Hacker Newsnew | past | comments | ask | show | jobs | submit | cupcakestand's commentslogin

tl;dr

British billionaire Jim Mellon wants to launch a company named Juvenescence with plans to make a big splash in anti-aging research.


Why should I discover content on Medium?

This job is done by HN and Reddit and unfortunately Facebook if I am on FB. I don't have time to get sucked in another feed.


Specific content. For example I follow the Laravel tag on medium, I've not seen anything regarding Laravel on HN in a fair while.


Because most aggregators become biased and limited silos either intentionally to sell advertising or otherwise via community moderation, power users, etc. The same stuff always bubbles to the top or is buried at the bottom. Variety is key.


Agree that for a western English audience NY is the best location reaching both the US and EU with a ok latency.


> Simple VPS, webserver, letsencrypt, static site generator, cloud flare for CDN duties and a little CSS.

Why didn't you host on Github Pages and you could have saved all that but having the same flexibilty regarding a SSG use case?


GitHub suspended my account because I used it to host a "for-profit" website, which led to the website being down for several days before I checked.

I believe the last time I checked the TOS, no such mention was made for Github pages.


> GitHub suspended my account because I used it to host a "for-profit" website, which led to the website being down for several days before I checked.

What kind of website were you hosting?


A website for a car rental organization.


That's a hugely popular option for sure, and great for a ton of static site use cases, but you do sacrifice the flexibility of deploying whatever the hell you want on your own VPS. Who knows what I might want to do in the future?


> Why didn't you host on Github Pages

Tried that once, fell down a 3 day rabbit hole trying all the static blog generators.. ended up picking WordPress.


Same. For all the times this is recommended, I haven't found a static blog generator that really works well. WordPress is still the easiest and most flexible to get running, but it's a PITA to keep it maintained.

After trying Lektor (fine for general use, but hard to theme; no pre-baked templating/theming system at all) and Hugo (tbh I don't remember the exact thing here, but I got really frustrated trying to do something and ended up deciding that themeless Lektor was better), I'm now holding out hope that Ghost, while not a SSG, will be good and simple enough that I can self-host a "blog" without falling back to raw HTML editing or WordPress. Started out by having to edit the package dependencies to versions that work on recent versions of Node... :\


Write your own generator, it's pretty easy! Blogs don't have to be fancy, most programming blogs need just paragraphs, images, and code blocks. Then you can add fancier features (interactive diagrams, fancy animations, a nice deployment pipeline, etc.) as you go along.


We developed Strattic to solve the issue of difficult to maintain and optimize WordPress websites. Strattic publishes WordPress websites as static sites, and the origin site sits behind a login so it's only accessible to you. No server issues etc. Strattic is currently in beta: https://strattic.com.


Try Metalsmith if you want to poke around and write code. Try Hugo if you want something that works out of the box but that doesn't have a plugin system.


> Tried that once, fell down a 3 day rabbit hole trying all the static blog generators.. ended up picking WordPress.

I'm still falling through the very same hole. What made you pick WordPress?


Honestly no real reason other than it's the one I know and didn't want to spend any more time on finding software to write with instead of spending the time on writing

Essentially it was the default option


Heh I can empathise. Two of my biggest reasons for wanting to move to a static site generator are actually to have actual control over the content (in markdown files, and because WordPress and its plugins do a fair bit of processing on the raw content), and because I'm more comfortable hacking in languages that aren't PHP.

That seems to have stopped me from writing though.


> more comfortable hacking in languages that aren't PHP

Oddly enough it was the opposite of that for me. Each of the static site generators I looked at were non-PHP (which is fine but I lack the experience)

I tell you what, if someone ever had a gun to my head and said "make this Ruby software work first time without errors" they might as well just pull the trigger


I disagree. It doesn't matter what kind of coder you are, at some point you always deal with a shell. And setting up a blog on bare metal is done in few minutes on a shell. If you have never faced a shell or a remote server as a developer then you should find a toy project to do so. It's not the worst skill to have.

Moreover, I wouldn't see basic HTML or CSS as specific knowledge. This is something every coder knows (like English), of course not advanced layout stuff like Flexbox but the basic of HTML and CSS are common knowledge. And you do not need React, Webpack or any advanced stuff to make a simple SSG blog. You even don't need to understand HTML or CSS if you just stick with the standard templates. Just Markdown.

Edit: Why the downvote?


