bzr:// vz nosmart+bzr://

bzr:// vz nosmart+bzr://

[Eli Zaretskii]

Maybe you know, maybe you don't, but the bzr server on savannah was
upgraded to the latest version 2.4 a few days ago.  As result of this
(or maybe as result of some other changes on savannah), the initial
"bzr branch" command became approximately twice faster than it was
before.  It takes about 18 min on a gnu.org machine (probably on the
same LAN), and 1 hour on a very remote machine.

What's also important, it seems to be no longer true that using the
nosmart+bzr:// protocol for the initial "bzr branch" is faster than
the normal "smart" bzr://.  In fact, nosmart+bzr:// seems to be about
30% _slower_ now.

Would people please time the following 2 commands and see if this
observation is correct in general?  If it is, I'd like to update the
instructions on the wiki, because describing 2 different protocols
there might unnecessarily confuse new users.

  bzr branch nosmart+bzr://bzr.savannah.gnu.org/emacs/trunk/
  bzr branch bzr://bzr.savannah.gnu.org/emacs/trunk/

Please note that this should be done outside of the directory where
you have the Emacs bzr repository, i.e. where you typed the original
"bzr init-repo" command (should be the parent of the `trunk' directory
if you were following the instructions on the wiki).  Otherwise, bzr
will not bring stuff from the remote server, but rather reuse what it
already has in the repository.

TIA

Re: bzr:// vz nosmart+bzr://

[Frank Schmitt]

Eli Zaretskii <eliz@gnu.org> writes:

> Would people please time the following 2 commands and see if this
> observation is correct in general?  If it is, I'd like to update the
> instructions on the wiki, because describing 2 different protocols
> there might unnecessarily confuse new users.
>
>   bzr branch nosmart+bzr://bzr.savannah.gnu.org/emacs/trunk/
>   bzr branch bzr://bzr.savannah.gnu.org/emacs/trunk/

All I get (in both cases) is 

bzr: ERROR: Connection error: failed to connect to
bzr.savannah.gnu.org:4155: Connection refused

-- 
Have you ever considered how much text can fit in eighty columns?  Given that a
signature typically contains up to four lines of text, this space allows you to
attach a tremendous amount of valuable information to your messages.  Seize the
opportunity and don't waste your signature on bullshit that nobody cares about.

Re: bzr:// vz nosmart+bzr://

[Eli Zaretskii]

> From: Frank Schmitt <ich@frank-schmitt.net>
> Date: Mon, 10 Oct 2011 11:04:35 +0200
> 
> >   bzr branch nosmart+bzr://bzr.savannah.gnu.org/emacs/trunk/
> >   bzr branch bzr://bzr.savannah.gnu.org/emacs/trunk/
> 
> All I get (in both cases) is 
> 
> bzr: ERROR: Connection error: failed to connect to
> bzr.savannah.gnu.org:4155: Connection refused

There's an unrelated problem on savannah, please try these later when
the problem is fixed.

Re: bzr:// vz nosmart+bzr://

[Ari Roponen]

Eli Zaretskii <eliz@gnu.org> writes:

> Would people please time the following 2 commands and see if this
> observation is correct in general?  If it is, I'd like to update the
> instructions on the wiki, because describing 2 different protocols
> there might unnecessarily confuse new users.
>
>   bzr branch nosmart+bzr://bzr.savannah.gnu.org/emacs/trunk/
>   bzr branch bzr://bzr.savannah.gnu.org/emacs/trunk/

[nosmart]$ time bzr branch nosmart+bzr://bzr.savannah.gnu.org/emacs/trunk/
Branched 106042 revision(s).

real	23m21.398s
user	9m32.193s
sys	0m8.146s


[bzr]$ time bzr branch bzr://bzr.savannah.gnu.org/emacs/trunk/
Branched 106042 revision(s).

real	12m27.467s
user	7m25.061s
sys	0m10.856s

-- 
Ari Roponen

Re: bzr:// vz nosmart+bzr://

[Eli Zaretskii]

> From: Ari Roponen <ari.roponen@gmail.com>
> Cc: emacs-devel@gnu.org
> Date: Mon, 10 Oct 2011 12:36:48 +0300
> 
> [nosmart]$ time bzr branch nosmart+bzr://bzr.savannah.gnu.org/emacs/trunk/
> Branched 106042 revision(s).
> 
> real	23m21.398s
> user	9m32.193s
> sys	0m8.146s
> 
> 
> [bzr]$ time bzr branch bzr://bzr.savannah.gnu.org/emacs/trunk/
> Branched 106042 revision(s).
> 
> real	12m27.467s
> user	7m25.061s
> sys	0m10.856s

Thanks!

Re: bzr:// vz nosmart+bzr://

[Sven Joachim]

On 2011-10-10 10:43 +0200, Eli Zaretskii wrote:

> Maybe you know, maybe you don't, but the bzr server on savannah was
> upgraded to the latest version 2.4 a few days ago.  As result of this
> (or maybe as result of some other changes on savannah), the initial
> "bzr branch" command became approximately twice faster than it was
> before.  It takes about 18 min on a gnu.org machine (probably on the
> same LAN), and 1 hour on a very remote machine.
>
> What's also important, it seems to be no longer true that using the
> nosmart+bzr:// protocol for the initial "bzr branch" is faster than
> the normal "smart" bzr://.  In fact, nosmart+bzr:// seems to be about
> 30% _slower_ now.

