- Always run apt update before any other apt command. (This fixes
incremental builds failing if a remote package is updated.)
- Only copy dependency lists before installing dependencies. (This means
editing code doesn't force all dependencies to be re-downloaded.)
- Delete cache in the same layer that it is created. (Otherwise,
deleting cache *increases* the size of the image on non-squashed
builds.)
- Move the installation of some static dependencies to *before* Mastodon
code is imported to Docker.
https://www.ruby-lang.org/en/news/2018/03/28/ruby-2-4-4-released/
> This release includes some bug fixes and some security fixes.
>
> - CVE-2017-17742: HTTP response splitting in WEBrick
> - CVE-2018-6914: Unintentional file and directory creation with directory traversal in tempfile and tmpdir
> - CVE-2018-8777: DoS by large request in WEBrick
> - CVE-2018-8778: Buffer under-read in String#unpack
> - CVE-2018-8779: Unintentional socket creation by poisoned NUL byte in UNIXServer and UNIXSocket
> - CVE-2018-8780: Unintentional directory traversal by poisoned NUL byte in Dir
> - Multiple vulnerabilities in RubyGems
Both of yarn and npm are used in Mastodon, but the combined usage requires
a redundant dependency and may lead to data inconsistency.
Considering that yarn has autoclean feature which npm does not have,
this change replaces all npm usage with yarn.
This change requires documentation update. Most notably, the following
command must be executed before assets precompilation if any system
dependency of node-sass has changed:
yarn install --force --pure-lockfile
When volumes are declared, but the corresponding directories don't exist, permissions for those directories will be root:root instead of mastodon:mastodon..
This changes makes sure the permissions of the volume directories are as expected.
System default libiconv of Alpine Linux only supports some charset (e.g. UTF-8).
Therefore, the preview card of the page which is not UTF-8 will be broken in the Docker environment.
Using GNU libiconv!
* Specs for language detection
* Use CharlockHolmes instead of NKF
* Correct mistakes
* Correct style
* Set hint_enc instead of falling back and strip_tags
* Improve specs
* Add dependencies
- improve docker_entrypoint.sh
- serve static files with puma by default
- sort packages list
- use virtual package for build deps
- show how to assign UID/GID
Compact Language Detector v3 (CLD3) is the successor of CLD2, which was
used in the previous implementation. CLD3 includes improvements since CLD2,
and supports newer compilers. On the other hand, it has additional
requirements and cld3-ruby, the FFI of CLD3 for Ruby, is still new and may
be still inmature.
Though CLD3 is named after CLD2, it is implemented with a neural network
model, different from the old implementation, which is based on a Naïve
Bayesian classifier.
CLD3 supports newer compilers, such as GCC 6. CLD2 is not compatible with
GCC 6 because it assigns negative values to varibales typed unsigned.
(see internal/cld_generated_cjk_uni_prop_80.cc) The support for GCC 6 and
newer compilers are essential today, when some server operating system
such as Ubuntu Server 16.10 has GCC 6 by default.
On the one hand, CLD3 requires C++11 support. Environments with old
compilers such as Ubuntu Server 14.04 needs to update the system or install
a newer compiler.
CLD3 needs protocol buffers as a new dependency. However,it is not
considered problematic because major server operating systems, CentOS and
Ubuntu Server provide them.
The FFI cld3-ruby was written by me (Akihiko Odaki) for use in Mastodon.
It is still new and may be inmature, but confirmed to pass existing tests.
Sandro Jäckel