I can't see what the.. program:: directive does in the documentation preview. It would be good to see screenshot(s) to show rendering of this section before/after your changes.

@vsajip these are the changes ( with screenshot attached )

Before:

After:

ref:https://cpython-previews--130699.org.readthedocs.build/en/130699/library/venv.html#cmdoption-venv-system-site-packages

So we've duplicated the options information in the bit below "The command, if run with -h, will show the available options:" , in a slightly different formatting? What has that really achieved? And what did the.. program:: venv directive, specifically, render to? I'm not seeing anything in the documentation preview. Maybe I've missed it - please spell it out for me. I have to say, I'm not especially in favour of this type of "tinkering around the edges" - was the previous presentation lacking in some way? What's your motivation for making this change,@Mr-Sunglasses ?

So we've duplicated the options information in the bit below "The command, if run with -h, will show the available options:" , in a slightly different formatting? What has that really achieved? And what did the.. program:: venv directive, specifically, render to? I'm not seeing anything in the documentation preview. Maybe I've missed it - please spell it out for me. I have to say, I'm not especially in favour of this type of "tinkering around the edges" - was the previous presentation lacking in some way? What's your motivation for making this change,@Mr-Sunglasses ?

we've duplicated the options information in the bit below

@vsajip You're absolutely right that the options information is now duplicated in two slightly different formats:

• First, in the output of the-h help text (which is a direct copy-paste from the command-line output).
• Second, in the.. option:: directives that follow.

Now for the second question

What has that really achieved?

The.. option:: directives make the options stand out more clearly in the documentation. They are formatted in a way that makes them easier to scan and reference, especially for someone skimming the docs. and it also allow for better cross-referencing in the documentation. For example, if someone is reading about a specific option elsewhere in the docs, they can easily link back to its definition here. and also it maintains a consistency with how other Python modules document their CLI options (e.g., argparse), making the documentation more consistent across the board.

That said, I totally get your point about duplication. If you feel the duplication is unnecessary, we could remove the raw-h output and rely solely on the.. option:: directives.

About the.. program:: venv Directive

It tells Sphinx that the following options (.. option::) belong to thevenv program. This is useful for generating more accurate and linked documentation and it allows other parts of the documentation to reference thevenv program and its options more effectively. In the rendered documentation, the.. program:: venv directive itself doesn’t produce any visible output. It’s more of a behind-the-scenes marker that helps Sphinx organize and link the content. If you’re not seeing any visible change, that’s expected—it’s working as intended.

ref.https://www.sphinx-doc.org/en/master/usage/domains/standard.html#directive-program

What's your motivation for making this change,@Mr-Sunglasses ?

My motivation for making this change comes from a few observations in the issue#130160 :

• The previous presentation of the CLI options was functional but not especially easy to scan. The raw-h output is great but can feel dense in the docs.

• Many other Python modules (e.g., argparse, pip) use.. option:: directives to document their CLI options. Adding them tovenv with this style makes the overall documentation more consistent.

• Using.. program:: and.. option:: makes it easier to add cross-references or expand the documentation in the future without needing to restructure everything.

That said, I completely understand your perspective. If the previous presentation wasn’t broken, why fix it? My goal wasn’t to tinker for the sake of tinkering but to subtly improve the docs in a way that aligns with broader Python documentation practices.

If you’re not convinced that these changes add enough value, I’m happy to roll them back or refine them further. Please Let me know what you think!

You're absolutely right that the options information is now duplicated

Well, that's not good, in my view. I have no particular opinion on the.. program::venv if it doesn't render to anything. The benefit of the current setup is that if the command line options change, it's easy enough to runpython -m venv -h and paste the output of that back into the documentation - easy to keep the docs in sync with the CLI. With theoption directive, you have to update the docs whenever the CLI changes, but with your setup, you have to type in the option directive changes one by one and copy and paste the changed information for the option from the CLI-h run. More work for maintainers, for a slightly better presentation (bold rendering of options etc.)

Many other Python modules (e.g., argparse, pip) use .. option:: directive

Well, I didn't see any inargparse, andpip is an external program ... but sure,ensurepip uses:: option::.

But the duplication has to go, IMO. Note that if you remove the-h output to remove the duplication, you'd have to replace at least some of theusage: ... information that appears above the options in the-h output.

I see#130160 is trying to do this across various modules in the stdlib, so fair enough, I suppose ... my main concern is the extra work for maintainers (for this module, that's often me!) when the CLI changes.

By the way, you've said "fix" in the OP, which would close the issue if this PR is merged ... that's wrong IMO. Because the issue covers multiple modules, just changingvenv wouldn't be enough to close the issue, would it?

