mkiser / wtfjht Goto Github PK

Statements like

Ted Cruz leads Beto O’Rourke by one percentage point in their Texas Senate race. Cruz leads O'Rourke 38 to 37%.

are misleading without reporting the margin of error of the poll. This particular poll had a 4.4% margin of error. Ted Cruz would be ahead 42.4% to 32.6%, or O'Rourke could be ahead 41.4% to 33.6%.

Would also be good to indicate that a single poll is seldom a good representation of the race, because of different biases inherent in how polls are structured and it is best to look at aggregated polls.

This is such a helpful project!

It'd be nice to give each item (numbered or bullet point) an HTML anchor to enable hyperlinks directly to specific items.

I noticed that the WTFJHT search misses certain search inquires.

Does the VTFJHT search only show 'relevant' hits? or everything?

This arises from Issue #15 and this gist that shows formatted html for a typical wtf day.

Here's the Webpagetest result for this page.

While this shows a good nominal score, there are an excessive number of JS requests (33) which could be reduced to two.

I have a plugin that generates these, but there is some issues with the formatting in the XML when FB ingests it. I believe it's related to the blockquote and/or Twitter embeds.

This is false.

http://progressive.org/dispatches/republican-states-turn-over-kobach-voter-files-even-while-cl/

• Category: information architecture
• Sub-Category: Future discoverability, digests, and SEO
• The Problem: Currently the site does not use HTML semantics as well as it could. This is easy to fix within Jekyll templates.

Discussion: For full-value in the future, it would be helpful if the site used more than one level of heading tags to distinguish semantic structure . Currently h2 headings are used to delineate days, and that's it.

It will help search (and therefore discoverability) if search bots are able to better infer and structure site contents, and header nesting is one way to convey meaningful semantics.

Proposal: Daily sub-segments should be built with h3 tags, and things like Updates should use h4 tags. Consider using HTML 5 aside elements to structure the sidebar-like sections.

The following is captured using Chris Pedrick's Web Developer Toolbar Extension (under Information ~ View Document Outline). This is the current structure of the site:

Looks like the file https://d3ur8zm5qs6awd.cloudfront.net/uploads/Day-819.mp3 is being referenced for the inline player but not available.

This is next-level.

• Category: information architecture
• Sub-Category: WTF as fungible data.
• The Problem: Small adaptations can transform a "wall of text" into fungible, structured data. See WikiMedia's Wikidata Initiative, for example.

Discussion One way to do this is integrate schema.org markup into the Jekyll templates.

Some candidates:

• Article
• News Article, a specialized form of Article
• Report
• Potentially several others.

Take Note: there are examples of the markup at the bottom of each schema, and this is all well supported by the Google Structured Data Testing Tool among others. Here's a primer.

I have a plugin for the Google AMP Pages, but I'm not sure it's configured correctly. I think there's both some scheme issues and some formatting issues.

Something in the markup of the "Day 543" article appears to be preventing the RSS feed from rendering correctly.

This shows https://whatthefuckjusthappenedtoday.com/rss.xml in Firefox:

And here's the specific article in my newsreader - note the only two characters rendering in the article body are ."

I've not had time to pinpoint the issue much further.

Hook into the commit history for the file to add which user made the last commit (edit) to the file.

Hi- I was thinking of starting this exact same service, but then a friend told me about your very helpful posts. For about six years or so in the mid-90s it was my job to compile information on the anti-environmental movement and digest it into a monthly newsletter to activists across the country. I did a lot of it before the internet was in full swing, meaning I used actual press clips (gasp!) and good old opposition research like phone calls and newsletter subscriptions. (It was called the CLEAR Project if you want to look it up- we lost funding eventually because the powers that grant grants found us "too political" and the people we watched "too fringe." Of course those people are in charge now, so nobody listened to us. It is a most unsatisfying "I told you so".) After that I did some freelance work starting the stub pages for the Center for Media and Democracy's Congresspedia and setting up Greenpeace's Exxonsecrets.org, which is all about climate change and corporate funding to deniers. The point is, I have experience compiling a bunch of information into digestible bits and have a decent understanding of the players. If there is any way I can help, I am only working part time right now and would like to volunteer my services. I have up to 10 hours a week to give to your project. I've also done copy writing and editing, so my writing doesn't need a lot of fixing. If you google "Emily Headen" and "Clear Project" you will unearth some ancient history, but enough to know I am not full of shit. Please let me know! -Emily Headen

Love the search function, but would be awesome if you could sort by date too.

