From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from localhost (localhost [127.0.0.1]) by turing.freelists.org (Avenir Technologies Mail Multiplex) with ESMTP id 917162AA01 for ; Thu, 4 Oct 2018 06:49:53 -0400 (EDT) Received: from turing.freelists.org ([127.0.0.1]) by localhost (turing.freelists.org [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id cq67Isln2v4M for ; Thu, 4 Oct 2018 06:49:53 -0400 (EDT) Received: from smtp1.mail.ru (smtp1.mail.ru [94.100.179.111]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by turing.freelists.org (Avenir Technologies Mail Multiplex) with ESMTPS id 4F7552A8EB for ; Thu, 4 Oct 2018 06:49:53 -0400 (EDT) Content-Type: text/plain; charset=utf-8 Mime-Version: 1.0 (Mac OS X Mail 10.3 \(3273\)) Subject: [tarantool-patches] Re: [PATCH 2/2] sql: add test on changes/total_changes From: "n.pettik" In-Reply-To: <881CB143-0DC2-468A-90CA-0459E9EE7B15@gmail.com> Date: Thu, 4 Oct 2018 13:49:50 +0300 Content-Transfer-Encoding: quoted-printable Message-Id: <2DF5F7A1-C504-4B82-B394-0F16DF025D13@tarantool.org> References: <7BEF7461-8C36-472F-95D7-6E46CD99E956@tarantool.org> <9C1DDEBD-C37D-4B61-96AC-832BC4DDD12A@tarantool.org> <881CB143-0DC2-468A-90CA-0459E9EE7B15@gmail.com> Sender: tarantool-patches-bounce@freelists.org Errors-to: tarantool-patches-bounce@freelists.org Reply-To: tarantool-patches@freelists.org List-help: List-unsubscribe: List-software: Ecartis version 1.0.0 List-Id: tarantool-patches List-subscribe: List-owner: List-post: List-archive: To: tarantool-patches@freelists.org Cc: Alex Khatskevich Accidentally, I sent last mail from wrong address, so it didn=E2=80=99t = arrive to mailing list. Original letter below: > On 2 Oct 2018, at 16:40, =D0=9D=D0=B8=D0=BA=D0=B8=D1=82=D0=B0 = =D0=9F=D0=B5=D1=82=D1=82=D0=B8=D0=BA wrote: >=20 >=20 >>>> I have expected the same thing. >>>> However: >>>> 1. sqlite works in the same way. >>> It doesn=E2=80=99t mean anything. SQLite features a wide range of = really strange things. >>>=20 >>>> 2. implementing this just need adding (nChanges) to savepoint >>> Seems so, but it shouldn=E2=80=99t be complicated. >> I suppose it is not complicated. >>>> 3. I suppose that we should not implement this. >>> Why? I guess it is matter of discussion. >> I think that it is not important, because it is relatively useless = counter. >> 1. No one need to count exact number of changes. >=20 > Without proofs such statements worth nothing. >=20 >> 2. Changing the api require solid understanding of what total_changes = is >> and why we need to change this. >> It strange for me that it do not do rollback on rollback, however, = it meant to work this >> way and it also makes sense. >=20 > It only means that SQLite developers decided to do so, nothing else. >=20 >>> Anyway, I propose to decide what to do with TOTAL_CHANGES within = this patch, >>> since original issue sounds like: >>>=20 >>> "Fix if easy, remove otherwise.=E2=80=9D >>>=20 >>> Adding initially wrong tests is likely to be huge mistake. >> This thing works. Tests proof that. >=20 > =E2=80=98Works' and =E2=80=98works and gives wrong result' are sort of = different things. > In this regard, rollback doesn=E2=80=99t work, ergo this counter works = well > only in half cases. >=20 >> I believe that those tests are the artefact of this issue. >> If you want change something, you need to file an issue and explain, >> why this change is important. >=20 > I guess there is no matter of discussion: it is very likely to be = obvious > that rollback should return old total_change counter (i.e. value = before rollback). > If you strictly disagree, you are welcome to discuss it with other = members.