On Fri, Jul 01 2022, Oswald Buddenhagen wrote:
> On Fri, Jul 01, 2022 at 06:03:15PM +0200, Yuri D'Elia wrote:
>>Can the bigger buffering performed by ssh be an issue with the pipeline
>>size, or it's unrelated?
>>
> the only way it could matter is if changed chunk size or timing
> sufficiently to t
On Fri, Jul 01, 2022 at 06:03:15PM +0200, Yuri D'Elia wrote:
Can the bigger buffering performed by ssh be an issue with the pipeline
size, or it's unrelated?
the only way it could matter is if changed chunk size or timing
sufficiently to trigger an otherwise unused and broken code path.
On Fri, Jul 01 2022, Oswald Buddenhagen wrote:
> On Fri, Jul 01, 2022 at 05:34:05PM +0200, Yuri D'Elia wrote:
> make sure CRLF or other characters aren't mangled (i recently failed
> piping a binary file through ssh from android to linux, so this is a
> very real concern).
>
> you can try reverting
On Fri, Jul 01, 2022 at 05:34:05PM +0200, Yuri D'Elia wrote:
Socket error: receive buffer full. Probably protocol error.
make sure CRLF or other characters aren't mangled (i recently failed
piping a binary file through ssh from android to linux, so this is a
very real concern).
you can try r
I'm just trying a combination I've never tested before: connection via
Tunnel.
Whenever I use Tunnel, synchronization fails after a few hundred
messages with:
<...>
Synchronizing...
Socket error: receive buffer full. Probably protocol error.
<...>
However, I highly doubt it's a protocol error. T