The site looks great on mobile. Except it's a little difficult to tap the correct day at the top. I just found your site and want to read it from day 1.

I think a next/previous button would be a good idea at the bottom of each post. Just to make things easier if someone else wants to read through all the days.

On https://github.com/mkiser/WTFJHT/blob/master/_posts/2017-05-09-Day-110.md:

I admit I'm tired this morning so maybe my internal parser is failing:

"Sean Spicer downplayed acting Sally Yates’ warnings"

Acting Sally Yates?

It's currently 6:08pm PDT, but somehow the main page says "06/15/2017 Updated: 06/15/2017 07:09:27 PM PDT "... an hour in the future!

Would be cool if WTFJHT used anchor links like the following on their website.

https://github.com/theoryshaw/WTFJHT/blob/370fe465ca5f4bf79e96bccecdaefc049cbb9f2b/_posts/2018-05-10-day-476.markdown#2-rudy-giuliani-trump-wasnt-aware-that-michael-cohen-pitched-his-access-to-the-president-to-potential-clients

I've often wanted to link just one story blurb, vs. the entire blog post.

Just an idea.

Currently https://whatthefuckjusthappenedtoday.com/atom.xml can't be parsed, fails with "XML Parsing Error: not well-formed".
There's a non-printing \x10 in the frontmatter description of the post.

Pull request #942 should resolve it by removing that character.

Experts don't believe the North can't make a nuclear warhead small enough to be mounted on ICBMs, yet.

One of the big goals for WTFJHT is to collaboration to the news in a transparent and open source way. I started that mission by using Jekyll and git/GitHub to save the project, which means anybody is free to make a pull request to create, update, destroy code && content.

The problem. Jekyll, GitHub, and markdown are all a little too high of a barrier to entry for most users in the WTFJHT community.

The goal. Use Wordpress' world class CMS, auth, and user management tools for content creation, editing, and publishing as a headless CMS with the underlying Jekyll and git framework left intact.

Two tools look particularly promising for accomplishing this:

1. https://github.com/mAAdhaTTah/wordpress-github-sync

2. http://v2.wp-api.org/

I'm looking for help building this service. A few high-level requirements in the meantime:

• open source the resulting service
• import the existing markdown posts and pages from Jekyll into Wordpress. Convert markdown to html (most likely using WP Markdown) and map all the metadata, titles, images, etc. to their WP equivalent parts.
• be able to create new posts and pages in WP. Support for drafts and publishing needed. Convert the publish date, title, etc from WP HTML to Jekyll-friendly markdown (YAML required).
• automatically sync changes in WP to github (the current repo is setup to rebuild and deploy with a Travis hook on commits to master - we'd need to revisit bc building on every save of a draft will be overkill; probably needs to be on publish)
• use WP's media manager for attaching social images/cards to the markdown yaml

...and probably a bunch of things i'm not thinking about.

Problem: Running WTF through webpagetest.org yields some actionable insight. In my view none greater than the involvement of genius.com.

Factors to consider: This webpagetwest.org result from today (February 20, 2017) on the home page – Day 32. Specifically, the results shown in the domains tab.

In short, Genius is responsible for 74% of page weight, 18% of this page's 43 HTTP requests, and greatly affects page timing – see the genius lines in the waterfall chart in the details tab.

For starters here's a screen cap from the domains tab.

The normal person's guide to internet security is great, but I think it should mention ways to keep messaging like SMS, IMs, or emails more secure. Even a short list of recommended services could be very helpful.

If you want to expand the list beyond six things I can fork and submit some PRs if you'd like.

GREAT stuff, Matt!!! If it's not too late, you might wanna add (to yesterday's list) the GOP's rollback of the Obama Admin's "Stream Protection" rule.

Here's an Alternet article about it:

http://www.alternet.org/environment/republicans-wipe-out-stream-protection-rule-giving-gift-big-coal-while-threatening

And a little less "biased" one from Forbes:

http://www.forbes.com/sites/kensilverstein/2017/02/03/will-undoing-the-stream-protection-rule-really-help-coal/?partner=yahootix#759710b9e785

Keep up the good work -- bet it's not easy to keep up with!!!

