Hacker News new | past | comments | ask | show | jobs | submit login
Why ElementaryOS Left Medium (elementary.io)
315 points by amaccuish 32 days ago | hide | past | web | favorite | 130 comments



Yeah, so basically modals, banners, reminders to use the app etc. (vide https://www.webdesignerdepot.com/cdn-origin/uploads/2019/06/...) are new toolbars of the late 2010s (http://i.imgur.com/X7ipc.png).

I don't like the direction where the Web is currently heading. Shutting down open APIs, forcing users to use the one and only official app (Facebook Messenger, Instagram, Reddit's starting to do so more and more), ubiquitous tracking, aggressive ad targeting and so on.

How can we - users, programmers, hackers - fight this?


Well, the fundamental problem is that Capital comes in and builds a really great Thing, a Thing that may not have been built without Capital, but then Capital wants a Return, which causes the Thing to decay. The thing Hackers can do, as in the OPs article, is to salvage the good ideas that the initial efforts of Capital managed to prove, and leave out the rest.

This won't work for individuals, in general, because writing good software is hard. Even if you rely heavily on 3rd party libraries like Jekyll, you still have to know how to pick those libraries, and how to use them, how to deploy them.

Personally, I think it's cool that Capital is willing to get into the fray, even though it so often ends badly. Often it does contribute something, even if it dies. People get mad at Google for axing projects, like Reader, but why? Consider all the market validation they did, for you, for me, for free. Consider all the product design work they did, for all of us, for free.

I don't even think it's bad that devs at Medium are, almost by definition, currently working on nothing but anti-features. It's just part of the Capital-driven software lifecycle, and they are feeding their families, and learning a lot. I can't even hate the managers who made unethical product decisions, because it was always going to be that way, and better to participate in the Capital driven software lifecycle and contribute something then to never have participated at all.

There are amazing, sustainable, ethical, Capital-driven software ideas still waiting to be done, and that couldn't be done without Capital's involvement, but clearly that's not Medium and its ilk. But as Medium declines and eventually dies, remember that this is the fate that waits for every project, no matter how successful, and they leave behind valuable information for us.


I like to look at it this way: bad ideas, or bad executions of good ideas, or bad fits for VC, that somehow get funding and die before going public, or even better, get purchased by some bloated old company with too much money; the end result is a wealth transfer from the ultra-wealthy to the middle class. Paychecks get paid out. Stock options liquidate for something usually. Services get purchased to support the business even though they may not be worth supporting.

Right now, the VCs behind Medium are doing the world an amazing service. They're going to see almost no return on investment, but why would any of us "normal people" care about that? Its their money they're wasting. They're paying employees, probably a great wage. Probably providing a ton of benefits to attract talent, including lots of options, great insurance, snacks in the office, unlimited PTO. They're producing a product that is actively teaching competitors like write.as and ghost what Not To Do. I'd bet even most of their employees hate the direction they're going, which means they're learning as well.


>Consider all the market validation they did, for you, for me, for free. Consider all the product design work they did, for all of us, for free.

I have a feeling that their intent wasn't to do it for free. What happened is that they couldn't gather enough information about their users to turn a profit. Yes, that's what capital does, but the end result was not free for any of its users - they gave up personal data to use it.


> and they leave behind valuable information for us

Leaving information so far :

1. https://svbtle.com

2. https://posthaven.com

3. https://www.blogger.com

...

Not sure, but self-hosting was ( is ) the only way for me to sleep safe.


> But as Medium declines and eventually dies, remember that this is the fate that waits for every project, no matter how successful, and they leave behind valuable information for us.

This statement makes me so angry with Medium's corporate communications. It makes total sense to me that you'd think this, but I also think it's dead wrong.

The problem is that Medium is not clear that they've just straight up changed. So it's not that they are the big free open platform they used to be with an annoying "upgrade to pro" ad model on top of that.

Rather, Medium is primarily a subscription service. Even though they aren't saying this, it's best to think of them as an open-platform version of Netflix where anyone could post an episode or show and get paid for it.

That switch came with a switch in their incentive structure. The incentives for posting on Medium used to be that Medium will bring in a lot of traffic from their network. But they now limit that to articles that are published inside their subscription service. So there's not really any reason to post outside of their subscription service.

This is what I think you're experiencing: the free, non-paywalled portions of Medium are dying because the incentives for publishing went away and the aggressiveness at advertising their actual product (subscribe for, they hope, great content) has become more aggressive.

Meanwhile, while the experience that ElementaryOS was using Medium for is dying, there's a new experience that's quite good and I think plays a nice balancing role in the overall Internet ecosystem. Medium is making a place that rewards quality writing, tries to eliminate writing for the purpose of marketing, and is pouring money into authors and editors so that they can invest more quality into their articles.


* > Medium is making a place that rewards quality writing, tries to eliminate writing for the purpose of marketing, and is pouring money into authors and editors so that they can invest more quality into their articles.*

Can you explain this a little? I don't see much evidence that Medium rewards quality - unless you consider popularity to be a proxy for quality, which we've all learned, painfully, is false. Certainly Medium doesn't offer traditional publisher-style quality improvement services like copy-editing, editing, etc. So I'm curious what you mean.

