In the above note, you mean Lightroom, not Aperture.
I meant Aperture; "I believe its possible that Adobe released LR as a public beta in response to the release of Aperture," is what I was saying there.
However: I may be wrong about the lack of enthusiasm, it's just something I've sensed in reading about every post I can find about Lightroom. And -- again -- I may be wrong. So I took your advice and re-read the Shadowland story (which I'd originally read in January), and if YOU read it, you'll find that Schewe more than hints at a lack of company drive behind the project. At one point, he even credits the announcement of Aperture as having saved the Lightroom project, as late as a year ago.
I'd say there's a difference between Adobe corporate, and the people working on LR. Corporate may not have had much interest until Apple announced Aperture, but I'm fairly certain the people working on LR are very enthused about what they are making. To me, that is what counts, just as long as corporate lets them do their thing.
Given the number of features requests Adobe has gotten, and the comments about it in posts I've seen, I'd assumed that we'd have six or seven beta versions before the final, with a final version coming out sometime this fall. And in January, I was left with the impression that Windows would be out shortly. It didn't show up until summer. I don't care about that personally, because I'm a Mac user, but it made me think that things were not moving at high speed.
Software takes time. Personally, I'm not surprised at the rate betas have been released. Nor am I surprised LR took as long as it did to come to Windows considering how such a thing practically requires a re-write to achieve the level of exactness between platform versions Adobe is keen on maintaining. LR for Win is more than a port, it's been built for the operating environment.
Now that The Win version is up and running things should start advancing more quickly.
-- is optimizing for speed essentially a trivial task that can be done in a couple of weeks?
Optimization is a combination of debugging and running the app through a debugger application (pardon me if the term is wrong, I'm certain it is, it's been a wile). Basically, the debugger software executes every command the software being built can do and looks for portions of the code that are using more CPU process or memory than it should. Or, they use the program and have tool running that analyzes the processes in real-time looking for issues (you can do this yourself on a Mac with the developer tools on the install dvd. Shark I believe the app is called). The developers take the information they gathered, dive into the millions of lines of code that make up the developing app and then try and problem solve to make it more efficient. This can either be just changing some parameters, or re-writing portions of the app. Then you have to test again to make sure what has changed didn't negatively effect something else. Lather-rinse-repeat.
(this is just one part of the process as I understand it, maybe someone who has actually done some development can elaborate more)
Depending on the size of the application, the complexity and what needs to be optimized, this process can take days, weeks or months. If new things are added, or things in the app are changed, the process starts over.
Are the features still outstanding the kind of thing that are essentially plug-in modules that may be ready right now, in the Adobe labs, and don't really need much field testing?
I'm sure it would hard for anyone who would know that to answer that due to NDAs. At best we can speculate. From what I've read/heard along the way I know a new zoom level(s) are coming, it seems clear some toys from RSE are making their way in (sharpening/noise reduction anyone?), they might bring in cloning tools by the 1.0 release, selective edits (something they are certainly considering) probably wont make it until 2.0, importing will have some more added to it, and finally, a more robust system for managing the collection across numerous computers and hard drives (this includes collaboration efforts) is coming.
Most of these are not simply modules and effect the core of the application. Some are simpler than others to implement, some aren't. One of the things that can take time more than actually adding a feature is figuring out how that feature should be added and work both mechanically and aesthetically. Once something is set by 1.0, it's practically set for good (such as how the RSE additions will work with everything else, or how selective edits should be handled in the interface) or at the least it'll be difficult to change latter (they talk about this in the podcasts).
LR is supposed to be a simple, and intuitive application. That may sound like a simple and common sense thing to strive for but it can be extremely difficult to achieve the delicate balance between simplicity and functionality; a lesson I've learned in building and designing my website. I can only imaging how much more difficult an application would be. It's far easier to add a feature and throw it into a menu (ala Photoshop) than to take that feature and integrate it seamlessly into an interface that stays out of the way of the user yet is available and natural to use. The best designs are the ones you don't notice; which tend to be the ones that take the most time and effort to develop.