From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Date: Fri, 17 May 2019 18:29:52 +0300 From: Konstantin Osipov Subject: Re: [tarantool-patches] Re: [PATCH v1 2/3] schema: extend _func space to persist lua functions Message-ID: <20190517152952.GB20341@atlas> References: <0162c0c953de0f34121e5deb3ebe54aa6ed752fa.1557839195.git.kshcherbatov@tarantool.org> <20190514182103.GA3801@atlas> <92b6b7a0-7d8d-c239-6a40-b72f03f17e06@tarantool.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <92b6b7a0-7d8d-c239-6a40-b72f03f17e06@tarantool.org> To: Kirill Shcherbatov Cc: tarantool-patches@freelists.org, vdavydov.dev@gmail.com List-ID: * Kirill Shcherbatov [19/05/17 15:37]: > > While we are at it, I would also add 'type' - function or > > procedure, and other related to SQL procedures fields. > > > > See for example here: > > > > https://mariadb.com/kb/en/library/mysqlproc-table/ > > (not all of these fields are relevant, and perhaps we need an > > 'options' map, to make the whole thing easily extensible in the > > future, but it's better to put as many fields as possible into > > columns, not options). > > Thank you for this reference. > I think we'll need 'is_deterministic' flag soon enough (in scope of > functional indexes), so I've introduced it. > To make the process of introducing such options extensible, > @locker and I decided to put 'is_deterministic' and other new > flags in the 'opts' map field. I've reused existent opts_decode > mechanism in scope of this patch. I wrote earlier: it's better to put as many fields as possible to columns, not options. If you disagree, please put forward your arguments first. I don't mind options, but htey are hard to use, so should be avoided if possible. -- Konstantin Osipov, Moscow, Russia, +7 903 626 22 32