Maybe consider using lombok only in connectors modules

Lucifer Tyrant <[email protected]> 于2023年12月3日周日 23:10写道:

> Hi hailin,
>
> Thanks for your advice, In response to this proposal, I would like to
> clarify the following points:
>  1. It is not to deny that lombok is a difficult tool to use, but at
> present there are certain thresholds for the use of lombok and some
> rules for the use of Lombok, and people who are not familiar with
> Lombok may misuse it;
>  2. Due to the large number of people involved in and contributing to
> open source projects, these lombok rules will not be passed on or
> complied with by everyone. From the perspective of open source project
> operators, it will bring unnecessary maintenance costs and
> interpretation costs. It is important to unify the standard and
> specification, either use lombok, or use Lombok.
>  3. At present, the frequency of use is not very high in the whole
> project, and some people like to use it and some people don't like to
> use it, which leads to the code style of the whole project will look
> strange.
>
> As mentioned above, I still think it's better to remove it.
>
> Chao Tian
>
> hailin0 <[email protected]> 于2023年12月3日周日 22:28写道:
> >
> > What I would like to suggest is not to use @Data casually as it contains
> > too much hidden code
> >
> > hailin0 <[email protected]> 于2023年12月3日周日 22:06写道:
> >
> > > Lombok does have some problems, but it undeniably brings some
> convenience
> > > and I don't think it should be avoided.
> > >
> > > Especially since we have many connectors, each of which involves data
> > > model and configuration definition, using lombok can indeed bring some
> > > benefits.
> > >
> > > Regarding the conflict between @Buliber and @NoArgsConstructor, the
> > > general solution is:
> > >
> > > ```java
> > >      @Tolerate
> > >      public yourConstructor() {
> > >      }
> > > ```
> > >
> > >
> > > Lucifer Tyrant <[email protected]> 于2023年12月2日周六 12:00写道:
> > >
> > >> Hi ZhangJian,
> > >>
> > >> 1. In past weekly meetings, it was mentioned that @Data can cause
> > >> unexpected issues with the equals and hashcode methods, reference:
> > >> https://github.com/apache/seatunnel/issues/4150
> > >>
> > >> 2. When use @Builder,@NoArgsConstructor will conflict with it.
> > >>
> > >> 3. The readability of the code will become very poor.
> > >>
> > >> Chao Tian
> > >>
> > >> ZhangJian He <[email protected]> 于2023年12月2日周六 11:33写道:
> > >> >
> > >> > Hi, There are also well-known projects using lombok, like Apache
> Pulsar
> > >> > etc. Could you please share more details about lombok's problems.
> > >> >
> > >> > Thanks
> > >> > ZhangJian He
> > >> >
> > >> >
> > >> > On Sat, 2 Dec 2023 at 11:21, Carl-Zhou-CN <[email protected]
> >
> > >> wrote:
> > >> >
> > >> > > +1
> > >> > >
> > >> > >
> > >> > > 121
> > >> > > [email protected]
> > >> > >
> > >> > >
> > >> > >
> > >> > > &nbsp;
> > >> > >
> > >> > >
> > >> > >
> > >> > >
> > >> > > ------------------ 原始邮件 ------------------
> > >> > > 发件人:
> > >> > >                                                   "dev"
> > >> > >                                                                 <
> > >> > > [email protected]&gt;;
> > >> > > 发送时间:&nbsp;2023年12月1日(星期五) 晚上10:33
> > >> > > 收件人:&nbsp;"dev"<[email protected]&gt;;
> > >> > >
> > >> > > 主题:&nbsp;[DISCUSS] Remove lombok in SeaTunnel
> > >> > >
> > >> > >
> > >> > >
> > >> > > Hi developers,
> > >> > >
> > >> > > In the recent development process, I discovered an issue with
> Lombok.
> > >> > > Lombok can cause overloading of multi parameter constructors to
> fail,
> > >> > > resulting in generated code that may not meet expectations and
> loss of
> > >> > > readability. After comparing with other well-known projects such
> as
> > >> > > Flink and Kafka, I found that Lombok plugins were not used.
> Therefore,
> > >> > > I would like to propose to remove all Lombok plugins from the main
> > >> > > repository
> > >> > >
> > >> > > What do you think about this? If you have any advice, please let
> me
> > >> know.
> > >> > >
> > >> > > Chao Tian
> > >>
> > >
>

Reply via email to