For those curious, DiscoverDev is a fully static site hosted on Netlify. It's all handcrafted vanilla HTML and CSS. (The first site I've ever developed, zero design experience, so nothing fancy). Zero lines of Javascript as well :)
1. I've written my own crawler+parser which parses selected blogs and publications and then displays new articles to me in a chronological manner in a GUI.
2. I select the articles myself, and tag them.
3. Then a script sees all the selected articles and generates a JSON file.
4. I've written my own static site generator which consumes this JSON and spits out the updated website and RSS.
5. I push the changes to my git repo.
6. Netlify listens to the git repo and updates my build.
7. People see my updated webpage and RSS feed within seconds.
Thanks for taking a look.
-----
Subscribe to the newsletter if you're looking forward to receive a weekly digest of interesting articles and resources. You can unsubscribe anytime, I keep my content to the point.
And as a result, the site is fast as if made of magic. Thank you for doing it right.
> The first site I've ever developed, zero design experience, so nothing fancy
Design is good too. It is utilitarian, just at it should be. I'm increasingly convinced that getting "design experience" at some point starts to cause problems. At least if one's getting that experience without getting some extra "common sense" and "be nice to other people" experience.
One design bit I would change here is the monospaced fonts. I realize that monospace signals that this is a "coder's" site, but I find it hard to read. I don't even use monospaced fonts when I write code, as I prefer the readability of proportional fonts.
As a quick test, I changed the font to Arial, removed some of the bold and italics, and reduced some of the font sizes. (I like big fonts, but you don't need as large a font size in a proportional font. And there is probably a better font than Arial, it was just the first one that came to mind. )
One proposal regarding utilitarian design: As this already uses old school blue links, it would be nice to also have visited links in purple. Like in the olden days :) Makes a collection of links much more useful.
It's not about brutalism. It's like that "make it, make it good, make it fast" adage for optimization, for design it should be - "make it, make it good (for the job it's supposed to do), then make it pretty". Alas, I see the trend is to start with "make it pretty" and then often not even continue to "make it" part.
I don't think there's anything wrong, per se, with the styling as-is on the site. It's clean and it works. That said, a few typographic tweaks (increased font size and spacing between items, introduction of a scale and vertical rhythm, etc.) could improve readability a bit without introducing unnecessary clutter or impacting performance. That's especially true with a monospace typeface.
In any case, there shouldn't be a dichotomy between "make it pretty" (however you want to define "pretty" for a project) and "make it good." If a designer is creating that dichotomy, they're doing something drastically wrong. That's true for any project, IMO.
I like the site and the idea of it, so I took ~15-20 minutes and threw together a simple user style for my own use that keeps to the site's overall aesthetic while adjusting for readability a bit.[0] As the screenshot shows, the changes are mostly just typography tweaks with the biggest change being a dark design (though that's mainly just personal preference). I'd still check the site without the changes, but I do think they help a bit.
"Make it" without "make it pretty" leads to things that are raw and unadorned; French for raw is brut; the design school-of-thought based on rawness (no design is still a design) is called brutalism; you just described it perfectly; thank you and good night.
Hm. Fair enough. I guess I still had in mind the recent "brutalism on web" articles that showed stuff which really had nothing to do with actual brutalism-the-architectural-style.
For folks who want a different font, firefox's, "reader view" is helpful - but doesn't work here. This apparently means something wrong with the site's html etc.
https://www.howtogeek.com/268116/is-there-a-way-to-force-ena...
In Android firefox, when I force it with
Thanks for your ongoing work!
Would it be possible for you to create a duplicate rss feed where the articles are the root and not hidden in the description ? It would help keeping up with the pace by speeding up the scanning process (I only check for my feeds once a week or so, I have tons of links to go through).
I realize this would hide the 'daily' aspect of your effort, and I would totally understand if you declined.
Thanks again!
Oh hey! I used to do this before but a lot of people started complaining that they then had duplicate items in their RSS feed, like the one from the website and one from me as well.
Plus it will be hard to figure out which ones are coming from DiscoverDev, no?
In Feedly the source feed is always displayed, I thought it was the same in others feed readers as it seems a basic info.
I understand the duplicate issue for those following both DiscoverDev and the discrete websites, I wouldn't mind as it is intentional but that's just me.
Oh would I love to do hardware (as I got into software through hardware itself)! But I feel the project will become too broad, and wouldn't be for any one group.
I do include a bunch of hardware/maker resources in the newsletter though!
Well, I wanted to automate it, but the "AI" just so easily got fooled, though if anyone is interesting in working on it with me I'd be happy. Now that I've been running it for an year, I have some unique insights that would be helpful.
If you are planning to eventually automate this with some sort of recommendation engine, then I hope you're at least capturing all the data and analytics currently to help drive that in the future!
A small, subjective point of improvement: #fff and #000 are very high contrast, I find pages with a slight muting of these two colors (ie, #fcfcfc and #030303 - an offset of 3 in rgb) to be much easier to read. Regardless, I love the site. I've bookmarked it and will visit alongside my morning routine.
I maintain a similar site here, but it’s all hands off, RSS feeds + hn style voting interface: https://engineered.at problem is no one votes so it’s basically chronological.
The curation quality of your site is fantastic! Keep it up!
Wow, the design is amazing -- I am really not good with design but it is very simple and extremely very clean. I might steal it at some point. I also really like the no javascript approach. You even have a RSS reader, which is really helpful but I have one little nickpick, the date format is: "Fri, 11 May 2018 00:00:00" which is slightly wrong (you forgot +0000 as the time zone). Good job anyways!
There is a lot to like here, but I wish viewing by tags had a more obvious sort applied.
Some items included aren't of interest to me, so naturally I'd like some way to filter those out. Viewing a tag is a good way to do that, but then I lose the chronological organization.
Not sure about this, but the items on tag pages appear to be in reverse chronological order. It might help if there was a date displayed beside each item.
I like HN for what it is, I get to learn form a variety of topics I wouldn't have encountered otherwise. Discover dev was made to solve another pain point - I prefer tech only news, like a bunch of in depth articles without any commentary.
The idea was that engineering teams all over the world produce some really great content on their blogs, but find it hard to get any traction given all the SEO optimised blog spam, so I wanted to create a resource that would promote the good articles to those who seek it.
Later I also expanded to include a couple of tech publications like Hackernoon, code mentor, etc (but you'd be surprised at the number of blog spam and blockchain articles that get produced there - I filter out all but the interesting / in-depth ones).
1. I've written my own crawler+parser which parses selected blogs and publications and then displays new articles to me in a chronological manner in a GUI.
2. I select the articles myself, and tag them.
3. Then a script sees all the selected articles and generates a JSON file.
4. I've written my own static site generator which consumes this JSON and spits out the updated website and RSS.
5. I push the changes to my git repo.
6. Netlify listens to the git repo and updates my build.
7. People see my updated webpage and RSS feed within seconds.
Thanks for taking a look.
-----
Subscribe to the newsletter if you're looking forward to receive a weekly digest of interesting articles and resources. You can unsubscribe anytime, I keep my content to the point.