-
-
Notifications
You must be signed in to change notification settings - Fork 17.9k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
DOC: collect feedback on new sphinx documentation theme #28952
Comments
@jorisvandenbossche where do we add suggestions for typos? |
Typos are most probably typos in the actual doc sources, and not in the theme. So for those, you can create a PR fixing them if you want, or otherwise open a separate issue. |
Hi, thanks for the work on the new theme, it looks modern and clean.
|
I agree with @tritas 2nd point about some sort of margin added to the copyright notice. |
Thanks for the work! My first impression is that the left/right split for the contents navigation is a little confusing - its not instantly obvious how the right navigation relates conceptually to the left one. Also the fact that it disappears with narrower windows is maybe not ideal, it's useful so why hide it sometimes? |
Thanks for the feedback!
@tritas I am no expert in this, but from your question I assume the answer is "no" ?
Yes, that's certainly something to improve. For pandas, we probably want the same footer as for the main website, but also in general a better footer by default for the theme would be nice. I opened pydata/pydata-sphinx-theme#42 to track this
@mangecoeur There is an issue here about the content of the right sidebar: pydata/pydata-sphinx-theme#12. I image we could add something like "On this page" as a title on the right sidebar. Would that make it already clearer that the right sidebar is the local table of content for the page you are looking at?
That's part of the responsive aspect provided by bootstrap. If it would not disappear, other elements (the body, the left sidebar, which have priority over the right sidebar) would not have enough space anymore. |
Hi Joris, looks good, much better than the previous theme. Would it be possible to collapse the header to something much smaller (half the size?) when scrolling down? This leaves more room for the actual content, which is important for smaller screens ( = my laptop). |
Much modern and cleaner, definitely an improvement! However, I think you should have a max-width of 900px -1000px in the middle column. To difficult to read when maximized in wide screens. In general, I got the feeling that I could skim quicker to the info that I wanted with the old template but maybe I am just used to it |
Thanks again for the feedback!
@kvanderwijst Yes, that's certainly something to consider (the logo should also shrink then). Another option would be to hide the header when scrolling down, and eg only show it again when scrolling up or at the top. Or just make it smaller to start with.
@kavvkon Yes! I agree with that, and have been meaning to do that, but didn't come to it yet (pydata/pydata-sphinx-theme#17 as an issue for it)
@TomAugspurger gave similar feedback a few days ago: pydata/pydata-sphinx-theme#39 Feel free to give more detailed feedback on that issue!
@kavvkon Being used to it is probably part of the reason, but if you think later of more specific reasons, certainly interested to hear them. |
Yes I mean within a page.. I guess the above issues on headers and a better contrast balance among headers, code boxes and info boxes will resolve this. |
Yes, that part of the theme (styling the elements within the body: the headers, code boxes, links, etc) did not yet get much attention, so a lot of room for improvement there. |
Great stuff! I checked the API reference and for longer methods, like I did something extremely basic to illustrate my idea:
https://dev.pandas.io/docs/reference/api/pandas.read_csv.html |
Looking at https://dev.pandas.io/docs/reference/series.html
|
On point 1, in pydata/pydata-sphinx-theme#63 (comment) we deliberately increased the isze of h1 and h2 to distinguish them from h3 headers. Not saying the the values I picked are best, but I want to keep that h2 vs. h3 distinction clear. |
"BUG,DOC: responsive/mobile page width and left/right scrolling" #30910 |
It would be nice to be able to switch between docs versions more easily. Not everyone is already using the latest released pandas version (might take some time to port code), and google usually brings you to the latest version. Usually at readthedocs there is a version switcher in the bottom left of the page (bottom of the side bar). I have been desperately looking for something similar for Pandas docs... What makes it worse it that if landed on the current docs, it is absolutely impossible to click your way through the correct location where to find previous docs. You need to know that you need to manually go to pandas homepage (https://pandas.pydata.org/) (not the docs homepage) by editing the URL for example, and then look in the "Previous versions" right sidebar for the docs link of previous releases. Very tedious. The docs page are akin to a black hole in terms of version switching... |
Could a few rules in Python or a config file specify which subpath of a git
repo to copy _build/html to, generate a JSON file mapping versions to URLs
for a dropdown control, copy the latest JSON version-URL map to every
subpath, and update the /en/latest/ symlink (in order to store multiple
versions of the built HTML docs in one repo)?
"How to build [Sphinx docs locally] with the ReadTheDocs Docker container"
ammaraskar/sphinx-action#9 (comment)
( Sphinx-action creates pull request reviews with inline Sphinx errors
(with GitHub Actions; which could build the docs with the appropriate
version of pandas and all dependencies already-installed in a container))
…On Tue, Feb 18, 2020, 11:10 PM zertrin ***@***.***> wrote:
It would be nice to be able to switch between docs versions more easily.
Not everyone is already using the latest released pandas version (might
take some time to port code), and google usually brings you to the latest
version.
Usually at readthedocs there is a version switcher in the bottom left of
the page (bottom of the side bar). I have been desperately looking for
something similar for Pandas docs...
