FORUMS: list search recent posts

Apple and databases

COW Forums : Apple Final Cut Pro X Debates

<< PREVIOUS   •   VIEW ALL   •   PRINT   •   NEXT >>
Aindreas Gallagher
Apple and databases
on Mar 27, 2013 at 3:04:57 pm

Interesting article on how hilariously incompetent Apple are at managing databases. FCPX doesn't use a database does it?

http://www.theverge.com/2013/3/26/4148628/why-doesnt-icloud-just-work

choice quotes:

iCloud apparently chokes hard on the databases it’s supposed to be so proficient at handling.

Many problems stem from the fact that Apple doesn’t account for edge cases where users do unexpected things,

The problem is partly due to the fact that Apple only had four people leading the company’s work on Core Data as of last year, a source close to Apple has told me. The company has simply not expressed any desire to fix Core Data syncing.

"Our only recourse is to throw bugs into the black hole that is Apple’s Radar bug-filing system and hope that they fix stuff," another top iOS developer told me


mmm. say, that FCPX database people.... it doesn't use core data at all does it?

Oh it does? it's all based around core data?

Ah. right. mmmm.

http://vimeo.com/user1590967/videos http://www.ogallchoir.net promo producer/editor.grading/motion graphics


Return to posts index

David Lawrence
Re: Apple and databases
on Mar 27, 2013 at 4:06:08 pm

Yikes! Sounds like a complete clusterf&$k.

_______________________
David Lawrence
art~media~design~research
propaganda.com
publicmattersgroup.com
facebook.com/dlawrence
twitter.com/dhl


Return to posts index

Chris Kenny
Re: Apple and databases
on Mar 27, 2013 at 4:25:35 pm

[Aindreas Gallagher] "Interesting article on how hilariously incompetent Apple are at managing databases."

The article isn't about databases per se, it's about iCloud's ability to transparently sync object graphs between multiple instances of an application across devices. Providing this as a generalizable system-level service is new, ambitious, and it's honestly not surprising that it's taking some time to work out the bugs.

[Aindreas Gallagher] "The company has simply not expressed any desire to fix Core Data syncing."

The article that contains this quote also notes that many bugs have been fixed. I guess they magically fix themselves with no work on Apple's part?

Core Data syncing is quite important to the multi-device computing model that seems to represent Apple's vision of the future. The idea that Apple intends to neglect is is not especially credible.

[Aindreas Gallagher] "Oh it does? it's all based around core data? "

You're mixing up Core Data, which has been around for eight years now and is quite solid, with Core Data syncing, which is only about 18 months old and still has some bugs to be worked out. FCP X does not presently use any of Core Data's syncing features, although it's possible they could, in the future, form the basis for collaborative editing capabilities.

Nice try with the FUD, though.

--
Digital Workflow/Colorist, Nice Dissolve.

You should follow me on Twitter here. Or read our blog.


Return to posts index


Aindreas Gallagher
Re: Apple and databases
on Mar 27, 2013 at 6:05:32 pm

I get you on the specifics of the core syncing thing - but you're skirting around the fact that there are four guys in total working on advancing and bug fixing core data full stop nevermind the syncing issue. That was one of the specific points the developers made - that they view it as a resource/does apple really care/ issue, given that there are just four guys representing the entire company effort on core data.

I can't remember where we heard that kind of thing before.

http://vimeo.com/user1590967/videos http://www.ogallchoir.net promo producer/editor.grading/motion graphics


Return to posts index

Charlie Austin
Re: Apple and databases
on Mar 27, 2013 at 7:01:23 pm

[Aindreas Gallagher] "but you're skirting around the fact that there are four guys in total working on advancing and bug fixing core data full stop nevermind the syncing issue. That was one of the specific points the developers made - that they view it as a resource/does apple really care/ issue, given that there are just four guys representing the entire company effort on core data. "

It seems to me, in the context of that statement that those 4 guys referenced anecdotally by an unnamed developer are working on Core Data synching, not Core Data itself. In any case Core Data (not synching) has been around for quite some time and is a solid, well regarded framework. You're grasping at straws.

-------------------------------------------------------------


~"It is a poor craftsman who blames his tools."~
~"The function you just attempted is not yet implemented"~


Return to posts index

David Lawrence
Re: Apple and databases
on Mar 27, 2013 at 7:37:54 pm

[Charlie Austin] "In any case Core Data (not synching) has been around for quite some time and is a solid, well regarded framework. You're grasping at straws."

Actually there's a bigger point in the article that I think some of you are missing
"Many veteran developers have learned their lesson and given up on iCloud’s Core Data syncing entirely. "Ultimately, when we looked at iCloud + Core Data for [our app], it was a total no-go as nothing would have worked," said one best-selling iPhone and Mac developer. "Some issues with iCloud Core Data are theoretically unsolvable (stemming from the fact that you’ve put an object model on top of a distributed data store) and others are just plain bugs in the implementation,"
So can these issues ever be solved? "[Apple’s] approach to the problem was very novel and interesting, and perhaps they will ship a version of it that works – but it functions very differently than typical sync solutions in that there is not a central server hub that maintains the ‘truth in the cloud,’" Pierce told me. "Because of this there is a lot of fragility to the implementation, and I’m not sure it will ever scale well to larger data sets,"

What these developers are saying is that Core Data syncing is potentially unscalable and ultimately unfixable because the model itself is broken.

