[9fans] snarf buffers under P9P and synergy, acme vs rio

[9fans] snarf buffers under P9P and synergy, acme vs rio

[Paul Lalonde]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I use synergy to share my keyboard and mouse across a Mac and a Linux  
box.  On the linux box my WM is rio, and I do most of my editing in  
acme.  I've had no problems snarfing text in acme and pasting it over  
to my mac.  But the reverse is not true.  In any given session the  
first snarf from my mac pastes correctly into acme.  After that, I  
can paste into a 9term window, and get the correct paste, but only  
the first text I pasted goes to the acme window :-(.  I can paste  
into the 9term window, and cut/snarf from there and then dump it into  
acme without problems, but it's annoying that I have to paste via a  
9term.

Any clues to what I need to fix?

Thanks,
     Paul

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (Darwin)

iD8DBQFFHECjpJeHo/Fbu1wRAttuAJ0eVkw7phT7GbpMZH/XEeWte315MACfT53z
4Yf7jkyXwaPbozk6gLtRGMI=
=jsLR
-----END PGP SIGNATURE-----

Re: [9fans] snarf buffers under P9P and synergy, acme vs rio

[Russ Cox]

At least if you've updated in the past two months,
acme and 9term should be using identical code
to read the snarf buffer from the window system.

Is the 9term window on Mac or on Linux?
If it's on the Mac, then that's the difference.

One possible solution is to run snarfer in the
background on whichever box the 9term is
currently helping out on.  Snarfer is like xclipboard,
without the gui.

Russ

Re: [9fans] snarf buffers under P9P and synergy, acme vs rio

[Paul Lalonde]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Magic words - "update in past two months."  I'm clearly much too  
accustomed to software that isn't on a continuous improvement cycle :-)

Mea maxima culpa.

The 9term is on the Linux box.

I'll try the update, and if that fails I'll try the snarfer.

Thanks Russ.

Paul

On 28-Sep-06, at 2:54 PM, Russ Cox wrote:

> At least if you've updated in the past two months,
> acme and 9term should be using identical code
> to read the snarf buffer from the window system.
>
> Is the 9term window on Mac or on Linux?
> If it's on the Mac, then that's the difference.
>
> One possible solution is to run snarfer in the
> background on whichever box the 9term is
> currently helping out on.  Snarfer is like xclipboard,
> without the gui.
>
> Russ

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (Darwin)

iD8DBQFFHEWlpJeHo/Fbu1wRAkJjAKDPM7Gh8Id+VEpMBG8H/x0q5QGhrQCdE6I5
/0KDZpelnqxc8Gl1a4Rkdfo=
=BpZZ
-----END PGP SIGNATURE-----

Re: [9fans] snarf buffers under P9P and synergy, acme vs rio

[Russ Cox]

> Magic words - "update in past two months."  I'm clearly much too
> accustomed to software that isn't on a continuous improvement cycle :-)

It's a rather discontinuous cycle, but I rewrote the
graphics code to be single threaded on July 12,
and that may have fixed a bunch of weird races.

Russ

Re: [9fans] snarf buffers under P9P and synergy, acme vs rio

[Paul Lalonde]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Would that re-writing graphics single-threaded were in my life.
I'm doing the exact opposite right now - taking a single-threaded  
(3D) graphics engine and multi-threading it properly (I hope...) to  
run on multi-core architectures.  If you want weird races...  We have  
a little technology around not putting races in in the first place,  
but I find it difficult to trust that technology not to be buggy.

Paul

On 28-Sep-06, at 3:04 PM, Russ Cox wrote:

>> Magic words - "update in past two months."  I'm clearly much too
>> accustomed to software that isn't on a continuous improvement  
>> cycle :-)
>
> It's a rather discontinuous cycle, but I rewrote the
> graphics code to be single threaded on July 12,
> and that may have fixed a bunch of weird races.
>
> Russ

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (Darwin)

iD8DBQFFHEg1pJeHo/Fbu1wRAvguAJ0cSHzzQrRztPWmXMdg7eQx9tt+kQCg4Uzk
8QG1tQiTHtxJj7GbdmbSnoE=
=wHPy
-----END PGP SIGNATURE-----