No one should need JS to see the soups when that could be handled perfectly fine with CSS. I wish restaurants would just make their homepage a PDF of the menu.
No one should need an entire PostScript interpreter to see the soup of the day, either. A restaurant menu is text and images. HTML and CSS are perfect for text and images.
I agree with no JS, but why PDF over HTML? Hard-wrapping for letter-sized paper (ok, a PDF doesn't need to be letter-sized, but most menus are approximately that) with crapshoot reflow options for soft-wrapping in certain viewer apps is pretty dicey on a phone, mitigated only slightly by rotating the phone sideways.
The only benefit I can think of is if it leads to more frequent updates by the restaurant, due to limited skillset.
I vastly prefer looking at a PDF menu over an HTML one nearly all the time. PDFs are usually nicely formatted, and I don’t mind zooming and panning to see everything. HTML is frequently terribly formatted, interspersed with ads, slow, etc
If the restaurant doesn't have anything besides a menu, /index.pdf is fine—no web design required; reuse the menu they're printing anyway.
The trade-off is that they'll have to pinch/zoom if they have a small display. It's a minor inconvenience to make the exact information they want available instantly.
index.pdf won't tend to play nicely with screen readers and also sucks for people on crappy mobile networks, so it's a minor inconvenience for some, and straight up unusable for others
The complexity between the modern web and a pdf is marginal. PDFs do get printed for menus. Editing a PDF and uploading it to the site, integrating prices and syncing between the site, online ordering, PDF menus is just part of the business. There are lots of platforms that help with this such as Slice.
Translating anything that renders on my screen is the same two clicks to open an LLM with the screen contents. I expect that will become an increasingly universal experience as LLM features get shoved into every nook and cranny of tech.
There's been a translate button for years which hooks deep into every nook and cranny of the website's HTML. It works great, it's built in and many restaurants even advertise it for tourists, because it's a zero-effort translation of their existing menu. Plus, it's low-data when you're inside a 1-bar basement restaurant.
Using an LLM to translate the visible part of a PDF on a mobile... seems like the worst possible solution to the problem.
Translating PDFs is more complicated than that because the strcture of a PDF document doesn't lend itself well to this kind of thing.
For example: if there's a dish name with a 2 line description below it and some allergy symbols below that, in HTML you can imagine the document structure that produces that. In PDF terms that might be 4 separate objects and, in particular, the eyes can see the two lines are adjacent so they fit together but the document structure doesn't really represent it taht way, necessarily.
This might also not work with translation because the lines are set for the size of the text they contain. Same for resizing the font.
Put another waay, PDF should be viewed as a typeset and layout format, not a document format.
I think you're misunderstanding what I'm describing. It's getting a screenshot of the visible portion of the rendered document, not the document itself with all the tags and nastiness inside. The same feature works with a photo of handwritten text, where obviously no digital document exists. It's not perfect, but usually adequate for menu translation.
All of my static sites that I've built lately have been done on Netlify. Super easy to hook up to Github and the form handling is a breeze. I've known Mathias going back to when he was personally answering emails and promoting JAMSTACK so you can say I'm a bit biased. lol
Netlify is a great company that I'll always support.
PDF is an enormous pain in the tits to view on a phone and has significant accessibility issues for people using assistive technologies.
It's not even about blind people. People with ADHD or dyslexia use assistive technology, which frequently makes an absolute horlicks of interpreting PDF. It's one of the reasons I'm trying to move a lot of documentation at work away from PDF and onto just straight HTML.
Plain old HTML, with thin CSS on it to make it not be black-and-white Times New Roman. Kicking it oldschool.
It's pretty sad how there doesn't seem to be any decent free options for websites which are easy to use. Squarespace and such cost a fortune which isn't worth it if you aren't trying to run a full ecommerce site. Plenty of services offer free hosting of static content but don't have any way a normal person can use them. Having to use a static site generator is too hard for non programmers.
I'm just surprised we haven't seem some app that can act like a wordpress admin page but generating a static output you can host for free or very cheap somewhere.
Google Sites exists. Buy a domain, point it at a free Google site. So easy a religion major can make a site that looks pretty decent (ha ha, only serious; I thought he'd used Wordpress at first) for just the cost of domain registration.
Squarespace is like $20/mo for a basic site promoting your Brick and Mortar business. That includes domain, hosting, and a template/CMS. It's not that pricey.
It's not pricey if you are a serious business making good money. It's a huge price if you are say a part time artist just wanting somewhere to store a price list, gallery and contact form.
I'm just surprised there is nothing that fills the gap between github pages and a full hosted solution with a ton of junk you don't need. All it really needs is maybe a locally running app that can handle generating the static pages and uploading them for you.
The barrier to create a website using Astro + a Template + telling an LLM like Gemini what you want is very low nowadays. So still, if you work with code some technical knowledge is required, but it will only get easier, probably.
There is very little chance a non developer would make it through that. The current options are Instagram/Facebook page which is free and easy. Or a website which is either expensive or requires you to be a developer.
Yeah, with Statue the barrier is literally just running one terminal command. And funny you mention the LLM part, we're actually working on integrating that directly into the workflow so you won't even need to copy-paste between tools.
Check it out: https://statue.dev/
That’s what I’ve been working on for the last few years: https://palacms.com (MIT)
Not ready to announce the V3 RC yet (need to fill out the starter sites) but it’s working well if you wanna give it a spin - runs & deploys on Railway’s free tier.
A webapp or gui WYSIWYG static generator with basic git support abstracted away would go far for many. Just let it push to some private repo which cloudflare pages or similar would deploy off of.
It really feels like the only part of a non-static site most want is an editor. I absolutely loathe the matter but I do see why some restaurants only maintain a facebook page for their online presence.
This is what Netlify does. Hook up a private repo and deploy. Make a commit and it auto builds and you have a CI/CD pipeline. This is what I build all of my static sites with. You can do almost any JS framework like React, Angular, Vue, etc.
Netlify does way more than this, but it makes hosting static stuff super easy.
This is exactly what we're trying to build with https://github.com/accretional/statue - you can email me or hit me up on Linkedin to get early access to our free static site hosting (which our new site for the project at https://statue.dev runs on, and which will Soon™ have a public-facing product doing exactly what you just asked for)
Basically you'll be able to edit the markdown for your site in a souped up version of our lightly reskinned vscode IDE at https://brilliant.mplode.dev and instantly publish/preview the changes in the same browser tab in a pane. Brilliant comes with a full Linux environment running in a container on our cloud platform, and building a Statue static site is already a one-command operation. The little UI we're working on let's nontechnical people skip that and just edit files and click buttons to make changes and publish it, though.
Here's a one-liner that will get you an entire static site with content (not the landing page yet, though) you can edit via markdown:
> I'm just surprised we haven't seem some app that can act like a wordpress admin page but generating a static output you can host for free or very cheap somewhere.
Hey, we're actually building something very close to this! We have a template system where you can style everything with Tailwind (already included) or add other UI packages. You can host both frontend and backend for free using Cloudflare D1 and similar options. We're also working on a pretty elegant solution for the backend side of things.
Here's a working example built with our project: https://yc-game.pages.dev/leaderboard
Check it out at https://statue.dev/ — if you're interested, you can follow the project on GitHub or find us on X: @semihpvlat (me) and @FredAccretional (CEO).
Honestly, there are some easier ways out there now, although of course no solution is perfect.
For non-technical people I'd recommend the Hostinger Website Builder, Obsidian Quartz or Astro Starlight.
Although as a front-end dev I'd choose building a custom page with Astro, which has now become much easier though with good templates available + LLM assistance.
I'm not seeing anything very specific in the code - feels like this could be just another Jekyll theme and still work the same. There's some custom front-matter in markdown files, but change that to regular YAML and it will just work.
There is some code for looking up the geo lat/log of locations at compile. not sure how you would do that. But yeah outside of that your are mostly correct.
I used elixir because thats what I know and love so it was mostly just a personal choice rather than a technical one.
First, the site generator is MIT licensed but I don't see a link to the license. If someone forks this generator, would they be in compliance with MIT license requirements?
Second, the images linked in this site are quite nice. I can imagine someone choosing to use some of them as is. Are they yours to share?
Third, it appears that you are targeting non-developers. I would think about how to make it as easy as possible to customize. Decisions like putting images in "priv/output/images" seems a bit confusing.
Third: Yeah that's the challenge I'm working on at the moment.
Thanks for the feed back.
I do plan on cleaning up the repo so that you are not starting with the example and also plan on making a small tutorial video to show how much effort it takes to setup.
No one should need JS to see the soups when that could be handled perfectly fine with CSS. I wish restaurants would just make their homepage a PDF of the menu.
No one should need an entire PostScript interpreter to see the soup of the day, either. A restaurant menu is text and images. HTML and CSS are perfect for text and images.
Nobody should need 60 million lines of code (Linux Kernel 30M + Chromium 30M) to render some text and images ;)
I agree with no JS, but why PDF over HTML? Hard-wrapping for letter-sized paper (ok, a PDF doesn't need to be letter-sized, but most menus are approximately that) with crapshoot reflow options for soft-wrapping in certain viewer apps is pretty dicey on a phone, mitigated only slightly by rotating the phone sideways.
The only benefit I can think of is if it leads to more frequent updates by the restaurant, due to limited skillset.
I vastly prefer looking at a PDF menu over an HTML one nearly all the time. PDFs are usually nicely formatted, and I don’t mind zooming and panning to see everything. HTML is frequently terribly formatted, interspersed with ads, slow, etc
> HTML is frequently terribly formatted, interspersed with ads, slow, etc
You can put ads into terribly formatted PDFs too
If the restaurant doesn't have anything besides a menu, /index.pdf is fine—no web design required; reuse the menu they're printing anyway.
The trade-off is that they'll have to pinch/zoom if they have a small display. It's a minor inconvenience to make the exact information they want available instantly.
index.pdf won't tend to play nicely with screen readers and also sucks for people on crappy mobile networks, so it's a minor inconvenience for some, and straight up unusable for others
But you can easily serve a desktop version or a small screen version.
Because they can make one nice pdf formatted to get printed out in the restaurant and then reuse it to display on the website
The complexity between the modern web and a pdf is marginal. PDFs do get printed for menus. Editing a PDF and uploading it to the site, integrating prices and syncing between the site, online ordering, PDF menus is just part of the business. There are lots of platforms that help with this such as Slice.
PDF:s are not great on mobile. And you can’t easily translate them (I often translate restaurant menus when they are on a website with just 2 clicks)
Translating anything that renders on my screen is the same two clicks to open an LLM with the screen contents. I expect that will become an increasingly universal experience as LLM features get shoved into every nook and cranny of tech.
There's been a translate button for years which hooks deep into every nook and cranny of the website's HTML. It works great, it's built in and many restaurants even advertise it for tourists, because it's a zero-effort translation of their existing menu. Plus, it's low-data when you're inside a 1-bar basement restaurant.
Using an LLM to translate the visible part of a PDF on a mobile... seems like the worst possible solution to the problem.
It's the worst solution, apart from the fact it works better than all the other solutions.
Translating PDFs is more complicated than that because the strcture of a PDF document doesn't lend itself well to this kind of thing.
For example: if there's a dish name with a 2 line description below it and some allergy symbols below that, in HTML you can imagine the document structure that produces that. In PDF terms that might be 4 separate objects and, in particular, the eyes can see the two lines are adjacent so they fit together but the document structure doesn't really represent it taht way, necessarily.
This might also not work with translation because the lines are set for the size of the text they contain. Same for resizing the font.
Put another waay, PDF should be viewed as a typeset and layout format, not a document format.
I think you're misunderstanding what I'm describing. It's getting a screenshot of the visible portion of the rendered document, not the document itself with all the tags and nastiness inside. The same feature works with a photo of handwritten text, where obviously no digital document exists. It's not perfect, but usually adequate for menu translation.
To be fair this project uses zero 3rd party npm modules for runtime. The total runtime JS it uses is 1.76kB in size.
I agree. There are lots of free AstroJS themes for restaurants that generate static html that you can host somewhere like Firebase hosting for free.
- https://astro.build/themes/details/astropie/
- https://astro.build/themes/details/astrorante/
- https://astro.build/themes/details/tastyyy-restaurant-websit...
All of my static sites that I've built lately have been done on Netlify. Super easy to hook up to Github and the form handling is a breeze. I've known Mathias going back to when he was personally answering emails and promoting JAMSTACK so you can say I'm a bit biased. lol
Netlify is a great company that I'll always support.
I was going to recommend the same! Astro + Astro theme + an LLM will get you very far these days.
I used to be all in on Jekyll. Now all I use is Astro + Tailwind + Claude, and it’s magic. No need for a theme with this combination.
I love Astro; there is so much you can do with it.
Remember during Covid where every restaurant's menu was a QR code on the table that linked to a PDF in S3?
Remember how after Covid that didn't go away in tons of places
PDF is an enormous pain in the tits to view on a phone and has significant accessibility issues for people using assistive technologies.
It's not even about blind people. People with ADHD or dyslexia use assistive technology, which frequently makes an absolute horlicks of interpreting PDF. It's one of the reasons I'm trying to move a lot of documentation at work away from PDF and onto just straight HTML.
Plain old HTML, with thin CSS on it to make it not be black-and-white Times New Roman. Kicking it oldschool.
A PDF can't get the user halfway through the delivery process before seeing the soups.
No one is browsing the internet without JS today (within margin of error). Whether or not this "should" be the case, it is.
This is the wrong way of looking at it.
Making a website's basic functionality work without JS isn't just for the random users who switch off their browser's JS runtime.
It's also for the people who have a random network dropout or slowdown on a random file (in this case a JS file).
> It's also for the people who have a random network dropout or slowdown on a random file (in this case a JS file).
Does that really apply when the javascript is only ~2kb?
Yes, any request can get stuck at any time.
That is what's happening any time you've seen a website that randomly decides to load without styles, or with a missing image.
The good thing is that it's very apparent when that happens and you can just reload the page.
But it's not immediately obvious when it happens with a JS file.
That's half the reason why you shouldn't re-implement css features in a js file. (the other half is performance)
Nobody should need a PDF renderer to see the soups.
Actually, nobody should need an XML parser to see the soups either.
PDF is a terrible experience on mobile
The soup shows for me without JS.
No one should need PDFs to see the soups when they can be handled perfectly fine with CSS scoped to print and save to PDF....
/s
It's pretty sad how there doesn't seem to be any decent free options for websites which are easy to use. Squarespace and such cost a fortune which isn't worth it if you aren't trying to run a full ecommerce site. Plenty of services offer free hosting of static content but don't have any way a normal person can use them. Having to use a static site generator is too hard for non programmers.
I'm just surprised we haven't seem some app that can act like a wordpress admin page but generating a static output you can host for free or very cheap somewhere.
Google Sites exists. Buy a domain, point it at a free Google site. So easy a religion major can make a site that looks pretty decent (ha ha, only serious; I thought he'd used Wordpress at first) for just the cost of domain registration.
Squarespace is like $20/mo for a basic site promoting your Brick and Mortar business. That includes domain, hosting, and a template/CMS. It's not that pricey.
It's not pricey if you are a serious business making good money. It's a huge price if you are say a part time artist just wanting somewhere to store a price list, gallery and contact form.
I'm just surprised there is nothing that fills the gap between github pages and a full hosted solution with a ton of junk you don't need. All it really needs is maybe a locally running app that can handle generating the static pages and uploading them for you.
Macs used to come with iWeb. https://www.apple.com/welcomescreen/ilife/iweb-3/
These days you can buy paid software to do this:
- $110 https://blocsapp.com
- $90 https://realmacsoftware.com/rapidweaver-classic/
- $80 https://sitely.app
- $30 https://bootstrapstudio.io
- $0 https://www.silex.me
- $0 https://wordpress.org/plugins/simply-static/
- $0 until recently, https://web.archive.org/web/20240410200646/https://grapesjs....
RapidWeaver Classic calls itself a subscription and sets up autopay, but you can immediately cancel and keep that version forever, like Jetbrains.
iWeb was an utter monstrosity. Worse than MS Front Page, which I didn't think was possible.
The barrier to create a website using Astro + a Template + telling an LLM like Gemini what you want is very low nowadays. So still, if you work with code some technical knowledge is required, but it will only get easier, probably.
There is very little chance a non developer would make it through that. The current options are Instagram/Facebook page which is free and easy. Or a website which is either expensive or requires you to be a developer.
Yeah, with Statue the barrier is literally just running one terminal command. And funny you mention the LLM part, we're actually working on integrating that directly into the workflow so you won't even need to copy-paste between tools. Check it out: https://statue.dev/
That’s what I’ve been working on for the last few years: https://palacms.com (MIT) Not ready to announce the V3 RC yet (need to fill out the starter sites) but it’s working well if you wanna give it a spin - runs & deploys on Railway’s free tier.
A webapp or gui WYSIWYG static generator with basic git support abstracted away would go far for many. Just let it push to some private repo which cloudflare pages or similar would deploy off of.
It really feels like the only part of a non-static site most want is an editor. I absolutely loathe the matter but I do see why some restaurants only maintain a facebook page for their online presence.
This is what Netlify does. Hook up a private repo and deploy. Make a commit and it auto builds and you have a CI/CD pipeline. This is what I build all of my static sites with. You can do almost any JS framework like React, Angular, Vue, etc.
Netlify does way more than this, but it makes hosting static stuff super easy.
https://www.netlify.com/
Not exactly what you’re asking for, but this is nonetheless an interesting approach: https://getpublii.com/
This is exactly what we're trying to build with https://github.com/accretional/statue - you can email me or hit me up on Linkedin to get early access to our free static site hosting (which our new site for the project at https://statue.dev runs on, and which will Soon™ have a public-facing product doing exactly what you just asked for)
Basically you'll be able to edit the markdown for your site in a souped up version of our lightly reskinned vscode IDE at https://brilliant.mplode.dev and instantly publish/preview the changes in the same browser tab in a pane. Brilliant comes with a full Linux environment running in a container on our cloud platform, and building a Statue static site is already a one-command operation. The little UI we're working on let's nontechnical people skip that and just edit files and click buttons to make changes and publish it, though.
Here's a one-liner that will get you an entire static site with content (not the landing page yet, though) you can edit via markdown:
yes | npx sv create . --template minimal --types ts --no-add-ons --install npm && npm install statue-ssg && npx statue init && npm install && npm run dev
This isn't exactly the point. He said easy to use. Yours requires developer skills which is not what he is looking for.
> I'm just surprised we haven't seem some app that can act like a wordpress admin page but generating a static output you can host for free or very cheap somewhere.
https://news.ycombinator.com/item?id=44391535
(I'm not affiliated with it)
Hey, we're actually building something very close to this! We have a template system where you can style everything with Tailwind (already included) or add other UI packages. You can host both frontend and backend for free using Cloudflare D1 and similar options. We're also working on a pretty elegant solution for the backend side of things. Here's a working example built with our project: https://yc-game.pages.dev/leaderboard Check it out at https://statue.dev/ — if you're interested, you can follow the project on GitHub or find us on X: @semihpvlat (me) and @FredAccretional (CEO).
Honestly, there are some easier ways out there now, although of course no solution is perfect.
For non-technical people I'd recommend the Hostinger Website Builder, Obsidian Quartz or Astro Starlight.
Although as a front-end dev I'd choose building a custom page with Astro, which has now become much easier though with good templates available + LLM assistance.
I wrote a comparison of less-technical ways to build a website here with more details: https://webdev.bryanhogan.com/start/ways-to-build/
Have you looked at micro.blog? Plans start at $1 per month.
I’m building something like this…
NextJS + Git + Vercel.
how about something like feather.so? publish a website / blog from your Notion...
haven't used it, but looks like a great idea!
I'm not seeing anything very specific in the code - feels like this could be just another Jekyll theme and still work the same. There's some custom front-matter in markdown files, but change that to regular YAML and it will just work.
There is some code for looking up the geo lat/log of locations at compile. not sure how you would do that. But yeah outside of that your are mostly correct.
I used elixir because thats what I know and love so it was mostly just a personal choice rather than a technical one.
Link at the bottom of your example page results in 404. For me, anyway.
ah yeah, sorry work in progress. There is a link to the repo https://github.com/Local-Cafe/localcafe-lite
I'm getting NXDOMAIN, and various online resolution tools [0] show the same.
[0] https://mxtoolbox.com/SuperTool.aspx?action=a%3alocalcafe.or...
A couple of things:
First, the site generator is MIT licensed but I don't see a link to the license. If someone forks this generator, would they be in compliance with MIT license requirements?
Second, the images linked in this site are quite nice. I can imagine someone choosing to use some of them as is. Are they yours to share?
Third, it appears that you are targeting non-developers. I would think about how to make it as easy as possible to customize. Decisions like putting images in "priv/output/images" seems a bit confusing.
First: https://github.com/Local-Cafe/localcafe-lite/blob/main/LICEN...
Second: pixabay
Third: Yeah that's the challenge I'm working on at the moment. Thanks for the feed back.
I do plan on cleaning up the repo so that you are not starting with the example and also plan on making a small tutorial video to show how much effort it takes to setup.