Andrew Richards
Forum Replies Created
-
[Chris Harlan] “I’m sure you can understand the skepticism since none of these things were part of the initial release, and many of the step forward appear to people as workarounds or bandaids. I think Roles is an interesting idea that, despite its cleverness, screams “afterthought.” “
I can, and that goes with the territory of shipping a product with holes in it and filling them in as you go. I don’t think Roles are an afterthought though, they were there in 10.0, just with no apparent use. There is a rhyme and reason to a metadata-centic organization, but they shipped that feature half-baked and the casual observer can be forgiven for their skepticism.
Steve Jobs is famous for pushing for a shipping product and then iterating that product over time. They’ve followed that ethos since he came back to Apple in almost everything they’ve done. The saying is, if you wait for it to be done, it is never done.
[Chris Harlan] ” I just find it troubling to try to second guess Apple’s intentions. My gut guess is that Apple is having the same problem.”
I bet there are epic arguments in the halls and meeting rooms in Cupertino daily. May the best idea win!
Best,
Andy -
[Simon Ubsdell] “Surely the fact that Automatic Duck OMF export was something you could buy at the start of this week and now no longer can says quite a lot about how fragile the third party support model actually is – especially since AD is/was about the only third party solution yet available. “
That is the big risk, yes. The caveat I would point out though is that AD had exclusive early access to the hooks for delivering OMF. Essentially a monopoly position, reflected in their price. With FCPXML out there in the wild, anyone with the time and the talent can make an OMF export widget and sell it for whatever they think is fair. Nearly all third party integration depends on that XML capability, so the way things were for the first three months likely won’t be the way things will be moving forward.
Risk is certainly there, but then HP could decide to get out of the PC business, Microsoft could completely reskin Windows, and a tsunami could wipe out SR delivery. Unless we somehow go open-source, we’re always at the mercy of larger forces. Adobe is completely dependent on NVIDIA for CUDA. Apple is completely dependent on Intel for new Xeons to build a new Mac Pro around. Avid is still in the red (though they seem to be slowly clawing their way out). It’s a crazy world out there.
Best,
Andy -
[Walter Soyka] “Andrew, I think I am coming across as way more argumentative with you than I really mean to be, and I apologize for that. I agree with your general point that all this is coming, but suggesting that interchange is a problem that’s totally solved today isn’t accurate. Soon — but not now.”
No worries. Soon, very soon. My main point was that Apple held up their end of the bargain by rolling out XML interchange and by including some high-profile partners for early access so they could add the functionality to their apps very soon after FCPX could print out its XML.
[Walter Soyka] “I like Craig and Tim’s suggestion of “Industry Trends.” I’d love to see this level of interest and debate broaden significantly. I’ve learned a ton here.”
“Industry Trends” would be a lot less combative, wouldn’t it? We certainly live in disruptive times. It dovetails with the Business and Marketing forum too.
Best,
Andy -
[Chris Harlan] “Certainly, they managed to just deliver on the “this summer” promises, but failed on “in a few weeks…” promise. Also, there was a fairly strong implication that i/o was just around the corner. Now its next year.”
We only saw an update “this summer”, but in order for Blackmagic, Autodesk, and Square Box to ship launch day FCPXML support for Resolve, Smoke and CatDV, they almost certainly had access after “a few weeks”. But I’ll grant that the FAQ sure made it seem like public access to the new XML would come in a few weeks, not a few months.
[Chris Harlan] “I disagree with your argument that we can make behavioral assessments based on the last few months. Also, I think I have to disagree with your definition significant critical functionality. I just see a few steps forward, which are welcome ones, but small.”
Fair enough, it isn’t an adequate sample size to really know for sure. It is just so far, so good.
As to critical functionality, I’m thinking of that as the aspects of the software that make it able to operate in a modern post workflow. Aside from what they have already delivered (XML, Roles, SAN support) and what they have repeatedly stated they will add (Multi-cam, Broadcast I/O), what else is missing? The XML feature makes all manner of interchange possible, even if FCPX itself never exports OMF or EDL files. Others hate the idea of third parties being responsible for features essential to their particular workflows, but I’m OK with it. As I look through the lens of the last shop I worked at, the next rev will deliver everything needed to tick all the requirements boxes for their workflow.
Best,
Andy -
[Walter Soyka] “What “exclusively professional” media handling are they building into the OS?”
CoreMedia specifically lists codecs like ProRes and DVCPRO HD and container formats like XDCAM. This is deep OS-level stuff, brand new in Lion. Any developer can write an app for Lion with native 64-bit GPU-accelerated support for broadcast acquisition formats. There is no consumery web use case for XDCAM.
[Walter Soyka] “This doesn’t change the fact that Apple has some really wonderful technologies upon which professional applications may be built — but isn’t it also plausible that they’re just good, forward-thinking engineers without a deep professional media master plan?”
