Ok, lots of thoughts. I’ve reviewed the website and started drafted this as a Nino Doc. Haven’t dug into the more complex functionality yet.
# Tactical Feedback
On iOS, my password manager doesn’t recognize the Nino app with the Nino.app domain, so autofill doesn’t work at login. Not a problem, just a little friction.
Breaking problem on iOS: I cannot type the letter “m”. It appears to trigger a keyboard shortcut that toggles a reorder paragraph block mode. I had to abandon Nino and finish this write up in Notes.
In iOS Nino Doc, it’s not obvious how to “Select All —> Copy”. Moving from Nino to Notes I had to copy each paragraph block individually.
Concur that it would be good to have some templates or something to orient on first app open. I tried “Notebook” first because I assumed it was like “Notepad”, a very light text editor. Seems to be more like a Jupyter notebook?
Maybe an easy way to address this would be to add a tag line for each document type. Just a couple words to orient folks beyond the name. Perhaps add a “compare to X” as well.
# Strategic Feedback
First, the “too many tools” problem is real. Lotus Notes was good. There is appetite for consolidation in some large enterprises.
It’s impossible to get a big organization to switch over to this directly. You need to target smallish teams (10s of people) that can make their own decisions.
Security is everything. Recommend you review Sandstorm’s capability security model. That was the most promising productivity revolution I’ve seen in a while. It failed in part because they expected other people to write applications to their security model. If you can deliver the fine grained security AND a broad swath of functionality, you have something really special.
End to end encryption is excellent. Let me bring my own keys.
Some more specifics on security:
Block level security would be a killer feature. Especially if it goes beyond RBAC to ABAC. In my world, you frequently want to share a document with someone who is only authorized to view part of it. Automatically redacting the parts they aren’t supposed to see would be amazing. Then allow those blocks to be reused, with the same security labels, in many different contexts and app paradigms?
Honestly it would be amazing and I could sell it.
Next thing on security is data sovereignty. Absolutely respect going single region GCP to start. However, self-hosting will be an immediate ask in my world. Several tiers of this, all of which you can absolutely up charge for.
First, single tenet instances. Give me assurance that when bugs show up, they won’t leak my data to other customers.
Second, enterprises will want to run it within their VPC/Tenet that they already use / trust / audit. The major cloud vendors have regions focused on different industries, you need to be able to run there without calling home to your central instance.
Third, direct self hosting. Especially for smaller teams with poor connectivity back to the cloud. Can I run this on a NUC out of a trailer to help coordinate disaster recovery efforts? Can I host it in a VM on dedicated hardware where I understand the cost, capacity and backup story? Google and Microsoft can’t do this, so you can differentiate. See Next Cloud for a good approach to this.
Finally, interoperability. You have many different app views into the core block abstraction, but there will always be other tools and databases in play. Make it super easy for those external things to read and write blocks. Don’t write the integrations - just expose the API.
Thanks for finding the letter "m" bug, an embarrassing one for sure! Also thanks for bringing up the issue of select all on mobile, I'll think about how to improve that. Adding a tag line for each module is also a good idea.
Interesting that you mentioned fine-grained security. Currently block-level access control can be achieved with block embeds, but it has to be done one by one. I looked into enforcing access based on a query for page embeds (so allowing only portions of the page), it turned out to be non-trivial! Might try to implement in the future.
Thanks for breaking down the different levels of self-hosting. I'll keep them in mind.
# Tactical Feedback
On iOS, my password manager doesn’t recognize the Nino app with the Nino.app domain, so autofill doesn’t work at login. Not a problem, just a little friction.
Breaking problem on iOS: I cannot type the letter “m”. It appears to trigger a keyboard shortcut that toggles a reorder paragraph block mode. I had to abandon Nino and finish this write up in Notes.
In iOS Nino Doc, it’s not obvious how to “Select All —> Copy”. Moving from Nino to Notes I had to copy each paragraph block individually.
Concur that it would be good to have some templates or something to orient on first app open. I tried “Notebook” first because I assumed it was like “Notepad”, a very light text editor. Seems to be more like a Jupyter notebook?
Maybe an easy way to address this would be to add a tag line for each document type. Just a couple words to orient folks beyond the name. Perhaps add a “compare to X” as well.
# Strategic Feedback
First, the “too many tools” problem is real. Lotus Notes was good. There is appetite for consolidation in some large enterprises.
It’s impossible to get a big organization to switch over to this directly. You need to target smallish teams (10s of people) that can make their own decisions.
Security is everything. Recommend you review Sandstorm’s capability security model. That was the most promising productivity revolution I’ve seen in a while. It failed in part because they expected other people to write applications to their security model. If you can deliver the fine grained security AND a broad swath of functionality, you have something really special.
End to end encryption is excellent. Let me bring my own keys.
Some more specifics on security:
Block level security would be a killer feature. Especially if it goes beyond RBAC to ABAC. In my world, you frequently want to share a document with someone who is only authorized to view part of it. Automatically redacting the parts they aren’t supposed to see would be amazing. Then allow those blocks to be reused, with the same security labels, in many different contexts and app paradigms?
Honestly it would be amazing and I could sell it.
Next thing on security is data sovereignty. Absolutely respect going single region GCP to start. However, self-hosting will be an immediate ask in my world. Several tiers of this, all of which you can absolutely up charge for.
First, single tenet instances. Give me assurance that when bugs show up, they won’t leak my data to other customers.
Second, enterprises will want to run it within their VPC/Tenet that they already use / trust / audit. The major cloud vendors have regions focused on different industries, you need to be able to run there without calling home to your central instance.
Third, direct self hosting. Especially for smaller teams with poor connectivity back to the cloud. Can I run this on a NUC out of a trailer to help coordinate disaster recovery efforts? Can I host it in a VM on dedicated hardware where I understand the cost, capacity and backup story? Google and Microsoft can’t do this, so you can differentiate. See Next Cloud for a good approach to this.
Finally, interoperability. You have many different app views into the core block abstraction, but there will always be other tools and databases in play. Make it super easy for those external things to read and write blocks. Don’t write the integrations - just expose the API.
Good luck!