Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: Slyde – A new privacy-focused general social network (slyde.network)
62 points by PaybackTony on Dec 1, 2020 | hide | past | favorite | 64 comments



I see comments here asking is it open source, is it federated, why not mastodon etc.

Why would you build an open source, federated, mastadon like product? None of these things have led to great success for social networks have they?


The issue in my eyes is that things like Mastodon are pretty well designed for what it is. A main feature of ours is the personalization and separation which we could have hacked into such a protocol but ends up being a worse experience for the majority of the users. Personally I wouldn't want to use a fragmented network to connect with people I know in person (Friends, family, co-workers), but I have no issue with it for online communities. In the end though, I think fragmentation doesn't help simplicity and simplicity is what the majority general social network user is looking for.


> Why would you build an open source, federated product?

Is your question "Why would you make your users able to verify for themselves and make sure using your service is actually private when you sell it as a private service ?"


I have to ask - what's the revenue model? I didn't see a payed account option (which is nice), but I do wonder how you're going to sustain this without deciding after 5 years that maybe selling a carefully selected subset of information would not be privacy invasive (just projecting my worst fears here)?


This is in the air but there are a couple things we are keeping in mind. Ads are still on the table, just not hyper targeted ones. The only targeting being to people who follow certain subjects on slyde (cars, weather, etc). Outside of that, not much more. This will limit the amount of advertising revenue for sure, but this leads to the second point. We want to also allow creators of premium content to create premium channels on slyde, which would be a way for content creators to fund themselves, similar to how YouTube does it. Ultimately we have to accept the fact that we just aren't going to be able to monetize users to the same extent as facebook or google does, and that's okay with us.


Using ads will destroy the privacy and trust for sure. The ad networks will track anything and it all starts again. I think users will need to pay a very small fee. That gives them control to be a legal owner of their account and data's. That is actually the argumentation of most other platforms to use and sell user data's, because they have no legal rights on their data's. By paying they have. If you would offer an additional email address in a paid model, that would also secure the path from the platform to the user. Like 1$/month for Slyde. 2$/month Slyde with email address. Second can be a an option to receive funds or donations. Third users and platform keep fake accounts under control when users need to pay something. Also you should allow users optional to ID themselves by a third party service. That is very important for some public persons and many others. Later you can offer services between these users which bring you some extra income.


Has anyone at Stackexchange written open about their revenue model? Seems similar.

You do also need to keep in mind that you need to make sure the ads themselves aren't serving tracking garbage or outright malware, even if you do not offer precise behavioral targeting or even store behavioral data.

You can maybe also sell fun premium features like Discord does with Nitro.


Our idea at the moment is to act like reddit ads do. Advertisers essentially make a post without anything more than a title, description, image / video and a url, with the image / video hosted and scrubbed by us. This will most definitely take careful design but your point is valid. Right now it's a delicate balance of not getting ahead of ourselves but also not setting ourselves up for failure due to poor planning.

Premium things is something Snapchat tried out for a bit with bitmoji, though to be honest I haven't looked deeply into any available data on how that worked out for them.


Have you considered 2 revenue modes:

1) Free with ads that may be targeted to varying degrees.

2) Paid, with zero ads and zero data collection and zero targeting.


A YouTube premium but without the data collection type model is something we've talked about. Hopefully the users will guide us in the right direction with this. We have to worry about fighting the stigma that there is even a paid option for it, and would that stifle growth? Not answers we have but something we've questioned.


Perhaps Spotify offers a good example. Free tier is funded via ads. Premium tier, no ads + other benefits.

I expect you'd find that there is a significant segment of users who would be willing to pay for the service in order to not be targeted with their specific characteristics and use patterns. Brands can build trust, perhaps make that an objective to support subscription sales.


The problem with ads is you start optimizing for engagement which is basically the big issue with FB.