It's also not clear to me how they dissuade "writing for the purpose of marketing" - plenty of people seem to use Medium as a straight-up content marketing platform, e.g. consultants with newsletters driving people into their funnel, and they want a URL to park their content, for bookmarks, and a place for shared responses.

If Medium actually did these two things, I'd say that they were indeed doing valuable service. But it feels a little like the difference between ideal and real value proposition that the Google Play Store and Apples App Store have: theoretically those walled gardens can enforce higher quality, safer software for end users, and lower friction to revenue for devs, but in reality, devs pay a straight 30% of revenue for that low purchase friction, and whatever illusion of value the platform can plant in users minds with good marketing and PR.

Maybe you are saying that Medium can, rather than improve actual quality, build a reputation for focusing on quality. Then writers can enjoy a similar halo effect as devs on the App Store? It's probably worth trying, from Capital's perspective, but not from mine, because I care more about innovative product design (which Medium has, in abundance) than rehashing a well-worn, slightly devious, business strategy.


I think you're proving my point that Medium needs to clarify their brand. They are actually doing all the things you mentioned.

Here's what I can see them doing.

1. You can't get distributed by their algorithm if your article has blatant sales or a call to action.

This removes the incentives for a lot of content marketers. Medium literally has a human read and filter every article behind their paywall. That's thousands of articles every day. It's not very sophisticated, but it's good enough to push back on most of the marketing pieces.

2. Medium's promotion algorithm relies heavily on their Topic feature and topics are all manually curated by editors.

Topics, i.e. Programming, Data Science, etc, trigger most of Medium's distribution paths. They get an article promoted to people in the app and they get the article in the mix to be promoted over email.

3. Medium's Owned & Operated pubs (their term for their internally run ones) do very heavy story edits.

A copy edit is when you do light grammar and spelling and occasionally wording changes. But a story edit goes much deeper often taking several hours. Many of these articles are original commissioned pieces, but I know Medium often "elevates" (their term) an article from the general platform. Here's an example of how deeply an elevated article gets edited: https://twitter.com/tonystubblebine/status/11533445070871633...

4. Medium is increasingly paying for copy editing

I run two publications that do copy edits. They are my copy editors, but are paid for by Medium. These pubs are community aggregators, i.e. they publish a lot. I think there's some value in this even though I know the highest quality articles require a lot more time and editing. Anyway, we copy edit about 500 articles a month.

5. Medium is experimenting with paying for external story editors.

I think my original pub, Better Humans, is the only partner publication they have that gets paid to do a story editor. We have someone with subject matter expertise spend 4-5 hours on each piece reading each sentence for "is this true?" and "could someone follow this advice?" And we do all the normal editing things like sending a piece back for revisions.

---

In sum, I think Medium is trying to have it both ways.

One way is a high volume of "better than the rest of the internet" content. That's how my two new pubs operate. We just want to clean them up and filter out the worst ulterior marketing motives. But we are definitely volume first.

And then in a different dimension, they want to have prestige publications that feel must-read. I think this dimension is much harder, but it looks like they'll get there. (And I know they want to).


> Medium is making a place that rewards quality writing, tries to eliminate writing for the purpose of marketing, and is pouring money into authors and editors so that they can invest more quality into their articles.

As someone who by principle wont ever pay that subscription, my experience is quite the contrary having all those metrics you say, going to the shit, starting from quality content...

Maybe some great articles get hidden behind the paywall but mostly to me it seems all low-quality click-bait shit now, and on top of it they put the modals and stuff...


We made the transition from just publishing anything that got submitted to doing much deeper editing. Maybe not the topic you care about but here's an example of what we were able to do with an editorial budget: https://link.medium.com/uMM1k1cgjZ


Just curious, which principle is that?


> Well, the fundamental problem is that Capital comes in and builds a really great Thing, a Thing that may not have been built without Capital, but then Capital wants a Return, which causes the Thing to decay.

I feel like most of the examples of this are the other way around: Great Thing is built without a lot of capital (other than a bit to get started, if even that), then gets a bit of traction, and then Captial comes in and the demands for increased ever revenue starts, even if it might not lead to profits.

We really need to have a concept of a sustainable startup, not just "hyper-scale or bust". That is what normal people call a "business".


I have to politely disagree about Reader, it was already proven and working well, what they did was a bit of a scorched earth tactic, other companies couldn't compete with free and died. Then they pulled out and left a vacuum which has lead to a significant decline in in the usage of RSS in the modern web.


> How can we - users, programmers, hackers - fight this?

By doing exactly what ElementaryOS did here: stop using these platforms. And if you have to use them (as a user), try to use them in the most open way possible: Never via a dedicated app, preferably via RSS or similar for the platforms where it is suitable and where it is supported (like youtube).


This is correct. The readers should not be the ones paying for this, it should be the publishers who use the platform.


Improve existing platforms: Wordpress, Ghost, etc.

There's none of that nonsense on those open source platforms.

An aside: Tumblr isn't hostile like Medium is. At least not last I used it. I prefer seeing Tumblr to medium you can really just do any kind of blog you want whether it's a meme feed, or some serious blogging.