By the way, you've said "fix" in the OP, which would close the issue if this PR is merged ... that's wrong IMO. Because the issue covers multiple modules, just changingvenv wouldn't be enough to close the issue, would it?

Yeah, we did not want to close it until we update the targeted docs, I'll update it, thanks@vsajip

You're absolutely right that the options information is now duplicated

Well, that's not good, in my view. I have no particular opinion on the.. program::venv if it doesn't render to anything. The benefit of the current setup is that if the command line options change, it's easy enough to runpython -m venv -h and paste the output of that back into the documentation - easy to keep the docs in sync with the CLI. With theoption directive, you have to update the docs whenever the CLI changes, but with your setup, you have to type in the option directive changes one by one and copy and paste the changed information for the option from the CLI-h run. More work for maintainers, for a slightly better presentation (bold rendering of options etc.)

Many other Python modules (e.g., argparse, pip) use .. option:: directive

Well, I didn't see any inargparse, andpip is an external program ... but sure,ensurepip uses:: option::.

But the duplication has to go, IMO. Note that if you remove the-h output to remove the duplication, you'd have to replace at least some of theusage: ... information that appears above the options in the-h output.

I see#130160 is trying to do this across various modules in the stdlib, so fair enough, I suppose ... my main concern is the extra work for maintainers (for this module, that's often me!) when the CLI changes.

You're absolutely right that the options information is now duplicated

Well, that's not good, in my view. I have no particular opinion on the.. program::venv if it doesn't render to anything. The benefit of the current setup is that if the command line options change, it's easy enough to runpython -m venv -h and paste the output of that back into the documentation - easy to keep the docs in sync with the CLI. With theoption directive, you have to update the docs whenever the CLI changes, but with your setup, you have to type in the option directive changes one by one and copy and paste the changed information for the option from the CLI-h run. More work for maintainers, for a slightly better presentation (bold rendering of options etc.)

Many other Python modules (e.g., argparse, pip) use .. option:: directive

Well, I didn't see any inargparse, andpip is an external program ... but sure,ensurepip uses:: option::.

But the duplication has to go, IMO. Note that if you remove the-h output to remove the duplication, you'd have to replace at least some of theusage: ... information that appears above the options in the-h output.

I see#130160 is trying to do this across various modules in the stdlib, so fair enough, I suppose ... my main concern is the extra work for maintainers (for this module, that's often me!) when the CLI changes.

Thanks for your suggesstions,@vsajip .

so fair enough, I suppose ... my main concern is the extra work for maintainers (for this module, that's often me!)

Just my two cents, it will need a little bit of extra work while adding new features but it will make the docs overall better.

tagging@picnixz for his views, as he is helping me a lot in this issue and he originally created the issue. TiA.

• If we're okay with it, I can close this PR, if it does not impact things a lot.

There are benefits for having theprogram directive and options. The reason is that it allows us to have alink to them, which otherwise is hard. Note that it's normal that the directive produces no output:

Likepy:currentmodule, this directive produces no output. Instead, it serves to notify Sphinx that all followingoption directives document options for the program called name.

If you however think that it's not worth the change, we can skip that one. I can understand that it's easier to C/C if you already have an ArgumentParser outputting you the options. But for some modules, the option description may use backreferences.

their opinion on this discussion and their opinion on this PR

I thought I'd made that clear, but to reiterate:

• Thanks for the contribution, I understand the sentiment behindUse.. program:: and.. option:: directives for modules with a documented CLI #130160 and your desire to help address it.
• Although there might be a slightly increased maintenance burden than just typingpython -m venv -h and pasting in the results, it's probably worth it if the resulting presentation is better (the CLI shouldn't be especially volatile).
• I've no problem with you using theprogram directive, even though it doesn't render any output.
• The duplication of options is unsightly and should be addressed, by replacing the pasted help output with the bits that are missing from theoption directives, i.e. showing the invocation line with appropriate markup.

I hope that all makes sense!

@Mr-Sunglasses This is the last PR to remain before closing the parent issue. May you address the comments please? (or we can also skip the venv module or take over the PR if you don't want to work on this)

@Mr-Sunglasses This is the last PR to remain before closing the parent issue. May you address the comments please? (or we can also skip the venv module or take over the PR if you don't want to work on this)

Hey@picnixz , Sorry I forgot about this PR, I'll work on it and resolve the issues.

A Python core developer has requested some changes be made to your pull request before we can consider merging it. If you could please address their requests along with any other requests in other reviews from core developers that would be appreciated.

Once you have made the requested changes, please leave a comment on this pull request containing the phraseI have made the requested changes; please review again. I will then notify any core developers who have left a review that you're ready for them to take another look at this pull request.

I have made the requested changes; please review again.

cc.@picnixz

Thanks for making the requested changes!

@picnixz: please review the changes made to this pull request.