Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[SMatson]

I currently use the media server to send my library to a different computer in my house (connected to my home theater).  Why not just point the home theater computer to my networked library file?  Are there any advantages of the server for home use?  Aren't I just using resourses running the server on my upstairs computer?

Thanks,

Scott

[John Gateley]

You can corrupt the database if two PCs access it at the same time (using Windows sharing).  This can't happen with Media Server.

j

[JustinChase]

Any chance this is getting a rework in 11.1?

[John Gateley]

Yes, there's a chance, we're always working towards this. But it's a pretty big hurdle, so no promises.

j

[GHammer]

I'd think using a DB like Microsoft SQL Server 2000 Desktop Engine (MSDE 2000) would take care of the DB issues.

Why reinvent the wheel?

"MSDE 2000 is a royalty-free, redistributable database engine that is fully compatible with SQL Server. MSDE 2000 is designed to run on Microsoft Windows® 98, Windows Millennium Edition (Windows Me), Microsoft Windows NT® Workstation version 4.0 (with Service Pack 5 or later), and Windows 2000 Professional as an embedded database for custom applications that require a local database engine."

Of course, if you don't care about Win 9x support then Microsoft SQL Server 2005 Express Edition is a better tool.

"Microsoft SQL Server 2005 Express Edition (SQL Server Express) is a powerful and reliable data management product that delivers rich features, data protection, and performance for embedded application clients, light web applications, and local data stores. Designed for easy deployment and rapid prototyping, SQL Server Express will be available at no cost, and be free to redistribute with applications."

[hit_ny]

I'd think using a DB like Microsoft SQL Server 2000 Desktop Engine (MSDE 2000) would take care of the DB issues.

Why reinvent the wheel?

I'm guessing
- it will be slower than what's implemented currently. Also that's a black box (can't look inside), won't be able to speed it up.
- the dB is the heart of MC, uncomfortable to not have 100% control of it.

[GHammer]

I'm guessing
- it will be slower than what's implemented currently. Also that's a black box (can't look inside), won't be able to speed it up.
- the dB is the heart of MC, uncomfortable to not have 100% control of it.

SQL Server is one of the speed champs in test after test. It is unlikely that something made to massage terrabytes of data is going to be slow.

Peek under the hood? Why? You want to store and retrieve data. It does, and for the original question, knows how to share access to that data.

Will it happen? Nope. Should it? Well, it would certainly remove the care and feeding of the database engine from the developers.

[John Gateley]

SQL Server is one of the speed champs in test after test. It is unlikely that something made to massage terrabytes of data is going to be slow.

Custom databases can take advantage of the structure of the data, while generic database are designed to work well on ANY kind of data, and will not beat a (well designed) custom database in any circumstance.

j

[bbrip]

John,

has the library server problem in 11.1.39 been fixed in 11.1.40?

Thanks
B

[johnp]

I personally couldn't care less how it is implemented, I just want to be able to massage the database from any PC in my house.  This includes adding titles or removing them,  riping CDs, creating playlists, etc.  Simply streaming files from the master to a client is not satisfactory.

The application model of having a central media server (perhaps even rack mounted in an equipment room) that is not conveniently located with a console to work with is the rule, not the exception.  To that end most users expect to be able to access and manipulate the data from any PC on the network.  JRiver MC falls well short of that expectation

It seems a reasonable expectation to me and one I have myself.  Couldn't you just lock the database while it is being modified by a client and issue an update to all other connected clients when the update is complete?

Or why not have MC be service based and use an applet to log in and manipulate the data?  If a session was open you could have it automatically close if the same user opened the new session.  That way only one user would have access to the database at any given time.

As I said earlier I am not really interested in how it is implemented and I am sure that my suggestions are an over simplification, but this feature has been needed since 8.x and it is still not off the ground.  Very frustrating for an integrator like myself.  Especially when I know there are database solutions out there that offer these features.

Best,

John

[johnp]

I have another thought.   What about making an automateded way to update multiple unique independant libraries.  This actually has a nice side effect advantage by keeping multiple copies of a the library, so if one gets corrupt, you have any number of client machines with a recent backup copy.  Think of this as a kind of library sync feature.

