From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <lvasiliev@tarantool.org>
Received: from smtp21.mail.ru (smtp21.mail.ru [94.100.179.250])
 (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 E3A644696C3
 for <tarantool-patches@dev.tarantool.org>;
 Wed, 22 Apr 2020 19:17:18 +0300 (MSK)
References: <cover.1587334824.git.lvasiliev@tarantool.org>
 <6d673bc4-c85b-219e-6ff4-79efc7627275@tarantool.org>
 <20200420080547.x5tttvnqcmplksp6@tarantool.org> <20200421190325.GA3677@atlas>
From: lvasiliev <lvasiliev@tarantool.org>
Message-ID: <9f449c31-a47f-7249-349f-7734ee84b59f@tarantool.org>
Date: Wed, 22 Apr 2020 19:17:17 +0300
MIME-Version: 1.0
In-Reply-To: <20200421190325.GA3677@atlas>
Content-Type: text/plain; charset="utf-8"; format="flowed"
Content-Language: en-US
Content-Transfer-Encoding: 8bit
Subject: Re: [Tarantool-patches] [PATCH V6 00/10] Extending error
 functionality
List-Id: Tarantool development patches <tarantool-patches.dev.tarantool.org>
List-Unsubscribe: <https://lists.tarantool.org/mailman/options/tarantool-patches>, 
 <mailto:tarantool-patches-request@dev.tarantool.org?subject=unsubscribe>
List-Archive: <https://lists.tarantool.org/pipermail/tarantool-patches/>
List-Post: <mailto:tarantool-patches@dev.tarantool.org>
List-Help: <mailto:tarantool-patches-request@dev.tarantool.org?subject=help>
List-Subscribe: <https://lists.tarantool.org/mailman/listinfo/tarantool-patches>, 
 <mailto:tarantool-patches-request@dev.tarantool.org?subject=subscribe>
To: Konstantin Osipov <kostja.osipov@gmail.com>, Kirill Yukhin <kyukhin@tarantool.org>, Vladislav Shpilevoy <v.shpilevoy@tarantool.org>, tarantool-patches@dev.tarantool.org



On 21.04.2020 22:03, Konstantin Osipov wrote:
> * Kirill Yukhin <kyukhin@tarantool.org> [20/04/20 11:08]:
>>> It is still unfinished because of underdesigned traceback
>>> feature, because of payload absence. IMO, MP_EXT is also an
>>> overkill. Tuples live fine as MP_ARRAY, and they are the most
>>> used type. We should have gone for simple MP_MAP, without
>>> MP_EXT. Just a map.
>>
>> We already broken connectors by introducing UUID (w/ MP_EXT).

I’ll add for clarity: "MP_ERROR doesn't break connectors because
IPROTO_ERROR doesn't use MP_EXT, MP_ERROR used at IPROTO_BODY after
client do session_settings update."
> 
> UUID change was planned. This was not. UUID is a fundamental data
> type. struct error is just an object.
> 
> This was a rushed up decision. Better revert it and go back to
> MP_MAP.
>>> It is worth mentioning separately, how hard it is to use the
>>> error marshaling now, because of this session setting. And
>>> there still is no way to enable the feature without touching
>>> the session, even if all my connectors support it. As I
>>> mentioned, enabling it for every session manually is a
>>> non-trivial task for a user.
>>
>> I guess we can switch it on by default in future releases.
>>
>>> But we have releases coming, so lets push all in whatever
>>> state it is, of course.
>>
>> --
>> Regards, Kirill Yukhin
>