- rpmlint checks return:

W: clamav-exim no-url-tag
W: clamav-exim no-documentation
E: clamav-exim incoherent-logrotate-file /etc/logrotate.d/clamd.exim
E: clamav-exim non-standard-uid /var/run/clamd.exim clamexim
E: clamav-exim non-standard-gid /var/run/clamd.exim exim
E: clamav-exim non-standard-dir-perm /var/run/clamd.exim 0750
E: clamav-exim non-standard-gid /var/log/clamd.exim exim
E: clamav-exim non-root-group-log-file /var/log/clamd.exim exim
W: clamav-exim dangerous-command-in-%post chmod
^ If you really need a log file, perhaps create in %install?

E: clamav-exim init-script-name-with-dot /etc/rc.d/init.d/clamd.exim
E: clamav-exim no-status-entry /etc/rc.d/init.d/clamd.exim
W: clamav-exim no-reload-entry /etc/rc.d/init.d/clamd.exim
E: clamav-exim subsys-not-used /etc/rc.d/init.d/clamd.exim
W: clamav-exim incoherent-init-script-name clamd.exim


- license (GPL) OK, need text in %doc.

?
- No upstream, is this okay?
- macro use not consistent between %var and %{var}

Good:

- package meets naming guidelines
- package meets packaging guidelines
- spec file legible, in am. english
- package compiles on devel (x86)
- no missing BR
- no unnecessary BR
- no locales
- not relocatable
- owns all directories that it creates
- no duplicate files
- %clean ok
- code, not content
- no need for -docs
- nothing in %doc affects runtime
- no need for .desktop file

Thanks very much for the review.  I should have mentioned that I knew rpmlint
barfed lots, mostly due to this being a very weird package.

> W: clamav-exim no-url-tag

There isn't a relevant URL.

> W: clamav-exim no-documentation

There are no docs, though maybe I should write a short README.

> E: clamav-exim incoherent-logrotate-file /etc/logrotate.d/clamd.exim

Is this the filename (i.e. clamd.exim) that it's complaining about? If so, I
know it's not the same as the package but the convention wasn't invented by me
but does make sense.

> E: clamav-exim non-standard-uid /var/run/clamd.exim clamexim
> E: clamav-exim non-standard-gid /var/run/clamd.exim exim
> E: clamav-exim non-standard-dir-perm /var/run/clamd.exim 0750
> E: clamav-exim non-standard-gid /var/log/clamd.exim exim

I don't really understand these; the uid/gid/perms are intended and correct.

> E: clamav-exim non-root-group-log-file /var/log/clamd.exim exim

So, clamexim.root then? Howveer, users in the "exim" group might conceivably
want read-access to the logs.

> W: clamav-exim dangerous-command-in-%post chmod
> ^ If you really need a log file, perhaps create in %install?

I could do. However, this stuff all came from the original spec fragments that
were checked into the clamav package by David Woodhouse (dwmw2@redhat). Given
that David is more experienced than me I was deferring to his judgement. Any
other thoughts from others would be welcome.

> E: clamav-exim init-script-name-with-dot /etc/rc.d/init.d/clamd.exim

See above discussion about the clamd.exim convention.

> E: clamav-exim no-status-entry /etc/rc.d/init.d/clamd.exim
> W: clamav-exim no-reload-entry /etc/rc.d/init.d/clamd.exim
> E: clamav-exim subsys-not-used /etc/rc.d/init.d/clamd.exim

These are all bogus considering that the init script is essentially just a
pointer to the main clamav one.

> W: clamav-exim incoherent-init-script-name clamd.exim

See above discusion about the clamd.exim convention.

> license (GPL) OK, need text in %doc.

Well, there is no text "upstream" (because there isn't an upstream) so this
isn't essential.

> No upstream, is this okay?

Well, there's no upstream to have, this package is really not much more than
metadata for packaging consistency

> macro use not consistent between %var and %{var}

I'll review, though I didn't think there was a problem with %x and %{x} where
used for clarity.

Thanks again.