> Moreover, I wouldn't see basic HTML or CSS as specific knowledge. This is something every coder knows (like English), of course not advanced layout stuff like Flexbox but the basic of HTML and CSS are common knowledge

i work in a building full of embedded developers who wouldn't know a css selector if it bit them on the ass.

what use is it to any of them?


> i work in a building full of embedded developers who wouldn't know a css selector if it bit them on the ass.

Maybe, maybe not. Even embedded coders face HTML/CSS at some point when their embedded devices incorporates a webserver for some admin interface which sends HTML/CSS to the world (e.g. a router). Or they have create an API doc for their newly created embedded system, on the web. There are many examples.

What I am saying is that HTMl/CSS is ubiquitous in tech and besides, not that hard to grok compared to code in Assembly or C for an embedded system. And again they don't need HTML or CSS anyway. They just need to know a bit of bash, git, Markdown and a SSG of their choice.


> Maybe, maybe not.

uh it isn't some sort of weird toss-up? i'm referring to specific people who i know.

> Even embedded coders face HTML/CSS at some point when their embedded devices incorporates a webserver for some admin interface which sends HTML/CSS to the world (e.g. a router). Or they have create an API doc for their newly created embedded system, on the web. There are many examples.

there's a whole world of embedded stuff that doesn't have admin interfaces, doesn't have APIs, etc. you can concoct increasingly strained examples all day long, and there are still folks doing work you didn't realize was even a thing, who'll have long and profitable careers without touching a line of CSS.


Ok, it's hard to imagine that such people exist but I believe what you say. You are right, they should rather use Medium.


I work with a lot of kernel dev type people, you may find it hard to imagine but there are a lot of people that only program in c, maybe python.

I even know a lot of fortran only programmers. If you think html and css are common you might want to expand your scope of programmers, there is a lot of people that don't need or want to know about web programming.


I'll use my Father as an example. He works for Autodesk, before that he worked for Revit (when they were a tiny startup before Autodesk bought them out). He writes code in C++ mostly (I think some C# these days). He doesn't use HTML or CSS, and has never learned either. Not because he can't, just because there's not been any point and because the amount of effort that would take is large enough to not be worth it. Why spend several hours or days of extra effort just for a blog post? Not to mention the extra money needed to buy a domain and hosting.


I don't understand this, because C++ or C# are both high-level programming languages. HTML, on the other hand, is simple markup language. I'd think it would take somebody on that level a weekend to learn it, possibly a few weeks or a month to be confident writing it.


I think you are missing the point. He is not saying that his father can't do it, he is saying there is no point in his father learning it.

Edit: words go in a specific order.


I haven't touched HTML since somewhere around v. 3.x... I haven't ever bothered to really learn CSS. I am retired, but I once write software to model traffic.

Just another datapoint for you.

I can install Wordpress and a theme. I have no idea how to really customize said theme. I could learn, I am not feeling compelled to do so.


HTML is really not all that ubiquitous. The breadth of the technological world is wide... In any case, from my point of view the question is not "could I put together a virtual server with some kind of blogging software on it?" (I'm sure I could) but "is that the most interesting/important/worthwhile thing for me to do with my limited time?" and "do I want to give myself the ongoing maintenance chore of ensuring that the server and software stack is up to date with security updates?". Those two things are why I host my rather-intermittently-updated technical blog on wordpress.com.


Having a Webserver is not that simple. Even if you knew HTML and CSS good enough to make decent looking blog, there are many things that are not so simple:

- installing the basics. Why would I know if I should use Apache or Nginx. The internet is no help for that and while both are easy to use, it is not obvious how to publish a blog. You might also need other systems such as a MySQL db if you want to use Wordpress to publish. And I wouldn't expect a Postgres to know how to use MySQL

- securing the server. I would not expect everyone to know how to secure a server. There are some many things that are hard. Even iptables is not easy if you use it for the first time (don't get my started on firewalld). More advanced things such as AppArmour or SELiunx are freaking overwhelmingly when you use it for the fist time and only use a shell. Those things take time to learn and get right. It is useless if you set up the ip tables correctly but save them incorrectly so they flush after a reboot.

- keeping the server up to date. It requires discipline to active look for updates and act once they are discovered. Why would a embedded programmer want to bother.

