Urgent: Marks and *very* old nnml groups

Urgent: Marks and *very* old nnml groups

[Kai Großjohann]

My boss has some *very* old nnml groups.  He needs to move to another
machine.  I was thinking it should be possible to just tar up the
~/Mail directory (or its subdirs) and unpack them on the new
machine.  Thanks to the new .marks files, the marks should be
transferred, too.

But I looked and I found some (459...) directories in ~/Mail which
did not have .marks files in them.  Inspection shows that some of
those can be explained because they are not groups, just parent
directories of other groups.  For example, he has groups nnml:pers.john
and nnml:pers.jane, so there is a directory ~/Mail/pers which does
not have a .marks file.

However, I'm afraid that this might not explain all of them.

So, here come my questions:

* Will the .marks code create the files even for groups that have
  never been entered?

* If not, is there a way to make sure that the .marks files will be
  generated from the data in .newsrc.eld?

Thanks a lot.
-- 
Ambibibentists unite!

Re: Urgent: Marks and *very* old nnml groups

[Simon Josefsson]

kai.grossjohann@uni-duisburg.de (Kai Großjohann) writes:

> * Will the .marks code create the files even for groups that have
>   never been entered?

Nope.

> * If not, is there a way to make sure that the .marks files will be
>   generated from the data in .newsrc.eld?

Maybe process marking all groups and pressing M-g works?

Re: Urgent: Marks and *very* old nnml groups

[Kai Großjohann]

Simon Josefsson <jas@extundo.com> writes:

> kai.grossjohann@uni-duisburg.de (Kai Großjohann) writes:
>
>> * If not, is there a way to make sure that the .marks files will be
>>   generated from the data in .newsrc.eld?
>
> Maybe process marking all groups and pressing M-g works?

I've now done that, and then I got another idea for checking: I
printed a list of all directories, then looked for ones that had
.overview but not .marks.  Checking the first half of the 25 results
appeared to indicate that they were all somewhat bogus, so I think
losing marks on those is not really a problem.

Thanks, Simon!
-- 
Ambibibentists unite!