I would argue that this may also be true for the FCPX magnetic timeline in its current form.

_______________________
David Lawrence
art~media~design~research
propaganda.com
publicmattersgroup.com
facebook.com/dlawrence
twitter.com/dhl


Return to posts index


Charlie Austin
Re: Apple and databases
on Mar 27, 2013 at 7:42:29 pm

[David Lawrence] "What these developers are saying is that Core Data syncing is potentially unscalable and ultimately unfixable because the model itself is broken.

I would argue that this may also be true for the FCPX magnetic timeline in its current form."


I'm interested in how you make the connection from "Core Data syncing is potentially unscalable and ultimately unfixable" ... to the magnetic timeline?

-------------------------------------------------------------


~"It is a poor craftsman who blames his tools."~
~"The function you just attempted is not yet implemented"~


Return to posts index

Aindreas Gallagher
Re: Apple and databases
on Mar 27, 2013 at 7:47:52 pm

try this I linked it below:

There is no inherent superiority in FCPX due to its use of Core Data, and indeed I think it contributes to a lot of the sluggishness observed when a project gets complex and necessarily performs a lot more I/O on its SQLite database for each new timeline input.

thats from a dude who really knows. Thats how the database problem pertains to the FCPX timeline model in operation.

Clearer for you now? :)

http://vimeo.com/user1590967/videos http://www.ogallchoir.net promo producer/editor.grading/motion graphics


Return to posts index

Charlie Austin
Re: Apple and databases
on Mar 27, 2013 at 8:10:33 pm

[Aindreas Gallagher] "thats from a dude who really knows. Thats how the database problem pertains to the FCPX timeline model in operation.

Clearer for you now? :)"


Um... no. That's the opinion of a guy who knows, which may well be correct. I don't know, and neither do you.

-------------------------------------------------------------


~"It is a poor craftsman who blames his tools."~
~"The function you just attempted is not yet implemented"~


Return to posts index


David Lawrence
Re: Apple and databases
on Mar 27, 2013 at 8:33:33 pm

[Charlie Austin] "I'm interested in how you make the connection from "Core Data syncing is potentially unscalable and ultimately unfixable" ... to the magnetic timeline?"

The connection is here:
"Some issues with iCloud Core Data are theoretically unsolvable (stemming from the fact that you’ve put an object model on top of a distributed data store) and others are just plain bugs in the implementation..."
"Because of this there is a lot of fragility to the implementation, and I’m not sure it will ever scale well to larger data sets..."
It's not about Core Data syncing, the problem may be a fundamentally broken design. If true, this is a very big deal. They'll have to start from scratch.

I believe the same is true with the magnetic timeline's hierarchical, single primary, parent/child, relative time model. I don't think it will scale. Without absolute, external time, I don't see how it can ever really support parallel, independent media streams such as multi-channel audio in a DAW. I don't think roles will ever be enough. I believe the absolute, external time frame of reference is essential.

Like "an object model on top of a distributed data store", the magnetic timeline may turn out to be wrong model for a truly universal time-based editing system.

_______________________
David Lawrence
art~media~design~research
propaganda.com
publicmattersgroup.com
facebook.com/dlawrence
twitter.com/dhl


Return to posts index

Steve Connor
Re: Apple and databases
on Mar 27, 2013 at 8:40:32 pm

[David Lawrence] "Like "an object model on top of a distributed data store", the magnetic timeline may turn out to be wrong model for a truly universal time-based editing system.
"


It's an interesting theory, but as it stands, in terms of real projects, what does the lack of "absolute" time stop you doing in FCPX?

Steve Connor

There's nothing we can't argue about on the FCPX COW Forum


Return to posts index

David Lawrence
Re: Apple and databases
on Mar 27, 2013 at 8:49:37 pm

[Steve Connor] "It's an interesting theory, but as it stands, in terms of real projects, what does the lack of "absolute" time stop you doing in FCPX?"

Using FCPX like a track-based DAW for editing. You can do it, but it's a PITA and it forces you to work against the timeline design. I would not want a DAW with a magnetic timeline. Would you?

_______________________
David Lawrence
art~media~design~research
propaganda.com
publicmattersgroup.com
facebook.com/dlawrence
twitter.com/dhl


Return to posts index


Steve Connor
Re: Apple and databases
on Mar 27, 2013 at 8:59:05 pm

[David Lawrence] "I would not want a DAW with a magnetic timeline. Would you?"

Not one that functions in the same way as FCPX does at the moment, I use Logic Pro a lot and I'm assuming the new one won't be following the same model.

I still think Apple may come up with something to tempt Editors who much prefer to work the way you do, I would imagine the previously discussed "fixed lanes" idea would be the most obvious. I don't think it would be impossible to implement this, despite your concerns about core data.

The fact you can almost set the "fixed lanes" now as well as the addition of the tilde key modifier which breaks the parent child link, albeit temporarily, means it is possible in theory.

Whether Apple wants to do it is a whole other question

Steve Connor

There's nothing we can't argue about on the FCPX COW Forum


Return to posts index

David Lawrence
Re: Apple and databases
on Mar 27, 2013 at 9:31:14 pm

[Steve Connor] "Not one that functions in the same way as FCPX does at the moment, I use Logic Pro a lot and I'm assuming the new one won't be following the same model."

Yes and from the hints so far, it looks like it's still a regular, tracked timeline model. I think there's a reason for this. I think this reason is why some editors like myself don't feel comfortable with the magnetic timeline.