- price: while 5 bucks is not much for many of us, it adds up, Especially when you have many such little things. Secondly, such a cheap server might not be enough. You need backups which requires additional work. Furthermore, if you do not use a static website, the hacker news front page could quickly overwhelm such a little server (it might even be hard with a static website).

These are all reasons why one should not use their own server for that. If you want to learn a lot about web servers, having a blog is a good and valuable start but I would never expect anyone to do that


That's why we developed Strattic: it allows users to enjoy the flexibility and power of WordPress without the need to maintain and secure servers. Strattic publishes the site as static and serverless, and the origin site is only available to the site owner. It's like a static site generator, but for WordPress. Strattic is currently in beta: https://strattic.com.


Totally agree. Code embedding is also limited on Medium, I think they still lack syntax highlighting.

Btw, great tip wih Netlify. How is your experience with them, what are pros cons compared to Github Pages and can I run server code like Node for free?


OT, but syntax highlighting is hardly the weakest part of displaying code on a webpage. Nobody's disrupting horizontal scrollbars or margin-busting, but the first person who does is going to be a hero.


I was so close to doing this when I was making my new blog, then I found out that the CSS element responsible for making the code block scollable isn't supported on mobile.


Any time I have to embed code in Medium, I just type the code in a free Github Gist and post the link - Medium automatically embeds the Gist in a nice, syntax highlighted format, with a link to the original gist.

e.g. (towards bottom of this article I wrote - syntax highlighting in HTML, Javascript and Ruby snippets): https://hackernoon.com/building-a-face-recognition-web-app-i...


> Totally agree. Code embedding is also limited on Medium, I think they still lack syntax highlighting.

Limited how? I find embedding GitHub Gist's works perfectly fine and is automatic -- just paste the link to the Gist in your post and you get an embedded object.


To the point criticism of Medium, very detailed and the OP is picking the right issues of Medium.

While Medium looks so beautiful and clean at first glance it really disappoints when you use it on a daily base both as a creator and as a user. Everytime I use Medium, I am surprised that Medium is successful. Its appearance feels definitely premium and significantly of higher value than any other blog system but the usability is a nightmare.

So we are back to square one. Which blog system should we use? SSG on Github Pages?


As a reader I avoid Medium based articles.

For one thing, the UI for wildly different blogs are the same. This means that I am unable to place anything I read in context. As a visual learner, one of the ways I remember stuff I read is by remembering the entire context (for blogs this is usually the look of the blog). Medium completely eliminates that crutch for me.

I feel the overuse of annoying funny GIFs in the middle of articles is far more prevalent on Medium pages. I suspect because it's one of the only ways for authors to personalize their content visually.


Medium blogs all have different colored nav-bars and logos.


That's so minute. Different fonts, typography, page layout, structure, etc all improve uniqueness and memorability.


I guess I'm in the camp of people who prefer uniform blogs. The decorative accents on medium are more than enough for me to be able to distinguish them. I enjoy the fact that the layout is familiar and I can dive right into a piece without the distraction of a different or unique layout. To each her own!


> Which blog system should we use?

Hosting your own WordPress site? It's simple and easy. Why did anyone ever want to be on Medium in the first place? Most likely for the name association and the visibility that comes with it.

I think we have a responsibility to promote self-reliance as part of technological culture. If we want people to move away from centralization (which I do), we have to teach by example (which I do).


It's the visibility, yes. I just moved my various blogging endeavors to one domain (coyotetracks.org) rather than the mishmash I've been using, which includes a few articles on Medium. But with only those few articles, I have over 400 followers on Medium, and my most popular article has 59K reads and 175 recommendations. (The next most popular has 39K and only 10 recommendations, because it's a criticism of "Mr. Money Mustache" and most people are hate-reading it from a link from his web site. Article #3 drops down to a mere 2.3K views, but 51 recommendations.)

I moved to hosting my own WordPress site; as much as I think WordPress's internals are a garbage file (actually, a series of garbage files passed around as globals, but never mind), it's so easy to make work--and the ecosystem is so big--that it almost makes up for its nastiness. But when I write an article that seems "Medium-esque," there's a very good chance I'm going to crosspost it to Medium to attempt to get those views. (With links back to my main web site, of course.)


"Hosting your own WordPress site? It's simple and easy."

I would think that, for at least 90% of those publishing on Medium, that's neither simple nor easy.

Even ignoring that, hosting your own Wordpress site means taking responsibility for keeping it up to date, security-wise. That hurdle alone probably makes most bloggers use some third-party blogging provider.


