The collapsing quality of dev.to (2021)

It's not just dev.to, it's most online developer content sources these days.

A lot of people noticed that you can become 'famous' online which will gain you the ability to sell a course, earn from content or get better jobs and opportunities due to 'influence'. And it works especially good if you can find a new framework/language/platform that you can write content about, since there is less content and you can get 'famous' faster.

And as such, it has become an endless race of trading off content quality for eyeball quantity - people write bullshit blog posts and ask silly or provocative questions on twitter that are easy to engage with, thus gaming the algorithms and getting more eyeballs on their content.

Then, in a nice little reference loop from hell, the 'lower ranked' influencers will like and share this content, leave comments such as 'Omg great article!' not because it is good, but because they know that by engaging with another creator with 'more influence' they can get more eyeballs on their content, making the creator think 'hey this gets eyeballs' so they create more content like that. And then others will copy what works for the 'higher ranked' influencers, thus creating a stream of similar content which the 'lower ranked' influencers will copy from and so on and so on. With so many cyclic references and no garbage collection, all we are left with is garbage itself.

The problem with dev.to, and pretty much every blogging platform like it, is that it encourages the mindset that the articles about problems you have yourself are the best articles to learn from, and that things that aren't directly applicable to other people won't get promoted. That leads people to think they should write boring, generic things about language fundamentals in order to become 'famous'. People end up rewriting the language docs in their own words. How dull.

The reality is that most interesting dev articles are about relatively unusual problems that a dev solved in an innovative way. You can learn a lot about dev work from those (approaching problems, thinking about things, some cornercase dev tools, algorithms, etc) even if you can't cut and paste code from them into your own project. This exposes a further problem though - you can't write those articles until you're sufficiently experienced to face those problems, and knowledgeable enough to solve them. If you're a new dev starting out with ambitions to be a 'thought leader' and wanting to get to the top fast, you're bound to end up writing another "Why Array.pop is the best array method ever!" article.

It's worth noting that it's a shadowbanned domain on HN: https://news.ycombinator.com/from?site=dev.to

I have left Dev.to long ago after being one of the early adopters. The quality of articles is going down every time I visit the site. And this is absolutely not surprising: Dev.to encourages clickbait and low effort articles. Likes seem to be the thriving factor for pushing articles and you can even earn achievements for post with the most likes and views. A possible solution could be to ditch likes and view counts entirely, so people focus on the content instead of trying to publish articles with the most likes. Fight the root cause and not the symptoms.

I never understood how dev.to worked from an engagement perspective.

Back in 2018ish I mirror'd some of my blog posts from my personal site to dev.to. Within a few months I had over 7,000 followers by doing nothing other than copy a few posts there with canonical URLs pointing back to my domain. I never tried to gain a following there, all I did was mirror my own posts and answer any questions folks asked in the comments of those posts.

I'll admit I suck at Twitter but I've had an account there for 10 years and while I don't post a ton, I do at least try to tweet something every few days and I have about a third of the followers there as I do on dev.to current day. I don't know, it makes me think something doesn't line up. I have a hunch almost all of the followers I have on dev.to aren't humans because why would so many people follow me on a platform I don't post on much but on Twitter I'm lucky to get ~10 new followers a month? At the same time I've had a few direct chats with the co-founder of dev.to (Ben) over the years and he's a really genuine dude. I can't make sense of the situation. Maybe dev.to is really just super optimized for making it easy for someone to follow you?

As for content, I tend to only create posts and videos on things I encounter in my day to day from any topic related to developing and deploying web apps. That could be anything from using shellcheck's -x flag to live coding a pull request for a third party project. I find it difficult to build a readership with these styles of posts because it's usually some obscure thing I learned while actively working on something, not "10 reasons why XYZ sucks" or something that will get a lot of interaction or be heavily optimized for organic search. The posts are more to reflect on something I've learned so I don't really focus on "gaining an audience", but it would be nice to grow large enough to be able to make courses full time. The process of learning something new, using it in production for a while while understanding it in depth and then distilling that into a video is really fun to me.

I think this will always be the case with these adtech platforms. There is a conflict of interest at play. In the end, the ad revenue always wins.

The only way to really filter this stuff is through self-hosted blogs. These can also be connected, but in a decentralised way e.g. https://indieweb.org is a good intro to some of the protocols in use.

Dev.to does allow downvotes and blocking bad content. You need to become a moderator to do that. I'm such an external moderator and flagged a lot of abuse in the system. I've noticed an improvement in the past couple of months where I have less content to flag. So they probably improved their spam account detection.

There's a lot of beginner content which is fine. Some marketing/SEO content which is to be expected. It's still a better experience overall when compared to medium. Right now only dev.to and hashnode are significant unmoderated alternatives to medium (there are smaller players like tealfeed etc.). Both are doing a better job than medium and I blog on all of them.

