should not store indexes in ~

Bug #77768 reported by Jerome Haltom
6
Affects Status Importance Assigned to Milestone
Beagle
Won't Fix
Medium
beagle (Ubuntu)
Invalid
Undecided
DaveThacker

Bug Description

Binary package hint: beagle

By default Beagle should not store indexes in a user's home directory. In *many* cases, home directories are not local file systems, but either NFS or AFS mounted. The data Beagle uses (including it's own socket) is cache data. It can be destroyed without harm. The data is also system specific.

It would be better of Beagle stored it's data in a directory like /var/cache/beagle/indexes/$username.$rnd using basic temporary directory name allocation methods: find an existing directory $uid.*, check owner, if not create a new one. This would allow beagle to maintain it's own data per-system the user logs into, as well as making it part of the system policy that the data can actually be deleted.

It would also allow beagle to function out of the box on remotely mounted ~ directories.

Revision history for this message
DaveThacker (dthacker9) wrote :
Changed in beagle:
assignee: nobody → dthacker9
status: Unconfirmed → Confirmed
Revision history for this message
Kevin Kubasik (kkubasik) wrote :

Please don't confirm your own bugs, even if your 100% certain, its generally a better idea to let someone else do the confirmation, or for the package maintainer to do it, since its often a means of keeping their bugs orgainized.

After a short chat w/ dBera upstream, this doesn't really seem like a bug to me, should their be massive public consensous or demand, we can look into it, but there are already environmental variables (like BEAGLE_HOME and BEAGLE_STORAGE) that alow individual users to customize index locations.

Changed in beagle:
status: Confirmed → Rejected
Revision history for this message
DaveThacker (dthacker9) wrote : Re: [Bug 77768] Re: should not store indexes in ~

On Wednesday 03 January 2007 11:19, Kevin Kubasik wrote:
> Please don't confirm your own bugs, even if your 100% certain, its
> generally a better idea to let someone else do the confirmation, or for
> the package maintainer to do it, since its often a means of keeping
> their bugs orgainized.

Sorry, I'm new to triaging. Thanks for the guidance. I don't think that is
stated in the triage guide. Should it be added?
>
> After a short chat w/ dBera upstream, this doesn't really seem like a
> bug to me, should their be massive public consensous or demand, we can
> look into it, but there are already environmental variables (like
> BEAGLE_HOME and BEAGLE_STORAGE) that alow individual users to customize
> index locations.
>

It looked like a feature request to me, and I was about to reject, but I got a
second opinion in IRC that said it was a reasonable request and suggested I
file it upstream. To improve the quality of my triage efforts, could you
tell me:

1) Are *any* feature requests ever opened upstream as I did today?
2) If so, what criteria should I use to make that decision.

Again, thank you for the courteous feedback.

Dave

Changed in beagle:
status: Unknown → Rejected
Changed in beagle:
status: Invalid → Won't Fix
Changed in beagle:
importance: Unknown → Medium
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.