There couldn't be more truth to this. Ads are such a slippery slope. If for arguments sake Slyde grew to be something, it would be due in no small part to the "take it or leave it" approach to ads, if they ever become a thing. Any deviation from this would essentially be biting the hand that feeds you. I can't say we have all or many of the answers to combat this problem but I can say we are weary of it and is the primary reason we want to develop other monetization strategies that are a win-win for us and the users.


I think there's something wrong with your home page: https://monosnap.com/file/qJKDEaHP9VMzogjo9NvzCdBWYYtWNJ

Lots of extra scrollbars. Latest Firefox on macOS 11.


I am using Chrome 87 on macOS 10.15.7 and I'm not seeing any scroll bars at all on your site, even though I have my Mac preferences set to always show scroll bars.


This is a UI bug due to formatting meant for mobile devices . The scrollbars are tiny in these instances. This isn't by design on desktop and is getting fixed as we speak.



Yeah, it seems like a firefox-specific issue. Replicated on windows here.

Works absolutely fine on Edge though.


They're so focused on privacy, they only tested it on a browser that spews 'telemetry.'


We've tested on a number of browsers but not often enough on some of them - like Firefox. We don't like all the data Chrome collects on it's users, but it makes up for the vast majority of usage on desktop and Android so it's important to test it most thoroughly on the most thoroughly used browser. We did the same on Safari which is a bit more private than Chrome, but not totally. We need to test more, perhaps something like browserstack needs to make it's way back into my toolchain.


It looks just fine on my Firefox instance on Android 11.


Thank you for the screenshot. Will get these kinks worked out asap.


Hi. Found a couple bugs in the web app (also using Firefox). I can upvote my own post in the feed (which may or may not be a bug), but then if I open that post and it slides open from the right, it doesn't show the upvote there unless I refresh the page. Also I can upvote my own post in the feed a second time, Refreshing removes the upvote, so that's just a UI bug.


Yup, same for me. FF on Linux


Is the protocol open and federated? (If not, how is this not yet another silo?)


It appears to be another silo for (as the 'owner' put it elsewhere) reducing 'complexity' (i.e. mastadon is too hard because users have to choose an instance). Instead they want you to use their walled garden. It's a bit silly to claim any focus on privacy when you still have to absolutely trust one group of random people on the internet with no option to host it yourself, etc.


Give me a social network that sets dark pattern "themes":

* default-- show me a single page posts of my friends by date

* sadness-- choose post order and timings to maximize my sadness

* filter bubble-- choose posts to maximize my outrage about alternative viewpoints

* rabbit hole-- capture maximum amount of my attention on the platform by feeding me conspiratorial and junk content

* gamblevision-- infinite scroll, maximize my time clicking upvote/like buttons, karma building, interacting with a/b tested time-wasting games, spending karma points on loot box type bullshit, etc.

* human bioweapon-- fake news injections, with an intersection between my inferred personal tastes and a slowly increasing menu of anti-scientific covid conspiracies

Then give me, the user, an analytics page where I can see my engagement, sentiment analysis, and measurements of my changing viewpoints for each theme.

If you ask, "Why would anyone ever use such a manipulative, unhealthy platform?" well...


Imagine this engine but reverse who it's for.

I would like my social network to make me feel:

Happy, excited, inspired,


First we have to learn why it's dangerous to touch the burner on a hot stove.

Otherwise a company can flip some bits and you'll be less happy and more agitated without understanding why.


I see the satire in this. Things like the "Can we make people more sad or happy by being selective on the feed" experiment facebook did is just one of the many reasons we wanted as pure a social feed as possible. People you know ordered by who's posted most recently and a way to traverse their posts if you wanna know a little more. I don't think a "Top" algorithm has a place on a social feed, or you end up with satire but in real life.


shall be built


Why would I use this and not mastodon?


I think mastodon is a great concept, and I like the idea of a community of communities. I think the difference is mainly that slyde was built to have a little lower barrier to entry -- you don't have to pick an instance, for example - and that it's core principle is to get you to connect with everyone you know and share freely with them as opposed to be selective or have to add them to arbitrary lists.


