On 13 Nov 2016 at 8:03, Antony Polukhin wrote:
2016-11-13 4:35 GMT-08:00 Bjorn Reese <breese@mail1.stofanet.dk>:
On 11/09/2016 07:45 PM, Antony Polukhin wrote:
The Boost.Process library is accepted.
Unconditionally?
Conditional acceptance means that the mini-review is required. I can see no issues that require fixing in mini-review: * There's no big unsolvable technical issues at this point and the library author is going to address all the comments before the library would appear in Boost.
The way I've understood these reviews is that if there are unsolvable technical issues with the proposed design, the library is rejected. If there are solvable technical issues commonly agreed by the community, the review manager lists them as conditions for acceptance which as you mention implies a subsequent mini review or at least that the author asks for a quick look by the community to verify those problems have been addressed before the library enters the main Boost distro. If everyone agrees there are technical issues but nobody can reach consensus, it's up to the review manager to guide the community to a consensus in my opinion. I had taken the unconditional acceptance as meaning you felt no one in the community reviews had identified as a consensus problems you felt serious enough to be technical issues - there were only a few i's to be dotted and a few t's to be crossed, otherwise the library was Boost quality. I found that quite surprising personally, I feel the library as proposed as fundamentally flawed as Boost.Iostreams which is a travesty of bad design in my opinion. But others whom I know do know this domain very well did not object as substantially as I did (they also do not object to Iostreams as I do), so I defer to their judgment. Do you think a consensus list of issues needing to be fixed can be drawn up? Perhaps Klemens has been keeping a list he can simply copy and paste in here? Niall -- ned Productions Limited Consulting http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/