review of new tcpcrypt port...

[Bryan Drewery]

On 11/12/2014 7:11 PM, John-Mark Gurney wrote:
> Bryan Drewery wrote this message on Wed, Nov 12, 2014 at 18:36 -0600:
>> On 11/12/2014 6:16 PM, John-Mark Gurney wrote:
>>> John-Mark Gurney wrote this message on Sat, Nov 08, 2014 at 02:07 -0800:
>>>> I'd like someone to review the attached port for me to commit...  It
>>>> contains a daemon to run on divert sockets to implement the tcpcrypt
>>>> protocol.  I have tested this port w/ HEAD and it works fine w/ the
>>>> attached patch...
>>>>
>>>> The included patch has been submitted upstream and committed, but a
>>>> new release has not yet been released.
>>>>
>>>> portlint -A looks fine.
>>>
>>> As I haven't received a review, in a couple more days, probably Friday,
>>> I'll commit the port...
>>>
>>> Thanks.
>>>
>>
>> Can you include the security/Makefile bit in the patch too please?
> 
> Hmm... Thought I did, but I clearly did not...  Fixed...  I've attached
> the latest version of the port..
> 
>> Have you ran it through poudriere? At the very least 'env DEVELOPER=1
>> make stage' and 'make check-plist' are required. Portlint is not
>> sufficient, nor can it even be trusted in all cases.
> 
> I didn't do poudriere, but I did all the building w/ DEVELOPER=1
> set, which did fix a few things..

Well 'make check-plist' must be ran manually. I think I'll enable it
automatically soon. I've avoided it because it would break Redports even
more.

> 
> Looks like port test from porttools as documented at:
> https://www.freebsd.org/doc/en/books/porters-handbook/testing-porttools.html
> 
> Is wrong.. you need to be in the port directory for it to work...
> 

Ugh. That's a bad page. I'm working on a blog article for this case to
get out soon. I may get it into the Journal as well.

>> If you check those and get back to me I'll approve it.
> 
> Thanks...
> 
>> Usually you'd need a port commit bit to commit anything, or an approval
>> (which is different than a review). I hate our process, but that's how
>> it is. Poke people in #bsdports on EFNet, or me on IRC, for
>> approvals/reviews in the future.
> 
> Well, I'm fine w/ someone else committing the port, just trying to save
> someone the work, and trying to get it into the tree in a timely
> manner..  I viewed the review as approval...  Guess I should have made
> it more clear in my original request...
> 
> btw, the included patch has been upstreamed, but they haven't tagged a
> new rc yet...
> 

3 comments:

1. No need for commented line in the Makefile
2. Only 1 WWW for pkg-descr
3. If 'make check-plist' passes then I approve it and you can commit
yourself. I am mentoring several src committers for ports and getting
them bits after enough work, so just ping me whenever you have more.

Thanks,
Bryan

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 488 bytes
Desc: OpenPGP digital signature
URL: <http://lists.freebsd.org/pipermail/freebsd-ports/attachments/20141112/a0dbdd61/attachment.sig>