The big difference is at the corporate level, Apple has decided to have the Core OS team shoulder load that can only be useful to the Pro Apps team (at least internally). The old way was silted QuickTime teams, OS teams, and Pro Apps teams all working in parallel, instead of in concert. I’m not saying that Apple’s market target for FCP hasn’t changed, I’m just saying that there is clearly some level of commitment to pro video when you have support for handling XDCAM at the OS level.
Best,
Andy -
[Andreas Kiel] “Hmm, at least for this point I can’t say it’s solved. It’s there, but that doesn’t mean it’s working as expected by some users.
I think it’s good step, but no solution.
If you can’t do a XML round trip from FCP X to FCP X that doesn’t sound like a solution – at least for me.”Apple held up their end of the bargain, anyway. And why can’t you round trip? FCP X writes and reads FCPXML. I don’t have details for Smoke and Resolve round tripping, but CatDV 9.0.1 round trips. That many users don’t really understand what XML is and in turn projected a false hope for native XMEML/FCPXML interchange is just the natural outcome of that (reasonable) ignorance.
[Andreas Kiel] “Regarding the shop and updates I got a mixed feeling. If I’m in a region where internet connection is slow and expensive it’s bad – very bad. My temporary provider tried to charge me about 4000 USD for the FCPX download (took 10 days).”
This could be a VAR opportunity. But like they did with Lion, I think it would behoove Apple to offer alternative distribution methods for larger app payloads. Not everyone is on unlimited broadband and multiple GB downloads are impractical for a lot of people.
Best,
Andy -
[Walter Soyka] “What can I do with FCPXML today? What other apps support it? Can FCPX successfully roundtrip an edit to itself via FCPXML?”
You mean other than Smoke, Resolve, and CatDV? That is just the launch-day list. AJA and BMD are hot on the trail.
[Walter Soyka] “Where do I buy this? Not from https://www.automaticduck.com — that’s a news release about AD’s new relationship with Adobe.”
Yeah, curious, that. Adobe and Automatic Duck need to crap or get off the pot. Is this an acquisition? Are the products going away? Apple isn’t the only company with questionable corporate communications.
[Walter Soyka] “Isn’t this extremely limited?”
It is, but we are one week hence from the release of 10.0.1. I’d be very surprised if more folks don’t try to attack this problem. The demand is certainly there.
[Walter Soyka] “Of coures, by then, we may also have a better idea what to expect from CS6 and MC6. I don’t expect it to get dull around here anytime soon.”
We’ll need to lobby for renaming this forum “NLE Wars” or something broader than the FCPX debate.
Best,
Andy -
[Franz Bieberkopf] “You have just listed – almost in its entirety – all communications Apple have made about their plans. It may take another sentence or two to complete the picture.”
When Avid had their preview event for MC6 where they started off with a huge disclaimer stating nothing they were showing was a promise. How much more communication do you need? Does Randy Ubillos need to host a weekly call-in show? The FAQ said everything they needed to say about the immediate future of FCPX, and then they walked the walk.
[Franz Bieberkopf] “One assumes they’ve taken down the FAQ because that communication is no longer relevant to their plans.”
I assume they took it down because they did everything they said they were going to do in the FAQ, save for two more features they still list publicly as being forthcoming with a timeline. Discussing future product plans is a dangerous exercise. Go mention “Longhorn” in Redmond and see what kind of reaction you get…
[Franz Bieberkopf] “It is ridiculous that you have to qualify “candid” (“for them”) and you don’t see the irony …”
What irony? Everything is relative. They are habitually mum on new products and with FCPX they have been unusually forthcoming. I’m not sure what standard you are holding Apple to, but I can’t find the blogs on Adobe’s site telling us all about the ongoing development of CS6 or on Avid’s site telling us all about what they are working on for MC6.
[Franz Bieberkopf] “Apple is a company that doesn’t communicate – it uses marketing.”
Everything a company does in public is marketing.
Best,
Andy -
[Marvin Holdman] “Just because minor changes may come more frequently, it doesn’t mean overall functionality will do the same. Here’s a quote straight from Apple, via Townhill, “the first foundation stone in a building that’s going to be assembled over the next ten years.” Sounds like a slow train to me.”
Except I would argue that the one minor rev we have so far added significant critical functionality and it did so in a very short time (by software development standards). The long tail Townhill is referring to is software that can grow with Moore’s Law for a long time, not that they will drag their feet getting features added. They’ve already proven they can deliver big features on a short schedule. If the updates slow to a crawl after the next promised rev in “early 2012”, then I’ll eat crow. For now, they are doing just what they said they would.
[Marvin Holdman] “Look at Townhill’s revisionist statement regarding the matter, “The professional [editor] is critical to Apple, and it’s a customer we don’t want to lose.” Pretty much just blows off any criticism regarding the release. Sure it’s a statement, but it really doesn’t say anything other than “What problem?” I wouldn’t call spin “candid”. “
They have been claiming to love the pro all along, it just rang (rings?) hollow when you see the product. The candidness I was referring to was the FAQ and the bottom of the new feature page. Apple almost never talks about future product like that. They published their timeline and they hit their first deadline.
