> cvs server: Updating .
Simon> cvs server: cannot open directory
/pack/anoncvs/XEmacs/packages/xemacs-packages/mail-lib: No such file or directory
Simon> cvs server: skipping directory
Simon> [jas＠dhcp128 mail-lib]$
--
Adrian Aichner
mailto:adrian＠xemacs.org
http://www.xemacs.org/
8:21 p.m.
Adrian.Aichner(a)t-online.de (Adrian Aichner) writes:
Simon> cvs.sunsite.dk and tanko.sk.tsukuba.ac.jp doesn't
have the
Simon> same "Repository" path. The following illustrates the
That's right. I'll try to get a symlink created on SunSITE.dk to fix this.
Thanks!
But that's not a sync problem.
Can you give an example of what you called "doesn't seem to be tightly
synced with tanko either"?
If I use tanko as CVS Root, I get the patch I committed today, if I
use sunsite I don't get them. But you said that they are synchronized
every day, and that's probably OK by me (except for mail-lib though..).
8:41 p.m.
>>>> "Simon" == Simon Josefsson
<jas(a)extundo.com> writes:
Simon> Adrian.Aichner(a)t-online.de (Adrian Aichner) writes:
Simon> cvs.sunsite.dk and tanko.sk.tsukuba.ac.jp doesn't have the
Simon> same "Repository" path. The following illustrates the
>
> That's right. I'll try to get a symlink created on SunSITE.dk to fix this.
Simon> Thanks!
I'm afraid this will have to wait. It's not as simple as it looked.
> But that's not a sync problem.
>
> Can you give an example of what you called "doesn't seem to be tightly
> synced with tanko either"?
Simon> If I use tanko as CVS Root, I get the patch I committed
Simon> today, if I use sunsite I don't get them. But you said
Isn't this the issue with the relative paths in .../CVS/Repository?
Later,
Adrian
Simon> that they are synchronized every day, and that's probably
Simon> OK by me (except for mail-lib though..).
--
Adrian Aichner
mailto:adrian＠xemacs.org
http://www.xemacs.org/
9:39 p.m.
Adrian.Aichner(a)t-online.de (Adrian Aichner) writes:
>> Can you give an example of what you called
"doesn't seem to be tightly
>> synced with tanko either"?
Simon> If I use tanko as CVS Root, I get the patch I committed
Simon> today, if I use sunsite I don't get them. But you said
Isn't this the issue with the relative paths in .../CVS/Repository?
No, I don't think so. If I change Root+Repository to be correct for
sunsite, sunsite still doesn't contain the latest changes from tanko.
This is not a problem, a 1 day delay is OK.
9:53 p.m.
>>>> "Simon" == Simon Josefsson
<jas(a)extundo.com> writes:
Simon> Adrian.Aichner(a)t-online.de (Adrian Aichner) writes:
> >> Can you give an example of what you called
"doesn't seem to be tightly
> >> synced with tanko either"?
>
Simon> If I use tanko as CVS Root, I get the patch I committed
Simon> today, if I use sunsite I don't get them. But you said
>
> Isn't this the issue with the relative paths in .../CVS/Repository?
Simon> No, I don't think so. If I change Root+Repository to be
Simon> correct for sunsite, sunsite still doesn't contain the
Simon> latest changes from tanko. This is not a problem, a 1 day
Hi Simon!
I did a manual rsync earlier today on SunSITE.dk and your mail-lib
checkin of today did get synced over to SunSITE.dk!
bash-2.03$ cd /pack/xemacscvs/packages
bash-2.03$ find . -mtime -1 -ls
30351 1 drwxrwsr-x 2 cvs xemacs 1024 Nov 8 20:29
./xemacs-packages/mail-lib
32338 42 -r--r--r-- 1 adrian xemacs 42185 Nov 8 18:23
./xemacs-packages/mail-lib/smtpmail.el,v
32337 40 -r--r--r-- 1 adrian xemacs 40803 Nov 8 18:23
./xemacs-packages/mail-lib/ChangeLog,v
bash-2.03$
Again: Please state exactly a few files which are not in sync NOW
(i.e. showing up as locally modified because the repository is
behind).
Thanks in advance,
Adrian
Simon> delay is OK.
--
Adrian Aichner
mailto:adrian＠xemacs.org
http://www.xemacs.org/
9:43 a.m.
On 8 Nov 2001, Adrian Aichner wrote:
I did a manual rsync earlier today on SunSITE.dk and your mail-lib
checkin of today did get synced over to SunSITE.dk!
Thanks.
Again: Please state exactly a few files which are not in sync NOW
(i.e. showing up as locally modified because the repository is
behind).
Only ChangeLog and smtpmail.el in mail-lib, but you fixed this.
(The problem is that the files don't show up as locally modified, they are
reverted to the version on the server. So if I commit something on tanko,
then go to sunsite, my changes are reverted until sunsite gets them. As I
said, I can live with this.)
12:41 p.m.
>>>> "Simon" == Simon Josefsson
<jas(a)extundo.com> writes:
Simon> On 8 Nov 2001, Adrian Aichner wrote:
> I did a manual rsync earlier today on SunSITE.dk and your
mail-lib
> checkin of today did get synced over to SunSITE.dk!
Simon> Thanks.
> Again: Please state exactly a few files which are not in sync
NOW
> (i.e. showing up as locally modified because the repository is
> behind).
Simon> Only ChangeLog and smtpmail.el in mail-lib, but you fixed
Simon> this.
Great! This confirms rsync to work right for xemacs packages too.
Simon> (The problem is that the files don't show up as locally
Simon> modified, they are reverted to the version on the server.
Simon> So if I commit something on tanko, then go to sunsite, my
Simon> changes are reverted until sunsite gets them. As I said, I
Simon> can live with this.)
OK, I'll concentrate on getting this fixed for good.
This won't happen over this weekend though.
Best regards,
Adrian
--
Adrian Aichner
mailto:adrian＠xemacs.org
http://www.xemacs.org/
8210
days inactive
8217
days old
24 comments
6 participants
participants (6)
-
Adrian Aichner -
Adrian Aichner -
Ilya N. Golubev -
Karl Kleinpaste -
Simon Josefsson -
Stephen J. Turnbull