> it's core principle is to get you to connect with everyone you know and share freely with them

Assuming everyone you want to talk to is on Slyde, right?


Or at least willing to try it out. Critical mass is THE issue and we have a couple ways we're looking to get past it. Ultimately people will have to want to connect with people the way slyde lets you, and want to share it with those people to boot.


> Critical mass is THE issue and we have a couple ways we're looking to get past it

And then? Let's assume (for the sake of argument) you get enough users and <big corp> dangles millions, then what?

Privacy? We'll never sell/share your data? Users are just supposed to take your word for it? The page doesn't even have a link to something like T&C or Privacy Policy - just few highlighted words.

I don't think we should trust any social network that's not "open" by design and implementation.


When we started building this we were adamant about making something we wanted to use. There are a number of open and successful options out there, they just didn't work the way we wanted it to. Building a general network that appealed to us and our parents / co-workers meant we had to make sure the experience was just right and is why we went for a ground up approach. Unfortunately that puts users in the position that they have to just trust our words when it comes to privacy. Trust is earned and we just got started. I can make promises about not selling out but until we prove it they will just be words. We are users ourselves, we don't want to feel like we're being watched and have every grain of data out there for advertisers to see. In time we hope to prove we want that for everyone else too.

This is not to say we haven't been actively looking at ways we can open up or become compliant with a few of the open protocols that already exist as long as it fits with the personal privacy we want for our users on top of the data privacy.


Well that's the thing. Transparency is a dependency of trust. And as it stands, I for one ain't seeing a lot of the transparency, seeing as how (AFAICT) we can't see "under the hood". The lack of a published privacy policy is another red flag; publishing that policy - and including the necessary provisions to ensure that Slyde or any future owners/operators thereof will always put user privacy first - would go a long way toward establishing that transparency and therefore trust.

That said, it's a similar problem with DuckDuckGo: how do we know they're respecting users' privacy? DDG has it easy, since they're a search engine; absolutely nothing about their product requires storing any info about their users whatsoever, and thus (as far as anyone knows, and as far as they say) they don't. Slyde, however, is a social network, which makes the collection and storage of data much harder (if not impossible) to avoid.


I don't think I could argue with these points. The Privacy Policy right now is only on the Register page and is a standard Termly one. I think some better effort needs to be put into this on our side to make a simple one and have it more clear, on the homepage. DDG had a couple bumps at the beginning but they made it through and through time established trust. But like you said, Slyde is a social network and has a greater responsibility given the kinds of data people are putting on it, to keep it private. Long story short I appreciate this take and agree with you.


In any case, I wish y'all the best of luck. Even if Slyde never achieves a perfect privacy story, it'd at least be an improvement, especially if coupled with a strong privacy policy and a solid picture of how things work behind the scenes.

Another huge help on the transparency front would be to at least ensure the source code for any apps is publicly available. Even if not necessarily under a FOSS license (though that'd be a bonus), the apps being as easy as possible to independently audit would alleviate a lot of my own concerns with these sorts of social networking apps doing scummy things like using the microphone for ad targeting.


Google+ tried something similar with their 'Circles' concept, didn't they?


Arguably Google+'s Circles were pioneered by XMPP thanks to roster groups (https://tools.ietf.org/html/rfc6121#section-2.1.2.6), which can be used for PubSub to only allow members of a given group to retrieve elements from a given node. So you'd publish stuff for your friends group, for your family group, ...


They did, and we felt like it didn't work because it gave people too many choices. They had some suggested circles, but we felt that if we simplified the logic of separating your life, it would make it easier to use and require less thought when connecting with people.


One of the first things I come across in channels is flat earth society- and worse, no way to block it from my feed. Of course... at this point free speech is just a dog whistle for a zone that is so polluted with bullshit that its impossible to have meaningful connection or discussion.