It's just an example, there are many choices, some of which are presumably even simpler.

But yes there will be some responsibility involved, I think that falls into healthy technical culture as well. It's the difference between taking care of something and getting someone else to take care of it. At the end of the day, someone has to do it, and we shouldn't always simply trust others with it just because it's easier.


WP can update itself though.


Wordpress is for sure powerful, has a huge ecosystem and is for teams a safe bet because of very good and battle-tested workflows.

But sorry, I don't want to setup a stack I did 20 years ago. However, I haven't looked into Wordpress for years.

For simple personal blogs I like DB-free SSG blog systems. But I am still wondering which one is the best. Just Hexo (Node) and Hugo (Go) come to my mind.


As I mention in my other reply, I meant it as an example. The issue to me is more "self-reliance" versus "paying others to do it". I'm certain that an increased trend of self-reliance such as hosting your own blog would go a long way towards demystifying the internet and therefore getting people more interested in the very real political issues behind the tech.


I agree, there's an air of "my gardener moved away, how ever will my grass get cut now?"


How easy do you think it is? Have you done it?

I found that self hosting wordpress is incredibly complex and expensive.

Wordpress.com does a great job though, nothing to manage.


Use wordpress.com

Works wonder. 2 decades of great service.


> Which blog system should we use?

Hugo works extremely well for me. Very fast and reliable.


They have. But still, every founder wants to avoid such situations. The founding team's reputation went from sky to hell within few weeks. Not nice and nobody wants to loose their face like this.


> It's the ground-zero meat space for their content corpus

What do mean with this sentence and content corpus??

That burning man matches their DNA?


Are you being purposefully obtuse? He's saying that one could make the argument that meeting new people at the epicenter of your prospective user's culture would help growth ultimately. This is also referred to as "boots on the ground" marketing.


No, I find his sentence just odd. I googled 'content corpus' and couldn't find anything meaningful, thus my comment.

I assumed something in the direction what you wrote, btw way better articulated, but I wanted to get it confirmed.


Late response, but I simply mean the collection of content on the site. Burning man is a major center for electronic music. You probably wouldn't be surprised by the head of Netflix going to Cannes, by contrast.


Best comment in the thread. It shows that one of SoundCloud's key problem is an oligopolostic industry they are dealing with.


I'm not so sure that's true. Their key problem is that they didn't understand how much value there is in up-and-coming, amateur, and underground music. The record labels only have power when it comes to their catalogue. SoundCloud could have truly disrupted the labels by providing an alternative way to promote, monetize, and discover music. But they somehow got off that track.


And that is the big loss. Having a place where one could discover new music has so much value over being served the same songs by the same few artists


Does it though?

I'd like this to be true, but in practice it seems like most people (me included) strongly prefer hearing songs they already know most of the time, enjoy songs more with repetition, at least up to a point, and are more than willing to learn to like almost anything with high enough production value.

Some people are music snobs as a hobby and most go on discovery quests from time to time, but it honestly seems like most people most of the time want music they don't have to think too much about or devote too much attention to. Indie-only channels can't really deliver that.


I think most people like variety. Yes, some people always listen to the same 50 songs, and some people never listen to the same song twice. But most people don't go their whole lives without expanding their musical tastes.

While iTunes and Google Play gave chart toppers the spotlight, SoundCloud turned amateurs into professionals. "Nobodies" could put their music on soundcloud and get them streamed millions of times, and that recognition could eventually get them a chance to sign on a dotted line


Why do these two groups of people need to be served by the same service? Why couldn't there be a site for young producers to easily publish their music online and connect with the fans? If I make a song, I can't even imagine what would it take to get it to Spotify. I'd probably have to pay some agency. SoundCloud with its community is the perfect tool for this.


> ...most people most of the time want music they don't have to think too much about or devote toouch attention to.

I agree with that statement - dedicating significant energy to finding new music is a fairly niche hobby.

However, I don't think easy-to-find necessarily equals label-owned music.

I often listen to music for hours at a time in the background while I'm working or relaxing at home. I'll aim for an artist or mood and then I'm happy to let 'related tracks' from artists I've never heard of play indefinitely. If there's a track I really like I'll look at who the artist is and make note. I listen to a ton of "indy" music this way. Mostly, I never know who the artist is. Sometimes I discover new artists. All of this is very low effort on my part.

