Hello Roman,

On 11 Feb 2014, at 20:37 pm, Roman Shaposhnik <[email protected]> wrote:

> Apologies for the late reply -- I am fine with you guys forwarding the
> vote to general@

No worries, Roman, thanks for getting back to us.

> To put my comments in context, here's my only bit of feedback to you:
> mentors are
> volunteers. They are not being payed or otherwise incentivesed to
> review releases
> in cases where it is not immediately obvious how to do a certain bit of 
> release
> verification. This is a bit of accommodation that you may consider useful to 
> get
> votes quicker. On the other hand, like you mentioned in the case of
> hashsums -- you
> seem to be following some kind of documentation. The fact that to this
> day, I don't know
> of a tool that would let me automate that check doesn't mean other members
> of the incubator community wouldn't be more creative.

This was exactly the feedback I provided to the community: please provide some 
kind of script that allows everybody to quickly validate such hashsums. There 
is no need for everybody to validate them by hand, I totally agree there.

Greetings, Marcel


> The only way to find out is to try the vote and see what happens.
> 
> Thanks,
> Roman.
> 
> On Tue, Feb 11, 2014 at 4:09 AM, Marcel Offermans
> <[email protected]> wrote:
>> I'm in favor for forwarding the vote, we need someone else to look at it, or 
>> Roman to answer to the responses given here. I tried pinging Roman last 
>> week. I think he must be very busy at the moment, so let's try to move ahead!
>> 
>> Greetings, Marcel
>> 
>> 
>> On 07 Feb 2014, at 9:52 am, Pepijn Noltes <[email protected]> wrote:
>> 
>>> Hi All,
>>> 
>>> I would like to propose to forward the release vote to the incubator
>>> mailing list We got two +1 binding vote and -1 vote, so we are one binding
>>> +1 short.
>>> There is still some comments from Roman, but I think there is always some
>>> room for improvement and again there is no  -1 vote.
>>> 
>>> I would like to known if any mentors see a problem with this approach. I
>>> don't want to step on anybody's toes, but would like to push the release
>>> forward.
>>> 
>>> Greetings,
>>> Pepijn
>>> 
>>> 
>>> 
>>> 
>>> On Tue, Jan 28, 2014 at 8:08 PM, Pepijn Noltes 
>>> <[email protected]>wrote:
>>> 
>>>> Hi Roman,
>>>> 
>>>> Could you have a look at the comments of Alexander? I known I'm pushing a
>>>> bit, but we are hoping to get the release ready :).
>>>> 
>>>> 
>>>> On Fri, Jan 24, 2014 at 12:11 PM, Alexander Broekhuis <
>>>> [email protected]> wrote:
>>>> 
>>>>> Hi Roman,
>>>>> 
>>>>> See my remarks inline below. I hope this gives you enough confidence to
>>>>> sign this release off.
>>>>> 
>>>>> 2014/1/24 Roman Shaposhnik <[email protected]>
>>>>> 
>>>>>> I know that some of the items are nits, but if we are to
>>>>>> re-cut an RC for Boost reasons -- I'd suggest we may
>>>>>> as well take care of them
>>>>>> 
>>>>> 
>>>>> The way I read [2], there is no need to add anything to the notice file at
>>>>> all. All third party sources we use have a header with the respective
>>>>> license information. At [2] it is even explicitly mentioned not to add
>>>>> anything unless legally required.
>>>>> 
>>>>> "Do not add anything to NOTICE which is not legally required."
>>>>> 
>>>>> So I don't see a reason why a new release is needed for Boost.
>>>>> 
>>>>>> 
>>>>>>> The checksum has been created with the command mentioned on the Apache
>>>>>>> Signing Releases page [1]. I don't see what is wrong with this.
>>>>>> 
>>>>>> There was an old discussion on that some time ago. Basically
>>>>>> the problem boils down to a fact that I can't verify it with shasum(1)
>>>>>> and thus can't sign off on it.
>>>>>> 
>>>>> 
>>>>> This was indeed an old discussion, but there has never been reached a
>>>>> consensus, and as stated before, I've explicitly used the method described
>>>>> on the Apache pages, which uses the gpg tooling to verify a checksum.
>>>>> Instead of using shasum, you can simply use gpg --print-md "filename".
>>>>> 
>>>>> If all I do is follow the official Apache document then what am I doing
>>>>> wrong?
>>>>> 
>>>>> I've had some discussion with Marcel on this topic as well, and in some
>>>>> other project where Marcel is involved, they use a script to compare the
>>>>> checksums. A similar solution might be implemented for Celix as well, I
>>>>> don't mind adding this to the backlog.
>>>>> 
>>>>> 
>>>>>> 
>>>>>>>>  * it would be nice to have version embedded into the name of the
>>>>> top
>>>>>>>>    level dir inside of the tarball
>>>>>>>> 
>>>>>>> 
>>>>>>> We have decided to leave it out since else there would always be an
>>>>> issue
>>>>>>> with the BUILDING instructions and the default directory. This was a
>>>>>> remark
>>>>>>> by someone on the first (0.0.1) release where we did have the version
>>>>> in
>>>>>>> the top-level directory.
>>>>>> 
>>>>>> Hm. I'm just curious -- was there a thread on this one?
>>>>>> 
>>>>> 
>>>>> This was a remark made by Marcel on our first release. See [3] for his
>>>>> message/the release thread.
>>>>> 
>>>>> 
>>>>> 
>>>>>> 
>>>>>>>>  * boost license is missing in NOTICES
>>>>>>>> 
>>>>>>> 
>>>>>>> Why should the boost license be in the NOTICES file? There have been a
>>>>>> lot
>>>>>>> of discussions on this file, and my understanding always has been that
>>>>>> if a
>>>>>>> license is in a header it is not needed to add it to the NOTICES file.
>>>>>> 
>>>>>> I honestly don't recall this. Care to point a thread?
>>>>>> 
>>>>> 
>>>>> I can't find the thread, but [2] gives a good explanation.
>>>>> 
>>>>> 
>>>>>> 
>>>>>> Thanks,
>>>>>> Roman.
>>>>>> 
>>>>> 
>>>>> 
>>>>> [1]: http://www.apache.org/dev/release-signing#sha-checksum
>>>>> [2]: http://www.apache.org/dev/licensing-howto.html#mod-notice
>>>>> [3]: http://incubator.markmail.org/thread/ot7cwepmcusdblqs
>>>>> 
>>>>> --
>>>>> Met vriendelijke groet,
>>>>> 
>>>>> Alexander Broekhuis
>>>>> 
>>>> 
>>>> 
>> 

Reply via email to