I can't recommend or use this until there is some kind of content control, what are your plans to prevent this from becoming another voat or parler?

Any plans for scheduling posts or a bot api?


We want to be a place where most anyone can feel at home, but that is a fine line. We don't want to be another politically biased or polluted forum. We have blocking for users and will be extending that to Channels and our upcoming Groups next week.

As for API's and scheduled posts, there aren't immediate plans for that until we nail down a proper OAuth implementation to roll out with the officially supported API's. Though there are 4 of us all together, I'm the only engineer of any kind working on it, with potentially some more help on the way here shortly.

Your questions are good ones. We asked ourselves very specifically how would we avoid being another parler yet at the same time not alienate one political crowd over another? Was a difficult question. The 4 of us span the political spectrum and I think that has helped keep us centered so far.


Maybe an odd comparison, but watching different social networks launch and failing reminds me of the MMO's that were produced after World of Warcraft's massive success. Most of them touted as the "WoW killer"

Don't get me wrong, I would really like to see a new social network take off - one focused on privacy and data ownership.


I think that's an apt comparison. Our goal is to make something that could replace facebook but not by being a facebook killer. Instead of focusing on being an anti-facebook we tried to take a fresh look and make up for some of the discrepancies between how we live life in real life and how we do it online without being idealistic about it. The success of slyde hinges on being a better tool for the job (among other things). Our hope is if we focused only on the job that needed to be done, and not what other tools people were using, we could make something different and useful.


I hope for your true to mission success.


Typo in homepage: yoruself instead of yourself


Nice! I actually toyed with similar idea a few years back, and I named it facesbook :) The selling point for me is the multiple profiles. It would reflect our real life relationships.


Exactly, we realized right away that we all already have a built in personal privacy policy. We just had to turn it into a UX.


I feel like it would be great to have a demo of some sorts, the landing page doesn't really give much information about how it looks or how a sample feed could look like.


We have a slightly outdated one, it's just not as obvious on that page as it should be. If you click the "Watch Slyde in Action!" it's actually a button. We have a pending update that embeds an updated version of that video.


FYI: The assistive technology I'm using doesn't recognize it as a button, nor does it seem to even have a tabindex.

If you can get in touch with the frontend person, please recommend that they use a link (A element) or button (INPUT element) to make the control, otherwise it is not recognized as clickable by some useragents.


Thank you for this. Will make sure this is addressed asap. We previously far overused divs (and still do) and have been going over everything making them buttons but still need to do a lot of work in making it more accessible.


+1 on the accessibility issues.

I'm using serenade.ai to control the page, and it won't even let me scroll down or click buttons/links


Slyde is to me the most promissing startup for a new social media platform.


Tl;dr; We built a new social network called Slyde where you categorize connections based on who they are to you: friends, work, family or public. It has anonymous likes / dislikes and profile customization among other things. We launched to friends and family a month ago and just opened publicly. Story is below.

---

Around a year ago a few friends and I were talking about how we disliked social media. We came to a quick realization that we kept our accounts for the same reasons we never post on them; our friends, family and co-workers were there and we had no way to express our personality. We couldn’t share some things with friends because of fear of family members or work friends seeing. We could use groups or filters but those were cumbersome so we just opted not to post at all. Turns out we weren’t alone.

I floated the idea to them that maybe we should make our own. It was clear that there was a need for something new. But how could we give people an outlet to display their personality and still maintain that privacy?

It didn’t take long for us to determine that in most cases we all have 4 different personas in our day to day life: Friends, Family, Work / School and how we are in public. This would be the base of Slyde; it was the natural way we already maintained our social lives, we just needed to bring it online. Keep it simple. Give people 4 categories they can put people they know into, and allow them to share with those categories. Google Plus was something we looked at as an example of how giving people too many options can be a bad thing.