[Steve Connor] "I still think Apple may come up with something to tempt Editors who much prefer to work the way you do, I would imagine the previously discussed "fixed lanes" idea would be the most obvious. I don't think it would be impossible to implement this, despite your concerns about core data."

Agreed. I think the "fixed lanes" idea is interesting. But I would go further and allow multiple primaries in a "fixed time" window. If they did that, they could keep everything else the same and it would work for me as well as for you.

[Steve Connor] "The fact you can almost set the "fixed lanes" now as well as the addition of the tilde key modifier which breaks the parent child link, albeit temporarily, means it is possible in theory."

Yes, but doesn't the tilde key modifier only affect the connected clips attached to the primary clip you're manipulating? Is there a way to make all connected clips stay put if you want? Something like Shift-tilde? That would make a big difference for me.

[Steve Connor] "Whether Apple wants to do it is a whole other question"

Indeed.

_______________________
David Lawrence
art~media~design~research
propaganda.com
publicmattersgroup.com
facebook.com/dlawrence
twitter.com/dhl


Return to posts index

Steve Connor
Re: Apple and databases
on Mar 27, 2013 at 9:34:17 pm

[David Lawrence] "But I would go further and allow multiple primaries in a "fixed time" window. If they did that, they could keep everything else the same and it would work for me as well as for you. "

Very good idea, will feature request that one

Steve Connor

There's nothing we can't argue about on the FCPX COW Forum


Return to posts index


Charlie Austin
Re: Apple and databases
on Mar 27, 2013 at 10:04:12 pm

[David Lawrence] "Agreed. I think the "fixed lanes" idea is interesting. But I would go further and allow multiple primaries in a "fixed time" window. If they did that, they could keep everything else the same and it would work for me as well as for you.

Fixed Lanes - or ability to assign a kind of Z order to clips - and grouped Roles would pretty much do everything tracks currently do... for me. If you need to add a crossfade and don't want to use the fade handles you can create secondary's, as you can now. They'd just stick with their group. In a perfect world, Roles become sort of "invisible" compound clips. In other words all clips for a role would stick together as a group, which could be routed to a mixer or FX send/return if needed. Click a "minimize" button and all the visible clips in the "container" would shrink to single, locked in time (compound) clip, but one which could be expanded *in the timeline* it resides in for further editing if needed.


[David Lawrence] "Is there a way to make all connected clips stay put if you want? Something like Shift-tilde? That would make a big difference for me."


That would be a good thing. They should have a role based option to enable/disable connection, like snapping, if you want. This would be handy, especially for music that's all chopped up and thus connected to random "parent" clips. Stuffing it into a CC is a workaround but you have to keep breaking it apart and remaking it to adjust cuts. The imaginary scenario above would also solve that since, once collapsed, the MX role would be locked to time.

Maybe they'll do it. I'm pretty sure FCPX isn't done yet. :-)

-------------------------------------------------------------


~"It is a poor craftsman who blames his tools."~
~"The function you just attempted is not yet implemented"~


Return to posts index

Marcus Moore
Re: Apple and databases
on Mar 27, 2013 at 10:21:51 pm

I have a giant blog post about Roles as a data structure for organizing the timeline that I'm desperately trying to finish before I head to NAB. I've been wanting to get these ideas out for almost 6 months.

But I'm too darn busy cutting in X!



Return to posts index

Franz Bieberkopf
Re: Apple and databases
on Mar 27, 2013 at 11:02:09 pm

[Marcus Moore] " I've been wanting to get these ideas out ..."

Marcus,

Please do.

Franz.


Return to posts index

Marcus Moore
Re: Apple and databases
on Mar 28, 2013 at 12:46:48 am

I shall try.

We have Easter up here in Canada this weekend, so I expect to have some time to pound out the second half of my ideas. The challenge is making the time to mock up some screen grabs. A lot of this is very difficult to explain just with words.

But to tease it a bit, to my mind, Apple has already solved their timeline problem in how they structure Events...



Return to posts index

Marcus Moore
Re: Apple and databases
on Mar 27, 2013 at 10:15:55 pm

[David Lawrence] "Yes, but doesn't the tilde key modifier only affect the connected clips attached to the primary clip you're manipulating? Is there a way to make all connected clips stay put if you want? Something like Shift-tilde? That would make a big difference for me."

How are you manipulating more than one primary clip at a time? Or are you just looking for a "tilde-lock" option?



Return to posts index

Charlie Austin
Re: Apple and databases
on Mar 27, 2013 at 10:21:46 pm

[Marcus Moore] "How are you manipulating more than one primary clip at a time? Or are you just looking for a "tilde-lock" option?"

Can't speak for David, but if you want to do a swap edit, like move a clip 7 shots earlier in a cut, the tilde key will disconnect things from the clip you're moving, but all the clips connected to the 7 shots that move in the swap move with them. Which, in my case, i didn't want them to. You can work around it, but it would be nice to temporarily "disconnect all" if you need to.

-------------------------------------------------------------


~"It is a poor craftsman who blames his tools."~
~"The function you just attempted is not yet implemented"~


Return to posts index

David Lawrence
Re: Apple and databases
on Mar 27, 2013 at 10:29:02 pm

[Charlie Austin] "Can't speak for David, but if you want to do a swap edit, like move a clip 7 shots earlier in a cut, the tilde key will disconnect things from the clip you're moving, but all the clips connected to the 7 shots that move in the swap move with them. Which, in my case, i didn't want them to. You can work around it, but it would be nice to temporarily "disconnect all" if you need to."