Tom Ledermann (248) 310-7647

The House passed a sweeping voting rights, campaign-finance and ethics reform package.
The legislation includes expansion of early voting, redistricting reform, making Election Day a federal holiday,

The referenced Politico article has a correction:

CORRECTION: A previous version of this story said the legislation made Election Day a federal holiday. That provision was removed from the bill before the final vote.

EDIT: the other reference (NPR) makes no mention of the federal holiday removal, so now I don't know what to believe...

I’d like to make a pull request, but this repo only seems to include generated files.

First of all, thanks for doing this. I had the same idea... I was going to call mine "What Did These Assholes Do Today?" because that's how it feels waking up in the morning to check the news. :-) Anyway....

Strongly suggest an HTTPS certificate. I'm sure you understand. Lots of ways to get free SSL certs, like GoDaddy or StartSSL or Let's Encrypt.

Your work on this project is invaluable. Don't change a thing. I like the simplicity and layout of this page. What about an app?

5/ - Second sentence needs re-write; "instead" suggests that something happened in the first sentence which will be different in the second. Change to "The Trump transition team ordered..." (removing instead)

Currently the root URL loads all the content in a single page. Would be better if it redirects to the latest day for faster load and less bandwidth.

Instead of displaying "Site updated: 02/02/2017 06:54:37 PM PST" in the site's header, it would be great if WTFJHT used the browser locale and/or moment.js to infer the correct local TZ and convert the date to a "time since" relative date, e.g. "Site updated: 3 hours, 12 minutes, 27 seconds ago"

And PS: THANK YOU for creating this!

Much like #640, from newsboat:

Which makes me suspect it's the unicode character following "Tuesday" on this line:

    <p><strong>Voter Guide</strong>: How, when and where to vote on Tuesday�. (<a href="https://www.nytimes.com/2018/11/05/us/politics/midterm-elections-voting-guide.html">New York Times</a> / <a href="https://www.bloomberg.com/news/articles/2018-11-05/midterms-speed-read-all-you-need-to-know-about-tuesday-s-vote">Bloomberg</a>)</p>

Confirmed broken in Firefox too.

mobile and desktop are going to be a disaster w/o a better of navigating to a specific day in the past.

proposing to add a simple drop-down to the top nav, moving the social icons into the top-right corner, and adding links for Today | Yesterday | About | Membership

I find myself checking the site a few times a day and looking for revisions. However, sometimes the revision history is a long list of commits with mostly small grammatical tweaks that show up as big change blocks like 7356892#diff-7906e7a1c37349acedf7342d5c65b9e0R28; so, its hard to see from the revision history when new stories were added.

My suggestion is that you add a semantic version to the merge commit for revisions. That way you it would be obvious what type of changes were made. The versioning could go something like: major (1.x.x) version bump for story added or removed, minor (x.1.) version bump for details added to story, and patch (x.x.1) for grammar changes / typo fixes.

You could also sneak the version next to the updated date to easily see what type of new updates have been added.

I use Algolia for site search. It's really good. I would like to use a few faceted searches to produce static WTFJHT topic pages when the site builds (i.e. russia, health care, travel ban, etc).

This seems like an appropriate task for a Jekyll Generator plugin. When the site builds, make a call to the Algolia API for each topic I want to create, and build a single page per topic with all the responses in reverse-chron order using a specific template file.

The Algolia API returns a raw response and is indexed on p tags. So, theoretically (?), we could assemble all that HTML on the page and we'd be good to go. A nice to have bit would be to remove the leading 1/, 8/, 15/ slug.

Why static? I want them to have a simple permalink, be indexable in Google and shareable on social without the need of a query string.

I don't really know Ruby and am unclear if you can even make an API call from within a generator, return the data, and then build the page. There is probably a much better way of doing this... Suggestions?

Suggest you offer a "sponsored" (aka "free") email version (with ads) and a "membership" (aka paid and no ads) version. Given you are up to 50K users with some pretty clear liberal demographics, you might be able to get some decent ad dollars to offset your costs. I would be fine seeing ads in my free version. Many advertisers would kill for the inventory you have, and I expect it will grow substantially as word spreads.

This is just a stub for a call for help on building a proper WTF native app. My intention is to hold this till the API is done as I believe a big part of any WTF app is to be more of a push-as-a-service thing than a feature-rich news app. Let's talk.

