Re: Improvement of installation document

On Tue, 26 Mar 2019 22:18:53 +0900 (JST)
Tatsuo Ishii <ishii(at)sraoss(dot)co(dot)jp> wrote:

> > One of our clients suggested that the installation document[1] lacks description
> > about requriements of installing *-devel packages. For example, postgresqlxx-devel
> > is required for using --with-pgsql, and openssl-devel for --with-openssl, and so on,
> > but these are not documented.
> >
> > [1] http://www.pgpool.net/docs/pgpool-II-3.7.4/en/html/install-pgpool.html
> >
> > I know the document of PostgreSQL[2] also lacks the description about openssl-devel,
> > kerberos-devel, etc. (except to readline-devl). However, it would be convenient
> > for users who want to install Pgpool-II from source code if the required packages
> > for installation are described in the document explicitly.
> >
> > [2] https://www.postgresql.org/docs/current/install-requirements.html
> >
> > Is it not worth to consider this?
>
> I am against the idea.
>
> Development package names could differ according to
> distributions/OS. For example, the developement package of OpenSSL is
> "openssl-dev", not "openssl-devel" in Debian or Debian derived
> systems.
>
> Another reason is, a user who is installaing software from source code
> should be familiar enough with the fact that each software requires
> development libraries.
>
> In summary adding not-so-complete-list-of-development-package-names to
> our document will give incorrect information to novice users, and will
> be annoying for skilled users.

OK. I agreed.

# From this viewpoint, it would not be so good that PostgreSQL doc[2]
# mentions readline-devel...., but this is noa a topic here.

>
> > BTW, the Pgpool-II doc[2] says:
> >
> > --with-memcached=path
> > Pgpool-II binaries will use memcached for in memory query cache. You have to install libmemcached.
> >
> > , but maybe libmemcached-devel is correct instead of libmemcached?
>
> I don't think so. "libmemcached-devel" is just a package name in a
> cetain Linux distribution. "libmemcached" is a more geneal and non
> distribution dependent term.

Thanks for your explaination. I understood it.

> Best regards,
> --
> Tatsuo Ishii
> SRA OSS, Inc. Japan
> English: http://www.sraoss.co.jp/index_en.php
> Japanese:http://www.sraoss.co.jp

--
Yugo Nagata <nagata(at)sraoss(dot)co(dot)jp>