Bingo!

_______________________
David Lawrence
art~media~design~research
propaganda.com
publicmattersgroup.com
facebook.com/dlawrence
twitter.com/dhl


Return to posts index

Aindreas Gallagher
Re: Apple and databases
on Mar 27, 2013 at 10:49:05 pm

[Charlie Austin] "it would be nice to temporarily "disconnect all" if you need to."

god rather yes please. I throw pies at the software as a hobby really at this point - but there is a lot I like.
that said - the timeline behaviour will honestly never be one of those things. It really does form a brick wall of no thank you please.

http://vimeo.com/user1590967/videos http://www.ogallchoir.net promo producer/editor.grading/motion graphics


Return to posts index

David Lawrence
Re: Apple and databases
on Mar 27, 2013 at 10:28:11 pm

[Marcus Moore] "How are you manipulating more than one primary clip at a time? Or are you just looking for a "tilde-lock" option?"

What I'm looking for is a tilde option that would be global to all connected clips, not just the one's attached to the selected primary. Like Charlie mentioned, this is important with music for example, where you may have a mix spread out along the timeline that always needs to stay in sync relative to itself, not the primary. Mixing in a secondary, spiked to frame one seems like a hack to me.

_______________________
David Lawrence
art~media~design~research
propaganda.com
publicmattersgroup.com
facebook.com/dlawrence
twitter.com/dhl


Return to posts index

Marcus Moore
Re: Apple and databases
on Mar 28, 2013 at 12:38:44 am

If that's what you're looking for, then I'd almost recommend you compound clip the audio, then within the compound clip all the audio elements would be attached to a primary slug.

I believe a lot of this would be covered with a "connect to time" override. This would also be useful for broadcast standard commercial slugs.



Return to posts index

Aindreas Gallagher
Re: Apple and databases
on Mar 28, 2013 at 1:08:38 am

[Marcus Moore] "I'd almost recommend you compound clip the audio, then within the compound clip all the audio elements would be attached to a primary slug."

please.

mate - nesting is not every answer to fundamental timeline flaws. not every day of the week at any rate. seriously- the amount of times CC comes up in X.
Its near placebo.

http://vimeo.com/user1590967/videos http://www.ogallchoir.net promo producer/editor.grading/motion graphics


Return to posts index

Marcus Moore
Re: Apple and databases
on Mar 28, 2013 at 3:30:37 am

Yup. Just another one of the MANY times I've suggested nesting so solve all timeline ills...*

*If you're looking for a non-snarky answer, see my answer below to someone who's interested in carrying on a decent conversation.



Return to posts index

David Lawrence
Re: Apple and databases
on Mar 28, 2013 at 1:34:39 am

[Marcus Moore] "If that's what you're looking for, then I'd almost recommend you compound clip the audio, then within the compound clip all the audio elements would be attached to a primary slug."

Sure, but as you know, this prevents you from being able to do any mixing in context of the overall program. Stepping in and out of CCs is a bag of hurt.

[Marcus Moore] "I believe a lot of this would be covered with a "connect to time" override. This would also be useful for broadcast standard commercial slugs."

Yes, something like that would be great. I just don't see a clean way to do it without making the window the time frame-of-reference.

[Marcus Moore] "But to tease it a bit, to my mind, Apple has already solved their timeline problem in how they structure Events..."

I very much look forward to your post.

_______________________
David Lawrence
art~media~design~research
propaganda.com
publicmattersgroup.com
facebook.com/dlawrence
twitter.com/dhl


Return to posts index

Marcus Moore
Re: Apple and databases
on Mar 28, 2013 at 3:58:24 am

[David Lawrence] "Sure, but as you know, this prevents you from being able to do any mixing in context of the overall program. Stepping in and out of CCs is a bag of hurt."

Exactly. And I've often wondered about the an option to play an compound clip "in context" of it's parent timeline. Not as an overall solution to the mixing problem, but just generally as a feature.

[David Lawrence] "Yes, something like that would be great. I just don't see a clean way to do it without making the window the time frame-of-reference."

I agree. It would seem to be "mixing metaphors". The thing is, for most purposes, I find connected clips to be unbelievably more intuitive than what I was doing in 7 (but not for visual organization at present). Of course, there are always exceptions- but I rate those on the scale of the whole. Overall I think the magnetic timeline structure is allowing me to work faster across a whole edit, weighing the pluses to minuses. Audio organization and Mixing is undeniably FCPX's weak point.



Return to posts index

Charlie Austin
Re: Apple and databases
on Mar 27, 2013 at 8:50:59 pm

[David Lawrence] "The connection is here:
"Some issues with iCloud Core Data are theoretically unsolvable (stemming from the fact that you’ve put an object model on top of a distributed data store) and others are just plain bugs in the implementation..."
"Because of this there is a lot of fragility to the implementation, and I’m not sure it will ever scale well to larger data sets...""


No, it's not. The statement is about iCloud Core Data. "Some issues with iCloud Core Data are theoretically unsolvable...". You are making the connection.

[David Lawrence] "I believe the same is true with the magnetic timeline's hierarchical, single primary, parent/child, relative time model. I don't think it will scale. Without absolute, external time, I don't see how it can ever really support parallel, independent media streams such as multi-channel audio in a DAW. I don't think roles will ever be enough. I believe the absolute, external time frame of reference is essential."

