I love this platform. I've been hanging around there ever since Medium became shit. So while it's great to see the platform all grown-up, it's also kind of disheartening to see some (most) of the issues that plagued Medium follow along.
Here's a personal take on issues that plague Dev.to. Please be reminded as you read that this does not intend for any moral judgement, but is a single man's opinion on the origins of the issues.
Over-representation of a certain type of development
Not a big issue, but certainly an issue. It feels like this platform is centered around a certain type of developers. Precisely, the "Junior fullstack javascript developer fresh out of bootcamp".
This is not a bad thing per se, but many developers outside of this category just have very little reason to browse around here. Most posts are centered around Javascript, Node.js and React. Very little talk about the rest of the ecosystem.
It would be a net gain for everyone to appeal to a broader audience.
Lack of quality
Dev.to is not Reddit nor Hackernews. Content is not organized by "best". Which is not to say that it should. But as with any platform that becomes big enough, more and more people are posting here.
Which would not be a problem if everyone was doing its best to write great, useful, and insightful articles. Sadly, this is not what's happening.
To be clear, I'm not trying to say that inexperienced developers should not write, quite the contrary I think they definitely should.
What I'm trying to say is that writing should be motivated by mostly altruistic reasons, rather than selfish ones.
Personally, I wrote some Ruby articles because I felt like they were missing while I was learning the language. I wrote what I would have wanted to read on the internet and that did not exist yet. I wrote so people around me would have it slightly easier than I did.
Writing a good article is hard and takes time. Sadly, the front-page seems stuck with the eternal same "5 Top VSCode extensions" or yet another guide to React hooks. Were the previous guides so bad that the internet needed yet another?
The shallow and repetitive nature of the articles wouldn't be an issue, if it felt like there always was a genuine / heartfelt attempt at producing quality content. But there is none of that, because a genuine effort is rooted in altruism, and that's not why most people seem to post here.
Personal and professional branding / advertisement
And here comes the bigger issue. Every post seems to either be motivated by personal branding, or is just plain business advertisement.
I'm not even criticizing anyone here, I'm just stating the facts: I understand launching your startup and being in need of free advertisement. I really do.
But ads don't make for good content at all. And this is one of the main reasons why good content is becoming rarer and rarer here.
I'm also not criticizing the fresh-out-of-bootcamps junior devs that want to be able to distinguish themselves - the job market can be quite a bitch for juniors. We've been repeating for years that junior devs should distinguish themselves by contributing, starting a blog, sharing their knowledge, and participating in open-source projects, if they wanted to have a chance at getting a job.
But the return on investment is much better by just rehashing the same knowledge over and over again.
There seems to be a dissonance between what the writers on Dev.to want, and what Dev.to need.
What do?
So how can a website with a very low bar of entry such as this one, do its best to promote the best possible content, without becoming yet another Reddit?
Should there be some kind of human curation happening ? The newsletter is trying to achieve this, but this doesn't help with the browsing part of the experience.
Are there community-driven ways to uphold ourselves to higher standards?
I don't really have a solution. Do you?
Latest comments (159)
funny that you mention reddit as a positive(?) example for "best" articles. But maybe it used to be a good site before my time, just like the mythical "tech twitter" that I never knew. Anyway, I am happy that you did not give up writing on DEV ever since.
A lot of sympathy here. I use dev.to as a stand-in replacement for an RST-based dev blog I host on my personal server, so the social/learning aspects are less important to me. More signal, less noise. Ultimately the "feed fusion" problem comes down to two specific challenges on top of "just a combination of RSS feeds", and it's still a somewhat-solved problem by most social networks (though Twitter's really been struggling lately as they experiment):
Frontier Selection: You want "popular" posts (in terms of interactions, upvotes, comments, etc.), and you want "recent" posts. These are two orthogonal dimensions, so you have a "multi-objective optimization", or MOO, problem. Classic solution is to start from one end of the space and march along the "frontier". Not clear to me that "popular" doesn't override/overweight in the dev.to feed fusion approach.
