Cleanup the documentation page.

[Carreau]

The Documentation page in pure html was bothering me with lots of extra
element, and not really DRY, so hard to extend.

This fixes all that (and also use orange as the main color... why blue
we never use blue ?).

I'm also tempted to remove all the /en/latest/ or /en/stable/ from the
rtd links as RTD should detect the default page; not the pointing links,
but that will be the subject of another PR.

Also I want to change the URL from /documentation.html to
/documentation/ but same; we'll do that another time.

[Carreau]

Note that all the whitespace and - changes in footer and header is to prettify the rendered html source ,and make it more compact. We had hundreds of blank lines with lots of spaces on each consuming extra bandwidth.

[Carreau]

I'd like the input of Documentation-Sensei (@willingc) as well.

[willingc]

Happy to comment from an editing and outreach standpoint. Documentation-Wizard (@jzf2101) will have good insights from her recent work on JupyterLab and has a better view of @ellisonbg's vision.

[willingc]

I've left my suggestions and edits on content. Only did a quick scan of the CSS/html commands.

[willingc]

Wondering if we should rename this last item as it implies that IRkernel and IJulia are project maintained.

[Carreau]

Let's do that later. See my other PR; I can add use the data source on both page and automatically build links to Kernel documentation if needed.

[willingc]

JupyterWidgets??

[Carreau]

Depends if you speak of the Js library or the Python bindings !

[willingc]

Shorten perhaps:

Browse the Jupyter universe for information. or Browse the Jupyter universe.

[ellisonbg]

Cool, thanks for refactoring, can you add a screenshot to make visual design review easier?

[Carreau]

Cool, thanks for refactoring, can you add a screenshot to make visual design review easier?

No change of visual design beyond current Blue->Jupyter Orange used everywhere else.

[Carreau]

Ok, it's been a week, merging.

[ellisonbg]

I would have liked to review this

…

On Wed, Jan 17, 2018 at 8:09 AM, Matthias Bussonnier < ***@***.***> wrote: Merged #247 <#247>. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#247 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AABr0KBtllPshCwIRfuyg064wy-m4RLwks5tLhspgaJpZM4RZiAw> .

-- Brian E. Granger Associate Professor of Physics and Data Science Cal Poly State University, San Luis Obispo @ellisonbg on Twitter and GitHub bgranger@calpoly.edu and ellisonbg@gmail.com

[Carreau]

Sorry it's been a week, and it's just a change from A-blue-we-use-nowhere to The-official-Jupyter-orange.

I understand that it can be hard to not be involved in everything, but if you don't trust others the community cannot grow. It's also by doing mistakes that we learn.

Also we can still revert.

[ellisonbg]

In general I am fine with a "merge and iterate" approach. At the same time, there are two reasons I think it is appropriate to have specifically asked me for review on this PR:

1. I recently made the change (card header color) you were modifying.
2. I am not involved in "everything" in the project, but I am heavily involved in design and in this particular repo.

I think some of the challenge is that we don't have a strong sense of "team" on this repo that helps folks to know who should look at what. I have ideas how to fix that and fill follow up separately.

[ellisonbg]

When I make design decisions, I try hard to do so with a reason. Those reasons may not always be obvious; I am always willing to share my reasoning. Here were my reasons for using blue for the card headers in the documentation page:

• In doing design with "jupyter orange" we have found that it is a great emphasis color, but a bit too strong as a layout color. Thus in JupyterLab, we use it in a few very carefully chosen places, but not as a core layout color. I initially tried jupyter orange here, but found it to be "too much".
• Much of our recent visual design work is based on material design. The "card" construct being used on this page comes directly from material design (https://material.io/guidelines/components/cards.html), and it is common for material design cards to have a colored header with white text.
• Blue is a complementary color to orange, so it goes well with the Jupyter orange used elsewhere.

I am totally open to other reasons that override my decision, but simply saying "we don't use blue" isn't enough to override the above reasons without discussion.

[Carreau]

Happy to revert to blue now that I know the reasons.

It was hard for me to know that this was done on purpose seeing the number of inconsistencies in the HTML and CSS I have found; In particular it's not the first time I've refactored a page that was completly hardcoded HTML, with each html block slightly different, which was not easy.
The CSS is regularly weird on viewport size handling, and in a couple of place seemed "hacked together" to fix some layout issue.

So I honestly though this was a left over from an unfinished page that did slipped through review.

The links are still blue, but blue header look (to me) strongly out of style with the rest of the site.

I'll send a PR reverting just the color.

[Carreau]

Another things that made me weight for a mistake, the #1976D2 blue is (was) no used anywhere on the website, unlike #337ab7 which has 25 occurrences.

[Carreau]

Oh wow ! Github renders colors now !

[ellisonbg]

Thanks Matthias. I am trying to put together why I used that particular blue. Normally I would have picked a material design blue, but this isn't one of them. Hmmmm. I am open to other blues or colors (jupyter grey?). Very cool, that GH renders colors in line, that is one of my favorite aspects of VS Code :-)

…

On Sat, Jan 20, 2018 at 3:45 PM, Matthias Bussonnier < ***@***.***> wrote: Oh wow ! Github renders colors now ! [image: screen shot 2018-01-20 at 15 44 47] <https://user-images.githubusercontent.com/335567/35189110-de09fefc-fdf8-11e7-95c1-07a8bc832741.png> — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#247 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AABr0Di722Y5w9V1_WvwbYkuzaxspru2ks5tMnqBgaJpZM4RZiAw> .

-- Brian E. Granger Associate Professor of Physics and Data Science Cal Poly State University, San Luis Obispo @ellisonbg on Twitter and GitHub bgranger@calpoly.edu and ellisonbg@gmail.com