That;s a lot of believing. :-) I beleive that X supports "parallel, independent media streams such as multi-channel audio in a DAW". You just can't run them through a "mixer" yet. Admit it, you just don't like the timeline. ;-)

-------------------------------------------------------------


~"It is a poor craftsman who blames his tools."~
~"The function you just attempted is not yet implemented"~


Return to posts index

David Lawrence
Re: Apple and databases
on Mar 27, 2013 at 9:33:17 pm

[Charlie Austin] "I beleive that X supports "parallel, independent media streams such as multi-channel audio in a DAW". You just can't run them through a "mixer" yet."

Not without a bunch of hackery ;)

[Charlie Austin] "Admit it, you just don't like the timeline. ;-)"

OK, I admit it! ;)

_______________________
David Lawrence
art~media~design~research
propaganda.com
publicmattersgroup.com
facebook.com/dlawrence
twitter.com/dhl


Return to posts index

Aindreas Gallagher
Re: Apple and databases
on Mar 27, 2013 at 7:43:04 pm

No - sadly wrong again there Charles :) -

it clearly states: The problem is partly due to the fact that Apple only had four people leading the company’s work on Core Data as of last year, -

a lot of people were surprised at that article. Even John Gruber was caught by it. And he's the biggest Apple booster in the business. That is a savage article about Apple's ability to execute and their reliability, "Devastating" as Gruber put it - no matter what way you slice it that article doesn't read well, and it features issues with manpower and competence.

Read Bieberkopf's linked piece from Andrew. Apart from anything else, its been floating around for a bit that the fundamental database structure for X may be inherently flawed and will always degrade once the project increases in complexity.

the people grasping at straws, that they then stick in the bottle of kool aid, and drink deep, are you poor guys. :)

http://vimeo.com/user1590967/videos http://www.ogallchoir.net promo producer/editor.grading/motion graphics


Return to posts index

Charlie Austin
Re: Apple and databases
on Mar 27, 2013 at 7:59:40 pm

[Aindreas Gallagher] "Even John Gruber was caught by it. And he's the biggest Apple booster in the business."

Yes, he was. "Devastating piece by Ellis Hamburger for The Verge on the calamitous state of iCloud Core Data syncing"

Didn't see anything about Core Data itself, but I guess I don't have the right glasses on.

-------------------------------------------------------------


~"It is a poor craftsman who blames his tools."~
~"The function you just attempted is not yet implemented"~


Return to posts index

Aindreas Gallagher
Re: Apple and databases
on Mar 27, 2013 at 9:26:16 pm

you don't have any truck with david's reading of the point that the fundamental issue might lie with the core data model itself?
not to mention that there's four guys and a cat at apple handling the whole thing?

http://vimeo.com/user1590967/videos http://www.ogallchoir.net promo producer/editor.grading/motion graphics


Return to posts index

Charlie Austin
Re: Apple and databases
on Mar 27, 2013 at 9:35:18 pm

[Aindreas Gallagher] "you don't have any truck with david's reading of the point that the fundamental issue might lie with the core data model itself?

Nope, because it doesn't say there's a problem with the Core Data model, it says there's a possible problem with the iCloud Core Data synching model.

[Aindreas Gallagher] not to mention that there's four guys and a cat at apple handling the whole thing?"

A cat?!? Crap! I guess I need to go re-read the article again...

-------------------------------------------------------------


~"It is a poor craftsman who blames his tools."~
~"The function you just attempted is not yet implemented"~


Return to posts index

Franz Bieberkopf
Re: Apple and databases
on Mar 27, 2013 at 9:58:26 pm

Aindreas, Chris, Charlie, David, et al.,

I'm going to go out on a limb and hazard a guess that we're all out of our depth here (except maybe Chris since he's used the secret password phrase "object graph") - basing this guess on the fact that no one has yet pointed out that Core Data is not a database:

http://maniacdev.com/2010/02/using-iphone-core-data-vs-a-database/

Which is maybe why Andrew Richards talks about Core Data in conjunction with SQLite in my previously linked post:

http://forums.creativecow.net/readpost/335/43894

Which is why I'd love to hear his take on this. But it isn't really apparent that it's relevant to the way FCPX works.

At least I've learned today that an object and data are not the same thing (though I don't really know what the difference is - I suppose objects need to be managed after being graphed, while data needs to be based).

Franz.


Return to posts index

Steve Connor
Re: Apple and databases
on Mar 27, 2013 at 10:03:35 pm

[Franz Bieberkopf] "Aindreas, Chris, Charlie, David, et al.,

I'm going to go out on a limb and hazard a guess that we're all out of our depth here (except maybe Chris since he's used the secret password phrase "object graph") - basing this guess on the fact that no one has yet pointed out that Core Data is not a database:

http://maniacdev.com/2010/02/using-iphone-core-data-vs-a-database/

Which is maybe why Andrew Richards talks about Core Data in conjunction with SQLite in my previously linked post:

http://forums.creativecow.net/readpost/335/43894

Which is why I'd love to hear his take on this. But it isn't really apparent that it's relevant to the way FCPX works.
"


Thank Franz, I can hear the sound of clutched straws snapping

Steve Connor

There's nothing we can't argue about on the FCPX COW Forum


Return to posts index

Aindreas Gallagher
Re: Apple and databases
on Mar 27, 2013 at 10:13:35 pm

Perhaps if you were just to re-read this section steve.