If they add "publications" which is the editorialized capability of medium they would solve the segmented content problem but that might make their monetization harder.

the real problem is that beginner oriented content is whats in top demand. I've written some advanced articles in the past on how to use more obscure locks and stored procedures in postgres in the past. it paled in popularity compared to random javascript content that only newbies wouldn't know. people are voting. just not for anything substantive.

This was the core motivation behind why I started https://www.discoverdev.io/

Ran this for about 4 years before I called it a day earlier this year. Got busy with graduation and job searching! I do plan to start it again after I find a job and settle down :)

What's the average lifecycle of any website since the 2.0 days ? It's impressive how they constantly pop up and then pop out rapidly. Only to be replaced by the new -r variant.

Feels like articles talking about the low energy of lightning. Impressive but too short to matter.

Dev.to is suffering from the same problem that sites like Hacker Noon and Quora have already accepted as their faith. For the time being, Google is still treating Dev.to with quite generous "authority", so I imagine their organic traffic is through the roof.

But this has consequences of having your site infested with spammers, poorly written promotional content, etc. For example, one of the things you can do on Dev.to is publish an article, and then add a Canonical link to the original article (your website), which is what a lot of people do.

Having said that, you can find some great articles on the site from people who simply use Dev.to as their preferred blogging platform of choice. Though, even at a glance, the ratio of good and poorly written articles is probably around 25/75 which isn't all that great.

I stopped reading where they lamented grammatical errors and then proceeded to write “queue” when they obviously meant cue.

I have visited it four or five times in the past 9 months or so..

twice for something I knew was there (some time before) and could not find it - instead some other 'articles' talking around the subject with no answer.

and a few times because search results brought it up higher than other sites which actual had the answer I was looking for.

Didn't this site takeover some resource, trying to remember, something like robottxt.org or something get bought or pushed into this thing?

Now I can't remember the one thing it was good for so long ago, and do not intend to return to have it confuse and waste time in the future, I'll always be looking further down the results.

I'm very glad that others are observing this as well. My last DEV post was literally the question of where I should hang out with possible future content to discuss on a more intermediate level.

It's great that beginners can start by diving into DEV but at some point you're just not a beginner anymore but because of all the noise your only option is to move on. Even if you put your 'experience level' to expert the relevant content shown in the DEV feed is for beginners. I think 'fostering a community for experienced developers' is really the only way for DEV if it wants to move beyond beginners.

I stopped reading after he admitted not actually reading the articles from his feed he is criticizing. He probably makes a valid point nonetheless but thats skipping your due dilligence in my book.

I pretty much agree with everything here. At the same time however, in the past 10 years official documentation for languages, frameworks, libraries, etc. has got _so_ much better. And while it’s true that googling stuff had become difficult, usually i can find the answer I need on how-to topics just by looking at the official docs.

I left dev.to awhile ago, largely in part because they encourage sexism, racism, etc. as prat of their code of conduct. It's truly sickening to me for a platform like this to not view all forms of sexism and racism as bad, not just the "correct" ones.

Dev.to simply is too focused on the open-source community and the web, it promotes and pushes up Javascript related content, presumably because a lot of the open-source repositories available on Github are JS or TS based.

It's not a true knowledge sharing platform, it's just a popularity platform where being the first to write about the new ANGULAR version 9999 is seen as a sign of great wisdom and being "on top of your game as a dev".

I wrote there a lot and stopped exactly due to the misalignment between what I wanted to consume and write about.

More generally, I find that ad-hoc aggregator websites like HN have a much higher hit ratio with my own interests. Also, if you're a professional developer actively involved in open-source and working within the JS/TS ecosystem you'll likely get a bit more value out of dev.to than I did, I guess it's just not my style...

NB: /has taken queues/s/queus/cues/

I too agree with lack of quality content. There are other platforms like Hashnode that you can definitely give a spin.

AFAIK, dev.to is shadowbanned here on HN.

Whenever I stumble upon dev.to, I can’t help but read everything with an indian accent. Is it just me?

I hate to say this but I feel like Dev had a nice start but went downhill fast.

Nearly all articles can be boiled down to “Look at this really basic thing, and follow me on A, B and C.

The first Dev talk I watched from the founder (Ben) was about how they gained a lot of traction being such a fast loading site, focus on content and less fluff. I loved the attention to performance, being open source and the customizations it enabled. Sounded great.

However the site got funding and started to focus on “diversity” and “inclusion” at all costs. Created a new CoC and the beginner content flooded in and has gone downhill since.

There is also a rampant spam issue and affiliate link farming.

I vowed never to use this site because I found certain elements of their code conduct to be questionable.

Why not prioritize everyone’s safety over anyone’s comfort?

And the bit about “reverse -isms” is wholly ambiguous.

It screams to never invest one second of effort in the site.