Edit: Forgot to add, also Tumblr lets you optionally advertise to non-logged in users to join Tumblr, and it's not a modal, just a button on the top right hand corner where the "Follow" button shows up to logged in users, or you can hide it entirely.


Haven’t you noticed a glaring difference between these platforms, namely the fact that Wordpress and Ghost have a completely different business model and basically no social layer to distribute the content?

That’s why they don’t have such problems – but this is also they reason most people don’t care about them.


My point somewhat still stands though. You've got Tumblr which has a social layer through likes, reblogging, and even "Ask'ing" and messaging other users.

Then you've got projects like Mastodon, and specs like ActivityStreams that have not been adopted by Ghost or WordPress yet, but there's no reason plugins can't be made, these would add a social element to these blogging platforms and turn them into federated blogging platforms. This is just a side thought, but it's not impossible for WordPress and Ghost to get there. WordPress is used more than most realize.


WordPress powers 1/3 of all sites on the internet . . .


And it's getting worse by setting the wrong priorities.

OPs point stands though, as there is no platform like YouTube where you can easily discover similar content. wordpress.com is not really the place for that, either.


Since the whole oath & GDPR thing happened Tumblr is actively hostile to EU visitors. That might not be you, but basically to see any tumblr content I have to agree to a kitchen-sink of data collectors, trackers and other stuff.

Medium still works like a charm, especially with JS turned off.


As users, a step in the right direction would be to vote with your wallet, not use nasty services and prefer respectful services wherever possible.

As engineers & designers, we need to stop being complicit and building this shit nor working for such awful companies. If the majority of engineers become ethical then companies like Facebook, Google, etc will have a hard time hiring people to do their dirty work.


> As users, a step in the right direction would be to vote with your wallet, not use nasty services and prefer respectful services wherever possible.

I'd love an app where I could define companies, people and perhaps countries I don't want to pay so that I'm notified when browsing to a site, or scanning a product to buy so that my time and money could be invested elsewhere.


Adblocking is a good start. For physical products .. it's a lot harder. I don't want to go full "no ethical consumption under capitalisim", since that's useless, but basically everything that isn't specifically marketed as ethical probably is doing something wrong in the supply chain somewhere, and half of the "ethical" claims are dubious too.


> I don't want to go full "no ethical consumption under capitalism"

True but one step at a time in the right direction is all it takes for change. Say if I got shafted by Expedia. I would like to add them to my "do not buy" list and have their other brands (e.g. orbitz, travelocity etc) added to the do not buy list. Could make it social so that if my sister got shafted by some company I could see that too and opt for an alternative. There are often stories about unethical companies (Monsanto for example which is now owned by Bayer). I'd like to avoid them, but the burden of doing so prevents it. Lower that burden, provide alternatives and maybe over time we'd see improved ethics and accountability.


Is there a way to restrict your card so you can’t buy things from certain vendors? Then you could subscribe to a list of boycotts based on which ethical questions are important to you.

I’m not sure how else to keep track of everything you want to vote-with-your-wallet against.


Monzo API might make this possible....


Groupthink, impostor syndrome and 'fitting-in-ism' rule the day Nextgrid, there's no chance you're going to get a fresh grad out of university fighting for 'ethics' when they're too busy jumping hoops for their management overlords


Or more simply just trying to make rent, school loan payments, etc. When you're broke, ethics are hard to keep when someone is promising lots of money.


agluszak, I've been around modern technology since the very beginning, before the Apple ][ days when the web wasn't much more than a gleam in Tim Berners-Lee's eyes and 'Money for nothing' by Dire Straits was the cutting edge of computer graphics and MTV was the closest thing anyone had to streaming media

I've witnessed the rise of the machines from a unique vantage point just across the y2k horizon which makes me uniquely qualified to describe the evolution of the so-called 'technology professional' born of the nexus of home-brew computing, video gaming, sci-fi and old corporatism -- the modern 'tech bro' if you will

I tell you for sure it's the attitude of the casually invested subtly egotistical tech pro intermingling with the traditional corporate stereotype that has brought us here and the only way out is to bootstrap ourselves out of the shadows and take over the system from the inside out by becoming sentient adults rather than code weaving pithicathropus. Our black monolith moment is way overdue


Upvoted for being pithy and beautiful. Where's our manifesto? Can I suggest you write it?


> How can we - users, programmers, hackers - fight this?

my_article.html

Uploaded into a public directory on a web server.


As users, I guess we can start with paying for good content, tools, apps etc. Most of us don't even pay for something as important as email, and then complain about Google's non existent support for GMail etc.

The other thing we could do is intentionally support good, honest indie makers over mega corps (but there is no guarantee they won't sell to mega corps).


Disincentivize company founders from cashing out to the detriment of their users by moving the cradle of these technologies to a place where you can respectably buy a house and raise kids without millions of dollars.

We did this to ourselves through decades of equating personal wealth to community and social value.

Which is ironic given the economic purpose of hacker news.


By use a diverse set of browsers and minimally using apps, we can slow down the silo'ing of the Internet. If users don't blindly buy into every app out there, big corps will have little choice but to interoperate on the open web.