That seems to be correct.

> Would people please time the following 2 commands and see if this
> observation is correct in general?  If it is, I'd like to update the
> instructions on the wiki, because describing 2 different protocols
> there might unnecessarily confuse new users.
>
>   bzr branch nosmart+bzr://bzr.savannah.gnu.org/emacs/trunk/
>   bzr branch bzr://bzr.savannah.gnu.org/emacs/trunk/

% time bzr branch nosmart+bzr://bzr.savannah.gnu.org/emacs/trunk/
Branched 106042 revision(s).                                                                       
bzr branch nosmart+bzr://bzr.savannah.gnu.org/emacs/trunk/  448,75s user 7,46s system 37% cpu 20:21,80 total
% mv trunk trunk.old
% time bzr branch bzr://bzr.savannah.gnu.org/emacs/trunk/ 
Branched 106042 revision(s).                                                                       
bzr branch bzr://bzr.savannah.gnu.org/emacs/trunk/  352,41s user 5,63s system 47% cpu 12:36,04 total

Cheers,
       Sven

Re: bzr:// vz nosmart+bzr://

[Tassilo Horn]

Eli Zaretskii <eliz@gnu.org> writes:

Hi Eli,

here are my results, with a pretty fast internet connection.

>   bzr branch nosmart+bzr://bzr.savannah.gnu.org/emacs/trunk/

bzr branch nosmart+bzr://bzr.savannah.gnu.org/emacs/trunk/
678.70s user 7.29s system 54% cpu 21:04.38 total

>   bzr branch bzr://bzr.savannah.gnu.org/emacs/trunk/

bzr branch bzr://bzr.savannah.gnu.org/emacs/trunk/
543.03s user 7.29s system 65% cpu 13:57.98 total

So your observation of the smart server actually being faster seems to
be true.  I use the bzr client version 2.4.1.

Bye,
Tassilo

Re: bzr:// vz nosmart+bzr://

[Eli Zaretskii]

> From: Sven Joachim <svenjoac@gmx.de>
> Cc: emacs-devel@gnu.org
> Date: Mon, 10 Oct 2011 12:17:22 +0200
> 
> % time bzr branch nosmart+bzr://bzr.savannah.gnu.org/emacs/trunk/
> Branched 106042 revision(s).                                                                       
> bzr branch nosmart+bzr://bzr.savannah.gnu.org/emacs/trunk/  448,75s user 7,46s system 37% cpu 20:21,80 total
> % mv trunk trunk.old
> % time bzr branch bzr://bzr.savannah.gnu.org/emacs/trunk/ 
> Branched 106042 revision(s).                                                                       
> bzr branch bzr://bzr.savannah.gnu.org/emacs/trunk/  352,41s user 5,63s system 47% cpu 12:36,04 total

Thanks.

Re: bzr:// vz nosmart+bzr://

[Eli Zaretskii]

> From: Tassilo Horn <tassilo@member.fsf.org>
> Cc: emacs-devel@gnu.org
> Date: Mon, 10 Oct 2011 12:53:23 +0200
> 
> >   bzr branch nosmart+bzr://bzr.savannah.gnu.org/emacs/trunk/
> 
> bzr branch nosmart+bzr://bzr.savannah.gnu.org/emacs/trunk/
> 678.70s user 7.29s system 54% cpu 21:04.38 total
> 
> >   bzr branch bzr://bzr.savannah.gnu.org/emacs/trunk/
> 
> bzr branch bzr://bzr.savannah.gnu.org/emacs/trunk/
> 543.03s user 7.29s system 65% cpu 13:57.98 total
> 
> So your observation of the smart server actually being faster seems to
> be true.  I use the bzr client version 2.4.1.

Thanks.

Re: bzr:// vz nosmart+bzr://

[Juanma Barranquero]

On Mon, Oct 10, 2011 at 10:43, Eli Zaretskii <eliz@gnu.org> wrote:
>  bzr branch nosmart+bzr://bzr.savannah.gnu.org/emacs/trunk/

00:32:13,66

>  bzr branch bzr://bzr.savannah.gnu.org/emacs/trunk/

00:23:10,48

So nosmart was about 39% slower.

This on Windows 7, with a 2.5b1 client.

    Juanma

Re: bzr:// vz nosmart+bzr://

[Eli Zaretskii]

> From: Juanma Barranquero <lekktu@gmail.com>
> Date: Mon, 10 Oct 2011 16:10:00 +0200
> Cc: emacs-devel@gnu.org
> 
> On Mon, Oct 10, 2011 at 10:43, Eli Zaretskii <eliz@gnu.org> wrote:
> >  bzr branch nosmart+bzr://bzr.savannah.gnu.org/emacs/trunk/
> 
> 00:32:13,66
> 
> >  bzr branch bzr://bzr.savannah.gnu.org/emacs/trunk/
> 
> 00:23:10,48
> 
> So nosmart was about 39% slower.

Thanks.