[phillyfan358]

Hi All,

I'm Dave, current Masters student at University of Pennsylvania, been a user of Kodi for a couple years now. Really excited about Kodi being a part of the Google Summer of Code as I've been looking for ways to give back to this awesome community. I have a bunch of ideas for projects and one that matched up well with proposed ideas was the Better Profile Support suggestion.

I have a good amount of experience with C++ and have worked with both relational and NoSQL databases extensively. My github has some of my projects that aren't work related.

From the wiki, this appears to be a pretty large undertaking, as they suggest multiple people to work on it and/or splitting the project into multiple chunks. My proposal has some overlap with poplap's proposal (here), which should be fine so long as we are still considering dividing this project into chunks.

Kodi Better Profile Support
Name: Dave Hand

forum/e-mail: phillyfan358 / davehand13 at gmail.com

Summary: Kodi profile support is not very modular/well organized at this point. There is little support for multiple users and for organizing profiles in a simple, easily accessible way. My project works to revamp how the profiles are stored in the database and also adjust how profiles are accessed, possibly through a framework, as poplap suggests.

How will I achieve this: First learn and understand how profile data is currently stored and accessed throughout Kodi. Then work to redesign the storage logic on the db end. Then work with poplap and others to combine a C++ framework with the database backend.

What will the project focus on: Profile support / organization, multi-user support

Benefits: All users of Kodi will benefit from better profile support and organization through a better app. Users with multiple profiles will especially benefit. This will make it easier and more modular for developers to do their thing without having to hack together profile support or access.

Goals: 1) Learn more about the under the hood workings of Kodi, especially in regards to profiles the db. 2) Redesign the profile system to be more modular/organized and less hacked together. 3) Meet current Kodi developers as well as other GSoC people 4) Contribute to one of my favorite open source projects.

What does it touch in Kodi: Just about everywhere where profiles are used in the Kodi app (will need to be revised to access through framework or db), Profile system, database storage, settings, addons

Requirements: C++ knowledge, DB knowledge, some architecture knowledge

Possible mentors: For you guys/gals

Let me know if there's anything missing or if I should flesh out my idea some more. Will also be submitting on GSoC.

Status: Student Did not complete
Code: Unknown

[poplap]

I would be willing to work with you on this though I think we would have to figure out what overlaps and what doesn't and try to break it up as much as possible.

[phillyfan358]

Definitely agree on that! Glad to hear you're willing to work together!

[colek42]

I have a persistent data model proposed in my Gsoc proposal. Might work well for this as well. Let me know. I can create stubs for testing very quickly.

[phillyfan358]

Hey colek, sorry for the delay. Your model looks like it could be applicable, so those stubs may come in pretty handy if the proposal is accepted. Thanks for reaching out!

[ludi]

Hi,

First of all, it is great to see that the Profile Support in Kodi is being improved; thanks for working on this.

I don't know whether the following is directly related, but when I think about Profile support, I think that it would be even greater if several Kodi installations could also use the same database stored for example on a NAS; kind of making the Profiles (or should I say the whole database) device independent. This would for example allow the user to start watching a movie in the living room, stop it, go to another device with Kodi and resume watching the movie there.

Of course, this is not part of your GSOC and I am aware the it is already quite late. Moreover, I don't even know whether it is related to what you are working on. I nevertheless wanted to mention it, in the case where it would be relevant for the enhancement you are currently working on and you wanted to take it into account, so that your work can later be enhanced to achieve what I mentioned above.

Cheers

[Razze]

Thanks for taking this challenge, I'm really looking forward to a real profile support.
I really don't like the fact that you have to rescan the folders for each profile on their own.

Maybe there is even a chance to improve parental control (in the long run), something like setting an birthday for the profile and it hiding the movies/shows that have been scrapped with an bigger age restriction.

Hi,

First of all, it is great to see that the Profile Support in Kodi is being improved; thanks for working on this.

I don't know whether the following is directly related, but when I think about Profile support, I think that it would be even greater if several Kodi installations could also use the same database stored for example on a NAS; kind of making the Profiles (or should I say the whole database) device independent. This would for example allow the user to start watching a movie in the living room, stop it, go to another device with Kodi and resume watching the movie there.

Of course, this is not part of your GSOC and I am aware the it is already quite late. Moreover, I don't even know whether it is related to what you are working on. I nevertheless wanted to mention it, in the case where it would be relevant for the enhancement you are currently working on and you wanted to take it into account, so that your work can later be enhanced to achieve what I mentioned above.

Cheers

You can already do that for movies, at least if your using trakt and the trakt addon.

[ludi]

@Razze

You are right: parental control would also fit in the design, even more than the feature I suggested above.

Thanks for the hint to trakt.tv. I did not know that the site was offering such a service. It even seems to offer more than what I was describing. But it also seems to require a subscribtion, probably among others because of the additional services it offers. Since I don't need these additional services at the moment, I won't subscribe for now. Thanks nevertheless for mentioning it.

[Razze]

@Razze

You are right: parental control would also fit in the design, even more than the feature I suggested above.

Thanks for the hint to trakt.tv. I did not know that the site was offering such a service. It even seems to offer more than what I was describing. But it also seems to require a subscribtion, probably among others because of the additional services it offers. Since I don't need these additional services at the moment, I won't subscribe for now. Thanks nevertheless for mentioning it.

You need a login, not a subscription. You can subscribe to get some thing, but their mostly visual, previews at upcoming features and a say for new features. So you can use everything without that subscription.

[zag]

Some requests for features here

http://forum.kodi.tv/showthread.php?tid=228415

[RockerC]

@phillyfan358, are you following the related discussion here http://forum.kodi.tv/showthread.php?tid=200911

You might want to try to get your input feedback of your requirements for to m.savazzi as soon as possible.

[m.savazzi]

I've build the new data model for the library to help have a better library and user management. I'm here to help you as I can!

[Razze]

As said at http://forum.kodi.tv/showthread.php?tid=...pid2090548
This sadly seems to be dead.

[iainmacleod]

There has been some discussion on this thread http://forum.kodi.tv/showthread.php?tid=...pid2086434 for profile functionality.

For me centralized profiles would be great, so I could log in and see watched and resume status on a per user basis.

[Stormregion0]

There has been some discussion on this thread http://forum.kodi.tv/showthread.php?tid=...pid2086434 for profile functionality.

For me centralized profiles would be great, so I could log in and see watched and resume status on a per user basis.

Fully agree. Would be interesting to know if this project is still alive?