There is no inherent superiority in FCPX due to is use of Core Data, and indeed I think it contributes to a lot of the sluggishness observed when a project gets complex and necessarily performs a lot more I/O on its SQLite database for each new timeline input.


As pointed out - its the core data model pointing to the SQL database that may be a source of problems - the larger question relative to the verge article is whether Apple have the manpower assigned or basic competence to resolve the issues arising.

http://vimeo.com/user1590967/videos http://www.ogallchoir.net promo producer/editor.grading/motion graphics


Return to posts index

Aindreas Gallagher
Re: Apple and databases
on Mar 27, 2013 at 10:03:48 pm

oh indeed -I'm totally making all this up myself. As you say - the only voice I would buy would be Andrew Richards.

In the quote he does cast doubt on the repeated calls to the SQL database from the timeline tho. there - don't I almost sound plausible there like.

http://vimeo.com/user1590967/videos http://www.ogallchoir.net promo producer/editor.grading/motion graphics


Return to posts index

Charlie Austin
Re: Apple and databases
on Mar 27, 2013 at 10:08:28 pm

[Franz Bieberkopf] "I'm going to go out on a limb and hazard a guess that we're all out of our depth here"

Heh... I think it's a very sturdy limb you're on. :-) I actually just opened an X project database in an SQLite database browser. I don't know what any of it meant, but it looked cool...

-------------------------------------------------------------


~"It is a poor craftsman who blames his tools."~
~"The function you just attempted is not yet implemented"~


Return to posts index

Franz Bieberkopf
Re: Apple and databases
on Mar 27, 2013 at 10:12:01 pm

Charlie,

The link within the link is sort of interesting (comparing Core Data and databases)

http://www.cocoawithlove.com/2010/02/differences-between-core-data-and.html

I also found some discussion on strengths and weaknesses of Core Data (generally positive, though it seemed to have some controversy attached to it circa 2010):
http://bluerope.org/?p=82
http://inessential.com/2010/02/26/on_switching_away_from_core_data

Franz.


Return to posts index

Aindreas Gallagher
Re: Apple and databases
on Mar 27, 2013 at 10:19:27 pm

two cents in - I bookmarked this one a few months ago too - interesting that the early impression was that they had based it on the FCP server engine. there are implications for project sharing in core data apparently. he also... does say some positive things about core data.

http://www.fcproxuniversity.com/FCPro_X_University/Blog_Archive/Entries/201...

http://vimeo.com/user1590967/videos http://www.ogallchoir.net promo producer/editor.grading/motion graphics


Return to posts index

Franz Bieberkopf
Re: Apple and databases
on Mar 27, 2013 at 11:17:17 pm

[Aindreas Gallagher] "... there are implications for project sharing in core data apparently"

Aindreas,

Thanks for the link, some interesting stuff there.

(Aside - the blog notes that Quicktime Pro 7 is still available from Apple, in 2013.)

Yes, from what I was reading it seems like Core Data is a good solution for dealing with a database that exists in one place and doesn't get up beyond some large number (which does seem to raise questions about using it for iCloud syncing, and multiple-users, and large projects ...).

Nonetheless, all interesting, not least since I haven't come across any info on how Avid or PPro projects are structured ...

Franz.


Return to posts index

Aindreas Gallagher
Re: Apple and databases
on Mar 27, 2013 at 11:33:58 pm

[Franz Bieberkopf] "PPro projects are structured"

badly is the answer. It's basically AE in effect - which is close to unusable in a modern editing system.
My completely ill informed perception is that its literally prehistoric project management architecture. One project at a time, relinking is a nightmare...

that said, a birdie told me that is about to very, very radically change.

Also If DRW is on the money, we're going to see an end to .5 releases - and a fundamental re-ordering of project sharing management in PPro.

If he is right, we go direct to PPro 7, and adobe lay down fairly serious cards.

that said - from the hints on here from reps - the audio channel oddities with assignment and audio channel based keyframing are here for a bit.

Also part of me isn't sure whether the benefits of anywhere are bedded into day to day single user basic software operation.
I would be really surprised if the current project limitations survive the advent of adobe anywhere tho.

I don't think they could leave that AE style edit project handling crud in the next version - I'm presuming anywhere does bleed down to all basic footage and project handling architecturally in 7, outside of the specifics of the anywhere proposition.

Project handling in PPro cannot stay as it is.

http://vimeo.com/user1590967/videos http://www.ogallchoir.net promo producer/editor.grading/motion graphics


Return to posts index

Franz Bieberkopf
Re: Apple and databases
on Mar 27, 2013 at 11:42:29 pm

Aindreas,

Your thoughts and misgivings, and the listed shortcomings of PPro mirror my own.

You seem to be closer with the birdies though.

I was referring more to database architectures and other things just over the horizon of my knowledge.

Franz.


Return to posts index

Aindreas Gallagher
Re: Apple and databases
on Mar 28, 2013 at 12:16:52 am

[Franz Bieberkopf] "You seem to be closer with the birdies though."

It's incredibly hard to get my one small birdie to speak, I barely have a birdie frankly - but the birdie is in a 50-100 seat place that just plumped for PPro.
the birdie says anywhere was a fundamental sway.

the birdie had been very strong on FCPX, investigating it thoroughly for nearly a year for corporate reasons - but then, after examination, that pressure went away.

