Database: Refactor MP1 database for multi seat usage (1 Viewer)

[CyberSimian]

As I said, getting information was like pulling teeth. It really seemed to me that it was bordering on being obstructive.

In my view, the reason that it has been difficult to get information about how these various features are supposed to work is because none of the current team members actually knows how they are supposed to work (sorry if that offends any of the team). Each of us knows about the small part of MP that we use, because we use it every day, but know nothing (or very little) about other parts of MP. For example, I am one of those weird people who cannot see the value in downloading additional information from the internet, and I have no idea how it is supposed to work (and cannot help with the testing of it).

You have done a tremendous amount of work in restructuring this fundamental part of MP, and from your descriptions of the changes, the modified code offers significant improvements. I for one would be saddened if this code were not incorporated into MP. But as @ajs has said, we must not remove function merely because we personally do not see the value of it.

So what to do? If there is an error in the code that is obviously a bug, then we fix it. But if it is not clear that it is a bug, then we need to preserve that behaviour. If you modify that code for other reasons, you need to ensure that the suspect behaviour is included and remains unchanged. Yes, this means doing additional work to preserve the suspect behaviour, possibly for no one's benefit. But without a long-term product architect who knows how each and every part of MP is supposed to work, we need to take that extra step and preserve that suspect code. This may involve more work than you want to commit to.

-- from CyberSimian in the UK

 

[framug]

Understood, completely.

OK then, I downloaded your code and, tried to apply.
As far as I can see, all parts are in MP client, not in TVEngine nor TVplugin, right ?
I tried to compile it like a beginner (it's on purpose) and, I have errors :



Also, by looking deep (but not too deep), I found you used framework 6.4.4.
What about ?
Are you sure 6.4.4 is installed on MP install ?

I had no more time for the moment but, this can give you some ways to follow for improving...

Also, you modified some plugins like minidisplay, etc... where, it could be minor and, delayed..
I keep believing that your work is great but, would be integrated step by step.
Hope it can help....

Regards.

 

[doskabouter]

That's because nuget packages were changed. You should fiddle around a bit in the nuget area of the project (can't remember exactly what but it was fairly obvious) and close/reload the solution.
Then it'll compile

 

[framug]

I was sure about a nuget story and, as "a purpose test" I didn't want to go deeper.
Therefore, it compile with MP 1.29 "official" code with all available nuget at the release time.
(Sure because I reverted entire folder).
Therefore, it's not the case with Tony code.

 

[doskabouter]

strange. I managed to compile it at my end fairly quicky (just copied over the new files (over master branch!) and did the nuget thing

 

[framug]

Will try nuget update later because of limited data download at this moment
Therefore, 6.4.4 and others (big things) could be delayed.
Step by step...
This is major what should be retained from my POV...
OFC, the main goal would be that finally, all Tony's (great) work could be integrated

 

[Anthony Vaughan]

strange. I managed to compile it at my end fairly quicky (just copied over the new files (over master branch!) and did the nuget thing

I only sent you that code so you could see what my coding is like. But it's already out of date. I have added whole loads to handle stacked movies and various other things.

My version uses Entity Framework. I did include the package fields to support that.

 

[Anthony Vaughan]

So what to do? If there is an error in the code that is obviously a bug, then we fix it. But if it is not clear that it is a bug, then we need to preserve that behaviour. If you modify that code for other reasons, you need to ensure that the suspect behaviour is included and remains unchanged. Yes, this means doing additional work to preserve the suspect behaviour, possibly for no one's benefit. But without a long-term product architect who knows how each and every part of MP is supposed to work, we need to take that extra step and preserve that suspect code. This may involve more work than you want to commit to.

Thank you for your comments.

To me, the best solution would seem to be to offer this as an option for people who don't want to use all of the old extensions but want to have a multi-user setup.

For example, I am one of those weird people who cannot see the value in downloading additional information from the internet, and I have no idea how it is supposed to work (and cannot help with the testing of it).

That is exactly what I feel. Links are provided to view actor/movie information directly from the IMDB website, so why store that data in MP? To me it's a major shortcoming because it makes scanning take forever. One of the changes I am going to make for my version is to make actor/movie data optional when scanning the first time around. This cuts down scan time from minutes to seconds. If you are interested, I will happily keep you apprised of what I am doing. I suspect we both use MP is the same way .

 

[emphatic]

Would it be possible to have all the databases on a shared volume or a NAS and point each client there?

 

[Anthony Vaughan]

Would it be possible to have all the databases on a shared volume or a NAS and point each client there?

Yes. That is really the whole point of my changes .