A couple of links that may be useful on the internet security page:

• How Secure Is My Password? - Find out how long it would take a computer to crack your password.
• Have I Been Pwned? - Find out if you have been included in any data leaks and subscribe to get updated when your email comes up in the future.

I see that you have a Creative Commons Attribution-ShareAlike 4.0 International License as the governing license on the WTFJHT site itself. That should be added to the repo as well.

Is there a way to have a feed that only publishes when the articles are done, versus when first published? I know it might go against the community editing procedures but seeing "more updates coming" at the bottom is totally harshing my drama buzz. :)

You got a double 'has' in section 5 sentence 4.

Hit me up sometime dude, I've got some free time on in the evenings whenever you need an extra. It would help to justify the Poli Sci degree that took me 16 years to complete...Go Dawgs!

-MB

I just wanted to reach out and thank you for doing this.

I'm working on something similar: https://garykac.github.io/trumpocalypse/index.html [1]
But we have completely different (complementary) approaches to our day-by-day logs.

Anyway, I know this project is a lot of work and I just wanted to drop off some encouragement.

[1] Github repo: https://github.com/garykac/trumpocalypse

In the Patreon update, you mentioned the newish podcast. Adding it to the Amazon Echo flash briefing might be a good way to get new subscribers. https://developer.amazon.com/public/solutions/alexa/alexa-skills-kit/docs/understanding-the-flash-briefing-skill-api I've done some bare bones Echo development before, so might be able to help out.

So that each post has a count of all time pageviews, similar to Gawker and Business Insider

Current API returns each day's update as a single response. Would like to find a way to break this down so each day has multiple sub-items representing each news blurb.

Note: editorial workflow needs to remain the same (building site via the static markdown file), so this would require a post-processing approach.

https://whatthefuckjusthappenedtoday.com/api/v1/pages.json

Mostly because Paypal just isn't cool. Never was.

Some weeks even these very helpful "executive summaries" are more than I really need to know, but I remain curious as to whether the overall status quo has changed significantly.

What do you think of adding an "API", which really could just be a static file that someone commits to maintaining, that would track a handful of key topics?

{
  "realDonaldTrump": {
    "still_president": true,
    "impeached": false,
  },
  "USAgov": {
    "defcon": 5,
    "aca_repealed": false,
    "daca_going_away": true,
    "travel_ban": ["TCD", "IRN", "LBY", "PRK", "SOM", "SYR", "VEN", "YEM"],
    "wars": […]
   }
}

Some of these are probably frought with peril (e.g. at what point do you say we're at/not at war with N.Korea? what if the Affordable Care Act gets revised but not exactly replaced/repealed?) and could use some honing, and I'm sure there's tons more that could be added…

The idea would be to support things like IFTTT, personal dashboards, single-serving sites for those who'd rather have a "push notification" for certain things rather than having to continually "poll" the news to see if we e.g. have/haven't nuked a roaring mouse.

Based on @StevenBlack's suggestion, I run an analysis on google's structured data tool and noticed the main schema.org name and headline fields are duplicated:
https://search.google.com/structured-data/testing-tool#url=https%3A%2F%2Fwhatthefuckjusthappenedtoday.com%2F

<script type="application/ld+json">
{
	"@context": "http://schema.org",
	"@type": "WebSite",
	"name": "What The Fuck Just Happened Today?",
	"headline": "What The Fuck Just Happened Today?",
	"image": "https://whatthefuckjusthappenedtoday.com/public/wtfjht-f.jpg",
	"description": "Today’s essential newsletter. Logging the daily shock and awe in national politics. Read in moderation.",
	"url": "https://whatthefuckjusthappenedtoday.com/"
}
</script> 

Setting up CI with Jekyll + Github is a relatively simple process that will let you deploy with a bit more confidence. Jekyll provides some documentation on how to set this up here: https://jekyllrb.com/docs/continuous-integration/

Those instructions show how to set up html-proofer to run automatically on every push, which validates your markup (i.e. W3C validation), checks for broken links, etc.

The first step would be to decide which CI service to use. Travis and Circle CI both have free options for open source projects, so I think either choice would befine.

I'm afraid I don't have the front-end chops to make a dent on this right off the bat. If someone can point me in the right direction I'm happy to help, but not sure where to help.

I have an iPhone 7 and 7 Plus on hand for testing, if that's needed at any point.