(In reply to comment #2)
> Thanks very much for the review.  I should have mentioned that I knew rpmlint
> barfed lots, mostly due to this being a very weird package.

Yeah, we'll have to muddle through... :-)

> > W: clamav-exim no-documentation
> 
> There are no docs, though maybe I should write a short README.

I think that would be good - explicitly stating the license as well as a copy of
the GPL.
 
> > E: clamav-exim incoherent-logrotate-file /etc/logrotate.d/clamd.exim
> 
> Is this the filename (i.e. clamd.exim) that it's complaining about? If so, I
> know it's not the same as the package but the convention wasn't invented by me
> but does make sense.

I think you need to use clamd-exim here and elsewhere (init.d).  "." usually
implies a meaningful suffix.

> > E: clamav-exim non-standard-uid /var/run/clamd.exim clamexim
> > E: clamav-exim non-standard-gid /var/run/clamd.exim exim
> > E: clamav-exim non-standard-dir-perm /var/run/clamd.exim 0750
> > E: clamav-exim non-standard-gid /var/log/clamd.exim exim
> 
> I don't really understand these; the uid/gid/perms are intended and correct.
> 

Yeah, looks like this is what clamav does.

> > E: clamav-exim non-root-group-log-file /var/log/clamd.exim exim
> 
> So, clamexim.root then? Howveer, users in the "exim" group might conceivably
> want read-access to the logs.
 
Well, others seem to do this too, and seems sensible.

> > W: clamav-exim dangerous-command-in-%post chmod
> > ^ If you really need a log file, perhaps create in %install?
> 
> I could do. However, this stuff all came from the original spec fragments that
> were checked into the clamav package by David Woodhouse (dwmw2@redhat). Given
> that David is more experienced than me I was deferring to his judgement. Any
> other thoughts from others would be welcome.

Okay, looks like this is what clamav spec does too.
 
> > E: clamav-exim init-script-name-with-dot /etc/rc.d/init.d/clamd.exim
> 
> See above discussion about the clamd.exim convention.

Yeah, should be clamd-exim  

> > E: clamav-exim no-status-entry /etc/rc.d/init.d/clamd.exim
> > W: clamav-exim no-reload-entry /etc/rc.d/init.d/clamd.exim
> > E: clamav-exim subsys-not-used /etc/rc.d/init.d/clamd.exim
> 
> These are all bogus considering that the init script is essentially just a
> pointer to the main clamav one.
> 
> > W: clamav-exim incoherent-init-script-name clamd.exim
> 
> See above discusion about the clamd.exim convention.
> 
> > license (GPL) OK, need text in %doc.
> 
> Well, there is no text "upstream" (because there isn't an upstream) so this
> isn't essential.
> 
> > No upstream, is this okay?
> 
> Well, there's no upstream to have, this package is really not much more than
> metadata for packaging consistency

I'll ask some questions on the list.

OK, I've looked into this. The /usr/share/clamav/clamd-wrapper script, provided
by clamav-server, hardcodes "clamd.XXX" as the service name in several places.
So either we stick with clamd.exim as the service name or have to file a bug
with Enrico and get him to patch the main ClamAV package.

I presume there was some reason why clamd.XXX was chosen.

Also, this package should possibly be renamed clamav-server-exim.

I think you should bail on using clamd-wrapper.  Use it as a base for a new
clamav-exim (or clamav-server-exim) script, but change it so you can use the
expected naming conventions for the init and logrotate scripts.

If "clamd.SERVICE" truly is wrong, I think it should be fixed "upstream" in
clamd-wrapper rather than just ignoring clamd-wrapper. I've emailed the Enrico
offline to solicit his comments.

(In reply to comment #6)
> If "clamd.SERVICE" truly is wrong, I think it should be fixed "upstream" in
> clamd-wrapper rather than just ignoring clamd-wrapper. I've emailed the Enrico
> offline to solicit his comments.

See also bug 190492

I've actually already added this support in a subpackage of exim: exim-clamav.

I'll go ahead and close this out.