But the big companies do everything that's possible to stop us from using browsers and plain, old websites. Users are not blind, but choosing not to conform to corps' will requires knowledge (why and how, because dark design patterns are called dark for a reason)


I quit using Imgur because of agressive promotion of their app. Actually you can't even upload an image, neither view a full-size image on mobile version now.


Also by setting good examples and not using this shit ourselves, even if it's the expedient option.


There needs to be an app that has the ability to connect ( via oAuth, whatever ) with DigitalOcean / AWS and that can generate a static website out of a WordPress(-like) editor, keeping the state in the cloud - again DO or whatever in a managed database.

The app should cost <20$ ( or be open sourced ) and it should have a huge eco-system of themes / plugins, including paid ones.

Sadly this app will never have an IPO or growth rate worth investing, but it will fix the problem ( for a while ).


I think we should code around them, instead of writing comments on HN.

Me included. I previously wanted to write an elaborate comment, but I should really focus more on productive endeavors.


To be honest, I find writing comments and discussing on this topic very important. Maybe on HN people are aware of such practices and know how to mitigate them, but the general public need to be educated, otherwise they will be completely helpless and at the mercy of corporations relentlessly trying to "monetize" them.


Avoid those sites, and when asked for feedback mention the problems (a great place for feedback is the customer serivce chat, followed with a FU, enter and a ctrl+w)


>How can we - users, programmers, hackers - fight this?

Pardon the interruption -> close window.


Well it is being fought on some level. More and more blogs have moved off of medium and hosting their own. I sense that rss is making a bit of a comeback as well


Yeah, but I meant the more general problem of "modalization" and monetization of the Web, biggest service providers like Facebook or Google shutting down their open APIs and pushing us not towards common, open web standards, but towards some sort of closed bubbles or silos. Rss may have its own niche, especially among the hacker community, but most non-tech people will rely on Facebook or Twitter for staying up to date with sites/topics/people they wish to follow. And both FB and Twitter are neither open nor free (as in freedom)


I believe the only way to combat this as users is to use agnostic tools. When a website, program or other service wants to force you to use a specific thing to access them just view in a web browser with ad/script blocking or choose to not view it at all if that isn't an option.

As programmers we can refuse to implement such nagware. There's not much worker solidarity in our field so quitting is usually the only option if the company has made up their mind on the matter.

Overall I don't believe it can be stopped at this point. This process is called privatizing the commons and it will continue to happen to every common good as long as capitalism is the system which we work under.


Your point on worker solidarity got me thinking. In some countries(such as mine) we have decent unionization rates for IT workers, would it be a good idea to have such ethical issues encoded into the union lines?


Absolutely. If you feel you are creating a product that has a negative impact on the people you are serving you should do something about it. Unions can greatly benefit society should they choose to.


But it appears that the issue of ethics in IT is almost non-existent. From a perspective of a CS student - everyone around wants to work at Google/MS/FB because that's the bleeding edge of technology, but no one gives a shit about corporation business practices and their consequences on privacy and digital wellbeing of other people.


Do you think we could do anything to change the industry for the better?


I wouldn't be surprised to learn that there is actually a code of ethics within my union, I'll try to seek it out and see what we may want to do. Could be cool.


If you do find it could you let me know about it? I'd love to see it.


I guess you meant late 2000s? We are in the late 2010s right now!


Yes, and I meant 2010s, because that's when all these aggressive modals urging you to log in/subscribe/etc. started to appear. This may be a bias, but to me the Web used to be a much more peaceful place when a simple adblock sufficed to browse it without pain


You know what's creepy ? "You read a lot. We like that."

Or install our mobile app and get more content, or signup and get more content we figured with our fancy algorithm for you.

Here's the thing, Medium never worked for me. I had over ~300 followers and if I published something none of them would see it, no reaction, no visit, no read at all and I repeated the same thing over and over again with no luck.

I'd only get visits and reads on my articles when I shared the articles on social networks.

It seems your followers, real or fake don't matter are spammed with Medium algorithm to read "recommended" content.

I'm a developer, I like simple thing, but I don't like using static site generators to build a blog out of rst or md files and push to github etc.

I want a community around my blog, I wanna nice simple UI, a simple commenting solution, a place to upload my files and attach in posts optimized for web.

Plenty of blogging platforms may or may not provide all of it out of the box, but each with their own weird ToS and privacy policy.

I simply had enough of it and created my own blogging platform, https://www.gonevis.com where I have my own blog https://alireza.gonevis.com

I'm done with weird platforms where things are shiny and promising until they're not.


All they would have to do, in order to have a functional and transparent recommender 'algorithm' is to rank order articles by how many of your followers liked, or read, an article... so if you have 100 followers and 50 of them happened to like another article 'B' then that article would be highly recommended to people reading your article. If they knew who was reading the article (logged in) then they could simply exclude articles from the list which the person had already viewed. That would help people browse through groups that shared similar interests.


Thing is they want to put more recommended content to your face as much as possible without knowing much about you, otherwise your scroll bar will hit the bottom quickly.

