After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 714094 - Optional downloading of attachments (or, don't automatically store attachments in filesystem)
Optional downloading of attachments (or, don't automatically store attachment...
Status: RESOLVED OBSOLETE
Product: geary
Classification: Other
Component: engine
master
Other All
: High normal
: ---
Assigned To: Geary Maintainers
Geary Maintainers
Depends on: 776309
Blocks: 714728 766581
 
 
Reported: 2012-11-19 07:43 UTC by Jim Nelson
Modified: 2019-06-02 10:09 UTC
See Also:
GNOME target: ---
GNOME version: ---



Description Charles Lindsay 2013-11-21 20:24:38 UTC


---- Reported by jim@yorba.org 2012-11-19 11:43:00 -0800 ----

Original Redmine bug id: 6086
Original URL: http://redmine.yorba.org/issues/6086
Searchable id: yorba-bug-6086
Original author: Jim Nelson
Original description:

Although there exist another ticket to deal with downloading attachments in
the background (#4462), at least one user has requested no automatic
downloading of attachment(s), only manual (much like displaying images in an
email). This could be something configured by attachment size so that small
attachments are immediately downloaded.

Requested here: https://bugs.launchpad.net/geary/+bug/1080177

Related issues:
related to geary - Feature #4462: fetch messages incrementally (Fixed)
related to geary - 6487: Write attachments outside of database
transaction (Open)
related to geary - 6512: attachment filename not verified before writing (Open)
related to geary - Feature #6791: Make attachments pretend to live in /tmp (Open)



---- Additional Comments From geary-maint@gnome.bugs 2013-04-03 18:11:00 -0700 ----

### History

####

#1

Updated by Jim Nelson about 1 year ago

  * **Tracker** changed from _Bug_ to _Feature_

####

#2

Updated by Jim Nelson 10 months ago

  * **Category** changed from _engine_ to _attachments_

####

#3

Updated by Jim Nelson 10 months ago

  * **Target version** changed from _0.3.0_ to _0.4.0_

####

#4

Updated by Charles Lindsay 9 months ago

I'd like to second the idea of not downloading and saving attachments without
user intervention. The current situation where we blindly download and store
anything sent our way is a security risk.

I don't think this should be configurable based on attachment size. The bottom
line is, unless I trust a file, I don't want it on my hard drive.

####

#5

Updated by Jim Nelson 8 months ago

  * **Target version** changed from _0.4.0_ to _0.5.0_



--- Bug imported by chaz@yorba.org 2013-11-21 20:24 UTC  ---

This bug was previously known as _bug_ 6086 at http://redmine.yorba.org/show_bug.cgi?id=6086

Unknown milestone "unknown in product geary. 
   Setting to default milestone for this product, "---".
Setting qa contact to the default for this product.
   This bug either had no qa contact or an invalid one.
Resolution set on an open status.
   Dropping resolution 

Comment 1 Britt Yazel 2016-05-17 17:39:58 UTC
I agree 100%

The way geary works now (from my understanding) is that it downloads all messages and attachments locally from whatever timeframe the user sets (2 weeks, month, etc).

I tend to have mine load everything, irregardless of time, so I can search through past emails. However, this leads to geary taking up many gigabytes of storage space, most of which is taken up by old email attachments. For which the grand majority are a waste of space.

I find this a security risk, a waste of space, and a waste of bandwidth. Can we not only store the attachments that a user explicitly says to download (as is how nearly all other email clients work)? Further, I suggest that we only hold onto attachments for a fixed window of time before we automatically clean them out, perhaps 1-2 weeks from the time of download. After which point, the user can just download the attachment from the email of choice again.
Comment 2 Bill Thornton 2016-06-20 15:38:05 UTC
This is definitely a security risk. I was shocked to discover that Geary is downloading ALL attachments, including those in the Spam folder. Unfortunately, I will not be able to use Geary with this issue present.
Comment 3 Michael Gratton 2016-06-21 00:56:09 UTC
Britt & Bill,

I agree that making lazy downloading of attachments would be highly desirable for some users to conserve storage space and minimise download volume. However I don't see where the security risk is.

My understanding is that since that unless an infected file is executed or interpreted in some way then any malware it contains cannot be executed. Since Geary stores attachments in a hidden, implementation-specific directory, they will not be found by desktop services like Tracker and will not be otherwise triggered without user intervention. Thus the only way that an infected file will be opened and hence the malware triggered is through user explicitly doing so, and then it doesn't make any difference if the attachment is already on the filesystem or has to be downloaded first.

I readily acknowledge there may be specific attacks that I am not aware of that makes Geary's current attachment handling a risk, however since security risks are very important concerns, making vague claims about them isn't very useful.

So, what's the specific details of the attack vector that constitutes a security risk here?
Comment 4 Britt Yazel 2016-06-27 21:30:41 UTC
Ok, I may have been in error when I claimed it to be a security risk. The paranoid, tin foil hat wearer in me always feels a bit anxious when things find their way on to my computer without my explicit permission. Notably being spam email attachments, etc.

While it may not pose an immediate risk unless it is executed, it still makes me feel uncomfortable that much of my precious ssd space is taken up by wasted attachements, in the order of tens of gigabytes.
Comment 5 Britt Yazel 2016-06-27 23:22:04 UTC
Fair enough, thanks for the information. Hopefully Geary will get the support it needs now that it has found a new maintainer anyway, and I will be able to just use it instead of Pantheon-Mail
Comment 6 Britt Yazel 2016-06-27 23:23:04 UTC
Whoops, please ignore my comment #5, I mixed up my responses in my chrome tabs. What I meant to say was:

I did lay out a couple observations and suggestions in general for cleaning up mail (and attachments) over time in bug 766581. Please see that one as it may pertain to how to address some of these issues.
Comment 7 Michael Gratton 2016-06-28 02:07:03 UTC
(In reply to Britt Yazel from comment #4)
> Ok, I may have been in error when I claimed it to be a security risk. The
> paranoid, tin foil hat wearer in me always feels a bit anxious when things
> find their way on to my computer without my explicit permission. Notably
> being spam email attachments, etc.

No problem, that's understandable. Apologies if I came off heavy handed before.

> While it may not pose an immediate risk unless it is executed, it still
> makes me feel uncomfortable that much of my precious ssd space is taken up
> by wasted attachements, in the order of tens of gigabytes.

Yep, I totally agree. IMAP supports this, so we would need to see what changes are needed in Geary for support it.
Comment 8 Kristian Rink 2016-09-19 06:16:48 UTC
Adding myself to this. We receive a vast load of mails including *large* binary files in our professional environment, and most of them I neither need nor want on my local device because they're just supposed to go into some automated processing pipeline.

As an additional idea here, AquaMail on Android offers the feature to automatically *just* download message headers in order to cut down both storage space and bandwidth and still be able to provide search for recepients, subjects or senders or navigate communication threads. Maybe this would be an interesting idea for geary too.
Comment 9 Michael Gratton 2019-06-02 10:09:34 UTC
Closing this in favour of https://gitlab.gnome.org/GNOME/geary/issues/453