permissions of detached / saved attachments
Nicolas Thierry-Mieg
I reported a bug back in december 2009:
https://bugzilla.mozilla.org/show_bug.cgi?id=533976
Excerpt:
Email attachments saved (or detached) using the context menu in the
message view are saved with restrictive permissions (0600).
Instead they should be 0666 & ~umask : that's the way they were until
2.0, and also the way they are if you "open" the attachment and then
save it.
This month Chiaki ISHIKAWA worked on this bug, and came up with a patch
that solves the issue. This patch was checked in today, so whenever
seamonkey becomes based on today's latest comm-central, all will be fixed!
One question I have is: when would that be? I did look around in the
mercurial repo and try to figure this out (or a least get an
understanding of the relationship between comm-central and SM versions),
but there are many tags and branches, and it's quite confusing to me.
In any case I was eager to use a fixed version, and I regularly rebuild
SM for my own use (as an x86_64 rpm for centos 5). So I decided to
include the patch in my build, until it becomes part of the official SM.
The original patch doesn't apply cleanly on the SM 2.0.5 source, I had
to adapt it a bit. I then built 2.0.5 with this patch, and can confirm
that it solves the issue.
It's a very trivial patch, easy to review and convince oneself that it
doesn't break anything. I'm putting it over here (it should be there in
an hour when the webserver syncs):
http://membres-timc.imag.fr/Nicolas.Thierry-Mieg/seamonkey_wrongSavedPerms_bz533976.diff
I hope it can be of use to others, and maybe help to get it included in
some not-too-distant release of SM!
Regards,
Nicolas
Jens Hatlak
> This month Chiaki ISHIKAWA worked on this bug, and came up with a patch
> that solves the issue. This patch was checked in today, so whenever
> seamonkey becomes based on today's latest comm-central, all will be fixed!
> One question I have is: when would that be?
With the next SeaMonkey trunk nightly build, and consequently both
SeaMonkey 2.1 betas and the SeaMonkey 2.1 release.
> In any case I was eager to use a fixed version, and I regularly rebuild
> SM for my own use (as an x86_64 rpm for centos 5). So I decided to
> include the patch in my build, until it becomes part of the official SM.
> The original patch doesn't apply cleanly on the SM 2.0.5 source, I had
> to adapt it a bit. I then built 2.0.5 with this patch, and can confirm
> that it solves the issue.
> It's a very trivial patch, easy to review and convince oneself that it
> doesn't break anything.
If you want to have it in SM 2.0.next, you could ask in bug 533976
whether the fix will be back-ported to the stable release branch
(comm-1.9.1). If your Bugzilla permissions permit, just set
approval1.9.1.11 or approval-seamonkey2.0.6 to "?" on the attachment's
Details screen. Otherwise just leave a comment on the bug.
Please note that if someone asks or grants approval for 1.9.2.x that
won't have any effect on SeaMonkey, only Thunderbird 3.1.
HTH
Jens
--
Jens Hatlak <http://jens.hatlak.de/>
SeaMonkey Trunk Tracker <http://smtt.blogspot.com/>
Neil
> Nicolas Thierry-Mieg wrote:
>
>> The original patch doesn't apply cleanly on the SM 2.0.5 source, I
>> had to adapt it a bit. I then built 2.0.5 with this patch, and can
>> confirm that it solves the issue.
>> It's a very trivial patch, easy to review and convince oneself that
>> it doesn't break anything.
>
> If you want to have it in SM 2.0.next, you could ask in bug 533976
> whether the fix will be back-ported to the stable release branch
> (comm-1.9.1). If your Bugzilla permissions permit, just set
> approval1.9.1.11 or approval-seamonkey2.0.6 to "?" on the attachment's
> Details screen. Otherwise just leave a comment on the bug.
>
> Please note that if someone asks or grants approval for 1.9.2.x that
> won't have any effect on SeaMonkey, only Thunderbird 3.1.
Huh? 1.9.1/2 makes no sense, since the patch applies to c-c, not m-c.
But you could either request approval-thunderbird3.0.6? on the patch or
maybe status-thunderbird3.0? on the bug. (Requesting SeaMonkey approvals
is pointless since the bug is in shared code.)
--
Warning: May contain traces of nuts.
Jens Hatlak
> Jens Hatlak wrote:
>> If you want to have it in SM 2.0.next, you could ask in bug 533976
>> whether the fix will be back-ported to the stable release branch
>> (comm-1.9.1). If your Bugzilla permissions permit, just set
>> approval1.9.1.11 or approval-seamonkey2.0.6 to "?" on the attachment's
>> Details screen. Otherwise just leave a comment on the bug.
>>
>> Please note that if someone asks or grants approval for 1.9.2.x that
>> won't have any effect on SeaMonkey, only Thunderbird 3.1.
>
> Huh? 1.9.1/2 makes no sense, since the patch applies to c-c, not m-c.
> But you could either request approval-thunderbird3.0.6? on the patch or
> maybe status-thunderbird3.0? on the bug. (Requesting SeaMonkey approvals
> is pointless since the bug is in shared code.)
I know it's shared code, that's why I thought it would make sense to
understand the Gecko flags there as the version that both SM and TB are
based on. But then you're right, the Gecko version is the m-c version.
But then, from a logical point of view, requesting SM approval would
make just as much sense as requesting TB approval since, as you said,
it's shared code, so one would think that both sides need to agree. But
I guess logic doesn't apply there, since your explanation suggests that
TB reviewers can deny changes to shared code but SM reviewers can't.
Anyway, I was just trying to be helpful. I'd just request any flag that
looked appropriate to me and trust that more knowledgeable people would
correct it. You might assume that I know better, but what about people
like Nicolas?
Greetings,
Nicolas Thierry-Mieg
Thank you both for your replies.
Knowing the patch will be in SM 2.1 is already great news!
Following your advice I have posted a comment on the bugzilla entry.
Regards,
Nicolas