The recommendation engine from technical point of view is not a rocket science, they're just simply trying to tune for maximum engagement even if it costs them repeated nonsense content you don't like which tbh defeats the purpose of "recommendation".

It seems there's not much of community in Medium. All viral articles are the ones that the author had a good influence over their twitter of facebook etc audience.

If you already have 10s thousands on your twitter follower list that you worked hard to find and bunch of people know you for some of your other work, no matter Medium or a static generated blog, your article will go hand to hand and viral in many cases.


> I had over ~300 followers and if I published something none of them would see it

i hate this. it's the same with youtube. if i "follow" an author, or "subscribe" to a channel, it's because i want that creator's content; not content by other creators related to the same topic.


but they even can't handle singups correctly if you have 2 google accounts in one browser - show them correctly, you select one, it redirects, and it's still not authenticated...


Every time I follow a link to Medium they seem to get more and more pushy. I shy away from clicking on links there anymore, so news like this is welcome.

I must say I don't get what unique value proposition Medium think they have that allows them to treat both publishers and readers like crap.


> I must say I don't get what unique value proposition Medium think they have that allows them to treat both publishers and readers like crap.

My take is that they had a unique value proposition at the beginning: "YouTube for text: audience and authors all at the same place, with recommendations and unified comment platform". But then it became apparent that they will not magically become profitable or bought, so they started with the crap.


The problem is, with a lot of companies trying to be "YouTube for X", they left out the monetisation stories.

The reason a lot of creators have bought into YouTube is 1) it's still expensive to host video elsewhere 2) there's constant noise around the earning potential (as misguided as it might be to listen to it).

Medium doesn't have the depth of content to become a destination wholly in and of itself as they don't have a critical mass of creation. All they needed to do was allow creators to make some decent cash, get a few wild "this author is earning a $1m a year, just on Medium!!11!1!1" stories into the mainstream, and suddenly you'll have an influx. Then, just don't be a complete arse to those creators and keep the cash pumping.

I don't really want more ads in the ether, but that would have been the most obvious way to juice it - have their own self-serve ad product from ad one, and the offer the options to remove down the line.

Instead, we get dialogue boxes.


I wonder if they could have started as a sort of readability/patreon style platform for text without taking any investment and build just enough audience to become profitable. I imagine that running their servers should not be massively expensive, especially if they scaled up slowly.


But then you look at Ev's history and realise that's probably not a direction he'd have been exploring unfortunately.


It's not like Youtube is not annoying and filled with crap (ads).


It's true, YouTube ads can be annoying in the sense that all advertising is annoying. But I personally find Medium far worse.


last-gasp monetization


It's easy to build an audience on Medium, that's what draws authors there


Predictable audience? Absolutely not.

It's easy for a story or two to go viral (as in, slightly easier than on other platforms). What's not easy is to retain that viewership. Similarly to YouTube, you could put weeks worth of effort into one article just for it to flunk spectacularly simply because the recommendation algorithm didn't like it in the first 12 hours or so since you've clicked "publish".


My problem with Jekyll (or other static generators) is that it's not as straightforward as anything with a web gui, like wordpress or ghost (or medium or tumblr or...). You still need to handle things manually (e.g. the yaml front matter), the preview is not "out of the box", different plugins/extensions may add unpredictable behaviour, you need to setup a build/deploy pipeline, etc. - the experience is just not streamlined.

I think that when you want to write, you shouldn't need to mess up with tech details of your system. It's distracting. Of course I'd like a static website for deploying, but I'd love a web gui which acts as a frontend for such generators. I think that Prose was something similar but I found it buggy/not updated, while another saas I tried(Forestry, I think?) was good but you couldn't customize everything.

Ghost is my ultimate choice until an official, core-developers supported web gui is created for jekyll, hugo, or whatever static generator lies out there.


The main issue with Jekyll is the blog post system. It requires all the blog posts to live under `_posts/<date>-<slug>.md`. So now the user has to invoke a command-line tool to create the proper slug. Then if the post doesn't get published on the same day or the title changes the user has to rename the filename. Images live in a different folder and get all put together.

A better implementation would be: render all the markdown files in the repository and map README.md to index.html. Now you can create your own file structure, have your images with the article, and optionally add them to the RSS feed by adding a `published_at` in the frontmatter. Or even better use the git history to find out when it was added (and changed) in the repo. You get all the "admin" and rendering for free from GitHub.

I guess my point is that there is a lot of friction inherent to Jekyll's blog post publishing system. Since I adopted that other approach it's much easier for me. I love to only focus on the text and not worry about the rendering. And be able to edit it with my favourite text editor.


As a proud HN commentor you should have built your own static site generator long time ago. Preferably in Assembler or some esoteric unknown language!


As much as I love Jekyll, I have to agree with your points. Jekyll isn't very accessible with it's specific front matter and file naming convention. However I don't think Jekyll will ever have a proper GUI.

You can have the best of both worlds though. You can use gulp.js to build Jekyll friendly files coming from Ghost using the Ghost Content API: https://david.darn.es/tutorial/2019/08/11/use-ghost-with-jek...