What makes it worse it that if landed on the current docs, it is
absolutely impossible to click your way through the correct location where
to find previous docs. You need to know that you need to manually go to
pandas homepage (https://pandas.pydata.org/) (not the docs homepage) by
editing the URL for example, and then look in the "Previous versions" right
sidebar for the docs link of previous releases. Very tedious.
The docs page are akin to a black hole in terms of version switching...
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#28952?email_source=notifications&email_token=AAAMNS3HALYVGYDAK4GTGUDRDSWLDA5CNFSM4JAF3XJ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEMGI6HQ#issuecomment-588025630>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAAMNSYKZC7O5RXHZILC7OLRDSWLDANCNFSM4JAF3XJQ>
.
|
Also, generating a /sitemap.xml which links to each subpath/sitemap.xml
might improve search indexing for /en/v0.9.X/ /en/v1.0.X/ /en/latest/
…On Wednesday, February 19, 2020, Wes Turner ***@***.***> wrote:
Could a few rules in Python or a config file specify which subpath of a
git repo to copy _build/html to, generate a JSON file mapping versions to
URLs for a dropdown control, copy the latest JSON version-URL map to every
subpath, and update the /en/latest/ symlink (in order to store multiple
versions of the built HTML docs in one repo)?
"How to build [Sphinx docs locally] with the ReadTheDocs Docker container"
https://github.com/ammaraskar/sphinx-action/issue
s/9#issuecomment-586657133 ( Sphinx-action creates pull request reviews
with inline Sphinx errors (with GitHub Actions; which could build the docs
with the appropriate version of pandas and all dependencies
already-installed in a container))
On Tue, Feb 18, 2020, 11:10 PM zertrin ***@***.***> wrote:
> It would be nice to be able to switch between docs versions more easily.
> Not everyone is already using the latest released pandas version (might
> take some time to port code), and google usually brings you to the latest
> version.
>
> Usually at readthedocs there is a version switcher in the bottom left of
> the page (bottom of the side bar). I have been desperately looking for
> something similar for Pandas docs...
>
> What makes it worse it that if landed on the current docs, it is
> absolutely impossible to click your way through the correct location where
> to find previous docs. You need to know that you need to manually go to
> pandas homepage (https://pandas.pydata.org/) (not the docs homepage) by
> editing the URL for example, and then look in the "Previous versions" right
> sidebar for the docs link of previous releases. Very tedious.
>
> The docs page are akin to a black hole in terms of version switching...
>
> —
> You are receiving this because you commented.
> Reply to this email directly, view it on GitHub
> <#28952?email_source=notifications&email_token=AAAMNS3HALYVGYDAK4GTGUDRDSWLDA5CNFSM4JAF3XJ2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEMGI6HQ#issuecomment-588025630>,
> or unsubscribe
> <https://github.com/notifications/unsubscribe-auth/AAAMNSYKZC7O5RXHZILC7OLRDSWLDANCNFSM4JAF3XJQ>
> .
>
|
@zertrin yes, a version dropdown is on the to do list: pydata/pydata-sphinx-theme#23, I also really would like to see this (note this was a problem with the previous sphinx theme as well, although now there is also an issue in finding the home page -> #30896) |
@jorisvandenbossche is this actionable/closeable? |
Are they currently source-ordered?
As a workaround (for what could be a conf.py setting) can you Ctrl-F for
the documented callable you're looking for?
…On Wed, Jul 7, 2021, 08:51 Glen Pike ***@***.***> wrote:
I'm curious why the sidebar navigation is not sorted alphabetically. I can
understand grouping methods in the index page for something like DataFrame,
but as a n00b to the library, it's really hard to find functions that
people mention in tutorials, etc.
[image: Screenshot 2021-07-07 at 13 50 08]
<https://user-images.githubusercontent.com/789846/124762086-66b9b880-df2a-11eb-974f-2b4837029d33.png>
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#28952 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAAMNS2UDO6EOPV4E5J5SSTTWREVPANCNFSM4JAF3XJQ>
.
|
Hi, yes, I can try to find them - which works if I know what the method is. Not sure about the ordering 🤷 Thanks :) |
As discussed in #15556 and activated in #28623, the development docs are using a new sphinx documentation theme (for now developed in https://github.com/pandas-dev/pandas-sphinx-theme/).
You can see it in action at https://dev.pandas.io/docs/user_guide/index.html
But as with all new things there will be some rough edges / certain workflows that were easier before / things you don't like / ... with the new theme.
So feedback is very welcome! We can use this issue to collect feedback in this repo (and otherwise specific issues can also be opened at https://github.com/pandas-dev/pandas-sphinx-theme)
For the next release, I think most structural blockers are solved (search works now, tables are not completely ugly, second navigation level in left sidebar is not hidden anymore). But there are certainly a lot of improvements to be made regarding styling (typography, styling / coloring of inline code and links, ... width of the page) and mobile responsiveness.
In pydata/pydata-sphinx-theme#32 I am trying out some basic styling for indicating the current active navigation items.
The text was updated successfully, but these errors were encountered: