[ZwartePiet]

The fundamental problem with Kodi is that media is difficult to access. Any potential Kodi user is going to have to put up with:

1) Sticking to Live TV
2) (Illegally) ripping personally owned DVDs / Blu Rays
3) (Illegally) ripping rented/friends/relatives DVDs / Blu Rays
4) (Illegally) downloading media
5) (Illegally) streaming media

In other words, Kodi's demographic is inherently limited to enthusiasts. Nate is right to emphasize simplicity, because UX/UI complexity further limits that demographic to ardent enthusiasts. The problem with ardent enthusiasts is that they've invested so much time and money into this hobby that their goals (maximize return with -- FEATURES!) add complexity, setting up a self-selecting positive feedback loop.

If Kodi wants to grow marketshare, it needs to cater to the casual enthusiast. This is the type of person who used to rip CDs for their iPods back in the day... and who probably follows a few tech websites today. Catering to this market means adopting an Apple-like mindset - focus on simple, robust and polished core functionality. Emphasis should be on:

1) Router-based storage
2) CEC-enabled, USB-powered clients
3) Full Blu Ray specification support
4) Fluid GUI/Playback
5) Scrapers returning professional metadata
5) Network tuners for Live TV
6) Full PVR support (w/o need of server)
7) Live TV / PseudoTV integration
8) Media jukebox for local media
9) Automatic syncing between clients
10) 24/7, 365 stability... stability, stability, stability

Master all ten and the market would be huge.

[nickr]

2) (Illegally) ripping personally owned DVDs / Blu Rays

Not illegal here, and not in most places.

1) Router-based storage

Are you crazy?

In fact most of the 10 are there.

[ZwartePiet]

Not illegal here, and not in most places.

Illegal here and other places. Hence the parentheses. The root of the problem is the time required, anyways.

Are you crazy?

No?

For modern low-power clients (e.g. Raspberry Pi), the major storage options are:

1) Direct-Attached USB HDD
2) Dedicated NAS
3) Router based storage

1) works fine, but it adds clutter to a device best suited to the back of a TV.

2) requires decent money. HDD size growth means that in a year or two, multi-slot NAS will only be needed for the most extreme of collectors. A 20TB drive will hold well over 500 Blu Rays. That money would be better served towards...

3) Buying a decent router. Anyone who has a cheap router knows this is the worst place to save money. Nice router's have sequential reads >80MBps and writes >40MBps, which are more than enough for an average use case. Write speeds are the most important and 40MBps is enough for the biggest bottleneck -- ripping a Blu Ray to disk. Back up is through mirroring.

In fact most of the 10 are there.

...but not very well done.

Linux desktop is mostly feature parity with Windows and OS X, but its usage share has been stuck ~1% for as long as I can remember. The best funded and marketed distribution (Ubuntu) is a fraction of that. The dream of Shuttleworth and many open source advocates of shifting those usage percentages higher hasn't come to fruition. Kodi and Linux are both trapped in the same positive feedback loop I mentioned above

If that's the audience both want to cater too, great -- keep at it. But if Kodi wants to increase its usage base, I think polishing the product would do the most good.

[nickr]

Lets go through them shall we:

1) Router-based storage

Works the same as any other network based storage, except most routers are too pathetic to do it well

2) CEC-enabled, USB-powered clients

Raspberry pi, probably a number of android machines, but hey, kodi is deliberately hardware and OS neutral as it can be.

3) Full Blu Ray specification support

Blurays play except for menus

4) Fluid GUI/Playback

Mine works very fine.

5) Scrapers returning professional metadata

Done years ago

5) Network tuners for Live TV

Done (see in particular the recent HDHomerun client addon)

6) Full PVR support (w/o need of server)

I don't think you are gonna get the kodi devs to develop a full recording PVR, but as above for livetv

7) Live TV / PseudoTV integration

As above, not sure what pseudotv is though.

8) Media jukebox for local media

I have no idea what you mean by this

9) Automatic syncing between clients

Say slowly, m y s q l


10) 24/7, 365 stability... stability, stability, stability


Improving all the time, i haven't rebooted my clients except to update openelec in months.