I used Jekyll for a long time before I got tired of it for the exact reason you mentioned. I used write.as for a while until I found rwtxt (https://github.com/schollz/rwtxt). You could self host if needed.

It's extremely minimal, but I found that it's perfect for my needs. Sample blog: https://rwtxt.com/kevinfiol


I tried Hugo before but came to the same conclusion. The simplicity draw me into trying it, but in my opinion the solution is not streamlined.

That's why i finally settled for a Flat File CMS [1], in my case Grav [2]. Like with hugo you get total control. And with most of them you get a single web interface to administrate the site, manage extensions and write blog entries. And you still have the possibility to git push everything on the dev server and git pull on prod, without managing a database. It's a dream.

They are not without problems though. You need to host your php site somewhere. Sometimes it's necessary to dive into php to fix some problems and creating something unique still takes some web skills (in contrary to wordpress for example).

[1] https://github.com/ahadb/flat-file-cms

[2] https://getgrav.org/


I would love something with a UI like Grav to edit and create my site, but with a static web page export to publish it. I'm fine with running the UI only locally on my computer so even a desktop app would be fine.


There's a plugin [1] for that, but I haven't used it before. It creates plain old html files.

[1] https://github.com/BarryMode/grav-plugin-blackhole


> Of course I'd like a static website for deploying, but I'd love a web gui which acts as a frontend for such generators.

We used to have that with MovableType and (less known) Bricolage.

Today there is Lektor (does it support Python 3?) and that's about it.


Yeah my dream blog is basically Ghost that deploys static posts when I hit Publish/Update. I’ve been slowly working on something like that for myself.


Have you looked into doing just that? There are a number of options and more arriving all the time.

https://ghost.org/docs/api/v2/gatsby/

https://ghost.org/docs/api/v2/eleventy/

https://ghost.org/integrations/netlify/


It feels a bit overkill to pay $29 per month just for a hosted blog UI. And if you're running Ghost yourself on a server or on your computer, it's not really a static site anymore.


From the OP:

> Of course I'd like a static website for deploying, but I'd love a web gui which acts as a frontend for such generators.

If you want such a gui on your local machine you're not much more than a few commands away

  npm i -g ghost-cli
  ghost install local


This looks promising. I'll check it out.


A good compromise is something like Netlify CMS[0], which adds a GUI on top of a static site generator so you can write posts through a web interface. AFAIK it works with a variety of static generators, so you have a few choices. Of course, you're tied to the netlify infrastructure.

[0]: https://www.netlifycms.org


I used it. Nice but incomplete, and still needs manual tinkering.


I'm using Pelican [0] and found it really easy to setup and use.


I use Jekyll + one admin plugin. Work locally, then do jk build and ./deploy.sh and everything goes smoothly. But well, it's not pretty interface to deploy.


I always liked Joel Spolsky's CityDesk software. It's way out of date now but the basic principles were good. I'd love to find a modern equivalent.


What about a local Wordpress instance + a script to scrap content and publish it as a static website? Best of the two worlds without too much fiddling.


Are you sure it's easy to 'scrap' ? What if the plugins render something conditionally depending on e.g. the user agent or the current time or other variables? Faking a static website through a mirror doesn't work nicely.


I thought this would be another rant, but this post was actually very well written and really informative. Key points are:

* Medium policy changes and how they affected readers and authors

* The importance of "owning" your own posts

* Possible alternatives to medium

* Issues with CDN wrt privacy

* Issues with commenting services

* Use of custom fonts

* Sharing to social media without being privacy invasive

* Google AMP, and how Google strong arms you into using their tech

... and much more. Highly recommended!!


I myself have left Medium and cancelled my subscription. When asked, I told them it was annoying to see the modal popup asking me to sign in every time I looked at Medium from my work computer.

The popups, as mentioned in the linked article are annoying. As is the increase in the number of articles you have to pay to read. Not to mention the fact that the trolls have discovered Medium, and the quality of writing there has gone down (One article prominently linked was about how one needs to “ditch loyalty” to have a healthy relationship).

For my personal blog, https://samiam.org/blog, I use a homegrown CMS that makes static content I wrote using UNIX shell scripts, with a bit of Perl and PHP, about a decade ago. I updated it to use web fonts once those became viable about five years ago, and updated its color scheme based on my wife’s wishes after she passed away.

For reading stuff online, I have a subscript to the New York Times which keeps me up to date on the news; I also read articles linked to here (and I haven’t seen Hacker News link to a Medium article in a while).


Your blog is a huge breath of fresh air. I genuinely miss blogs of this style. Easy to navigate, easy to read, quick to load. Love it.


I’m quite proud of it. Not only does it use 100% open source fonts for the rendering, it also renders on pretty much any browser made. The stylesheets are set up so the same HTML + CSS looks good on both desktop and mobile browsers (we use different CSS for mobile, but there’s no nonsense like m.samiam.org). The webfonts have been carefully hinted and subsetted to be of minimal size, and look good on pretty much any browser (I made sure they look good in Chrome + Windows, even on a low DPI display, which has had a lot of font rendering issues).

