From: engelbert g. <eng...@gm...> - 2022-06-21 22:25:20
|
Thanks to Günter and Adam several changes happened that made a release necessary RELEASE NOTES 0.19 * Drop support for Python 2.7, 3.5, and 3.6. * Output changes: HTML5: Wrap groups of footnotes in an ``<aside>`` for easier styling. The CSS rule ``.footnote-list { display: contents; }`` can be used to restore the behaviour of custom CSS styles. * After package installation, the CLI commands ``python -m docutils`` and ``docutils`` start the `generic command line front end tool`__. __ docs/user/tools.html#generic-command-line-front-end * Support parsing "Markdown" input with 3rd party parsers myst_, pycmark_, or recommonmark_. * The default values for the "pep-references", "rfc-base-url", and "python-home" `configuration settings`_ now use the "https:" scheme. The PEP-writer template's header is updated to fix links and resemble the header of official PEPs. * Various bugfixes and improvements (see HISTORY_). install with : pip install docutils --pre Release 0.19 is planned for july 5th cheers and all the best engelbert |
From: Guenter M. <mi...@us...> - 2022-06-22 11:48:34
|
On 2022-06-21, engelbert gruber wrote: > ... several changes happened that made a release necessary Thank you for the release. Some bugs are fixed in commits after the release: r9089 Stop ignoring the "input-encoding-error-handler" setting/option. r9090 Fix the wheel naming problem. r9094 Fix the System Messages in https://docutils.sourceforge.io/HISTORY.html Could you upload a fixed version to the web space? Regarding r9082: The script in ``sandbox/infrastructure/version_identifier_parsing.py`` can be used to update the `docutils.__version_info__`:: #> cd sandbox/infrastructure #> ./version_identifier_parsing.py --change-version-info ../../docutils/docutils/__init__.py This feature is fixed in r9096. It could be made part of the ``set_version.sh`` script. Open Points (best solved before releasing 0.19 (final)): - input encoding: * Announce a switch of the default to 'utf-8'? * Mark the handling of "input_encoding is None" as provisional? - tools / entry-points: * Announce upcoming changes? · name changes (drop extension *.py) · keep the scripts in /tools but install auto-generated scripts via entry-points (allows modern packager, supports Windows) · do not install "exotic" scripts in the binary PATH? - testing with Sphinx and other third party tools * contact concerned developers/maintainers Github tickets? Günter |
From: Adam T. <aat...@ou...> - 2022-06-22 23:50:55
|
>> ... several changes happened that made a release necessary > Thank you for the release. Seconded. > Open Points (best solved before releasing 0.19 (final)): > - testing with Sphinx and other third party tools > * contact concerned developers/maintainers Github tickets? Speaking in my co-maintainer of Sphinx role, Sphinx passes all tests with r9096 [1]_, although I still need to do visual testing, I will at the weekend. I will also respond to the other questions later on in the week. _[1]: https://github.com/sphinx-doc/sphinx/runs/7013501543 A |
From: Adam T. <aat...@ou...> - 2022-06-26 12:56:47
|
Dear Günter, Engelbert, all, > Open Points (best solved before releasing 0.19 (final)): > > - input encoding: > > * Announce a switch of the default to 'utf-8'? > * Mark the handling of "input_encoding is None" as provisional? I would be in favour of both of these, yes. > - tools / entry-points: > > * Announce upcoming changes? > > · name changes (drop extension *.py) We should announce this, yes. > · keep the scripts in /tools but install auto-generated scripts > via entry-points (allows modern packager, supports Windows) For now I think we should keep the scripts in tools/, but perhaps later consider deprecating and removing them. > · do not install "exotic" scripts in the binary PATH? Which would we keep? There is an argument for eventually only installing ``docutils`` into PATH. I agree we should cut down what we put into PATH regardless. Thanks, Adam |
From: Guenter M. <mi...@us...> - 2022-06-26 10:30:13
|
