[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: New proposed-updates diff: spamassassin 3.3.1-1.1



On Tue, 2013-03-12 at 13:08 -0400, Noah Meyerhans wrote:
> On Tue, Mar 12, 2013 at 04:14:37PM +0000, Adam D. Barratt wrote:
> > On 12.03.2013 16:00, Debian Queue Viewer wrote:
> > >+spamassassin (3.3.1-1.1) stable; urgency=high
> > 
> > "-1.1" seems a slightly odd version number for a maintainer upload.
> 
> I went with the version number generated by dch's --security option,
> since it doesn't have an option to generate an s-p-u version and
> security seemed logically closest. Happy to re-upload with something
> else if it matters.

Ah, I see. "--security" was designed for use by the security team, so
uses NMU versions (I'm not convinced they use it, but that's another
matter).

If the upload hadn't been made yet then I'd have suggested 3.3.1-1
+squeeze1, but I'm not sure it's worth a reject / re-upload cycle just
for that.

> > >+  * Fix the RCVD_ILLEGAL_IP rule to no longer consider addresses
> > >+    in the 5.0.0.0/8 range as invalid. (Closes: #696144)
> > >+  * Remove tests referencing the NJABL blacklist, which shut
> > >+    down as of 1 March, 2013. (Closes: #702839)
> > 
> > Thankfully this one isn't as urgent as most other RBL shutdowns, as
> > the NJABL folks opted to simply empty the zones, thus returning no
> > match for all lookups, rather than the more common approach of
> > matching everything.
> 
> Yes, although it's not clear how much longer the servers will continue
> to answer queries at all. At some point, the servers will shut down and
> queries will time out.

Indeed. That's still better than listing the world though.

> > One question: what happens if the user is running sa-update, either
> > via the supplied cron integration or some other method? Do the rules
> > simply re-appear until upstream remove them?
> 
> If upstream was still publishing njabl rules, then the ones received via
> sa-update would have priority over the ones from the package.  However,
> it looks like they've already published updated rules via the sa-update
> channels, so people using sa-update will not require a package update.

Ah, it was still being pushed in the updates last time I looked;
apologies for not having checked that was still the case.

Regards,

Adam


Reply to: