perladvent / perldotcom Goto Github PK

This can present problems for browsers/keyboard combos where no enter key is present. (Certain mobile devices.)

As a way to scan the article list it's a useful tool, but doesn't show on the front page?

This article should be ready to go but could use another set of eyes. It's in the nadim-data-dump branch.

So users can send pull requests wih fixes and/or raise issues.

(for new articles)

The old one looks tired. Use the Camel icon instead.

It would be even easier to suggest a correction if the above text were hotlinked to e.g. https://github.com/dnmfarrell/perldotcom/edit/master/content/article/a-hackerspace-webcam.md

Right now it's just the twitter feed for The Perl Foundation. I think we should hook up Perlybot to publish its feed to twitter.

e.g. the tutorials have a table but all of the entries are in a single row

To point to the official perldotcom account (or TPF).

Many of the articles must have broken links, scan for these and figure out
whether they should be updatred or removed.

This happens sometimes. Sorry i don't have more details, but it simply happens sometimes.

Got this in Opera v12.18 with JS off.

e.g. "development" is too broad, break articles having that category into
more specific, useful categories. Likewise "tutorials" isn't really useful
as much of the site content is teaching material. Maybe "tutorials" should
be a tag instead.

I thought it was a quicklist of articles on perl.com, but instead it's ... something else?

All of the old links should still work, but don't right now on the static site:

The path contains an extra /article component and is missing the .html ending

I'd very strongly prefer to see my old stuff under a plain CC BY-SA license. I really dislike the NC clause.

That said, I don't know if I have standing to request that, since I think my stuff was written for ORA way back when. If they licensed the content to TPF only under CC BY-SA-NC 3.0 then there's not much I can do about it.

But regardless, it might be good to support per-article licenses.

This program would check an article file to ensure everything is as it should be:

• The meta data are well-formed
• Meta data are complete
• The content is valid markdown
• The author data is in the repo
• The images are in the repo
• Any external links are good

extract_article_links - short codes
``` " tchrist quoting
` runaway code line (tchrist article) - unclosed (maybe limit length of stuff on the inside)

We should have mostly a handful of colors:

• white for background and empty space (done)
• charcoal for text (done)
• dark blue like perl.com for the header
• a light blue or grey for secondary color
• an even lighter version of secondary color

Would be good to have OpenGraph and Twitter Card headers, so that Facebook and Twitter show images when the site's content is shared.

I'll have a look to see how easy it is over the next couple of days.

Basically review and don't migrate the bad ones.

• merge similar tags
• remove author names from tags

The unicode cookbook and style guide index slides in particular may need
formatting

Trying to find articles by e.g. rough timeframe is only possible in a hacky way right now. There should be some kind of archive/calendar/??? view.

"Side note: Why aren't comments allowed on Perl.com? A blog without comments isn't a blog." - https://www.reddit.com/r/perl/comments/7r1b33/an_open_letter_to_the_perl_community/dsts4sb/

Basically, tell people Perl.com uses reddit (or whatever) for comments and explain that as a volunteer thing we don't have time or willpower to create and manage such a thing. Link to the reddit comments at the end of each article.

Jim suggests:

Can we add a small box above Community Articles with:

Perl Resources

• Get Perl
• Learn About Perl
• Get Perl Code
• Help Perl

At least for the most famous or still active perl programmers.

Get Sawyer's author info. I've already emailed him asking for a pull request.

• Picture static/images/author/sawyerx.jpg
• JSON data data/author/sawyerx.json

https://perl.com/article/monitoring-my-phone-s-internet-activity-with-netflows-and-perl/ has an image styled code segments. On the rss feed the image is elided, and the code segments are not styled.

And then apply perl.com styling to it

Right now clicking them doesn't lead to the browser loading anything, but instead just sitting there for a few seconds until JS does the thing. Bad ux. Especially on bad networks where things might just break silently in the background.

Visiting / should be the same as /article

• Create a new article specific directory under static/images
• Copy contents of graphics folder there (if exists)
• Copy any other locally referenced images there
• Update all HTML src to point to the new location
• Determine if there is an article thumbnail and set the object
  attribute if there is
• Update the serialize code to include an article thumbnail
• Update the hugo template to try and use the thumbnail first, then
  the article cover image, then the first author's image.

Particularly the "more" icons top right are bad about that.

e.g. data/author/tom-christiansen-nathan-torkington.json

should be two separate files. It probably comes from when Article.pm
extracts the author names from either the RDF or vcard in the article
html.

The latest article:

https://www.perl.com/pub/2014/02/perl-today-february-2014.html

Contains text in Russian. When it's translated, the text is corrupted:

http://localhost:1313/article/pub/2014/02/perl-today-february-2014/

to migrated perltricks articles

I expected an actual search, but got only DDG, this was an unwelcome surprise and in the spirit of not surprising the user it should make clear that it's that.

For now I think it can stay on perltricks.com, assuming brwosers don't
prevent it from loading on perl.com because of cross-domain requests.

Would need a google account to associate this with.

• Use a dark blue, rename the variable from "perltricks blue"
• Update class attributes to use the new name instead

This repo should be moved to the Perl organization, and the tpf.github.io
submodule removed. Additionally the perl.com directory should be deleted
as it duplicates many files, including several large media files.

https://www.perl.com/about/ links to https://github.com/dnmfarrell/perldotcom/CONTRIBUTING.md, which doesn't exist.

Looks like a whole bunch of full-size images for "behind the fold" stuff are automatically downloaded on first load.

Looks like you redirect to it via a non-GET request, leading to the search query not being retained in the url of the final page. Bad ux.