Dear Adam, On 2022-06-22, Adam Turner wrote: >> - testing with Sphinx and other third party tools >> * contact concerned developers/maintainers Github tickets? > Speaking in my co-maintainer of Sphinx role, Sphinx passes all tests > with r9096 This is good news. > although I still need to do visual testing, I will at > the weekend. I will also respond to the other questions later on in the > week. Another idea would be opening tickets at major contributing projects (MyST, furo or other CSS schemes, ...) and ask for testing with Docutils 0.19b there. Could you also have a look, if there are DeprecationWarnings when compiling a project? I could successfully compile "PyLit" documentation with sphinx-build 3.4.3 (from Debian/stable) and Docutils 0.19b (with warnings because of the "meta" element changes, but the result looks fine (although not polished yet, see https://milde.codeberg.page/PyLit/). Günter |
From: Adam T. <aat...@ou...> - 2022-06-26 12:47:22
|
Dear Günter, > Another idea would be opening tickets at major contributing projects (MyST, furo or other CSS schemes, ...) and ask for testing with Docutils 0.19b there. I opened several: - Sphinx (covered by me) - https://github.com/getnikola/nikola/issues/3632 - https://github.com/getpelican/pelican/issues/3017 - https://github.com/executablebooks/MyST-Parser/issues/591 - https://github.com/brechtm/rinohtype/issues/340 Themes: - https://github.com/pydata/pydata-sphinx-theme/issues/757 - https://github.com/mgeier/insipid-sphinx-theme/issues/101 - https://github.com/pradyunsg/furo/issues/462 - https://github.com/executablebooks/sphinx-book-theme/issues/578 - https://github.com/jbms/sphinx-immaterial/issues/113 > Could you also have a look, if there are DeprecationWarnings when compiling a project? > I could successfully compile "PyLit" documentation with sphinx-build > 3.4.3 (from Debian/stable) and Docutils 0.19b (with warnings because of > the "meta" element changes, but the result looks fine (although not > polished yet, see https://milde.codeberg.page/PyLit/). Sorry, I don't understand -- deprecation warnings from Docutils or from Sphinx? Do you mean expected warnings that we added or unexpected ones? Happy to test, my time during weekdays is limited at the moment though. Thanks, Adam |
From: Guenter M. <mi...@us...> - 2022-06-26 20:28:54
|
On 2022-06-26, Adam Turner wrote: > Dear Günter, >> Another idea would be opening tickets at major contributing projects (MyST, furo or other CSS schemes, ...) and ask for testing with Docutils 0.19b there. > I opened several: Thanks. >> Could you also have a look, if there are DeprecationWarnings when >> compiling a project? > Sorry, I don't understand -- deprecation warnings from Docutils or from > Sphinx? Do you mean expected warnings that we added or unexpected ones? > Happy to test, my time during weekdays is limited at the moment though. I mean deprecation warnings from Docutils. We expect some in "stable" Sphinx. The Sphinx development version should (ideally) not trigger any deprecation warnings but use the documented workarounds/alternatives. This ensures that the next stable Sphinx will be ready for the change... Viele Grüße Günter |
From: Adam T. <aat...@ou...> - 2022-07-02 11:00:00
|
>> Another idea would be opening tickets at major contributing projects (MyST, furo or other CSS schemes, ...) and ask for testing with Docutils 0.19b there. > I opened several: No one has reported errors with Docutils thus far on the ones I opened. > I mean deprecation warnings from Docutils. > We expect some in "stable" Sphinx. > The Sphinx development version should (ideally) not trigger > any deprecation warnings but use the documented workarounds/alternatives. > This ensures that the next stable Sphinx will be ready for the change... Ahh, yes. I will pick this up from the Sphinx side and fix the warnings. A |
From: Guenter M. <mi...@us...> - 2022-07-02 22:03:52
|
On 2022-07-02, Adam Turner wrote: > [-- Type: text/plain, Encoding: quoted-printable --] >>> Another idea would be opening tickets at major contributing projects (MyST, furo or other CSS schemes, ...) and ask for testing with Docutils 0.19b there. >> I opened several: > No one has reported errors with Docutils thus far on the ones I opened. No news is good news (hopefully). ... >> I mean deprecation warnings from Docutils. > Ahh, yes. I will pick this up from the Sphinx side and fix the warnings. Thank you. I am just working on an "input-encoding handling" proposal. > [-- Skipped Type: text/html --] > [-- Type: text/plain, Encoding: 7bit --] > [-- Type: text/plain, Encoding: 7bit --] |