Discovery: How do you find useful stuff (posts, users) that aren't already fused into your feed? Typically this is done by extrapolating (typically tag-based) from posts and users you have interacted with in some way (with FOLLOW being the strongest interaction of course). It's not 100% clear to me but some tag-extrapolations are much better signal-to-noise than others. Cloud < JavaScript < Kubernetes < Tensorflow < Python < C-Family < GPU, to give one highly subjective example.
Oh man, I certainly feel for you.
A friend of mine recently advised using this platform for blogging. I was looking for a place where I could start doing some write-ups of my past and future experiences. Not all, but many of them, would be very technical and require weeks or even months of preparation. And after looking at the platform for just a week it appears that this kind of content doesn't fit here.
I don't think anyone would like their well prepared content to get lost behind a plethora of relatively low-quality overviews, success/failure stories, ads, etc. Moreover, this content would be very far from web-development and also would be of interest only to the more mature group of our profession.
For now it looks like only the least technical articles for broader audience could fit here well. And for any more sophisticated stuff I am still to find the right place.
While I agree somewhat with the points made as far as some pitfalls that Dev.to may be falling into, I'm not sure I agree about the comparison with Medium. Dev.to has some huge advantages over Medium. Just the fact that you own your own content and it's not going to be put behind a paywall, make it enormously different than Medium! And infinitely more valuable. :)
100% agree. I loved Dev.to in the beginning and moved my blogs from Medium to here. Initially, I had great success and was motivated to produce quality content but then at some point Dev.to became a place for listicles and the only posts getting on the top pages (and even top 7 of the week) were listicles. I even experimented by creating a few listicles of my own (not very proud of that, as they look less than 10 minutes to create) and they did much better than the posts I really spent effort by researching a lot, making example codes, graphics etc. So decided to move blogs to a Jekyll and host on my own. Now its doing much better than the success I had with Dev.to, I still syndicate the posts here, since I gathered a good following here and didn't want to disappoint them.
Overall Dev.to needs to address the quality issue so that it doesn't become like Medium
Agree! Reddit, Stackoverflow, Hacknews, etc.. Unlike these sites, which are irreplaceable
This discussion reminds me of StackOverflow. There have been many discussions on StackOverflow meta about the dilemma how to prevent good content getting drowned in low-quality (or even factually wrong) contributions.
Their approach, an attempted meritocracy with upvotes, downvotes and a lot of beginner content getting deleted (too) quickly, seems to have several downsides: A gatekeeping mindset that discourages not only beginners but also advanced users making an effort to help beginners, reputation seekers that try to trick the algorithms instead of actually creating value, but maybe the worst disadvantage to me: They still don't succeed in prioritizing high quality content.
StackOverflow, despite all of its efforts to focus on high quality content, is full of outdated, irrelevant, misleading, and even factually false advice and code snippets, making it more of a museum of jQuery code than a go-to resource for getting things done in 2021. (Of course I am deliberately exaggerating to make a point.)
dev.to on the other hand, is one of the most open and welcoming online community for developers and programmers I have seen so far. I have found a lot of information, inspiration, disussion, as well as some entertaining fun stuff to read on dev.to. This is exactly what I love about this platform!
But still ... I keep seeing loads of articles that are either irrelevant to me personally (the mentioned beginners seeking discussion with fellow beginners, tutorials directed at beginners, and the overrepresentation of full-stack JavaScript and React) or spammy, excluding, gatekeeping and misleading (like "16 node modules EVERY developer MUST know in 2022") giving dangerous advice contrary to established consensus in the world of experienced developers, discussed over decades in mailing lists, issues trackers, conferences, books, and on Wikipedia.
I don't want to prevent innovation and thinking outside of the box, but I fear that spam and bad advice by intermediate developers, especially when they act like an authoritative source in their few lines of profile description, can confuse beginners and also harm dev.to's reputation altogether (for fellow developers, but also in the long run, for people making decisions how to prioritize trustworthy sources in search engine results). That that kind of content is annoying me personally should be the least of our worries.
