Documentation is too personalized

[serhiy-storchaka]

BPO	18280
Nosy	@gvanrossum, @freddrake, @terryjreedy, @pitrou, @ezio-melotti, @merwok, @JimJJewett, @serhiy-storchaka, @aixtools, @csabella
PRs	bpo-18280: Make documentation less personal. #21639
Files	Imemy.grep

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = 'https://github.com/freddrake'
closed_at = None
created_at = <Date 2013-06-22.08:28:34.438>
labels = ['easy', '3.7', 'docs']
title = 'Documentation is too personalized'
updated_at = <Date 2020-09-19.19:02:04.510>
user = 'https://github.com/serhiy-storchaka'

bugs.python.org fields:

activity = <Date 2020-09-19.19:02:04.510>
actor = 'georg.brandl'
assignee = 'fdrake'
closed = False
closed_date = None
closer = None
components = ['Documentation']
creation = <Date 2013-06-22.08:28:34.438>
creator = 'serhiy.storchaka'
dependencies = []
files = ['30665']
hgrepos = []
issue_num = 18280
keywords = ['patch', 'easy']
message_count = 13.0
messages = ['191636', '191637', '191685', '192008', '192036', '295141', '295164', '315557', '374279', '374326', '374330', '374363', '374443']
nosy_count = 11.0
nosy_names = ['gvanrossum', 'fdrake', 'terry.reedy', 'pitrou', 'ezio.melotti', 'eric.araujo', 'docs@python', 'Jim.Jewett', 'serhiy.storchaka', 'Michael.Felt', 'cheryl.sabella']
pr_nums = ['21639']
priority = 'normal'
resolution = None
stage = 'patch review'
status = 'open'
superseder = None
type = None
url = 'https://bugs.python.org/issue18280'
versions = ['Python 3.7']

Linked PRs

• gh-62480: Remove first-person usage for some docs #102295
• gh-62480: De-personalize "Partial mocking" section in unittest.mock examples #141321
• gh-62480: De-personalize "Mocking Unbound Methods" section in unittest.mock examples #141322
• gh-62480: De-personalize "Coping with mutable arguments" section in unittest.mock examples #141323
• [3.14] gh-62480: De-personalize "Mocking Unbound Methods" section in unittest.mock examples (GH-141322) #141324
• [3.13] gh-62480: De-personalize "Mocking Unbound Methods" section in unittest.mock examples (GH-141322) #141325
• [3.14] gh-62480: De-personalize "Partial mocking" section in unittest.mock examples (GH-141321) #141326
• [3.13] gh-62480: De-personalize "Partial mocking" section in unittest.mock examples (GH-141321) #141327

[serhiy-storchaka]

Some documentation files contain a number of I/my/me. Looks like they grew from personal modules and personal articles. Perhaps the official documentation needs more depersonalized style. Here is full list of such files:

Doc/c-api/exceptions.rst
Doc/c-api/long.rst
Doc/distutils/builtdist.rst
Doc/extending/extending.rst
Doc/extending/windows.rst
Doc/howto/argparse.rst
Doc/howto/curses.rst
Doc/howto/functional.rst
Doc/howto/regex.rst
Doc/howto/sockets.rst
Doc/howto/urllib2.rst
Doc/install/index.rst
Doc/library/audioop.rst
Doc/library/ctypes.rst
Doc/library/doctest.rst
Doc/library/heapq.rst
Doc/library/numbers.rst
Doc/library/ossaudiodev.rst
Doc/library/tk.rst
Doc/library/unittest.mock-examples.rst
Doc/library/unittest.mock.rst
Doc/reference/introduction.rst
Doc/tutorial/classes.rst

The list doesn't include FAQs where it may be appropriate and whatsnew files.

Andrew Kuchling recently has fixed Doc/howto/unicode.rst for this issue (as part of bpo-4153).

[serhiy-storchaka]

Here is a filtered results of

find * -name '*.rst' -exec egrep -n -w -B1 -A1 'I|me|my' '{}' +

[akuchling]

I've looked through the matches. "I/O" and the -I command-line switch are false positives. Many references in the FAQ ("How do I do X?"), but those don't need to be fixed.

I think personalized references are most problematic when they're expressing uncertainty ("I don't know if we implement all of the spec") or opinions. Sentences like "When I run this command under Linux, I see..." could be rewritten as "When *you* run this command...", but they don't seem worth fixing to me.

Files with personalized text are:

c-api/exceptions.rst
c-api/long.rst
distutils/builtdist.rst
extending/extending.rst
install/index.rst
library/audioop.rst
library/ctypes.rst
library/doctest.rst
library/heapq.rst
library/imaplib.rst
library/numbers.rst
library/ossaudiodev.rst
library/unittest.mock-examples.rst
library/unittest.mock.rst
reference/introduction.rst
tutorial/classes.rst

[terryjreedy]

I find some anonymous I references (Guido? 20 years ago?) off-putting when reading the doc as formal reference.

[pitrou]

The sockets tutorial deserves a good overhaul :-)

[csabella]

Would it be OK for me to tackle this?

[rhettinger]

Fred, do you want to opine on this?

In some cases, like heapq.py, the personal touch is an essential and beautiful part of the presentation and is a cherished part of Python. In other cases, it seems unnecessary or a little off-putting, so perhaps a few changes are warranted.

Personally, I've grown to really dislike the incessant stream of proposals to make broad sweeping trivial changes across the code or documentation to fix made-up problems (ones not reported or cared about by actual users). In particular, I worry about sending some new dev on a mission to rewrite documentation that was written by domain experts (Alex Martelli reported that copy-editors "wreaked havoc" on one of his books just prior to publication by subtly changing the meaning or correctness of his prose while applying grammar rules and minor style edits -- I wish to avoid the same for us).

Also, I place high value on text written by Guido and think we lose something every time someone wants to rewrite it to fit their personal tastes and views of the language. The tastes and views of module authors are more important are easily lost in style edits (especially those that change point of view, mood, or theme of presentation).

Another thought is that there should be different general rules for different sections. The standard library docs tend to be more formal. The language reference tends to be even more formal ("for language lawyers"). The tutorial tends to be personable. The how-to guides are often have a personal touch and are the only places where we attribute authorship back individuals (actual by-lines at the top of the file).

[Cheryl Sabella]

Would it be OK for me to tackle this?
You could, but I would really like to get you involved in more substantive work that involves thinking about real issues and real code. IMO, this project isn't worthy of you time and is not on the critical path to your stated goals. That said, feel free to volunteer for anything that interests you.

[mcepl]

What about WONTFIX here? I completely agree with rhettinger: this is a waste of time with potential for causing damage.

[gvanrossum]

Marking this as "easy". People are welcome to submit PRs that fix the wording in one or a small number of modules called out in Serhiy's list.