This is my thinking:

1.  Designate one PC as the master.
2.  Designate 1 to n PCs as clients
3.  Give the master a client list, or conversely give the clients a master list. so they know how to deal with one another.
4.  If you are about to make library changes from a client, pull master library from master (synchronize)
      I. In todays terms:         
             a. Backup master's library to network drive.
             b. Restore library on client
      II.  In dream world terms with a single command:
             a. Client requests master sync pull
             b. Client opens port with master and synchronizes library
5.  Then tweak local library to your hearts desire using local CD-ROM resources to RIP, burn, IPod Sync etc. etc.
6.  When finished, do a library push to master
      I. In todays terms:         
             a. Backup client's library to network drive.
             b. Restore library on master
      II.  In dream world terms with a single command:
             a. Client requests master sync push
             b. Client opens port with master and synchronizes library
7.  Additionally clients could do requested puls on a nightly schedule perhaps?
8.  Even more cool would be a library merge feature that generates a dif between the libraries and automatically syncs them.


Is there a way to implement this today through the current automation interface using girder or something?

Just food for thought.

[glynor]

I have another thought.   What about making an automateded way to update multiple unique independant libraries.  This actually has a nice side effect advantage by keeping multiple copies of a the library, so if one gets corrupt, you have any number of client machines with a recent backup copy.  Think of this as a kind of library sync feature.

This is exactly what I'm doing currently and it works fairly well.  Check this thread for a more detailed explanation.

I too would like to see a more robust, multi-user, network-aware database system (as I think JRiver would as well).  To put it in perspective, enterprise-level systems like that are not typically cheap or simple to implement.  At my company we are currently evaluating a number of enterprise level Digital Asset Management systems that offer these kind of features and they start in the $6,000 range and scale to the Millions (and most of them require that you already have an enterprise class database backend yourself).

[tofer]

C'mon guys.....

Let me get right to the point.

Until JR offers a true multiuser, multizone media management system, it will remain just a great <$50 media program competing with 100 others in an already crowded space.  If you haven't noticed - there's a rare opportunity sitting in front of you - JR's strength and reputation has been built on media management (most other systems just get in the way of managing media), you already have a handle on dealing with multiple media types, and you've got the underpinings of a multi-zone system.... why not finish the job and reset the bar when it comes to true media management.

With the advent of paid digital music downloads, we're no longer talking about a bunch of MP3's on a college student's laptop - were talking about serious money spent on media collections.  People are looking for serious tools to manage, archive, label and track their media - regenerating a failed library - or copying soon-to-be out of date library files all over a network is not the answer. 

As talented as your developers are, there is no reason they should be spending their time developing database management systems... especially single-user databases (almost an oxymoron these days).  There are plenty to choose from, their cheap or free, and they'll easily handle the relatively modest workload MC requires.

Fortunately for users.. someone will soon fill the void and jump at this opportunity.  I sincerly hope its JR.  It would reward and validate the collective participation of your customer-centric model for development and move "us" up to a new plane of media mangement.

Thanks for getting us here, now lets move on to the next level.

Tofer

[johnp]

What he said... 

[John Gateley]

Hi Y'all,

I hear what you are saying, and we're working towards this.

As talented as your developers are, there is no reason they should be spending their time developing database management systems... especially single-user databases (almost an oxymoron these days).  There are plenty to choose from, their cheap or free, and they'll easily handle the relatively modest workload MC requires.

It isn't as easy as this. A general purpose database is always slower than a custom crafted database. The kinds of things Media Center does won't dovetail into what general purpose databases do well.

We are working towards this, but it isn't as easy as everyone thinks.

j

[jgreen]

In talking about high-end software, I hear a lot about "Enterprise Class", and that interests me a great deal, as I am a longtime fan of Star Trek.  If they're offering lessons on it I'd like to know where to sign up.

Regarding the other bit, namely where jriver ought to point the mad scientists next, I agree with the posts in this thread that see the high-end db functions as the future for the high-end home media enthusiast and the SMB.  There is an awful lot of content being made available cheaply these days, and I think MC is in a class apart for dealing with it.  To a me, a non-programmer, non-engineer--pretty much non-anything that you could think of--it seems like a "slam dunk", although I don't play basketball, either. 

I'm sure there are technical details to be worked out, but that's what other people are for.  In the meantime I'll keep listening to great music using your great software.

[johnp]

It isn't as easy as this. A general purpose database is always slower than a custom crafted database. The kinds of things Media Center does won't dovetail into what general purpose databases do well.

I hope I didn't trivialize the effort to accomplish this in my comments.  It was not my intention.  I was just trying to point out that this is a very neccesary feature and try to find a work around using the current backup / restore functions.

Best,

John