[sarbes]

To extend the last post of nickr:

3) The specs are not free, so there will be no full implementation for now.
6) No server is needed, the backend can run on the HTPC. Configuration could be better though.
9) Sort of valid.

Small points like 9) could be implemented better for novice users, but that would be one of the last features to implement (IMO). Even without that, I can't think of a mediacenter software which could compete in terms of comfort and features. Users who expect a restricted and locked down Apple like platform will simply buy an AppleTV - there is nothing wrong with that.

The previous point of DVD ripping is irrelevant in almost every country in the world. Also, streaming from normal content providers like Youtube is not illegal.


To stay on the topic, I am happy with the progress of MVC for the Pi and the new version of FFMpeg with HEVC hardware acceleration. Sure, there are no huge additions like Live TV or (the still pending) RetroPlayer, but there will be enough reasons to upgrade.

[ZwartePiet]

Lets go through them shall we:

My pleasure.

1) Router-based storage

Works the same as any other network based storage, except most routers are too pathetic to do it well

It's clearly feasible, yet you suggested I was crazy for considering it. I disagree with that stance and think it's a disservice to malign it as an option. The number of users willing to maintain a multi-spindle storage solution is negligible... and Kodi has already cornered that demographic. Pushing for simple solutions expands the potential user-base.

2) CEC-enabled, USB-powered clients

Raspberry pi, probably a number of android machines, but hey, kodi is deliberately hardware and OS neutral as it can be.

I'd like to see usage numbers, but I wouldn't be surprised to learn that the vast majority of usage growth over the past few years has come from the Raspberry Pi. Kodi should remain platform agnostic, but this is the growth area and should continue to be emphasized as such.

3) Full Blu Ray specification support

Blurays play except for menus

Except no device has complete Blu Ray specification support. Show me a device that decodes/bit-streams h.264 (720p, 1080p/i), VC-1 (720p, 1080p/i), MPEG-2 (720p, 1080p/i), DTS-HD MA, True HD, 3D SBS, 3D MVC, .iso and has full menu support. AFAIK, the Pi comes closest...

4) Fluid GUI/Playback

Mine works very fine.

Great. That should be a universal experience.

5) Scrapers returning professional metadata

Done years ago

I strongly disagree. Fan art is inconsistent and low resolution. Plot synopses are poorly written, have inconsistent length (some scroll for pages) and are usually incomplete. Compared to IMDB, cast and crew lists are almost always incomplete. Actor fan art is especially inconsistent, fragmented and low quality. New releases have the best support. The other 99% of material is what suffers (especially TV series and Music).

Kodi also doesn't default to locally storing scraped meta-data, meaning the casual user needs to re-scrape after every reinstallation.

5) Network tuners for Live TV

Done (see in particular the recent HDHomerun client addon)

They exist, but they can be integrated much better. The HDHomerun app should be a back-end, not an add-on. It has an inconsistent UX/UI since it doesn't use the native skin. HDHomerun also includes APIs that would make recording without a server relatively simple. The primary emphasis for most PVR backends is on a server / client solution, which again very few potential users are going to deal with.

6) Full PVR support (w/o need of server)

I don't think you are gonna get the kodi devs to develop a full recording PVR, but as above for livetv

I'm not going to get anyone to do anything. I'm going to advocate my opinion.

7) Live TV / PseudoTV integration

As above, not sure what pseudotv is though.

The closest thing to integration is PseudoTV Live, but it doesn't integrate with native PVR functionality (especially skin support) because it's an add-on not a back-end. It's also extremely complex in its current implementation and thus subscribes to the hobbyist "more FEATURES!" mindset. As a result, it's not stable.

8) Media jukebox for local media

I have no idea what you mean by this

This is the core feature of Kodi and what it does the best. Emphasis should never veer from here, IMO.

9) Automatic syncing between clients

Say slowly, m y s q l

MySQL needs to be set up, requires usage of a CLI and a server. It also breaks if clients use different db versions by staggering updates. It's not automatic by any definition of the word.

It may currently be the best solution for client syncing, but it's not a good solution.

