Please help me renumber articles !

Please help me renumber articles !

[Aharon Schkolnik]

Hi.

After a while, exiting a group gets slow. In the past, I was told that
this is because of the large range of article numbers in the group -
caused by ticked articles. I was told that the thing to do is to copy
all the articles from the group to itself. That is what I did - using
#, and G M. Everything goes fine, but it doesn't solve the
problem. Here are the group parameters before and after:




;; Type `C-c C-c' after editing the group info.

("nnfolder:mail.osf" 2
 ((1 . 2573))
 ((tick
   (1 . 73)))
 (nnfolder ""))


;; Type `C-c C-c' after editing the group info.

("nnfolder:mail.osf" 2
 ((1 . 2573))
 ((tick
   (74 . 146)))
 (nnfolder ""))


I think what I want is:

("nnfolder:mail.osf" 2
 ((1 . 73))
 ((tick
   (1 . 73 )))
 (nnfolder ""))



gnus-version's value is "Gnus v5.2.25"


Can anyone out there help me ?


TIA.


-- 
  The day is short, and the work is great,     |   Aharon Schkolnik
  and the laborers are lazy, and the reward    |   Aharon@Health.Gov.IL
  is great, and the Master of the house is     |
  impatient. - Ethics Of The Fathers Ch. 2     |

Re: Please help me renumber articles !

[David Moore]

Aharon Schkolnik <aharon@health.gov.il> writes:

> After a while, exiting a group gets slow. In the past, I was told that
> this is because of the large range of article numbers in the group -
> caused by ticked articles. I was told that the thing to do is to copy
> all the articles from the group to itself. That is what I did - using
> #, and G M. Everything goes fine, but it doesn't solve the
> problem. Here are the group parameters before and after:

	If you want all of the numbers back to 1, you can go into the
group with `C-u SPC', turn off threading with `T T', sort the articles
in a reasonable order `C-c C-s C-n' then using `C-u ### B m' refile
everything into a _new_ group (where ### is large enough to cover all of
the articles).  Then you can go and rename & remove the original group,
and then rename the new group to the old name.

	If you just want the group compacted, but not starting at
article number 1, you can just refile into the same group rather than
a new one.


	Oh, and if you use the cache, be careful that everything works
right.  I think I've noticed that renaming a group doesn't always rename
the cache.


Anyone have better thoughts on this?  Considering how slow total-expire
currently gets with groups with lots of holes, maybe a safe compaction
routine should be added to Gnus itself.

-- 
David Moore <dmoore@ucsd.edu>       | Computer Systems Lab      __o
UCSD Dept. Computer Science - 0114  | Work: (619) 534-8604    _ \<,_
La Jolla, CA 92093-0114             | Fax:  (619) 534-1445   (_)/ (_)
<URL:http://oj.egbt.org/dmoore/>    | In a cloud bones of steel.

Re: Please help me renumber articles !

[Aharon Schkolnik]

>>>>> "David" == David Moore <dmoore@UCSD.EDU> writes:

    David> Aharon Schkolnik <aharon@health.gov.il> writes:
    >> After a while, exiting a group gets slow. In the past, I was
    >> told that this is because of the large range of article numbers
    >> in the group - caused by ticked articles. I was told that the
    >> thing to do is to copy all the articles from the group to
    >> itself. That is what I did - using #, and G M. Everything goes
    >> fine, but it doesn't solve the problem. Here are the group
    >> parameters before and after:

    David> 	If you want all of the numbers back to 1, you can go
    David> into the group with `C-u SPC', turn off threading with `T
    David> T', sort the articles in a reasonable order `C-c C-s C-n'
    David> then using `C-u ### B m' refile everything into a _new_
    David> group (where ### is large enough to cover all of the
    David> articles).  Then you can go and rename & remove the
    David> original group, and then rename the new group to the old
    David> name.


I don't know why it is necessary to turn off threading, and to sort
the articles. I skipped this step. I moved all the articles to a
temporary group (B m), renamed the original to something else, and
then renamed the temporary group to the original name - worked
fine. However, I was wondering if it would be possible to DELETE the
original after moving all the articles out of it, and then rename the
temporary group to the original name ? I guess the question is: is
there a way to DELETE a group ? (not just change its level)


    David> 	If you just want the group compacted, but not starting
    David> at article number 1, you can just refile into the same
    David> group rather than a new one.


