From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Date: Thu, 25 Apr 2019 18:32:11 +0300 From: Alexander Turenko Subject: Re: [tarantool-patches] Re: [PATCH v3 6/7] Add merger for tuples streams (C part) Message-ID: <20190425153210.kbwzjlq3fdjwogcn@tkn_work_nb> References: <963ad528ad35199943931150956c1d5e2c374c40.1554906327.git.alexander.turenko@tarantool.org> <20190425114343.GM29257@atlas> <20190425133227.mvell3bwcbhzqkyb@tkn_work_nb> <20190425134539.GA10572@atlas> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <20190425134539.GA10572@atlas> To: Konstantin Osipov Cc: tarantool-patches@freelists.org, Vladimir Davydov List-ID: On Thu, Apr 25, 2019 at 04:45:39PM +0300, Konstantin Osipov wrote: > * Alexander Turenko [19/04/25 16:37]: > > merge source or merger's source. I found it neat to have one prefix > > 'merger_' for all related structures and functions. > > What naming scheme other libraries for map/reduce use? Our merger is more like a specific reduce phase when we see on it in context of a general map-reduce framework. It is hard to find an intersection of our names with, say, hadoop. Once in the past we discussed how to better integrate general map-reduce API into net.box, but then decided to implement merger first as the concrete reduce phase w/o attempts to generalize it much. Anyway, I see the word 'input' is quite often in map-reduce libraries documentation. Should we use it instead of 'source'? merge_input or merger_input? WBR, Alexander Turenko.