Also, add script to create the keys, for easy creation of similar keys.

* fix_dsync:
  [dsync] Add missing string.h for strlen
  [dsync] Drop GNU MIN/MAX definition

* xz-translations:
  generate-i18n-Index: set -o nullglob to handle missing files
  Don't override compression in filewriter
  Apply suggestion to config/debian/dinstall.functions
  Fix NameError suite
  Use xz compression for Translation indices in unstable/experimental

* pr-184:
  update test output for 3000-throwaway-new
  add test 3000-throwaway-new
  Move throw-away NEW binaries to morgue
  add utils.move_to_morgue function
  Add option to throw away binaries for NEW uploads

* pr-187:
  dak/examine_package.py: improve colors for accessibility

* replace-gzip-file-is-not-multi-arch-same-safe:
  Revert "lintian tags: remove gzip-file-is-not-multi-arch-same-safe for now"

* pr-173:
  Change default action to abandon for empty notes.

* pr-190:
  testsuite: add reject to 0001-basic

* pr-181:
  Tests: add archive output
  Testsuite: check_all_suites: check content of archive directories
  tests: use scripts in 1000-external-signatures v1
  tests: split off setup of unstable
  Tests: split off 2002-c-s-versionchecks from 2000-dinstall
  Tests: split off 2001-nviu from 2000-dinstall

* pr-171:
  Per-discussion updates
  Grammar: noun vs. verb
  Move README.coding to docs/code-guidelines.rst.
  Add docs/README.rst
  Clean up documentation.
  Add a few more To Do items.
  Add some developer/tester documentation.
  Clean up formatting on readme file

This adds a newline to the end of the input prompt which prevents the
destruction of text displayed within prompts when the window is resized.
Signed-off-by: NMichael Lustfield <michael@lustfield.net>

Currently, only accept is actually tested, but reject isn't that
uncommon, so we might want to test that codepath as well.

Seems like operators have been dropped from GNU C++

dsync-flist was broken by the upgrade to Debian 10 (buster). Also the
file contained its own (outdated) checksum which is never correct.
I am not sure if we should fix this or just drop the file.

Signed-off-by: NMichael Lustfield <michael@lustfield.net>

Signed-off-by: NMichael Lustfield <michael@lustfield.net>

This moves the code standards to a place that is consistent with other
DAK-related documentation. This document is now rst-formatted as well.
Signed-off-by: NMichael Lustfield <michael@lustfield.net>

This is just a simple "readme" that directs new readers to the
documentation that will likely be most relevant to them.
Signed-off-by: NMichael Lustfield <michael@lustfield.net>

The leading "README." on each file seems excessive, especially
considering this is a "docs/" directory. This was removed to make
identifying specific files a bit easier. Appropriate file extensions
were also added.

Formatting was also cleaned up so nearly-rst documents could be
rendered properly.
Signed-off-by: NMichael Lustfield <michael@lustfield.net>

These are some frustrations I've encountered working through the queue.
Signed-off-by: NMichael Lustfield <michael@lustfield.net>

This additional documentation describes how to get a dev/test instance
of dak up and running.
Signed-off-by: NMichael Lustfield <michael@lustfield.net>

This file was nearly RST-formatted; now it is.
Signed-off-by: NMichael Lustfield <michael@lustfield.net>

In the past, I've accidentally created some empty notes by pressing
enter instead of Abandoning the note. This changes the default from Done
to Abandon only if an empty note was created.
Signed-off-by: NMichael Lustfield <michael@lustfield.net>

The colors for "main" and "contrib" sections (resp. yellow and aqua) are
barely readable on a white background, like on the package details on
https://ftp-master.debian.org/new/<package>.html.

This patch propose new colors with "equivalent psychological"
meanings that can be read on a white background.

This reverts commit 3ff66a42 as lintian
2.59.0~bpo10+1 was accepted in buster-backports:

  https://tracker.debian.org/news/1112252/accepted-lintian-2590bpo101-source-all-into-buster-backports/

The current lintian version in backports triggers false-positives for
this tag, see https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=954146

Remove this tag for now. It can come back when lintian is fixed.

Please note that this test assumes that uploads by buildds use a suffix
(like pkgnew_0.1-1_amd64-buildd.changes), to avoid filename conflicts
with the orignal upload.