-----BEGIN PGP SIGNATURE-----

iIUEABYIAC0mHE1vcml0eiBQb2xkcmFjayA8bW9yaXR6QHBvbGRyYWNrLmRldj4F
AmhIqS8ACgkQhCbXCZuMbdrfXwD8DUWShvCxLViGVceSz7bh0SBKYQmMeVZGMKr7
fdzOMRQA/iUAVInUt5FXHQ4JxrIPduHtiyHgHdBqzrHURbDXsVcF
=Gh9J
-----END PGP SIGNATURE-----

On Mon 09 Jun 2025 21:59:42, Peter Dobsan wrote:
> However, this or similar use of `*-cred-cmd` options work only if these options
> are re-evaluated when a connection fails. May I suggest to implement that?

Hi, just wanted to let you know that I've taken a look and at first glance
it does not look trivial to implement since a copy of the password is
kept in the worker. But it should be possible by injecting a function
that retrieves the password and using that instead.

-- 
Moritz Poldrack
https://moritz.sh

> Falling rock.

On Wed 28 May 2025 11:45:54, Eugenio Levi wrote:
> I am also trying to start with automatic folding, but there
> does not seem to be a corresponding option and the
> aerc-startup option does not work for this case.

Unfortunately, that is a known issue and due to how aerc runs commands.
An overhaul is due, but to my knowledge has not been started yet.

-- 
Moritz Poldrack
https://moritz.sh

> Shake well before using.

https://lists.sr.ht/~rjarry/aerc-devel/patches/59497 has been submitted
:)

-- 
Moritz Poldrack
https://moritz.sh

> To avoid suffocation, keep away from children.

On Thu 22 May 2025 12:51:09, Simon Martin wrote:
> diff --git a/commands/msg/delete.go b/commands/msg/delete.go
> index dcfc5ea1..3bafa2d2 100644
> --- a/commands/msg/delete.go
> +++ b/commands/msg/delete.go
> @@ -48,7 +50,31 @@ func (Delete) CompleteMFS(arg string) []string {
>  	return commands.FilterList(types.StrategyStrs(), arg, nil)
>  }
>  
> +// :delete is allowed if either:
> +//   - The `restrict-delete` setting is false, or
> +//   - The current directory is either the Trash or the Junk folder.
> +func mayDeleteOnSelectedDirectory() bool {
> +	selAccount := app.SelectedAccount()

On Tue 13 May 2025 10:47:36, Robin Jarry wrote:
> Moritz Poldrack, May 12, 2025 at 19:16:
>> +	var result strings.Builder
>> +	for i, v := range flags {
>> +		if i > 0 {
>> +			result.WriteString(", ")
>> +		}
>> +		result.WriteString(v)
>> +	}
>> +
>> +	return result.String()
>
> Why not use strings.Join(flags, ", ")?

On Tue 13 May 2025 08:21:26, Simon Martin wrote:
> It can, but I preferred to do it here to avoid putting the for in a big
> if, because we still want to run what's after it, and I find it less
> readable. I don't feel strongly about this though, so if you do, I can
> do this in a v2.

You could also just `return` :)

-- 
Moritz Poldrack
https://moritz.sh

> Many suitcases look alike.

Hi, thanks for the patch :)

On Tue 13 May 2025 08:08:55, Simon Martin wrote:
> b099981 added a call to updateDirCounts for deletions, and that function

The commit hash doesnt have to be referenced in the "prose", since its
been correctly referenced below. Makes it a bit harder to read in here.

> does not check the account store against nil before dereferencing it.
> This could happen for example in case of disconnection with the IMAP
> server during a very large deletion, and aerc crashes.
>
> This patch simply makes updateDirCounts bail out if the store is nil.

On Sat 03 May 2025 22:08:55, Bence Ferdinandy wrote:
>> [commands]
>> register=print
>
> Am I reading this right, that I would have  bunch of register=mycommand lines
> listed here?

No, just a single register=mycommand,myothercommand,thisdifferentcommand

The main idea was to avoid having "in development" commands show up in
your "production" aerc.

> Could we not infer these from the [command-mycommand] sections? Or by
> having a huge list (that I think we can write one-per-line in the

Hi, I've just taken a look at the current implementation and I dont
think that's possible/feasible with it

On Sat 03 May 2025 08:15:37, Matěj Cepl wrote:
> I can quite easily and speedily made another subselections and
> mark unnecessary messages as read.

Right now, the filtering is "flat" so there is no way to undo a single
filter. One way to change this, that I find interesting is by making the
filters into a chain of filter function, similar to mark-patch[0]. This
would allow predefining filters, go back only one layer, reapply the
filters, and probably more. But that would be a larger overhaul.

[0]: https://lists.sr.ht/~rjarry/aerc-devel/patches/58101#%3C20250314224613.7217-1-git@moritz.sh%3E+commands/msg/mark.go