Sounds like Heroku's business model and that works fine for a certain type of business/consumer. Of course you could spin up your own VPS or code your own Rails ecommerce app. But that's not always economical - up to a certain scale.
API changes breaking sites is another matter entirely than their subscription plugin model which could be solved while maintaining that structure.
This change by Spotify seems to be related to Mailchimp not providing a certain standard of service that they demand from other plugin-services on their platform. I could see that being a costly choice for any customer who has to now migrate to other services, but the intentions were largely good and pro-customer.
Hopefully there was some sort of data migration process in place and some upfront warnings before your email marketing system gets cut-off.
I think the problem I am describing goes deeper than just the API calls.
It is about support and accountability. As a business owner, I want as few parties responsible for my infrastructure as possible with healthy portability possibilities (to avoid lock-in).
With Shopify app store, the whole idea is insane to me - I have to now deal with Shopify + plugin developers individually to fix issues. Things break all the time, just go to Shopify forums for support. And then, there is the risk of a developer leaving Shopify ecosystem with a dangling plugin never to be maintained.
Sure but supporting your own infrastructure is exactly why people choose to use a service like this.
From my experience running your own services is far more fragile and subject to maintenance-overhead due to API changes than a service like Shopify or Heroku.
The main drawback is a lack of flexibility and customization - not so much the lack of stability in service.
I agree and I can see why supporting your own service would probably be more of a headache.
What about stores such as Squarespace? They don't have a market place of apps and everything they do is in their control. Ofcourse, now there is a risk of lock-in and unable to extend/scale your store if needed.
Comments
Sounds like Heroku's business model and that works fine for a certain type of business/consumer. Of course you could spin up your own VPS or code your own Rails ecommerce app. But that's not always economical - up to a certain scale.
API changes breaking sites is another matter entirely than their subscription plugin model which could be solved while maintaining that structure.
This change by Spotify seems to be related to Mailchimp not providing a certain standard of service that they demand from other plugin-services on their platform. I could see that being a costly choice for any customer who has to now migrate to other services, but the intentions were largely good and pro-customer.
Hopefully there was some sort of data migration process in place and some upfront warnings before your email marketing system gets cut-off.
I think the problem I am describing goes deeper than just the API calls.
It is about support and accountability. As a business owner, I want as few parties responsible for my infrastructure as possible with healthy portability possibilities (to avoid lock-in).
With Shopify app store, the whole idea is insane to me - I have to now deal with Shopify + plugin developers individually to fix issues. Things break all the time, just go to Shopify forums for support. And then, there is the risk of a developer leaving Shopify ecosystem with a dangling plugin never to be maintained.
Sure but supporting your own infrastructure is exactly why people choose to use a service like this.
From my experience running your own services is far more fragile and subject to maintenance-overhead due to API changes than a service like Shopify or Heroku.
The main drawback is a lack of flexibility and customization - not so much the lack of stability in service.
I agree and I can see why supporting your own service would probably be more of a headache.
What about stores such as Squarespace? They don't have a market place of apps and everything they do is in their control. Ofcourse, now there is a risk of lock-in and unable to extend/scale your store if needed.