My misgivings are whether or not anywhere represents a full intellectual re-fit of PPro - or a price proposition for editing SMEs.

again - I basically can't believe adobe could realistically leave the current basic architecture in place outside of anywhere.

Given the state of PPro project management, as it creeps towards primetime - Adobe have to swop that out before anyone notices how messed up it is.

http://vimeo.com/user1590967/videos http://www.ogallchoir.net promo producer/editor.grading/motion graphics


Return to posts index

David Lawrence
Re: Apple and databases
on Mar 27, 2013 at 11:59:21 pm

[Aindreas Gallagher] "Project handling in PPro cannot stay as it is."

Amen to that. I just got nailed by the Pr project handling stupids. I'll survive this project but it's lame and needs to change asap.

_______________________
David Lawrence
art~media~design~research
propaganda.com
publicmattersgroup.com
facebook.com/dlawrence
twitter.com/dhl


Return to posts index

Franz Bieberkopf
Re: Apple and databases
on Mar 28, 2013 at 12:02:24 am

[David Lawrence] " I just got nailed by the Pr project handling stupids."

David,

I've been very hesitant to dive in on anything large (quite aside from the relinking issue). Any details worth sharing?

Franz.


Return to posts index

David Lawrence
Re: Apple and databases
on Mar 28, 2013 at 9:51:14 pm

[Franz Bieberkopf] "I've been very hesitant to dive in on anything large (quite aside from the relinking issue). Any details worth sharing?"

Franz,

It's nothing you don't already know, I was just surprised at easily it reared its ugly head.

My Premiere project management workflow is straightforward. Instead of opening separate tabbed projects like in FCP7, I create a master project and import any source projects I may need. I set up the imported project bins as tabs in a Pr window. It works well enough and in some ways is even an improvement over FCP7's tabs (the Pr tilde key is pure win). So far so good.

So here 's what happened - I was copying clips from a source sequence into my new project sequence. Out of habit, I often cut clips from the source so I only have unused clips remaining. As I'm working I decided I want a clip that I deleted a while ago. In FCP7, I could go back and revert the source project. In PrP, I have to reimport the entire thing. And now I can't delete the reimported project once I get my clip back because the clip is linked to the newly reimported project.

There are many ways I could have avoided this scenario, but when working fast, it's really easy for things like this to happen. Like everyone says, PrP project management has a major case of the stupids. Hope it gets better in a couple weeks.

--dhl

_______________________
David Lawrence
art~media~design~research
propaganda.com
publicmattersgroup.com
facebook.com/dlawrence
twitter.com/dhl


Return to posts index

Dennis Radeke
Re: Apple and databases
on Mar 28, 2013 at 1:36:47 am

Next version ruminations are half the fun of NAB, aren't they?


Return to posts index

Walter Soyka
Re: Apple and databases
on Mar 28, 2013 at 12:27:22 am

[Franz Bieberkopf] "Nonetheless, all interesting, not least since I haven't come across any info on how Avid or PPro projects are structured ..."

I popped the hood on a few Premiere Pro project files to do a little bit of in-house development to work around what David L. so aptly calls "the project handling stupids."

Premiere Pro projects are essentially a description of the current state of every object in the project (assets, sequences, tracks, clips, etc.), all wrapped in XML.

It's pretty human-readable, and if you're the adventurous sort, manually and programmatically modifiable. (Danger, Will Robinson!)

Walter Soyka
Principal & Designer at Keen Live
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
RenderBreak Blog - What I'm thinking when my workstation's thinking
Creative Cow Forum Host: Live & Stage Events


Return to posts index

Aindreas Gallagher
Re: Apple and databases
on Mar 28, 2013 at 12:53:47 am

walter - given it kind of visibly resembles a fair bit of AE project structure, so from what you've dug into - is it stupid to think PPro is sharing some of the same project management architecture as AE?

import project into project - pretty bad re-linking, bar on multiple projects and that?

http://vimeo.com/user1590967/videos http://www.ogallchoir.net promo producer/editor.grading/motion graphics


Return to posts index

Walter Soyka
Re: Apple and databases
on Mar 28, 2013 at 2:13:14 am

[Aindreas Gallagher] "walter - given it kind of visibly resembles a fair bit of AE project structure, so from what you've dug into - is it stupid to think PPro is sharing some of the same project management architecture as AE? import project into project - pretty bad re-linking, bar on multiple projects and that?"

My guess? I don't think they share a project management implementation.

AEP files are binary; even Ae's XML exports are nearly entirely binary blobs wrapped in XML tags. (Sidebar: this makes them totally useless for automation -- good thing Ae has scripting, which is a much safer way of automating project changes than actually futzing with the project file itself as I have been doing!)

In my experience, AEPs are quite a bit less fragile from a re-linking standpoint than .prprojs, and goodness knows that an Ae-style Consolidate footage command [link] would go a long, long way in Premiere Pro.

I think the visible operational similarities are intentional design choices meant to unify the feel of the applications, and are not indicative of an actual common architecture.

That said, I do think that from a conceptual standpoint, they probably approach project management similarly, so if that's what you mean by "project management architecture" then I am inclined to agree.

I'll go a step further and suggest that FCP7 is in the same family. I think that Premiere Pro is two big tweaks away from an FCP7 level of project management: better understanding of "master clips" (thereby avoiding duplication) and improved re-linking. Of course, whether an FCP7-level of project management is a good goal or not is an open question.

But all this is only speculation. As always, I could be absolutely wrong.

