Show HN: A new image sharing service | Hacker News Reader

Show HN: A new image sharing service

itsosticky.com

42 pointsemily_b36 comments

You should read this:

https://drewdevault.com/2014/10/10/The-profitability-of-onli...

Disclaimer: I wrote it

For consumers this isn't too big of a deal. There are plenty of good-hearted people out there and "host my images" isn't something that requires a lot of personal trust (at least, for the kind of images I host). Some hosts don't scale much and can survive off donations. Others become too large and fall under their own weight.

When one service becomes poisoned (as you described it) - I move to another service. Once upon a time that was Imgur, followed by puush, followed by minus, followed by pomfse, and now I use vgy.me and mixtape.moe (when vgy.me's servers go down, which is rare but happens). There are no shortage of image hosts around to choose from.

Your writeup is right on the money. It isn't a service that scales well and any monetization tends to lead to users dipping to greener pastures. Imgur being a strange exception.

The internet doesn't need another ten thousand broken links when this one goes under, too.

That's thinking about a much bigger problem - and you're completely right about that.

Forgive the delay in responding.

Itsosticky was designed from the outset to be an image sharing service with community potential, contrasted with image hosting, it's a service that facilitates quick sharing of an image or screenshot with a friend or group.

When used for its intended purpose, direct sharing, as opposed to free hosting for hotlinking, there are features that make it worthwhile sharing an Itsosticky resource URI rather than a direct link to an image file. For example, zoom capabilities, that some people may find superior to those available natively in a web browser, albums and easy social sharing links, and not least, fast uploads from phones without the need for a dedicated app.

All users are invited to publish their images to the gallery for greater social participation or exhibition.

None of what you've said here makes you stand out from your biggest problem: Imgur. None of those features are groundbreaking or even original. And how are you going to monetize? Running this service won't be free.

Imgur isn't a problem, and this shouldn't be a problem for Imgur. There are similarities, but that's mostly due to necessity in good UI. If people prefer Imgur, they should certainly use Imgur.

I can't comment on what might be right or wrong with other sites, not in the capacity of someone promoting a competing product.

Itsosticky is fast (server hardware permitting), modern, has no advertising at all at the moment, and no third-party sharing widgets.

I can't discuss plans for monetisation - there are none as yet.

Out of curiosity, what were the main costs for the services you were involved in? Engineers? CPU for resize/transcoding? Storage? Bandwidth?

Just the hosting costs. Our costs were pretty low, actually, but it was out of our own pockets and never turned a profit.

Were they low enough that they could conceivably be satisfied by donations?

Certainly no startup but a self sustaining, non-commercial image host could be nice.

I think we need to think about why people might donate. Picture you use a service, it shows you an image... do you think to pull out your credit card and send them money? Why would you with all the other image services out there?

I mean, a few will, but I think you need to create something that actually solves someones problem in a meaningful way (that is unique compared to free services).

MediaCrush did accept donations and had ads, but the actual donor turnout and ad revenue was pretty pathetic.

https://web.archive.org/web/20150211095158/https://mediacru....

tl;dr: what about developers?

When I started building Theneeds, we wanted users to be able to post, like in Twitter. And soon, we wanted users to be able to attach photos or other media.

For media in the sense of an url that we then embed/display it's relatively easy. And even easier if you pay for an external service like embed.ly to do the job.

For images it's a mess. We had to take care about uploading, resizing, posting async to s3, invalidating caches, and of course displaying. If I had to do that again, I'd surely look for an external service (that I didn't find at that time).

Maybe this is a viable way to monetize? I guess that there are quite a good number of companies that need upload+display of UG images, or at least that they could do it if it was easier to plug in.

Just for reference, embed.ly costs 20$/mo for just display, so I guess you could at least ask for that + storage costs.

Next, you could add additional services such as cdn support, or anti spam/copyright infringement/adult content/etc.

Well, this is an incredibly difficult market to compete in. You need something really special to stand out and attract users, or devs won't bother with your service. Other sites like Imgur already have the mindshare and you need to somehow get huge and find an effective monetization strategy to pull it off. Big image hosting sites are expensive and you won't be able to turn a profit for a long time, and if you're paying for it out of pocket it's going to go south pretty quick. Especially if you don't pay for the high end infrastructure, your site will get slow as the load piles on (and it will pile on and you will see service-disrupting abuse) and most users will leave without a second thought if they have to wait more than a second or two to see their images.

imgix offers a solid solution in this space nowadays

Mind if I ask the name of your image hosting service?

It's mentioned in the post: https://sr.ht/

And previously:

http://web.archive.org/web/20150130040753/https://mediacru.s...

A new image sharing service

Itsosticky is a brand new image sharing service with some ambition, just launched after a year of development.

More info is available here https://itsosticky.com/about

There are several things that make this different to other image sharing services. Itsosticky will quickly and automatically rescale images that are above a certain size threshold, with auto-orientation of JPEGs uploaded from phones. The layout is fully responsive, allowing for consistent URIs when shared across devices. It's light weight, with no advertising technologies.

Many of the images currently displayed in the gallery are simply placeholders dropped onto the site by helpful beta users, and there are surely bugs to address here or there, but by and large, Itsosticky is ready for use.

Thanks for taking a look. Feedback from mobile devices would be particularly valuable.

Please upvote if you think this service is worthwhile and deserves to go somewhere.

How do you plan on enforcing the no-NSFW policy? That seems close to impossible pre-monetization unless you spent most of that year training a neural network.

Could always use https://www.projectoxford.ai or something like it. I know Google just came out with an API for generating text descriptions of images. Perhaps if scanned images passed through this api were tagged with certain words, the images could be marked for moderation and not shown until approved.

How do you plan to monetize this?

Keybindings would be great. Also, it'd be nice to see points on posts.

A full year of development?

Common engineering fallacy: "I could have done that in 2 months". Why not two weeks? Or "two days using frameworks x and y"?

I don't know anything about the author but if he worked on this nights and weekends for a full year it seems about right to me (and great job)!

Though, maybe I'm not 10x enough to know better. :-)

Looks extremely similar to Imgur, including the colors and modals. What would you say separates it from Imgur? Why did it takes a year to develop?

> Why did it [take] a year to develop?

I believe this is your answer (from the About page):

> Almost every piece of code is ad hoc, built from the ground up to be as lightweight and efficient as possible.

My opinion is they most likely built many things which were already available in the open-source world.

> ...built from the ground up to be as lightweight and efficient as possible.

This is my favorite part. So many sites start out this way, then become slow, unresponsive and bulky. Here's to hoping it stays slim and quick.

> What would you say separates it from Imgur?

It's not blocked by my company firewall yet, so that's nice.

Mine is blocked already...

Finding a way to provide viable (fast and high quality) image resizing on phones was one of the biggest challenges.

It's cool. How do you plan on getting people to use it? The reddit userbase basically propelled Imgur into an overnight success.

its ok. I still think imgur is slightly better to use. Just my common example : I will find images and using their url upload, upload the image so that I have an easily shareable link. On your page its a two step process. I had to first enter the url, press the www button (which wasn't immediately apparent ) and then pressed the upload button. On imgur its only enter link and upload.

You can drag and drop from other web pages or copy and paste via Ctrl-v. Might be a useful one-step alternative.

Wow that looks nothing like imgur :/

Infinite scrolling is broken. https://itsosticky.com/1srfcne

Scroll down, scroll back up, scroll down again.

Reply on news.ycombinator.com