2 ms·

I think there is an important difference between tools/services you need once and those you need continuously. Designer or developer tools are something you on

I think there is an important difference between tools/services you need once and those you need continuously. Designer or developer tools are something you only need while you build the product. If those suddenly go away, the code you've already written with them will continue to work, so while it will slow down future development until you find an alternative, it won't stop your current product from working. Services that your product depends on to work are a different matter; if you built your service on AWS InfiniDash and it's suddenly deprecated or they hiked the price beyond what you can afford, you are screwed - your existing, currently-working product will stop working. I don't see much of a problem in using third-party tools for the former but the latter is definitely dangerous and should only be used as a last-resort.

piaste 4y ago
I agree with this assessment. My philosophy as tech lead was this: every time we wanted to start using a SaaS product, up to and including cloud hosting itself, we needed to have a Plan B to migrate to in case the service suddenly went away for any reason. Eg.: we needed a message queue, we picked up an Azure product because it was the most convenient and covered all our requirements at low cost. But we coded against an interface and we implemented a proof-of-concept RabbitMQ implementation as a backup plan (also came in handy for local testing!). If we had to drop the Azure service for any reason, we knew we could install RabbitMQ on a VM and tune it up to production standards relatively quickly.
spookthesunset 4y ago
> Designer or developer tools are something you only need while you build the product. What startup isn’t constantly building their product? I’ve never worked at a startup, or any company, where the product was “done”. There is always something to add, something to fix, or something to improve.
- Nextgrid 4y ago
  Your revenue doesn't immediately stop if you stop incremental improvements. This gives you time to find an alternative tool. On the other hand, if the infrastructure your product is based on stops, your product is dead, revenue will stop and you may even breach your SLAs and have to pay them compensation.

piaste 4y ago

I agree with this assessment. My philosophy as tech lead was this: every time we wanted to start using a SaaS product, up to and including cloud hosting itself, we needed to have a Plan B to migrate to in case the service suddenly went away for any reason. Eg.: we needed a message queue, we picked up an Azure product because it was the most convenient and covered all our requirements at low cost. But we coded against an interface and we implemented a proof-of-concept RabbitMQ implementation as a backup plan (also came in handy for local testing!). If we had to drop the Azure service for any reason, we knew we could install RabbitMQ on a VM and tune it up to production standards relatively quickly.

spookthesunset 4y ago

> Designer or developer tools are something you only need while you build the product. What startup isn’t constantly building their product? I’ve never worked at a startup, or any company, where the product was “done”. There is always something to add, something to fix, or something to improve.

Nextgrid 4y ago
Your revenue doesn't immediately stop if you stop incremental improvements. This gives you time to find an alternative tool. On the other hand, if the infrastructure your product is based on stops, your product is dead, revenue will stop and you may even breach your SLAs and have to pay them compensation.