Walter Soyka
Principal & Designer at Keen Live
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
RenderBreak Blog - What I'm thinking when my workstation's thinking
Creative Cow Forum Host: Live & Stage Events


Return to posts index

Steve Connor
Re: Apple and databases
on Mar 27, 2013 at 10:16:24 pm

[Charlie Austin] "I actually just opened an X project database in an SQLite database browser. I don't know what any of it meant, but it looked cool...
"


Nice idea, I just tried it as well on a big event file, it was like something from "The Matrix"!

Steve Connor

There's nothing we can't argue about on the FCPX COW Forum


Return to posts index

Charlie Austin
Re: Apple and databases
on Mar 27, 2013 at 10:27:08 pm

[Steve Connor] "[Charlie Austin] "I actually just opened an X project database in an SQLite database browser. I don't know what any of it meant, but it looked cool...
"

Nice idea, I just tried it as well on a big event file, it was like something from "The Matrix"!"


That's what it looks like in a text editor, but in a database viewer it's all organized and pretty. And indecipherable to me. XML I can read, databases... not so much. My nerd-fu is limited :-) Here:

http://sourceforge.net/projects/sqlitebrowser/

don't open anything important! lol

-------------------------------------------------------------


~"It is a poor craftsman who blames his tools."~
~"The function you just attempted is not yet implemented"~


Return to posts index

Steve Connor
Re: Apple and databases
on Mar 27, 2013 at 10:30:00 pm

[Charlie Austin] "That's what it looks like in a text editor, but in a database viewer it's all organized and pretty. And indecipherable to me. XML I can read, databases... not so much. My nerd-fu is limited :-) Here:

http://sourceforge.net/projects/sqlitebrowser/

don't open anything important! lol"


That's what I used, it's when you drill down to some of the data it get's scary!

Steve Connor

There's nothing we can't argue about on the FCPX COW Forum


Return to posts index

Charlie Austin
Re: Apple and databases
on Mar 27, 2013 at 10:34:04 pm

[Steve Connor] "That's what I used, it's when you drill down to some of the data it get's scary!"

I'm going back in! Then i'm gonna change something! And open the project!!!! Or not. lol

-------------------------------------------------------------


~"It is a poor craftsman who blames his tools."~
~"The function you just attempted is not yet implemented"~


Return to posts index

Franz Bieberkopf
Re: Apple and databases
on Mar 27, 2013 at 5:15:01 pm

Aindreas,

Andrew Richards was talking about the database foundations of FCPX a while back. It would be interesting to hear if he has a perspective on this.

http://forums.creativecow.net/readpost/335/43894

Franz.


Return to posts index

Aindreas Gallagher
Re: Apple and databases
on Mar 27, 2013 at 6:10:30 pm

I remember that one -

There is no inherent superiority in FCPX due to its use of Core Data, and indeed I think it contributes to a lot of the sluggishness observed when a project gets complex and necessarily performs a lot more I/O on its SQLite database for each new timeline input.

some of the more fervent FCPX supporters do love to wax rhapsodic about the wonders of the database...

to my mind there is serious amounts of blind magical wishful thinking involved. as ever.

http://vimeo.com/user1590967/videos http://www.ogallchoir.net promo producer/editor.grading/motion graphics


Return to posts index

Bill Davis
Re: Apple and databases
on Mar 27, 2013 at 6:52:33 pm

[Aindreas Gallagher] "Interesting article on how hilariously incompetent Apple are at managing databases. FCPX doesn't use a database does it?"

SEE!

See that mole on Cindy Crawford's face? See it? STARE at it. It proves that she's hidiously ugly.

That's my opinion and I'm not going to hear anything else.

Ever.

Period.

That is all.

(except that I heard that Halle Berry once had a zit.)

digusting.

Know someone who teaches video editing in elementary school, high school or college? Tell them to check out http://www.StartEditingNow.com - video editing curriculum complete with licensed practice content.


Return to posts index

craig slattery
Re: Apple and databases
on Mar 27, 2013 at 7:49:27 pm

OMG, someone call broadcasting house, tonights programme about Danny Boyle BBC Two at 10pm, was cut on dodgy database, its got to be stopped!!!!


Return to posts index

Gary Huff
Re: Apple and databases
on Mar 27, 2013 at 8:04:36 pm

[craig slattery] "OMG, someone call broadcasting house, tonights programme about Danny Boyle BBC Two at 10pm, was cut on dodgy database, its got to be stopped!!!!"

And this is why it's hard to take fanboys seriously.


Return to posts index

Steve Connor
Re: Apple and databases
on Mar 27, 2013 at 8:06:49 pm

[Gary Huff] "And this is why it's hard to take fanboys seriously."

Because they have a sense of humour?

Steve Connor

There's nothing we can't argue about on the FCPX COW Forum


Return to posts index

Charlie Austin
Re: Apple and databases
on Mar 27, 2013 at 8:12:08 pm

[Steve Connor] "[Gary Huff] "And this is why it's hard to take fanboys seriously."

Because they have a sense of humour?
"


Clearly yes. Pointing out tenuous possible flaws in software you don't like is serious business!

-------------------------------------------------------------


~"It is a poor craftsman who blames his tools."~
~"The function you just attempted is not yet implemented"~


Return to posts index

<< PREVIOUS   •   VIEW ALL   •   PRINT   •   NEXT >>
© 2017 CreativeCOW.net All Rights Reserved
[TOP]