From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-lf1-f68.google.com (mail-lf1-f68.google.com [209.85.167.68]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by dev.tarantool.org (Postfix) with ESMTPS id 612BF46970F for ; Thu, 28 Nov 2019 15:31:22 +0300 (MSK) Received: by mail-lf1-f68.google.com with SMTP id a17so19904733lfi.13 for ; Thu, 28 Nov 2019 04:31:22 -0800 (PST) Date: Thu, 28 Nov 2019 15:31:20 +0300 From: Konstantin Osipov Message-ID: <20191128123120.GB29714@atlas> References: <20191127105150.GA7232@atlas> <20191127110532.GA16764@tarantool.org> <20191127111030.GB9233@atlas> <20191127112436.GB16940@tarantool.org> <20191127113917.GD9233@atlas> <20191127122138.GB29304@tarantool.org> <20191127130156.GA31045@tarantool.org> <73e0d8e6-3bd3-3807-732a-32dfe5790466@tarantool.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <73e0d8e6-3bd3-3807-732a-32dfe5790466@tarantool.org> Subject: Re: [Tarantool-patches] [PATCH 4/5] sql: replace control pragmas by SET List-Id: Tarantool development patches List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Vladislav Shpilevoy Cc: tarantool-patches@dev.tarantool.org * Vladislav Shpilevoy [19/11/28 10:29]: > > 3) If we move sysview out of the problem, where should we move it. > > To /dev/null. Together with all the other SQL. Wow, wow, easy please ;) It's always a question of tradeoffs, price/performance, etc. Even as it is now SQL is quite useful for BETWEEN, IN, open-bounded predicate + small limit queries. Someone will need to prepare a benchmark, of course, but eventually it will happen. There are a few remaining issues out there to make it more broadly useful, safe foreign keys, information_schema views and remaining issues with strong typing are high on my list. But here, the price/value ratio is ... high. Much faster and easier to replace keyword PRAGMA if someone doesn't like it with keyword SET, and be done with it. Besides, Mergen has prepared some cool SQL refactoring patches like removing some pragmas altogether which need to be looked at - they are bringing value regardless of whether there is consensus on the syntax for SQL settings. One must also admit there is some compose-ability value of SELECT over SET ; SELECT is usable in a join or a view. But you can easily work this around by lua('') generic built-in function, you can do lua('box.sql.exeucte("set