Would that speed up the expiration ? I think it still would leave one
big hole - no ?


    David> 	Oh, and if you use the cache, be careful that
    David> everything works right.  I think I've noticed that renaming
    David> a group doesn't always rename the cache.


_YES_ - it seems I had accidently cached an article in one of my mail
groups - that screws up the whole thing ! It seems that when you
rename your temporary group back to the original name, it finds the
cached article, and incorporates it. This can cause a large hole. I
think this may be considered a bug - I guess renaming a group should
rename its cache.


    David> Anyone have better thoughts on this?  Considering how slow
    David> total-expire currently gets with groups with lots of holes,
    David> maybe a safe compaction routine should be added to Gnus
    David> itself.


Yes Yes Yes. Please Please Please. I spend far too much time
compacting my groups (or alternately waiting for the expiration
process on groups with lots of holes). 


-- 
  The day is short, and the work is great,     |   Aharon Schkolnik
  and the laborers are lazy, and the reward    |   Aharon@Health.Gov.IL
  is great, and the Master of the house is     |
  impatient. - Ethics Of The Fathers Ch. 2     |

Re: Please help me renumber articles !

[St. Suika Fenderson Roberts]

Kai Grossjohann wrote:
>>>>> St Suika Fenderson Roberts writes:

>   Suika> [...] The problem with both of these is that this looses all
>   Suika> of the marks one may have. [...]

> When I move a ticked article it stays ticked.  Doesn't this work for
> you?

It didn't the last time I tried it, and from what I picked up, that is
the way it is supposed to be.  Since I can see reasons for wanting it to
work that way, I haven't done anything more with it, really.  Things are
still far faster than they were with 4.1, so I don't really mind.  I
don't use any of the auto/total-expire things, though.

Thanks,
	Suika
--
		wrobert2@mailhost.tcs.tulane.edu
Whenever I hear anyone arguing for slavery, I feel a strong impulse to see it 
tried on him personally.                -- A. Lincoln
   <a href="http://studentweb.tulane.edu/~wrobert2/">Suika no homepage</a>

Re: Please help me renumber articles !

[Kai Grossjohann]

>>>>> St Suika Fenderson Roberts writes:

  Suika> [...] The problem with both of these is that this looses all
  Suika> of the marks one may have. [...]

When I move a ticked article it stays ticked.  Doesn't this work for
you?

kai
-- 
I like both kinds of music.

Re: Please help me renumber articles !

[St. Suika Fenderson Roberts]

David Moore <dmoore@UCSD.EDU> writes:
> If you want all of the numbers back to 1, you can go into the
> group with `C-u SPC', turn off threading with `T T', sort the articles
> in a reasonable order `C-c C-s C-n' then using `C-u ### B m' refile
> everything into a _new_ group (where ### is large enough to cover all of
> the articles).  Then you can go and rename & remove the original group,
> and then rename the new group to the old name.

> 	If you just want the group compacted, but not starting at
> article number 1, you can just refile into the same group rather than
> a new one.

The problem with both of these is that this looses all of the marks one
may have.  If one is insanely patient one can pull up ones .newsrc.eld
and the folder/directory in buffers, then re-number the marks/mail by
hand as one goes along, although one will then need to modify the active
file and overview files when one is done.

> Anyone have better thoughts on this?  Considering how slow total-expire
> currently gets with groups with lots of holes, maybe a safe compaction
> routine should be added to Gnus itself.

I think that would be a wonderfull idea.  I would probably have it just
find the lowest unused article number, move the lowest higher numbered
article to that number, fix marks, modify the overview file, the active
file if it needs it, repeat.  I've contemplated doing it by hand to a
couple of my folders . . .

	Suika
--
		wrobert2@mailhost.tcs.tulane.edu
I hate quotations.                                     -Ralph Waldo Emerson
  <a href="http://studentweb.tulane.edu/~wrobert2/">Suika no homepage</a>