190285 - want way to cancel compaction

Status: NEW

(MailNews Core :: Backend, enhancement)

Description

[Mark]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.2.1) Gecko/20021130
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.2.1) Gecko/20021130

This is related to bug #90452 but a slightly different request.

I am using mozilla 1.2.1 on Windows 2000 on a somewhat underpowered laptop and
slow disk (P3 450Mhz) and do not encounter the exact problem of bug #90452, but
this is related.  I am using POP account exclusively - 5 of them.  When I answer
yes to the dialog asking me if I want to compact my folders, it starts
compacting everything.  I have a lot of email so it takes perhaps half an hour.
 (Guessing, as I havent had the patience to let it finish except right before I
leave to do something else time consuming.)  As described in bug #90452, it
takes up nearly all cpu and monopolizes disk access. (very annoying) I can just
barely still use the browser and email, but doesnt actually hang. It was just
very unresponsive.

My issue is that once this compaction starts, I can find no safe way to stop it
except letting it run to completion.  So I would like to the "Compact Folders"
menu entry change to "Stop Compaction" while compaction is running.  I ended up
having to exit mail in order to make it stop, and cross my fingers that it
wouldnt corrupt the folder it was working on.  Intuitively, I also tried
clicking on the progress bar at the bottom right of the mail window to be able
to stop compaction.  It didnt work, but that would be another place to allow the
user to stop the compaction in addition to the menu.  I would prefer the menu
entry if you can only implement one.  Apologies if Ive missed a way to stop it,
but in any case, it doesnt make sense to have the menu entry stay the same when
folder compaction is happening.


Reproducible: Always

Steps to Reproduce:
1.
2.
3.

Comment 1

[Boris Zbarsky [:bzbarsky]]

database

Comment 2

[David :Bienvenu]

not a mail db issue.

Comment 3

[Wayne Mery (:wsmwk)]

bienvenu, can you foresee any circumstance under which we would need this or want it, should a patch materialize?

Comment 4

[David :Bienvenu]

I'd rather spend effort on using a storage system that didn't require compaction, but in general, yeah, we should allow safe stopping of actions that can take a long time.