From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from smtp63.i.mail.ru (smtp63.i.mail.ru [217.69.128.43]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by dev.tarantool.org (Postfix) with ESMTPS id BEDEC445320 for ; Wed, 8 Jul 2020 13:35:13 +0300 (MSK) From: "Timur Safin" References: <1594199230-26036-1-git-send-email-alyapunov@tarantool.org> In-Reply-To: <1594199230-26036-1-git-send-email-alyapunov@tarantool.org> Date: Wed, 8 Jul 2020 13:35:12 +0300 Message-ID: <022401d65513$764e4970$62eadc50$@tarantool.org> MIME-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Content-Language: ru Subject: Re: [Tarantool-patches] [PATCH 0/2] Simplify alter.cc List-Id: Tarantool development patches List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: 'Aleksandr Lyapunov' , tarantool-patches@dev.tarantool.org Cc: v.shpilevoy@tarantool.org I'm very much infavor to use less verbose try-function blocks -=20 they significantly reduce verbosity of a C++ code which should=20 be in between C++ and C functions. So indeed - that's good idiom. Although I suspect `noexcept` in C++17 is not that much zero-cost=20 abstraction (it should be verified with all supported compilers) but = looks=20 like it's not adding any extra overhead which we would not have at the = moment. But I have some (debatable) notes about lambda usage. See in patches... Timur : From: Aleksandr Lyapunov : Subject: [Tarantool-patches] [PATCH 0/2] Simplify alter.cc :=20 : Since we use C++ we need to use in a way that benefits its advantages. : As I see, there are tons of try-catch blocks in alter.cc. We should = use : more modern techniques and approaches. :=20 : Aleksandr Lyapunov (2): : alter: use good c++ style : alter: use proper way to marry C and C++ :=20 : src/box/alter.cc | 418 = +++++++++++++++++++++++--------------------------- : ----- : 1 file changed, 177 insertions(+), 241 deletions(-) :=20 : -- : 2.7.4