As a personal consequence, I have joined many discussions (on dev.to, and also on StackOverflow meta), followed and liked developers that provide good work (including beginners that didn't show me anything I didn't know so far), but I have also started blocking people (on dev.to and twitter) and downvoting content (on StackOverflow).
I doubt that we could, or even should, try to establish a pseudo-meritocracy like StackOverflow. Maybe dev.to could change their metrics and algorithms. We can give likes, unicorns, bookmarks, we can report, block, or follow, and there are badges to reward achievements. Currently, most dev.to badges seem to reward quantity. For how long have I been a member, and how long do I manage to make a streak of continuous publishing every week.
Somehow, dev.to should rather reward quality, or stop rewarding quantity. Likewise, the number of likes does not indicate high quality. Maybe it doesn't even indicate popularity, I don't know how many people even have a dev.to login and bother to like something they read.
Filters could also be useful. There are hashtags like
#beginnersand topical hashtags like#javascript. Maybe dev.to could add a feedback option to suggest adding or removing hashtags for a post. Then there could be landingpages promoting the tags (there probably are, but I never used them) and maybe we could add filters so that an experienced PHP developer can opt not to see beginner JavaScript content in their feed.THANK YOU for this. I'm not crazy after all it seems. I was starting lose interest in this platform because of all the reasons listed a above. Until the Dev team finds a way to stop letting low quality posts run rampant and bury all the well thought out and creative posts, tag weights will help. I didn't realize it was a thing and I've been here for over a year! Here's more info on that for anyone curious - dev.to/devteam/changelog-adjust-th...
You are dead on here: Everybody has to start somewhere, and I think multiple takes on things are good, but this site has been SWAMPED with low-quality/poor content articles, or pieces on things that have been covered to death (looking at you, React hooks!).
I am junior js developer fresh out of bootcamp but I feel you on the 5 awesome vs code extensions posts part😅, every other day someone is writing such posts simply because writing blogs and having an online presence of some sort is kind of required for jobs or internships these days otherwise your profile looks bad, and hence instead to writing quality posts everyone is forced to write rehashed stuff.
You've outlined exactly why I've stopped coming to Dev.to more than a couple times a month.
To be fair, it was like this while I was most active. I tried to play the game for a little bit (my top post by a huge margin is just a listicle), but it's ultimately not a fulfilling way to go about things.
Occasionally I find some great content on here, but it is fairly difficult. Thankfully the mods who curate the weekly email of best posts do a good job of finding higher quality posts, but relying on that is a poor substitute to a good ecosystem.
A good solution is certainly difficult. Much of it has to do with finding better feedback loops for those who write higher level content or content that is not Javascript related.
Yup, got lots of the same thoughts.
Also, "Junior fullstack javascript developer fresh out of bootcamp" is almost entirely the only angel I take when writing here now as I realized a while ago that all my other content gets nowhere here. Overall though, I've largely just stopped writing and am angling to get some conversations started here.
I am starting to just leave my posts on my own blog instead.
Totally agreed, thank you for writing this @samuelfaure
I'd love to see more variation of content here. Personally, I just read articles by people I follow now unless there's a good article in the feed, which is rare as most articles in the feed are "My neighbors Cat's Favourite VSCode Plugin". Hopefully mods find a solution
More manual curation is indeed needed. Maybe daily editor's choice? With no mainstream article obviously.
I used to think much the same about the status of dev.to, although my view has changed a bit after discovering that I could change the weight of the different tags that I registered interest in.
With the default weights the feed was pretty bad, but after some adjustments to thee weights the feed quality did improve. I had to skip some tags, such as "productivity", which was overused in alll sorts of contexts.
The feed is better - not perfect and certainly some posts that I consider irrelevant. But much better than it used to be.
The various email digests sent out are not so good though, I do not read much of them.