Re: gnus crashes on threads deeper than 333 articles

[Chris Moore]

Chris Moore <address@hidden> writes:

> Chris Moore <address@hidden> writes:
>
>> One thing I noticed it that the new sort funcction changes the length
>> of the list of threads:
>>
>>   *** Welcome to IELM ***  Type (describe-mode) for help.
>>   ELISP> (length before-sorting)
>>   106
>>   ELISP> (length after-sorting)
>>   258
>
> That seems to be a red herring.  Running the new sort function on
> before-sorting gives a list the same length as running the old sort
> function does.  I tried using 'equal to compare the 2 outputs, but got
> a stack overflow error.

OK, so the 'loop' version of the sort function does seem to work, but
fixing that has shown up a new problem.

The '200' and '201' in this function limit the thread depth.
Increasing these numbers has allowed me to finally open my large
Increasing these mailbox!

  (defun gnus-make-thread-indent-array ()
    (let ((n 200))
      (unless (and gnus-thread-indent-array
                 (= gnus-thread-indent-level gnus-thread-indent-array-level))
        (setq gnus-thread-indent-array (make-vector 201 "")