Forum for discussing general topics related to Couch.
9 posts Page 1 of 1
Hello everyone and Happy New Year! I am contemplating putting some effort into developing a project but am not sure if Couch is the right platform. The ideal scenario - nothing revolutionary - would be for a client to sign up to use the project, log in to the Admin Panel and input data which would then be presented front-end to logged in users only. Ideally each client signed up would be using the same instance of Couch referencing their own set of database tables. I don't think that this is possible with Couch? Maybe I'd need to think in terms of a separate Couch installation for each client - set up as a sub-domain - so https://projectname.com/subdomain-xyz/

The other thing that would be a problem is the presence of Super Admin listed in Users. Because the client's data is confidential they wouldn't want to see a user listed who was not someone they personally have granted access to their Admin Panel. Yet, Super Admin is necessary ... ?
Hello and a very happy new year to you too :)

Regarding the use-case, I'll need more details to understand it.
I doubt this is the case but would there be a fixed number of clients? If not, I don't think creating a sub-domain for each would be feasible.

Please post the requirements in greater detail.

Thanks.
The app is related to care provision - so would involve personal data. A client would sign up/register (not sure how to collect a subscription - maybe by separate arrangement) to use the app ... they would begin by completing a profile of the person receiving care and various other information to share with a group of users - medication, contact information etc. The client would grant access to other members of a care team or health professionals etc. These users may have access to update only certain data e.g. a daily diary. Or they may only have access to view certain data. It may be that I would use data-bound forms and extended entities to achieve all this.Not sure!

The number of clients signing up to use the app would be any number and each client would require their own tables/database and secure data.

I hope this clarifies things a little @kk - thank you for giving it some thought!
The client would grant access to other members of a care team or health professionals etc.

Interesting, sound like a care house or a similar institution. I would ask clients to go to your umbrella website, then sign up / request account and provide some contact details. Next, your team talks to the client and sets up a completely separate server + domain + prefab couch installation. It is good for them to have their own point of entry (domain), separate server with sensitive law-protected med data.

I strongly discourage to manage everything in a single backend. Would be a quick real mess very soon!
potato wrote: The app is related to care provision - so would involve personal data. A client would sign up/register (not sure how to collect a subscription - maybe by separate arrangement) to use the app ... they would begin by completing a profile of the person receiving care and various other information to share with a group of users - medication, contact information etc. The client would grant access to other members of a care team or health professionals etc. These users may have access to update only certain data e.g. a daily diary. Or they may only have access to view certain data. It may be that I would use data-bound forms and extended entities to achieve all this.Not sure!

The number of clients signing up to use the app would be any number and each client would require their own tables/database and secure data.

I hope this clarifies things a little @kk - thank you for giving it some thought!


I still don't clearly understand what are you trying to do, but at a certain level could be possible. I have played with databound forms and I have developed some applications, so it depends how complex the app would be.
Probably will require separate subdomains or domains, in order to keep things separate and more safe.
thanks for your replies @atisz and @trendoman ... the app would be of use to individuals dealing with care in their own homes. I had hoped that a single domain name could be used and even a single instance of Couch. Imagine if there were a large number of clients and the app were to be upgraded or require a change it would take quite a lot of work to apply those changes to each domain. Maybe my hope of developing a single Couch-based 'app' involving clients signing up to an account is a step too far! I'll still consider the separate domain route/Couch installation per client though ...
potato wrote: Maybe my hope of developing a single Couch-based 'app' involving clients signing up to an account is a step too far! I'll still consider the separate domain route/Couch installation per client though ...


Going separate or combined really depends on more factors than were pronounced. I personally can't foresee 2 steps ahead without a plan. So my reach is pretty limited without a plan. I mean it is always best to create a backend plan based on realistic frontend mockup, with manually populated data. I'd see then where extra caching needed, where ajax based loading needed (loading values after selecting a value), whether best to use mosaics or repeatables, how to handle separate image uploading, how to overcome slow unresponsive page loading, notifications to mail/im and many many other bottlenecks that were/are commonly discussed in forums.

It is definitely possible to have everything sheltered together. For that to work out well, developer must create a DBF-based backend for clients/staff to log and manage their stuff in and keep Couch completely in the shadow, solely for private superadmin's business (data stronghold). It's therefore a huge app and if you want to have it running in 6-9 months at least, you'd need a lot of help which I gladly offer here.
thanks @trendoman - it sounds all a bit complex and risky to invest so much time/money at this stage. I'd be better off trialling the concept using individual installations of Couch and see how it works out.
potato wrote: thanks @trendoman - it sounds all a bit complex and risky to invest so much time/money at this stage. I'd be better off trialling the concept using individual installations of Couch and see how it works out.

Exactly. It is complex and risky everytime we go big. Did you have a positive experience with notejam-like apps?
9 posts Page 1 of 1