For example, the http → https redirect is set up so the redirect is not done in browsers which do not support Javascript (and may not support HTTPS); older versions of Internet Explorer which do not support modern TLS encryption also remain on the http website. The stylesheets do not “gracefully degrade” (in the sense I don’t try to have the site look mostly the same in antiques like IE6, IE7, or IE8), but the site is fully compatible with Internet Explorer 9 and above and any other mainstream browser from 2010 on. It is still completely readable (albeit with a different look) in older versions of Internet Explorer and in browsers without any CSS support (and, of course, it also can be read in the Lynx text-only browser).

I have tested it in Internet Explorer 5 and above, Opera 12 (the last version to use the Presto rendering engine before they moved on to WebKit), Dillo (depending on the version, it may need to have CSS disabled), Chrome, Firefox, Safari, you name it. It can be read by pretty much any browser that exists (the Unicode will look a little strange in stuff like Mosaic from 1994, but the site is still readable).


I never understood the hype of Medium. If I want to have a clean blog, there is plenty of off-the-shelve software to do so, especially free and open source ones.

Why did IT-affine people (i.e. people who can easily manage their own installations or which can easily move to other hosters) use Medium in the first place? Maybe [because] they never used the platform without being logged in? Because that's where all the cruelity ("you already read your 3 free articles this month...") starts.


Because of eyeballs and attention. Creating a brand new site and posting on it in this day and age is basically the equivalent of yelling into the void. No one is going to see it, and the only way to get it to come up on Google is by typing in "site:[myblog.com]), barring the few friends and family members who may give you a pity click when you post on your social media.

Almost nobody is writing anything 'just because'. They're writing because they want to be noticed, either as a "domain expert" or "thought leader", or simply as a stepping stone to a writing gig at HuffPo, Forbes.com, Inc.com, whatever replaced Gawker and the various pipelines of formulaic listicle generators out there that might pay a few pennies for the next "Elon Musk's top 10 secrets for Productivity" gruel.


Nice. I did the same exact thing last week with one of my own blogs.

But I was disappointed a little with some quirks with Jekyll. For example, I really needed to list blog posts in categories separately. Turns out there's no built-in way to do that with Jekyll, so I had to find a third-party plug-in to do that. Then there were some changes didn't trigger a rebuild with the local development server running, which I (eventually) realized required stopping the server, deleting the _build directory, and then restarting the server.

These things sound minor in hindsight, but they took a lot of time for me to figure out in the moment. "It's a static site...I picked a static site generator for simplicity!" I kept thinking, annoyed.

I chose Jekyll in particular because it's been around forever and expected it to have little things like this figured out, but was slightly disappointed.

Otherwise it's been good I guess.

For what it's worth, I will occasionally repost articles on my Medium account too, to take advantage of my following there, with the appropriate rel=canonical tag. No reason not to take advantage of it while it lasts.


I'm using Jekyll for my site[0] which currently has 240+ blog posts that are filtered by tag but I didn't have to use any plugins to get that functionality. It's really fast too.

The way I set it up was with tags but that's really no different than a category. For example if you click the #docker tag, you get back a bunch of posts I tagged with #docker. Each post has a `tags` item in its front matter where I can pass a list of strings (a post can belong in 1 or more tags).

It would be too much to type in here to cover everything but if you Google for things like "Jekyll filter posts by tag" you'll find a bunch of plugin-less examples.

What's neat is you can also define a custom page title / meta description for each tag's index page using nothing more than what's built into Jekyll.

[0]: https://nickjanetakis.com/


I had the same problems with Jekyll and switched to Middleman (also Ruby based). Never looked back.

https://middlemanapp.com/basics/blogging/


I'd definitely recommend checking out Hugo too.


I second this. I migrated from Middleman a while back because I couldn't be bothered to handle ruby dependency conflicts during an update anymore. Hugo is what I use now and it's great. Migration from one to other was actually surprisingly easy too.


Same here. Generating LaTeX statically is a pain in the ass.


I always thought it odd that a team like ElementaryOS with a clearly defined "ethical code and goal" would choose Medium, which is an anti-consumer and anti-privacy company (e.g. try accessing Medium on a mobile phone with a smaller screen).

Good on them for changing.


That is explained in the very first paragraph of the article (emphasis mine):

"In 2016, elementary moved to a Medium publication to host our official blog. At the time, Medium was touted as a simple, clean, and reader-focused host for writers. They supported custom domains, a robust API, RSS, rich formatting, and great image embedding. We had been largely happy with the experience—as were our readers—but something changed in 2017."


Yes -- good on them changing, and bravo to them for calling it out and saying why. In my opinion they put a lot of effort into making this post, and it shows.


> Write.as excels as a way to painlessly publish thoughts to the Internet—anonymously or otherwise. If you’re an individual looking for a clean, reader-friendly Medium-like experience, Write.as is a perfect fit.

That's the very same thing Medium took people away with from self-hosted. Rinse and repeat?


> While we understand and empathize with the struggles of ethical monetization—and applaud Medium for refusing to monetize via intrusive ads and trackers

This part is interesting. Assuming it isn't just polite PR-speak, what is Medium being applauded for, here? Trying to use user accounts openly and directly, instead of cookies and other tracking elements to track everyone? I've always thought of these Sign in prompts on Medium (and on Quora) as just an annoyance to get past, not a possible means to "ethical monetization somehow.