[Marvin Holdman] “To date, only a small handful of select vendors have received the API’s. This seems odd for an application that has continually stated the plans for 3rd party vendors.”
I’d rather see this stuff up on the Apple Dev site too. But practically speaking, we don’t really know who is and isn’t included on access to the camera SDK. If it turns out to be Sony, Panasonic, JVC, Canon, RED, and ARRI, will it matter the APIs weren’t opened up?
[Marvin Holdman] “…I’m not sure it’s the best way to create this type of software either. …The “whiffs” I’m getting are coming from the fact that the app store concept is an expanding one that goes beyond Apple. (Windows 8 anyone?)”
It’s just a distribution channel, and unlike iOS it isn’t the exclusive distribution channel. Third parties need not participate. If anything, the way Apple has limited the spec for what is allowed for sale on the App Store precludes much of the post world’s apps and plugins since they are not self-contained applications. Most of them wouldn’t qualify for sale on the App Store even if they wanted to be there under the present rules. Apple has bent itself over backwards to use the store for its own non-self-contained-app software (Lion, Xcode 4), but the rest of the place is very much slanted to the casual consumer. If they ever make the App Store the only way to get software on a Mac, then I get very upset.
[Marvin Holdman] “Most of the core functionality, like tape control, EDL, XML was handled in the program. This, combined with a single source for OS, hardware and app, made for a very competitive product.”
It also made for the bad old days when an innocent QuickTIme update for iTunes would bring FCP to its knees. Tape control in FCP was well known for its flakiness. XML is a moot point now, and really so is EDL (has anyone written an AppleScript to convert FCPXML to EDL yet?).
[Marvin Holdman] “What is bad about that? More pug-ins + more frequent upgrades + uncommunicative mothership = challenge to keep system current.”
The outsourcing of features strategy can be a double-edged sword, yes. However, we always needed drivers to be updated for AJA and BMD cards in the past (for instance), and they have almost always been johnny-on-the-spot with them following revs to FCP. Everything else takes place with XML, and just like XMEML before it, FCPXML will have versions with varying levels of features added over time. The addition of FCPXML 2.0 someday won’t break what works with FCPXML 1.0. The downside is users heavily reliant on third party stuff will need to be more cautious when upgrading, but again, how is that different from when Apple’s new QuickTime update would break Apple’s Pro Apps?
[Marvin Holdman] “I just wonder if this suggested architecture is “proper”. It’s a question that has been asked quite frequently in respect to timelines and metadata, I just wonder about it in terms of what Apple’s suggested re-definition of what a NLE is. In a day when all the other NLE seem to be moving towards native inclusion of functionality, Apple’s suggestion that it can all be outsourced is certainly radical. I just wonder about it’s practicality, based on history.”
I was thinking more of the underpinnings of the app, not so much the UI and its metaphor or the strategy to let third parties pick up any slack. But the underpinnings in the OS for handling video (AVFoundation, CoreMedia, and CoreVideo), for tracking project data (CoreData), and for rendering (shared engine with Motion and Compressor) mean less low-level tinkering by the Pro Apps team and more high-level feature work on a shorter timeline with fewer development resources. I wouldn’t be at all surprised to see them encourage third parties to jump in and solve problems and then add the features directly the app down the line. They’ve done it over and over before, so I doubt they will stop now.
As an aside, for anyone still doubting Apple’s corporate commitment to pro media, I ask why they are putting so much exclusively professional media handling capability into the core of OS X? They are going from having a motley crew of apps with different guts and development legacies to a suite of tools that all work from the same bones (I’m including Logic in that statement, anticipating Logic Pro X to be built on all the same bones). That’s the ten year plan: stop duplicating work and let the OS do the heavy lifting. This serves at least two purposes. It lets the Pro Apps teams focus on the visible stuff, and it acts as an incubator for the OS team to see firsthand how third party developers will interact with their APIs. OS X is on the verge of relevance in the IT world at large, and being ready with tried and tested development platforms is a good thing for the platform and ultimately for Mac sales.
I empathize with your concern over the whole outsourcing of features concept, but on the whole I think what they are doing is a net positive.
Best,
Andy -
[Aindreas Gallagher] “seriously?”
What were the critical workflow capabilities missing from 10.0?
- XML I/O = solved
- Multi-track audio output = solved (and very elegantly, IMHO)
- Shared storage support = solved (though not as wide open as it was with legacy FCP)
- OMF export = already supported by 3rd party (Automatic Duck)
- Legacy project support = already supported by third party (CatDV)
- Multi-cam = next rev
- Broadcast I/O = next rev
They also addressed some less than critical omissions like sequence starting timecode and connected clip transitions.
The theme here is that the debate must shift from what FCPX can’t do at all to what we think about how it does everything it needs to do in order to play ball. Plenty of valid points on both sides of that debate, as we’ve seen in some threads around here.
Best,
Andy