I usually feel a bit apprehensive when I see a new app pop up using the .so TLD (Somalia, operated by SONIC.so), since we’ve faced a lot of issues with it over the years. I end up repeating the same kind of comment about it (eg https://news.ycombinator.com/item?id=36825966), but I wonder if all the apps on it, including Notion, have changed the reputation of it with corporate firewalls by now.
Still, watch out if you host user-generated content there, it can be hard to undo reputation damage to .so domains if you get reported to various watchdogs. Notion’s longest hard-down outage happened after a takedown notice got sent to one of our upstream providers but they never forwarded it to us, and it took ling hours on the phone to various NICs and ISPs to get unblocked.
Also - the organizations that run the DNS servers of little top level domains are not necessarily competent. Back when I had a company on the .st domain, we had an 8-hour outage when something in the TLD infrastructure broke and their support team was asleep:
Yeah, this reminds me of our outage. When we finally got through to SoNIC at 2am their time, and they scrambled to find someone who could speak English, and we scrambled to find someone who could speak French. I think with high school level on either side plus Google Translate, eventually we worked out the issue was with an intermediary in Germany.
I lost access to all of our documentation at work for a week because my (small local) ISP banned all content from Somalia as being suspicious. I had to call their office and explain to them that Notion is a rather large and popular application.
Switching the product and marketing sites would take a lot of work for unclear/low benefit since at this point we’ve learned the .so lessons and have a good reputation.
Given the choice of projects like “scale infra so we stay online and fast-ish”, “improve search”, or “notion dot com”, Notion dot com ends up below the cut line.
if you have to use a TLD from some country that is not exactly the most stable and can at anytime shut down your domain because you feel it is edgy, maybe you should be rethinking that.
Maybe it's a generational thing. Do younger internet users not pause at the domain name if it looks sus at all? Then again, how many users are actually using domain names to get to websites? With everyone obfuscating the basic link with some cute button, or shortening within the address bar, or any other modern things to hide the grotesque string of text, do people even see the actual links in the day to day adventures on the web?
Maybe this is an everyone needs to get off my lawn moment, and I should just go find a cocktail?
> Why not just do the reverse, and have .so redirect to .com (since .so predated you owning .com)
There's an ocean of effort in the word "just" to do this without it being hugely disruptive to our users. If was "just" so easy, we would have done it! If I went into Cloudflare and fiddled a quick 302 redirect:
- Every Notion user in the world gets logged out, since their cookie is HTTP-only for origin `notion.so`
- Nobody using OAuth login like Google can log in on notion.com since we didn't add the new origin to various allow lists. Cross-origin embeds that allow-listed us break too.
- All our mobile and desktop app users apps will behave in some unpredictable way. Maybe they follow the redirect, maybe they treat any notion.com URL as an external web page and open it in the user's default browser or SFSafariViewController. App store ratings plunge from 4.5 stars to 1 star.
- Customers who added URL-based rules have their rules broken. Maybe this means our app doesn't load for a bunch of large enterprises with stringent security/firewall policies.
It would be the worst incident in Notion's history, it would cost us tens of millions of dollars of lost revenue, and our reputation might never recover.
I think I could fix all those issues and start testing in 1-2 months if it's all I worked on. But, right now I'm working on making Notion databases (the product feature [1]) faster. If I asked our customers to choose between .com and a faster Notion, I think 90% will choose faster Notion. If I think about our product roadmap, 0% of it depends on .com.
> Why take the risk when .com will never be retired.
Well, ccTLD retirement has a 5-year notice period, so if you gave me and SoNIC the notice that .so is retiring today, I still don't think I'd change Notion's roadmap for the next year at least, maybe we'd start working on it in 2025.
That said, I don't know everything going on at the company, and it'd be comparably easy to migrate just the "marketing" routes of our site like https://notion.so/product. Someone might already be working on it!
Just wanted to say I appreciate the time you took to write this up. Really shows the hidden reality behind the words "why not just..." which is so often the cause of unforseen headaches, which are usually a result of quick and shallow thinking without admitting there might be things not considered / unknown.
My point was, you clearly see value having .com otherwise - why go through the effort to acquire it … to then effectively not use it.
The reasons above sound far, but it also sounds like tech debt that is being punted and will only get harder to fix everyday you delay not addressing it.
Just recently started a new project with a .so domain, but looks like I'll have to pick up another one. No point introducing more work for ourselves just for a domain name.
What's the rationale behind using a .so tld in the first place? Unless it allows you to have a clever domain name (which neither notion.so nor shots.so are), I don't get the appeal.
Getting a nice named domain rather than having to do something like getX.com. If the name you want is taken on .com, .io, etc but none of them are actually in use or are completely unrelated to your industry, using a less common tld can make sense.
I honestly don't even look at tld's anymore, and I'm surprised anyone does.
We just got the .com for our re-brand, and my co-founder was excited because he hates .co/.net/etc etc.
I get why the .com is valuable, but does anybody really think that just because a site has a .so extension that the company has anything to do with somalia? How many people even notice that notion is a .so?
Please disregard all the naysayers in the comments.
I'm a technical product marketer, this is delightful and saves me so much time.
Wrong audience for the tool, since they focus on things like accessibility and browser support and what you promise to bring to the table in comparison to what you actually bring.
None of these matter - this is a TERRIFIC piece of tooling and I will use it starting literally next week for my work.
I almost clicked away not realizing there were actual mockups for phones / tablets / etc.. after uploading an image, seeing the results.. and scannign around the screen I thought oh this is super basic, gotta go.. and right before I clicked close tab, I noticed the almost invisible down arrow next to 'screenshot'..
and so I clicked and then, well now it's good.
So strongly suggest adding phone / tablet / computer / laptop /etc icons / graphics / words maybe.. I would even make the default very first screen when you upload your very first image.. also create one of each phone / laptop etc.. and display the results perhaps with a click for more of similar...
The whole thing is quite confusing at first for such a simple tool. The splash screen says "Create amazing mockups" and "Craft beautiful presentations", somehow that translates to "Frames and backgrounds for marketing images".
Surfing with browser set at less than 1702 pixels wide, I get a screen that says coming to mobile in the future.. get one of these other browsers.. but I'm already using firefox... so clicked to go full screen, 1920 px wide and things start happening..
Is Safari really that hard to develop for? Name one feature this site needs that won't run in safari. WebKit has been getting a lot better recently, I feel like not supporting Safari at this point is because of one of three things: you're too lazy to try it, don't have a Mac, or you are stuck to the opinion that Safari is somehow behind Chrome or Firefox in major ways.
I use safari daily and *never* have a problem with a site not working. Ever. So I doubt there's some new web api that this site needs that doesn't work on Safari.
It's been a few years since I did frontend dev, but Safari was _absolutely_ the most difficult browser we had to develop for, _especially_ on iOS.
With IE, you can polyfill missing APIs. With Safari, you have to put in extra work to ensure your polyfills override Safari's built-in implementations in all sorts of exciting ways based on which particular Safari version someone is running, because it's _obnoxiously_ common for Safari to have broken or incomplete implementations of web APIs. So if you're just checking for the presence of some particular global, it'll be there...but it'll be subtly broken in various ways. We had constant issues with Safari's implementations of LocalStorage and IndexedDB, and with its weird tendency to "ghost-hover" over elements after a completely different element in the page was tapped/clicked, and that's just what's off the top of my head from when I last did frontend dev consistently (about 7ish years ago).
The only way to repro any of Safari's weird bugs nonsense, of course, was to shell out for a Mac or iOS device. You can't repro any Safari issues on any other, less-expensive platform. It's classic Oracle-style vendor lock-in with their vertically-walled-off tooling and lack of test images.
Except at least Oracle never put out a constantly-broken browser, let alone one that had -- and still has -- monopoly capture of an entire platform.
If you've got a desktop Firefox or Chrome issue, you can launch Firefox or Chrome on a cheap Linux or Windows machine or VM and repro it, and then snapshot the VM for reproducibility.
If you've got a mobile Firefox or Chrome issue, you can launch Firefox or Chrome on a cheap android device (or emulator), and if for some reason it really truly proves to be something Google or Mozilla hasn't implemented correctly _and_ you can't polyfill it...then you can advise users to try a different browser as a workaround.
There _is no_ "different browser" on iOS. If Apple has screwed up yet another standard web API, your users are hosed unless you can get really creative at dancing around their broken pile of crap.
After switching user agents to Chrome and trying it out, I ran into a problem pretty quick. The dynamically generated "Magic" backgrounds use SVGs with the 'filter: blur' style which isn't working properly despite it being supported by Safari [1]
Most of the features work fine though. Website does state that it's in beta as well.
Someone doesn’t have to have anything against other browsers to prefer using their favorite browser without futzing around with browser-site combinations.
Also, web developers here are interested in any reports of cross platform problems, so they can avoid requiring their users to futz around.
> Someone doesn’t have to have anything against other browsers to prefer using their favorite browser without futzing around with browser-site combinations.
Yes, this is irritating. If I encounter a site/app with browser requirements, I’m going to find an alternative to that site/app, not use a different browser.
Here are some Safari-specific issues I have met when developing a website in the past two years:
* font being displayed way bigger than in the other browsers - had to add normalization rules specific of Safari
* Parsing dates in the format "YYYY-MM-DD HH:MI:SS" not working, had to convert to ISO-8601 using a regex (the dates in original format were received from a third-party website).
* Interactions in maps not working using leaflet library (version 1.7 at the time). Fixed by switching to maplibre, which turned out to be a great move for other reasons.
* HTML in SVGs (via ForeignObject) are always displayed on top of the SVG content, making it impossible to annotate HTML with SVG markup (arrows, circling stuff).
Plus some issues with `backdrop-filter:blur`, although this CSS is also problematic elsewhere.
All in all, it felt like having to deal with Internet Explorer 10+ years ago: the other browsers work well out of the box, and tricks and workarounds need to be identified for the bad browser.
Finding maplibre 'better' was more valid at the time than today, and is also subjective. The creators and maintainers of both libraries have done some great work (and are still doing so).
Back in January 2022, the stable version of leaflet, v1.7.1, was from September 2020, and was affected by some small bugs degrading the user experience. Although the release of following version seemed close, there was no clear schedule for it, and I had concerns about how maintained the library would remain.
As of today, the bug from 2015 where there is some white space between map tiles on fractional zoom levels [0] is still open.
Also, leaflet was a pain to integrate in Svelte Kit framework, because it depended on `window` and-or `document`, not available at server side.
Maplibre, on the other hands, with a feature set roughly equivalent to Leaflet, benefited from much more frequent releases, and seemed more stable across browsers and devices. It was also easier to make it work in Svelte kit.
It’s not. We’re now in the same situation we were in 20 years ago with Internet Explorer, where other browsers are measured by “does it act like Chrome?”
Yes, it does have bugs and shortcomings – but all browsers do. When a predominantly Chrome developer finds something that doesn’t work in Chrome, the perception is that the functionality is not ready for primetime; when they find something that doesn’t work in Safari, the perception is that Safari is bad; when they find something that doesn’t work in Firefox… ha, just kidding, they don’t test in Firefox. Firefox is the new Opera.
Try making a game for Safari and then tell me it's not hard. It's definitely not just a case of different browser quirks. Safari is so far behind Chrome and Firefox that it's not worth supporting.
Apple genuinely have some of the weirdest “fans” I’ve ever seen.
To sit there and try to tell an audience of largely web developers that Safari’s reputation as the new IE is totally unfounded as though they don’t know for themselves exactly what the last ten years were like themselves.
Or as though multiple governments across the globe didn’t just spend the past couple of years opening multiple antitrust investigations looking at what a problem Safari had become.
No… apparently the problem is just you and your perceptions and whatever other weird gaslighting that seems to come with these defences.
It's not just Safari, it's every webkit implementation, e.g. Orion, etc.
I too prefer the webkit rendering and use that daily.
People blaming Webkit for being behind should realise the amount of people working on Webkit vs. Blink and question whether they want another internet explorer.
I usually feel a bit apprehensive when I see a new app pop up using the .so TLD (Somalia, operated by SONIC.so), since we’ve faced a lot of issues with it over the years. I end up repeating the same kind of comment about it (eg https://news.ycombinator.com/item?id=36825966), but I wonder if all the apps on it, including Notion, have changed the reputation of it with corporate firewalls by now.
Still, watch out if you host user-generated content there, it can be hard to undo reputation damage to .so domains if you get reported to various watchdogs. Notion’s longest hard-down outage happened after a takedown notice got sent to one of our upstream providers but they never forwarded it to us, and it took ling hours on the phone to various NICs and ISPs to get unblocked.