I love the shout-out to Write.as. More people should check out that platform/software.


I liked that RSS support tops Elementary requirements list.


I can understand Medium wanting to make a profit but they've taken the wrong path (IMHO). Too many popups, distractions and recently reading limits.

I know I'm going to be downvoted for this but I'll give it a try :-)

Can we please-please have a native dark theme for Hacker News? Without plugins, browser specific hacks or other gimmicks, just a simple "Light|Dark" option.


At this point Medium looks like a dumpster fire.

I even have a subscription, but will cancel it.


Don't get me wrong, I'm on the Medium Hate Train along with many others here on HN, but aren't these stories getting a bit monotonous?

https://www.google.com/search?sxsrf=ACYBGNQoux3i-UEEDHgCPFks...


Aside from the aforementioned aggressive tactics of Medium, the reading experience is degraded by Medium's insistence of claiming a disproportionate amount of screen real estate for themselves. Scrolling between an oversized header and footer that promotes Medium is a miserable reading experience. Kudos to ElementaryOS. I look forward to keeping up with you in this improved reading experience and hope other follow suit. Cheers.


Good for them. A thousand deaths are not enough for Medium.


The one problem I have with Medium is the same as with YouTube. I have an account with both.

Whenever I click an article or watch a video, that must trigger such a strong certainty signal in their system that I then receive a lot (TOO MANY) related articles/videos as suggestions.

YouTube recently fixed that, but Medium has not. So I just don't click anymore. And I read a lot less on there.


Many articles on Medium are of low-quality. It used to be different. Don't know if this is caused by the monetization efforts.


It is the fate of most platforms, initially there are only early adopters that are selected, hand picked or invited, and as the platform progress there is an influx or more mainstream content producer, and the quality tends to be more average.


It used to be possible to visit medium.com and quickly find a few articles worth reading for free. Now its almost impossible to find anything on the front page without a star (paid sub required)


You could say the same for YouTube or any other "user-generated" content platform. Back in 2006 or so, YouTube was the only way to view copyrighted music videos, famous speeches, official movie clips, etc. While those still exist, the content ratio is more like 99:1 in favor of awful clickbait, "reaction" videos, obviously fake prank videos and other lowest-common denominator content. Just check the "Trending" tab and you'll see what I mean.


Well looking at the bigger picture it's not just Medium as a platform, but News/Writing/Journalism as a profession that's going hillside here.

I've long dreamed of a kind of Unsplash for news/text. Just don't have time/or care enough to do it, just would be great to have it available as a user.


I commented this 3 days ago: Medium is not for writers. They struggle to do a good business model. hurting non-profesional writers, so, it's obvious they leave Middle-Earth. If Medium only works for proffesional writers, there will be a tiny tiny portion of content compared they have today.


Medium is surprisingly usable if you turn off all JS (and I'm not usually part of the no-js crowd).


I think this happens when you don't have clear monetization strategy from the beginning.

Also, I always felt uncomfortable when reading a Medium post. It was always much more than a clear text that what I am interested.


Medium is an absolute disease and it makes no sense that anyone would use that platform for blogging. Its enforced and aggressive layers of pop-overs is a disgrace to the community.


Why should everyone care why X left Medium? There's at least a couple of articles like this every month. If you're not happy with a service, just stop using it.


Personally I like it when people and/or companies announce they've switched from service (architecture, etc) A to service B and then give a good account of why they made that decision, how they then established themselves on B, etc. Reading such articles helps broaden my knowledge about both A and B. Seeing responses to such articles on places like HN helps me deepen that knowledge.

I've seen Medium post links on various social media, clicked on them and read the articles (or got the paywall nag thing), but I never went beyond the articles to investigate what Medium itself was all about. Now - thanks to this article, and the comments it's generated on HN - I know everything I want to know about Medium, alongside some useful pointers for possible tools to set up my own static-site blog.

"Don't announce; just flounce" is often good advice for people who want the world to know what they're up to, just because they think they're important. But I'm happy to see people ignore that advice if they offer me a compelling backstory to go alongside their announcement.


Another not surprising shift away from Medium. When starting my own blog this year, I was debating between Medium and self-hosting solutions like WordPress and Ghost. When asked “why not Medium?” My answer was usually “I don’t want my content behind a paywall, I don’t like it when I’m a reader and I wouldn’t want the same experience for others.”

The other reason to me, is that you lose your own brand since your editing options are limited in terms of fonts, colors, domains etc.

I wonder now, how many blogs will follow suit? I like a lot of the content on Medium, but I don’t like the platform. It’s positive (to me) to see others migrating off.

I, personally, eventually settled on Ghost, hosted on DigitalOcean. WordPress felt frustrating to work with and something like Jekyll just wasn’t “enough” for me. My day job is writing code all day, a platform that is just text editing in the browser is a nice relief.


Thank goodness. I avoid Medium like the plague.


I hope that the Angular Blog will follow suit.


Up next: the .io domain




Applications are open for YC Winter 2020

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: