Live System User <[email protected]> writes:

> Hi,

Hi,

>   I cannot get it to fail again with my full configuratiom -- just that
>   one time (after the GZIP modifications on master but the error
>   ocurred on Emacs 25.2)
>>
>>>         A wild guess: did Emacs 25.2 use info that was cached/saved
>>>         from mater. perhaps from the `tramp-persistency-file-name'
>>>         file (though I haven't started up TRAMO on master for a couple
>>>         days)?

Well, I remember. The format of the `tramp-persistency-file-name' has
changed between Tramp 2.2 (part of Emacs 25) and 2.3 (part of Emacs
26). This explains the error, and that you couldn't reproduce it later.

>   OK I can do that -- maybe adding the Emacs version to the filemame
>   so that Emacs 14.2 doesn't use master's `tramp-persistency-file-name'.

Yes, you could do this. Or set `tramp-persistency-file-name' to nil.

>   Can you explain the puropse of the `tramp-persistency-file-name'?

It keeps some cached data for Tramp.

>   What's its main advantage and under what scenarios is it more
>   advantageous?

Faster startup.

> With a shred/master socket (controllmaster)?  Controllpersist?

No, it's not related.

>         What are its limitations?

None.

>         When is it best to use one?

Always.

>         When should you not use one?

In your case, when you switch between Tramp versions.

>         Are mulyiple separate SSH connections and
>         `tramp-persistency-file-name' files supported?

Sure, the file keeps separate data for every single connection.

>         Thanks.

Best regards, Michael.

_______________________________________________
Tramp-devel mailing list
[email protected]
https://lists.gnu.org/mailman/listinfo/tramp-devel

Reply via email to