10) 24/7, 365 stability... stability, stability, stability


Improving all the time, i haven't rebooted my clients except to update openelec in months.

Great. That should be a universal experience.

[mkortstiege]

5) Scrapers returning professional metadata

Done years ago

I strongly disagree. Fan art is inconsistent and low resolution. Plot synopses are poorly written, have inconsistent length (some scroll for pages) and are usually incomplete. Compared to IMDB, cast and crew lists are almost always incomplete. Actor fan art is especially inconsistent, fragmented and low quality. New releases have the best support. The other 99% of material is what suffers (especially TV series and Music).

And so i do. That's up to the community and strongly depends on the scraper you're using. There are more than the default ones we ship with.

Kodi also doesn't default to locally storing scraped meta-data, meaning the casual user needs to re-scrape after every reinstallation.

That's right, we do not pollute user dirs without explicitly being told so. If you choose to export your libraries you're asked where to save it. Being it a one-in-all export or export alongside your actual media. This metadata is used by the infoscanner and/or can be imported after reinstallation.

[Ned Scott]

I'm not sure why ZwartePiet decided to give general feedback in a thread about upcoming features, but I've split those posts out to this new thread.

[Ned Scott]

I'll throw my two cents in here.

In other words, Kodi's demographic is inherently limited to enthusiasts. Nate is right to emphasize simplicity, because UX/UI complexity further limits that demographic to ardent enthusiasts. The problem with ardent enthusiasts is that they've invested so much time and money into this hobby that their goals (maximize return with -- FEATURES!) add complexity, setting up a self-selecting positive feedback loop.

I don't think I totally agree with this, but even if I did, I don't see it as a bad thing. So what? Kodi is and has always been targeted towards the power user. People who just want to watch a single video file every now and then will just do so using one of the hundreds of other alternatives. We are not one-size-fits-all.

If Kodi wants to grow marketshare, it needs to cater to the casual enthusiast.

We don't want to grow "marketshare". We're not a company. We have no profits to make. Our current userbase is actually exploding at a point that we can't always keep up with the support burden, and it shows no signs of slowing down.

We do want to spread Kodi and open source in areas that it makes sense. That doesn't mean we try forcing Kodi in areas that don't make sense or are not the right tool for the job. If you want a turn-key video player then you probably don't even care about Kodi's core features. You don't want to dim the lights in the room automatically when a movie plays, or horde every movie you've ever seen on a mountain of hard drives and then easily find one really fast, or mix live TV with a "pseudo" TV channel made up of your own local video files.

That's not to say that we shouldn't try and improve in any way we can. Heck, if it's possible to cater to the causals and someone wants to do that work, great! But I won't loose any sleep if it doesn't happen.

[NEOhidra]

3) Buying a decent router. Anyone who has a cheap router knows this is the worst place to save money. Nice router's have sequential reads >80MBps and writes >40MBps, which are more than enough for an average use case. Write speeds are the most important and 40MBps is enough for the biggest bottleneck -- ripping a Blu Ray to disk. Back up is through mirroring.

From my exp (3 years back i suppose) a consumer router capable of supporting such speeds (on SAMBA share) would cost me almost as much as cheap AMD build (if there is such thing out there). Considering the wider choice of additional services it provides i ma going with AMD and the router stays strictly for routing.

[da-anda]

@PVR-backend
we won't write one, because writing a builtin backend would require to run on every single platform we support. The best we can do for a better out-of-the-box experience is having our own OS distribution (something like OE) with a preinstalled PVR backend running, one that can be configured via the Kodi UI some day (tvheadend might be an option). But there is no way to get a consumer ready product with only shipping the application itself as there are sooooo many pitfalls on OS side (drivers, ...)

[Dark_Slayer]

We don't want to grow "marketshare". We're not a company. We have no profits to make. Our current userbase is actually exploding at a point that we can't always keep up with the support burden, and it shows no signs of slowing down.

We do want to spread Kodi and open source in areas that it makes sense. That doesn't mean we try forcing Kodi in areas that don't make sense or are not the right tool for the job

Nicely put