We got great feedback across the board on those assumptions, but we still wouldn’t use something new just because we could separate our lives easier. Through all the faults of MySpace there were a couple pieces that always made us feel like we got to know someone better. Music and Style. If we could show people a little more about ourselves instead of just writing an about me, our profile would better represent us. Especially if we could adjust it based on who was looking at it.

Being the only engineer -- I have previously worked as a lead architect and senior engineer at Vacasa and Nike -- it was going to be a large task. The biggest elephant in the room was critical mass. The value of a social network is the network. Our critical mass answer wasn’t exciting, but early results may be showing it to be true: Make something people want to use.

If we made something we would want to use, others might too. Once they’re on they aren’t likely to just delete their account. If the appetite for something new is as big as we thought it was we would see small but steady growth, and on a micro level these pockets of people would reach their own critical mass and begin sharing it at a higher rate. We’re already starting to see this happen.

We knew at the beginning the best way to get to a critical mass was to target a smaller mass. We spent a good amount of time theorizing on how we can adapt our ideas to fit a smaller niche. The more we developed these ideas, the less we found ourselves wanting to use it. We didn’t personally have a need for something new in a smaller niche community. The product that we wanted was a general social network that all our connections could join that gave us our personalities and privacy for self-expression. That idea lost it’s value if it only pandered to a niche audience.

When we released it to friends and family we weren’t expecting much. We figured a few people would sign up, we would get some feedback and then iterate. Instead, we had hundreds of them sign up and share it, and received overwhelmingly positive feedback on the overall concept. This seemed to validate that there is a need out there, and that we might have even executed on that need.

Our biggest tell that we were possibly hitting a sweet spot was how many people signed back in every day. It was initially perplexing. Even with little content during our closed test and an average anywhere from 10 to 40 connections per user, nearly 40% of our users were signing in every day, with most of our users signing in every other day. After opening it up publicly a few days ago we had our answer. We might have built something people want to use. We’ve received the same positive feedback our friends and family gave us, from complete strangers. So far we’ve only had a small reddit post to test the waters. Starting with this HN post we’re ready to start telling the rest of the world about it.

We know that for true data privacy we have to give up ad retargeting and tracking. Privacy vs profit is a real discussion, however we believe we can have a viable business model without giving away data privacy. Less targeted advertising and things like premium channels are things we’ve discussed. Privacy is a core principle and guides us in every decision, however right now it’s about building something people want and worrying about business model if we’re lucky enough to get to worry about it.

We’ve possibly stumbled upon a couple of right answers, but we surely don’t have all of them. We wanted to post on HN for two reasons: to show others what we’ve done and to get feedback on our execution. Please fire away!


You've already identified the two big questions.

After signing up and clicking around a bit, here are my thoughts:

1. Button interactions need some kind of activation effect. Whether that's a sound, or the button changes color, or something, I need to know that the button has registered my click.

2. Is there a mobile app?

3. I'm a little confused by the feed. It looks like each post also has all of the person's other posts in a slideshow. Is that accurate?


1. Slyde was built mobile first, and there is still some work to be done on Desktop, this is definitely a big need and has been getting a lot more attention recently. We're definitely working on it.

2. Yes, for iOS and Android "Slyde Social" on the appropriate app stores.

3. That is exactly how it works. Instead of seeing someone's posts throughout your feed, you can browse through their feed from most recently active connection down. This feels much more natural on mobile, and getting tweaked and adjusted for desktop users.


Just checked out the app on iPhone. You're right, it feels a lot more natural than the desktop site. I'll continue poking around.

Also, I created the #StarWars channel. =P


Looks very promising, and I love concept. What's your policy on content policing?


Our terms are the standard termly terms for a social network, however, we have content rating and filtering. Pretty much we only care about public content and reported content if it is breaking government laws that we know about. Otherwise, what's private is private and we'd like to keep it that way.

[Edit] To clarify, we don't moderate based on opinion. Only when we are legally required to, otherwise we let dislikes / downvotes from other members do the talking.


Is this open sourced?




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

Search: