tarantool / tarantool/tarantool-python

Pack RPM/DEB for obsolete distros

未关闭
#257 0 条评论 0 个 reaction 已指派 0 人 在 GitHub 查看

还没有人认领这个 Issue。

backlog question
主要语言
Python
星标
108
派生
51
平均合并
1 天 8 小时
30 天内合并 PR
1

描述

Follows up #248

In PR #248 we had introduced RPM and DEB packages for distros which have Python 3.7 or newer in its package managers by default. Extending package testing and publish matrices is a complicated task. Why is that?

First, the code works with Python 3.6 or newer, so if system doesn't have it one (including our CI) must set it up manually. Second, running code with Python 3.6 requires dataclasses backport package. And here we come to the packages issue.

It seems rather useless to provide a package for a distro if distribution repositories (or some external popular repos like EPEL) don't have its dependencies. And the case is they mostly don't.

RPM with dataclasses backport is available for EPEL 8 only. It is not so hard to rebuild (see https://github.com/tarantool/tarantool-python/wiki/Building-RPMs draft), but we will need to provide and maintain external packages in Tarantool repo. I also haven't seen any DEBs for dataclasses.

dataclasses is not the only one troublesome package. For example, CentOS 7 (including EPEL 7) does not provide any pandas RPM for Python 3. Rebuilding pandas RPM from EPEL 8 on EPEL 7 is rather challenging task. We also shouldn't forget that pandas has many dependencies and it is possible that we would need to rebuild them to.

Distros planed in this ticket:

  • centos:7
  • debian:9
  • ubuntu:18.04 (and possibly 16.04)

centos:8 is EOL so we don't plan to provide a package for it.

贡献指南

这个仓库没有索引到贡献指南

从这里开始

  1. 先读完整个 Issue,再读项目的贡献指南。
  2. 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
  3. Fork 仓库,在一个分支上完成修改。
  4. 提交 Pull Request,并在描述里引用这个 Issue 编号。

调研方向

审查 PR #248 以及软件包测试和发布矩阵,然后查阅链接的关于构建 RPMs 的 wiki 草稿。检查 CentOS 7、Debian 9 和 Ubuntu 18.04(可能还包括 16.04)的依赖项可用性,包括 Python 3.6 的要求。完成的标准是:支持的 RPM/DEB 软件包能够针对选定的已停止支持的发行版,在其依赖项可用的情况下完成构建、测试和发布。

由索引模型根据 Issue 内容生成。

评估

技术栈
python
领域
build-system, release
Issue 类型
功能
难度
4/5
预计耗时
3-5 天
活跃度
停滞
描述清晰度
基本清楚
新手友好度
30/100

把新 issue 发到你的邮箱

精选适合新手参与的 GitHub issue 摘要。