Hacker News Reader

Comment by tinyprojects | Hacker News Reader

Hey! I'm the creator of Paper Website. This is the seventh tiny project I've built over the past 18 months - I've been documenting the whole process on my blog: https://tinyprojects.dev

I built Paper Website because I wanted a quick and easy way to start a daily blog. I get distracted when I write on my laptop, but I love writing using pen and paper - so I set out to build this.

I'm obviously very biased, but there's something about writing blog posts using pen and paper that is extremely fun and liberating. It's given me great joy just publishing my unfiltered thoughts onto the internet straight from my pen - it reminds me of all the unfocused blogs of the good old days.

I'm also weirdly very consistent at writing now - I've been posting for 40 days straight using this method (here's my daily blog: https://daily.tinyprojects.dev). It's not for everyone, but I love it.

Thanks for submitting this - if you have any questions I'm happy to answer them!

View on HN

I like the idea, it's cute and could be a whole new way for more physically minded people to get into publishing, but my qualms are: $8.25 seems a little steep for a simple blog. Perhaps if it was actually the price of a coffee ($3-5) it would be a bit more reasonable.

Second thing is there's no way to try it out. I don't really want to drop $99 on something without having any idea whether it's going to work for me, especially as handwriting recognition is so sketchy and I have no idea what it's going to be like.

Perhaps even just a demo mode or something to play around with?

Thanks so much for the feedback on this. There's a bit of a cost associated with paper to webpage conversions (and obviously hosting websites and images) - so I'm quite content with the current pricing.

You can try a Paper Website for 7 days for $1 here: https://paperwebsite.com/signup

> There's a bit of a cost associated with paper to webpage conversions (and obviously hosting websites and images)

Resources for OCR and hosting could be provided by the users themselves. You doing it for them could be the expensive option that justifies getting the credit card out :)

Remarkably good OCR is built into Apple imaging now, it’s hard for an iPhone user to not bring it.

Apple Notes of your note book, then Markdown “Exporter” app* into a static blog generator of choice does require that last thing, a statis blog generator, which may(?) be worth the $8/month of hassle alone.

* Exporter: https://apps.apple.com/us/app/exporter/id1099120373?mt=12

Lovely little app, lets Apple Notes be your blogging editor — support the dev!

> Perhaps if it was actually the price of a coffee ($3-5)

Anecdotal, but i've never seen a coffee that expensive. A coffee machine in France is usually < 1€ (more like 50c), and i've never paid more than 1.5€ for in an actual café.

Back on topic, i agree with the sentiment that this is the kind of tool you wouldn't use for business and therefore wouldn't get your credit card for immediately. A free-software, donation-based model may attract more people and make more money (and make grumpy librists like me happy).

Really cool stuff (on paper anyway)!

Looking through your blog, the sheer output you're able to accomplish as a 1-man show is staggering. I was reading through your MMO writeup and had to laugh at "Monday-Wednesday: Learn Unity". How do you keep yourself motivated enough to crank out a project in just two weeks?

Equally impressed. I feel like there’s a part of my brain when I try to learn some new tech that nags “are you sure this is this the best way to spend your time? There might be an alternative tech X that would be better to learn in the long run”.

It’s something I’m working on personally, but I suspect productive folks like OP have effectively silenced that for themselves :-)

I have to ask. How do you manage your time with this many projects under your belt? I have a number of side projects and it's always tough balancing fixing bugs, adding new features to existing projects, or creating new projects.

Part of the fun with building tiny projects like this is they're really low stakes. I build them as an mvp in a matter of weeks/months, launch them fast, then get a feel for if I enjoy running them. If I don't, I axe them, or end up selling them - (which is also a fun process).

This means I only really have a couple of things on the go at once, but I can try out way more ideas in a short period of time.

Do you support Markdown? I don't see any evidence of that, but it would seem to be a perfect fit, allowing users to skip a lot of editing/refining on the website and stick exclusively to paper.

Not yet - but I'm working on a tagging/markdown system for the exact reasons you described.

I love the concept of this, although I can't see it working out for me personally, as I often find myself meticulously re-writing text, so I'm hesitant about hand-written text (maybe that's the difference I need...?).

Separate question - Is there a feed of any sort (e.g. RSS / Atom) on the website that's produced?

I'm a messy writer. Just out of curiosity how does it handle this. If I scribble out text or turn it into a rectangle does it skip the text while scanning it? Does it support text that is deliberately struck through?

Neat project by the way!

It's using a combination of OCR and GPT-3. GPT-3 is basically being used as a glorified grammar corrector, trained on bad OCR sentences and their correct equivalents. If there's spelling mistakes, crossing outs, or bad OCR going on, GPT-3 will pick it up and correct it into a coherent sentence. There's also a web-based editor that you can go into and edit everything afterwards.

Love your stuff, I have been anticipating an update from you! I was wondering, how were you able to get access to GPT-3? AFAIK it's in closed beta and hard to get access to. Thanks!

Thank you! I applied as soon as it was announced - I probably waited about 8 months before I eventually got access. Even after that, there's hoops you have to jump through before you can go live. It's incredibly cool tech to have on an API call.

Also is the gpt3 an important part of the cost? Gave you tried if e.g. gpt2 was enough for your use case?

Reply on news.ycombinator.com