Re: 30.0.50; 'New' error message when using eglot

[mprodrigues@posteo.net]

So in my case the culprit was doom-modeline, I guess that is because it 
was trying
to call eglot--spinner, but that doesn't exist anymore. Anyway, I was 
expecting this
to be a warning on their side, not outright mess with eglot functioning. 
This has
already been reported on their side and I guess the fix is in the next 
release.

Thanks.

Best regards,
Mateus Rodrigues

On 28.02.2023 13:42, João Távora wrote:
> mprodrigues@posteo.net writes:
> 
>> Hello,
>> 
>> I have the same problem.
>> 
>>> emacs -Q --eval '(add-hook (quote python-mode-hook) (function
>>> eglot-ensure))' ~/tmp/coiso.py
>> 
>> If I use this it works, so maybe the bug is coming from interactions
>> with other packages?
> 
> It would surely seem so, as eglot.el doesn't have very many 
> plist-member
> calls, and none of them for the ':signatures' key.
> 
> So someone is going to go throught the trouble of making a reproducible
> error recipe (or bisecting a configuration), otherwise I can't be of
> much help.
> 
>> I did M-x toggle-debug-on-error so we can get a backtrace, and this is
>> what I got:
>> 
>> Debugger entered--Lisp error: (wrong-type-argument plistp [])
>>   plist-member([] :signatures)
>>   #f(compiled-function (jsonrpc-lambda-elem10) #<bytecode
>>    0xdcf9bf12d010fe5>)([])
>>   jsonrpc-connection-receive(#<eglot-lsp-server
>>   eglot-lsp-server-186e364> (:jsonrpc "2.0" :id 44 :result []))
>>   jsonrpc--process-filter(#<process EGLOT (mprodrigues/(python-mode
>>   python-ts-mode))> "Content-Length: 37\15\nContent-Type:
>>   application/vsco...")
> 
> It would also help to see what the the client request for :id 44 was in
> the first place.  You could find that in the buffer brought to you by
> M-x eglot-events-buffer
> 
> João