Live data from Hacker News

Show HN: A Reddit reader that looks like the frontpage of a print newspaper

unim.press

81–90 of 129 posts

Re: Show HN: A Reddit reader that looks like the frontpage of a print newspaper

#82
I went back reading a printed newspaper a few months back and the speed in which I get my news now is astonishing. It feels like the web is not made to update oneself on news efficiently. To skim and zoom in and out of articles. I think part is the display format, part is the organization. A physical newspaper has a very predictable structure and does not change while you are reading it. (I.e. no need to come back to the same section once you read it).

Typography and layouting of a newspaper is quite complex. If you look at a printed newspaper, the stories are arranged and cut to size with images, captions and inserts to fit neatly into the grid.

Depending on the newspaper the information flow from headline, subheadings, article is strict, optimizing for allowing for fast skimreading or to grab your attention.

To do this "right" on challenging unedited reddit posts will be a lot of work, but an interesting challenge.

Re: Show HN: A Reddit reader that looks like the frontpage of a print newspaper

#84

It hangs on "reading stories". Web app authors can be very creative, but one thing I don't like about web apps is that often their error handling is nonexistent. It's not about individual authors; behemots like Microsoft also do this in an irritating way: "Something did not work. Reload the page please" -- Is this really the best a billion dollar company can do? Well, I suppose this is still better than no error at a…

They (MS etc) know the error alright, but users get scared and overwhelmed by technical language and error messages. So they put some picture of a sad cat and a message like "zoinks, that didn't work, try again". And anyway, if you're going to reload anyway, why would you care about the exact error? Hopefully it got logged in their systems so the engineers can come back to it. Beyond nerd curiosity, why would a user need non-actionable detailed error messages?

Re: Show HN: A Reddit reader that looks like the frontpage of a print newspaper

#85

Text always says lorem ipsum. It could be composed of comments in the thread.

Depends on the type of the post. I suppose I could have fetched comments, but for now, it uses lorem ipsum if the post isn't a text post. If it's a self/text post, it uses that text as the article body.

I think using top comments to fill this up instead of using Loren ipsum would greatly improve it.

Re: Show HN: A Reddit reader that looks like the frontpage of a print newspaper

#86

Please do not use justified formatting for text unless you can hyphenate it. Otherwise you end up with large chunks of white space between the words, especially in narrower (then 6070 characters) columns. Browsers don’t hyphenate, so avoid justified text on the web.

Browsers do hyphenate, and this page enables it, with the "hyphens: auto" CSS rule. https://developer.mozilla.org/en-US/docs/Web/CSS/hyphens In Google Chrome, you'll see that it hyphenates "profes-sional" in the second story of the center column. It's still not enough, though, because it still leaves huge gaps in the justified text; it works best when you use width: 80ch. The page is currently using width: 600px, whi…

And if you don't see hyphenation in Chrome, probably you use Linux (or another unsupported platform). To feel how it should insert ­

    var Hyphenopoly = {
      require: {
        "en-us": "Supercalifragilisticexpialidocious"
      },
      setup: {
        selectors: {
            ".story": {}
        },
        CORScredentials: "omit"
      }
    }
    script = document.createElement('script')
    script.src = 'https://unpkg.com/hyphenopoly@4.5.0/min/Hyphenopoly_Loader.js'
    document.body.append(script)

Re: Show HN: A Reddit reader that looks like the frontpage of a print newspaper

#87
post #80

A note to those using Firefox who get stuck at "Loading stories...": Enhanced Tracking Protection (on by default?) blocks the request. To disable it for this page, click the little purple shield next to the URL.

The Firefox tracking protection is blocking https://api.reddit.com classified as "may external ads, videos and other content with tracking code". Odd.

Re: Show HN: A Reddit reader that looks like the frontpage of a print newspaper

#88

It hangs on "reading stories". Web app authors can be very creative, but one thing I don't like about web apps is that often their error handling is nonexistent. It's not about individual authors; behemots like Microsoft also do this in an irritating way: "Something did not work. Reload the page please" -- Is this really the best a billion dollar company can do? Well, I suppose this is still better than no error at a…

They (MS etc) know the error alright, but users get scared and overwhelmed by technical language and error messages. So they put some picture of a sad cat and a message like "zoinks, that didn't work, try again". And anyway, if you're going to reload anyway, why would you care about the exact error? Hopefully it got logged in their systems so the engineers can come back to it. Beyond nerd curiosity, why would a user…

How about the diagnostic tool you can launch in Windows. It works away for 5 minutes then says it couldn't fix it.

Re: Show HN: A Reddit reader that looks like the frontpage of a print newspaper

#90

It hangs on "reading stories". Web app authors can be very creative, but one thing I don't like about web apps is that often their error handling is nonexistent. It's not about individual authors; behemots like Microsoft also do this in an irritating way: "Something did not work. Reload the page please" -- Is this really the best a billion dollar company can do? Well, I suppose this is still better than no error at a…

It actually does handle errors properly -- the reddit api is probably blocked for you, which is why it hangs and can't do anything. I suppose I could have added a "it seems we can't reach Reddit" kind of message, but that would still have been an error state, and in rushing this out in an afternoon it wasn't much of a priority. Didn't realize until posting here that apparently Firefox blocks the reddit api from the b…

It doesn't load for me in Chrome either.

In the function fetchRedditStories, it looks like that the callback param in the catch is named `e`, but it tries to log `error`.

Checking the network panel I'm getting a 200, but I'm pretty sure it's failing anyway because of the ReferenceError

Post reply on HN