In fact, it would be significantly more effort to listen to megastar pop music or classics. I'd have to find a way to play the tracks - YouTube most likely, which means a ton of poorly thought out ads, or else paying for the songs directly, which usually is just slow and annoying.

I'd wager the biggest hours of music listening are for background music - at work, in stores and restaurants, while studying, while lounging around the house. I'm sure some people always want to listen to the same ten albums on repeat every day, but I suspect most people are driven more by ambiance. Indy music will do that job just fine.

Of course people will always want to play nostalgic hits. But I suspect that's a smaller driver of music consumption by hour of attention (or at least partial attention) than many people assume. That means the labels have way less power than people assume.


If that were true then MP3.com would have been a stunning success 18 years ago.


Wasn't it? It was taken down by a lawsuit over my.mp3.com.

https://en.wikipedia.org/wiki/UMG_Recordings,_Inc._v._MP3.co....


Everything was done by someone else before it was done in a way that people remember, sadly. Anybody remember Click-n-Run? No, just the App Store.


Probably because they realized that path would not be sustainable. Most people don't want to pay for music, period. Few would pay a monthly fee for a service that only offered relatively unknown artists. While SoundCloud does have a unique UI compared to other streaming services, it's not enough of a differentiator to compete with Spotify.


How should this work? Once somebody is trending on such platforms, labels will come, give them tons of money, a contract and take over. And take them away from these platforms as well.


They'll give them an advance as bait for an exploitative and crappy contract which may well be run in a dishonest way.

Given that most sales are downloads or streams now, the only thing big labels can bring to the table is a nuclear level of publicity. And by definition, that's going to be reserved for a tiny handful of multimillion-selling household names.

Today's smaller bands and artists - which will be the majority - get no particular benefit from signing to a big label.

They may get some smaller benefits from signing to a smaller niche label, but they certainly won't get a big pile of cash from that.

The situation is similar to book publishing, where a tiny handful of massively popular signed authors get most of the publicity and the cash, but there's a solid underclass of midlist writers who do better with self-publishing, because they earn more from direct sales and - ironically - have a more secure career too.

There was a massive opportunity for SoundCloud to become Amazon/Kindle for Musicians.

MySpace had that space for a while but lost it. BandCamp is close, but doesn't quite make it because the default artist pages suck, and it sees itself as a store, not so much as a marketing and PR outlet that also happens to sell downloads and physical media.

SoundCloud could have won that space, but apparently SoundCloud's management never understood that there's more money in hosting associated services - web pages, mailshots, blogs, and so on - than there is in providing hosting for audio files.

In fact YouTube seems to be quietly taking over for basic track hosting, with the added benefit of video. There's no reaction feature on the tracks, but I'm not convinced that was ever an essential USP on SoundCloud.

But I still think someone else with VC backing could win this space and clean up.


Yes. Someone please build this. Message me if you need a designer.


Most musicians who can figure out how to record and upload their own music to SoundCloud could just as easily create a generic website with squarespace or something similar to host a blog. The only value in a comprehensive social network for musicians would be the audience it could bring. A large audience is very hard to build.


Would love to chat! ranidu@theartistunion.com


Yes! This. SoundCloud should have focused on helping musicians promote themselves, connect with fans and monetize their music. And yeah, this space is pretty wide open right now. The ways people create and consume music are changing, and nobody is really taking advantage of it.

Man, now I really want to build what SoundCloud should have been.


Great. They leave, push their listeners over on the younger version of them.

It's like sailing... Every bit of air that hits your sail will end up behind you, dispersed. That doesn't mean your boat will stop. You travel by charting a course where you stay ahead of a pressure differential.

Same in music. You need to find a source of new artists that will continually replenish you as you lose artists to platforms that focus on the stage after you in maturity. The beauty of culture is it always replenishes.


This is exactly why SoundCloud should be a publisher. To take advantage of the gems their platform unearths.


yes.


SoundCloud should be a music publisher - a record label.


"track"


SoundCloud's key problem is that they couldn't stay small because of VC money, not the music industry.


They couldn't stay small due to user influx. They couldn't sustain their music offering without deals with the music industry.

VC money is just something they needed because they couldn't do any monetisation without deals with the industry (too much of their content is non-fair-use derivatives of the big 3's copyrighted content) and they obviously can't stream millions of hours of music without some sort of money.


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

Search: