Elvis Is Alive In My OS!

13»

Comments

  • Reply 41 of 55
    spartspart Posts: 2,060member
    <off topic>



    Clicking a menu and bringing the entire operating system to a screeching halt: Five seconds.



    Waiting painfully while front app decides what the hell it's doing: Twenty seconds.



    Crappy app locking up system, displaying system error, featuring non-functioning restart button: Two minutes restart time, a day's worth of work.



    Going deaf after being blasted in the ear by the sound of a thousand IM's due to craptacular multitasking: Timeless.



    </off topic>



    [ 10-17-2002: Message edited by: Spart ]</p>
  • Reply 42 of 55
    Oh, I just know you ain't talkin' 'bout my MacOS X, I just know it!
  • Reply 43 of 55
    wfzellewfzelle Posts: 137member
    [quote]Originally posted by Kickaha:

    <strong>To try and make this point as crystal clear as possible...



    Some of the current research I'm doing involves real-time tracking of objects in live video.



    Two parts of this problem are an algorithmic plugin system for tracking disparate objects (I have no idea what types of objects might be tracked later), and a video conferencing infrastructure. These are two separate, orthogonal problems, and are handled in separate orthogonal systems within the project. Either one can be used with or without the other.



    Does this mean that every video conferencing system I work on from now on out *HAS* to have the object tracking?



    Of course not.</strong><hr></blockquote>



    No, but it makes it far more likely that future systems you work on will have both (perhaps 0.0000001% -&gt; 0.001%, but still). Or that you will propose a similar object tracking system to a new company that you might one day work for. Of course, the relationship between object tracking and conferencing is much, much smaller than the relationship between file system metadata and file system journaling. I'd expect someone who worked on the BeFS to know quite a bit about both (which creates a relationship between the two technologies).



    Of course, you twist my argument by setting up strawmen. I only talked about probabilities, not about absolutes ("*HAS* to have").



    [quote]<strong>Does the fact that the video conferencing structure can be used without the video tracking, and has been used as such by folks, mean that I've never implemented the video tracking?



    Of course not.</strong><hr></blockquote>



    Do you agree that one has a finite budget of time? If so, every extra task someone does reduces the chance he is working on one specific task not already included. Example:



    Mr A is qualified to do 6 tasks, numbered from 1-6. Every task costs 1/3 fte (full time employee), so we know that A is doing 3 things. I'd like A to do task 5. Since we don't have inside information, A may be doing each task with probability 1/2. Now suppose that we know that A is doing task 6, the situation changes to:



    5 tasks

    A can do another 2 tasks

    Probability(task1-5) = 2/5 &lt; 1/2



    Q.E.D.



    We don't know how much work Dominic Giampaolo has/had to do on journaling. He might be able to do just one extra thing or even none. Of course, he may use the time he has left for something else than metadata altogether (there are enough useful features still missing in HFS+).



    [quote]<strong>Proximity does not imply a conceptual, causal, or even casual relationship between programmatic entities.</strong><hr></blockquote>



    Proximity is a relationship whether you like it or not. It means that the same group of people are going to work on something (they won't ask Quicktime engineers to implement metadata in the FS). Those people are going to have a big effect on the decision to implement something and the way in which it is implemented.



    [quote]<strong>This is one of the prime benefits of orthogonal systems design, which in turn is a foundation of OO programming.</strong><hr></blockquote>



    True, but orthogonal systems design has little to do with the fact that engineers are human (usually). That means they will have preferences, experience with certain technology, idea's about the future, etc.



    [quote]<strong>Just because the guy who worked on BeFS implemented the journaling filter for VFS doesn't mean that either he IS or IS NOT working on a metadata system. It has absolutely, utterly nothing to do with it, other than as a use of his finite time. Given this one fact, regardless of all others, you cannot draw any meaningful conclusions regarding what else he may or may not be working on... and you certainly can't draw any rational conclusions as to what *other* people may be working on. It's a logical fallacy to assume otherwise.</strong><hr></blockquote>



    Given the fact that an BeFS engineer works in the FS group, it becomes more likely that Apple will built a DB in the FS (because of his preferences, experience & idea's). Seeing that Giampaolo already works on a project, it becomes less likely. Knowing that Apple is using metadata in some apps, it becomes more likely. Considering the dislike expressed by Avie/Apple for the old metadata in favor of file extensions it becomes far less likely.



    Those are four conclusions that make sense. What doesn't make sense is that you decide to desregard all the evidence that opposes the conclusion you prefer.



    [ 10-18-2002: Message edited by: wfzelle ]</p>
  • Reply 44 of 55
    wfzelle, you make me laugh. Thank you for that.
  • Reply 45 of 55
    artman @_@artman @_@ Posts: 2,546member
    [quote]Originally posted by AirSluf:

    <strong>



    Because it's the same vocal minority that whine and complain about everything--they are beyond reason and reach. Spend a couple months on the boards and you can pick the same ones out from a mile away.



    </strong><hr></blockquote>



    True and I ain't one of them. I was only looking at the broad picture here with this (and technology in general). I have 768 MB RAM in that iMac and if I didn't I would be whining a lot. But I have more important things to do. This feature is all well and good if:



    1. You need it.

    2. Have a Mac that'll handle it, no matter how much resources it takes.

    3. You are waiting for the next big thing from Apple...if you can afford it.



    Continue... <img src="graemlins/smokin.gif" border="0" alt="[Chilling]" />
  • Reply 46 of 55
    kickahakickaha Posts: 8,760member
    [quote]<strong>



    Does this mean that every video conferencing system I work on from now on out *HAS* to have the object tracking?



    Of course not.

    </strong>





    No, but it makes it far more likely that future systems you work on will have both (perhaps 0.0000001% -&gt; 0.001%, but still). <hr></blockquote>



    THAT is 'far more likely'?!?



    [quote]Or that you will propose a similar object tracking system to a new company that you might one day work for. Of course, the relationship between object tracking and conferencing is much, much smaller than the relationship between file system metadata and file system journaling.<hr></blockquote>



    Really? You've still failed to show *any* relationship between metadata and journaling other than a circumstantial proximity in BeFS' implementation.



    [quote]I'd expect someone who worked on the BeFS to know quite a bit about both (which creates a relationship between the two technologies).<hr></blockquote>



    It doesn't create squat. Sorry.



    [quote]Of course, you twist my argument by setting up strawmen. I only talked about probabilities, not about absolutes ("*HAS* to have").<hr></blockquote>



    Actually, this is the *first* time you have discussed probabilities. Some previous quotes:



    <strong>On the other hand, now we know that the BeFS guy that Apple hired is not working on metadata.



    Some hoped he was working on metadata, but he clearly isn't.



    People wonder why Apple wants the one [journaling], but not the other [metadata].

    </strong>



    These are absolute statements. Don't backpedal and then claim that it's been your position all along.



    [quote]Do you agree that one has a finite budget of time? If so, every extra task someone does reduces the chance he is working on one specific task not already included. Example:<hr></blockquote>



    You didn't read to the end of my post before replying, did you?



    Specifically: "It has absolutely, utterly nothing to do with it, other than as a use of his finite time."



    &lt;Snip of example now moot&gt;





    [quote]We don't know how much work Dominic Giampaolo has/had to do on journaling. He might be able to do just one extra thing or even none. Of course, he may use the time he has left for something else than metadata altogether (there are enough useful features still missing in HFS+).<hr></blockquote>



    So basically you're saying that we have no facts about Giampaolo's workload, his time, what he may have been working on, may be working on, or may work on in the future, but from that you're willing to state that Apple isn't interested in metadata?



    Hey, whatever floats your boat.



    [quote]Proximity is a relationship whether you like it or not.<hr></blockquote>



    Proximity is a circumstantial casual relationship only. It is not a conceptual, causal, or implicatory relationship. You therefore cannot derive *ANY* conclusions between the two items. Logic 101, propositional logic.



    [quote]It means that the same group of people are going to work on something (they won't ask Quicktime engineers to implement metadata in the FS). Those people are going to have a big effect on the decision to implement something and the way in which it is implemented.<hr></blockquote>



    Still failing to see how this anything to do with *IF* Apple is working on metadata. This observation, while true, has more to do with the implementation details of *HOW*.



    [quote]True, but orthogonal systems design has little to do with the fact that engineers are human (usually). That means they will have preferences, experience with certain technology, idea's about the future, etc.<hr></blockquote>



    Agreed. Again, this has nothing to do with attempting to determine whether or not Apple is interested in metadata, may be working on metadata, or even may have Giampaolo working on it... unless you can state the chain of logic you're using.



    [quote]Given the fact that an BeFS engineer works in the FS group, it becomes more likely that Apple will built a DB in the FS (because of his preferences, experience & idea's).<hr></blockquote>



    True.



    [quote]Seeing that Giampaolo already works on a project, it becomes less likely.<hr></blockquote>



    Because Giampaolo has been working on something else, *Apple* is less likely to work on metadata? You are correct that: he's the most likely person to have on board for design decisions regarding metadata, and that his finite time has also been used for journaling (we presume from the eWeek article), so he might have had less time to devote to metadata, but jumping from his finite resources to the resources of Apple as a larger company (or even just the FS group) is a bit specious, I'd say.



    [quote]Knowing that Apple is using metadata in some apps, it becomes more likely.<hr></blockquote>



    Definitely.



    [quote]Considering the dislike expressed by Avie/Apple for the old metadata in favor of file extensions it becomes far less likely.<hr></blockquote>



    For the old (file system specific, fragile, non-portable) metadata, yeah. Says nothing about their opinions on metadata in general.



    [quote]Those are four conclusions that make sense.<hr></blockquote>



    Well, two at least.



    [quote]What doesn't make sense is that you decide to desregard all the evidence that opposes the conclusion you prefer.<hr></blockquote>



    No, I'm only pointing out that two of those conclusions are based on faulty assumptions, and even if they were completely accurate, the relative amount that they are likely to reduce the probability of *Apple* working on metadata is minimal compared to the other two empirically based facts. The weighting of these four conclusions is unequal.



    Apple IS moving the iApps towards metadata: Big +

    Giampaolo in the FS group: Big +

    Giampaolo may or may not be working on other things: Unknown... small to medium -

    'Apple hates metadata': Unfounded



    So (Big +) + (Big +) + (Medium -) = a negative result?



    Hmmm. I think there's something wrong with your probabilistic methodology.
  • Reply 47 of 55
    OK, this is a little off-topic, but I am real tired of people describing how they lost a days worth of work in this or that program thanks to an os 9 crash. Obvious exceptions include long program compiles and 3d renders, but :



    Computer science 101: periodically save your work!! If you spend all day writing the great american novel and never think to save it, you deserve what you get.



    The reality is that people save their work when they switch from one program to another-- it is good policy in any operating system, and it is just not that hard to do.
  • Reply 48 of 55
    wfzellewfzelle Posts: 137member
    [quote]Originally posted by Kickaha:

    <strong>THAT is 'far more likely'?!?</strong><hr></blockquote>



    10000 times more likely is indeed quite a boost.



    [quote]<strong>Really? You've still failed to show *any* relationship between metadata and journaling other than a circumstantial proximity in BeFS' implementation.</strong><hr></blockquote>



    Sigh. Hear yourself. No other relationship than [...]. The relationship you so easily disregard is a very important one.



    [quote]<strong>These are absolute statements. Don't backpedal and then claim that it's been your position all along.</strong><hr></blockquote>



    That was wrong. I said so before.



    [quote]<strong>So basically you're saying that we have no facts about Giampaolo's workload, his time, what he may have been working on, may be working on, or may work on in the future, but from that you're willing to state that Apple isn't interested in metadata?</strong><hr></blockquote>



    There is certain evidence which worries me. I never said that I know for certain that Apple isn't interested in seriously working on advanced metadata. I do disagree with all the post that seem to be sure of Apple's good intentions. I see no evidence that warrant such a level of confidence.



    [quote]<strong>Proximity is a circumstantial casual relationship only. It is not a conceptual, causal, or implicatory relationship. You therefore cannot derive *ANY* conclusions between the two items. Logic 101, propositional logic.</strong><hr></blockquote>



    Suppose that we have two shoe stores close to each other. Usual effect: an increase in sales for both. It seems to me that you don't understand that proximity can be an important variable, certainly when we are dealing with human behaviour.



    [quote]<strong>Still failing to see how this anything to do with *IF* Apple is working on metadata. This observation, while true, has more to do with the implementation details of *HOW*.</strong><hr></blockquote>



    Let's suppose that Apple's engineers have a say on the features that are produced. Example:



    Steve: Yo iTunes dude, any cool idea's?

    iTunes dude: I was thinking about metadata in iTunes. Can we perhaps make it application independent?

    Steve: Huh?

    iTunes dude: People can add their own metadata to mp3's and manage it in iTunes or any other app.

    Steve: Huh?

    iTunes dude: Ratings, only much better.

    Steve: I like ratings. Ok.

    Steve: Can you build it?

    iTunes dude: Nope.

    Steve: $#@&*grrr#%$#^

    iTunes dude: We've got this BeFS guy that might know something about it.

    Giampaolo: Metadata in the FS is so cool. Get a team together and ...



    If Giampaolo wasn't at Apple, the iTunes dude might not have gotten that idea, nor might he have known how to implement it (and abandoned the idea).



    [quote]<strong>Because Giampaolo has been working on something else, *Apple* is less likely to work on metadata? You are correct that: he's the most likely person to have on board for design decisions regarding metadata, and that his finite time has also been used for journaling (we presume from the eWeek article), so he might have had less time to devote to metadata, but jumping from his finite resources to the resources of Apple as a larger company (or even just the FS group) is a bit specious, I'd say.</strong><hr></blockquote>



    I assume that metadata in the FS is so complex that they *need* a guy with experience. It certainly took Be a while to get it right. IMO Apple would be stupid to disregard the experience gained by the BeFS engineers, so it's extremely likely that Giampaolo would be recruited for such a project (at least as an advisor).



    [quote]<strong>For the old (file system specific, fragile, non-portable) metadata, yeah. Says nothing about their opinions on metadata in general.</strong><hr></blockquote>



    I don't think that you can achieve a fully transparent, non-fragile metadata system. The extreme focus on portability that Avie seems to have is IMO a threat to an advanced metadata system. We might get a slow, bloated, non-functional piece of crap instead of something exceptional.



    [quote]<strong>No, I'm only pointing out that two of those conclusions are based on faulty assumptions, and even if they were completely accurate, the relative amount that they are likely to reduce the probability of *Apple* working on metadata is minimal compared to the other two empirically based facts. The weighting of these four conclusions is unequal.



    Apple IS moving the iApps towards metadata: Big +

    Giampaolo in the FS group: Big +

    Giampaolo may or may not be working on other things: Unknown... small to medium -

    'Apple hates metadata': Unfounded



    So (Big +) + (Big +) + (Medium -) = a negative result?



    Hmmm. I think there's something wrong with your probabilistic methodology.</strong><hr></blockquote>



    I weigh things differently. Having metadata in the iApps doesn't mean that Apple will create an app-independent metadata system. After all, I can point out thousands of apps with some kind of metadata. Furthermore, I do believe that Avie is a threat to the cause of advanced metadata, not in the least because he likes to overengineer things (Aqua) and has a very strong bias towards the command line.



    My estimate:

    small +

    big +

    medium -

    big -

    ---------

    Red alert



    [ 10-19-2002: Message edited by: wfzelle ]</p>
  • Reply 49 of 55
    kickahakickaha Posts: 8,760member
    [quote]10000 times more likely is indeed quite a boost.<hr></blockquote>



    Ah. You intended a 10000x boost to a infinitesimal quantity (resulting in an infinitesimal quantity) to attempt to illustrate 'far more likely'... working under the assumption that you couldn't have possibly meant that, I took the two numbers as a range... which was equally silly as a support for your argument, I'll grant.



    [quote]Sigh. Hear yourself. No other relationship than [...]. The relationship you so easily disregard is a very important one.<hr></blockquote>



    Then please, by all means, illustrate this. Continually stating it does not make it so.



    Illustrate how a casual proximity of the implementation of two filesystem entities means that they are *CONCEPTUALLY* related such that you can draw conclusions about the presence of one (or non-presence, I'm still not sure which side you're arguing here) based on the presence of the other.



    Explicitly describe the inference that you're attempting to use in your relationship, or just drop it, because without that description, this is a non-argument.



    [quote]There is certain evidence which worries me. I never said that I know for certain that Apple isn't interested in seriously working on advanced metadata.<hr></blockquote>



    You certainly stated that it was a given. (Instead of pasting them in here Yet Again, I'll just refer you to the later bolded area.)



    [quote]I do disagree with all the post that seem to be sure of Apple's good intentions. I see no evidence that warrant such a level of confidence.<hr></blockquote>



    Good intentions? Whoever said anything about good intentions? Sorry, my conclusion is based strictly on evidence I see with my own eyes, in this case. I see increased metadata coming out of Apple in the iApps, those apps have been produced *since* Jobs and Avie came back, I see Giampaolo being hired on, with his BeFS work being mentioned as a strong reason, I see rich hierarchical filetype metadata being introduced into the developer tools... these are all empirical evidence. There are no assumptions, no suppositions, no drawn conclusions at this point. Each of these is irrefutable, empirically provable and easily checked.



    The only conclusion that I am making is that Apple is interested in metadata, and that they are extremely likely to be working on it at the filesystem level, given the *filetype* metadata support in the dev tools... only the Finder is going to be using that type of data on a regular basis.



    I make no assumptions as to what Giampaolo may or may not be working on, whether his tasklist has any bearing on what Apple as a corporation may or may not be in interested in, how his presence affects Apple's larger concerns, etc, etc, etc.



    *THAT* is what I'm arguing against here. Much of your argument appears to be based on personal beliefs, assumptions, and chains of conclusions, and I'm attempting to point out the problems with that.



    I have no vague beliefs in what Apple's 'good intentions' are, nor am I drawing false conclusions based on little or no supporting data. My position is clearly given above, in total.



    Of the four items at the end of your previous post, two are drawn conclusions without much foundation other than your personal beliefs. If that is your position, then it's obvious that this discussion *has* been a waste of my time.



    [quote]Suppose that we have two shoe stores close to each other. Usual effect: an increase in sales for both.<hr></blockquote>



    In that context proximity is a variable to a function of shopper mobility.



    Provide for me a description of your function determining a relationship between orthogonal filesystem features, and we can discuss *that*.



    I can provide plenty of proximity-input functions of relationships... neutron scattering as a precursor to fission is highly spatially dependent. But in each case, the proximity is an input to a relationship function, not a relationship in and of itself.



    [quote]It seems to me that you don't understand that proximity can be an important variable, certainly when we are dealing with human behaviour.<hr></blockquote>



    Variable. Precisely.



    Describe the function that takes that variable, and we'll have a basis for discussion. Put it in context for *this* discussion regarding metadata and journaling implementations in filesystems. Stating that because proximity is used as a variable in other relationship functions means that it is automatically relevant here is a fallacy.



    [quote]Let's suppose that Apple's engineers have a say on the features that are produced. Example:



    [snipped]



    If Giampaolo wasn't at Apple, the iTunes dude might not have gotten that idea, nor might he have known how to implement it (and abandoned the idea).<hr></blockquote>



    Wow. Stunning bit of artificial suppositions slapped together into a single framework, I'll grant you that... but you didn't exactly show how this supports your assertion that Giampaolo's presence and workload (as affected by his presumed work on Elvis) affect the decision to pursue metadata at the filesystem level. I mean c'mon, I can come up with imagined conversations to support any position I care to come up with.



    I understand the point you are attempting to show, that Giampaolo's employment at Apple is likely to mean that they are pursuing metadata. I *agree* with that. What I don't agree with is your apparent assumption that he is the sole repository or source for metadata design and implementation, instead of perhaps being an advisor on a larger team, or elsewise using his resources and talents in a more widespread way. Nor do I agree with your assumption that without his presence at Apple, metadata would not be being pursued... his employment came after iTunes and iPhoto. This has obviously been under consideration for quite a while. Giampaolo is not the only person to have ever thought of rich metadata. He's not the only person to have *implemented* filesystem level metadata. He is the only person I know of that has implemented such a system essentially by himself, which shows real talent... but not uniqueness in the concepts involved. His presence at Apple is much more likely to mean that Apple is interested in his talents (metadata) than he is spontaneously going to infuse Apple with his interests.



    Besides, I have to admit I'm getting confused here... we were discussing proximity of metadata and journaling in BeFS, and how that proximity, you claim, means that the two orthogonal concepts are somehow tied at a conceptual level... and now we're discussing Giampaolo's proximity within Apple as a seed for meme distribution?



    Which was it again that you wished to discuss?



    [quote]I assume that metadata in the FS is so complex that they *need* a guy with experience. It certainly took Be a while to get it right. IMO Apple would be stupid to disregard the experience gained by the BeFS engineers, so it's extremely likely that Giampaolo would be recruited for such a project (at least as an advisor).<hr></blockquote>



    So, restating again that we know nothing about his workloads, etc... you're saying that because he (presumably - eWeek is the only source on this) worked on Elvis, that he couldn't even be an advisor on a metadata project? After all, that's the only way to support the following statements:



    <strong>On the other hand, now we know that the BeFS guy that Apple hired is not working on metadata.



    Some hoped he was working on metadata, but he clearly isn't.



    People wonder why Apple wants the one [journaling], but not the other [metadata].</strong>



    [quote]I don't think that you can achieve a fully transparent, non-fragile metadata system. The extreme focus on portability that Avie seems to have is IMO a threat to an advanced metadata system. We might get a slow, bloated, non-functional piece of crap instead of something exceptional.<hr></blockquote>



    'I don't think'... 'We might get'....



    Obviously a strong factual foundation there from which to argue your position.





    [quote]I weigh things differently. Having metadata in the iApps doesn't mean that Apple will create an app-independent metadata system. After all, I can point out thousands of apps with some kind of metadata.<hr></blockquote>



    You're correct, it doesn't. What it does illustrate is a newfound interest in metadata as an organizational tool for files... and one that has only shown up in Apple *after* the return of Jobs and the introduction of Tevanian and the ex-NeXT engineers. The *filetype* information in the dev tools would only really be useful to the Finder, in a general case, which would benefit most from filesystem level support.



    [quote] Furthermore, I do believe that Avie is a threat to the cause of advanced metadata, not in the least because he likes to overengineer things (Aqua) and has a very strong bias towards the command line.<hr></blockquote>



    As amply shown by the complete lack of interest in organizational metadata anywhere in Apple's shipping products... *sigh*



    [quote]My estimate:

    small +

    big +

    medium -

    big -

    ---------

    Red alert

    <hr></blockquote>



    Alright, this is obviously a useless argument to be having, since your beliefs are based on more assumption than fact. That's fine, but it certainly makes your position less arguable, which of course makes debate a waste of time. Perhaps we should move this to Fireside Chat? Seems at this point that it would fit in better there.



    Basically your arguments reduce down to a lot of beliefs, assumptions, and drawn conclusions without much empirical evidence to back any of them up. If that's the position you choose to take, that's fine, but I'm afraid I really don't care to continue this, in that case.



    [ 10-19-2002: Message edited by: Kickaha ]</p>
  • Reply 50 of 55
    a@rona@ron Posts: 201member
    This info is non first hand:



    Apple is indeed working on this and recent seeds of 10.2.2 are up to &gt;6F12 so my sources say. We should see this implimented in builds within the next week and a half. Just what I hear.



    A@ron
  • Reply 51 of 55
    wfzellewfzelle Posts: 137member
    [quote]Originally posted by Kickaha:

    <strong>Ah. You intended a 10000x boost to a infinitesimal quantity (resulting in an infinitesimal quantity) to attempt to illustrate 'far more likely'... working under the assumption that you couldn't have possibly meant that, I took the two numbers as a range... which was equally silly as a support for your argument, I'll grant.</strong><hr></blockquote>



    1 in 1000 is hardly infinitesimal. I would gladly pay $1 for a 1 in 1000 chance to become a millionaire. A 1 in 10,000,000 chance is far less enticing for me. Of course, I only used this as an example of how a proximity relationship can greatly increase a chance of something occuring. The values for the case under consideration are probably far greater, but it is helpful if you agree with the basic mechanism instead of argueing about the values I use to demonstrate a certain mechanism.



    [quote]<strong>Then please, by all means, illustrate this. Continually stating it does not make it so.</strong><hr></blockquote>



    I have. Then I hear things like: "Well, except for that, can you explain why this relationship exists". It gets kind of tiring if you keep disqualifying my arguments for no good reason.



    [quote]<strong>Illustrate how a casual proximity of the implementation of two filesystem entities means that they are *CONCEPTUALLY* related</strong><hr></blockquote>



    Why? The proximity causes certain real-life effects which I use to draw conclusions. The lack of a conceptual relation of the kind you seem to want (A has the same technical background as B) is no disqualification of such an argument.



    There is a fire in the house next door. I must evacuate. Or would you disagree since there is no relevant conceptual relationship between the two houses? Duh!



    [quote]<strong>You certainly stated that it was a given. (Instead of pasting them in here Yet Again, I'll just refer you to the later bolded area.)</strong><hr></blockquote>



    I'm working on three major projects right now. If the first is released, does that mean I haven't been doing anything else?



    Absence of evidence does not prove a negative.




    Journaling is not that easy to implement AFAIK. At the very least this new info makes it less likely that he is working on metadata. But my conclusion was probably too strong.




    I might have not have reversed my position, but I certainly told you that my previous conclusion was flawed.



    [quote]<strong>Good intentions? Whoever said anything about good intentions? Sorry, my conclusion is based strictly on evidence I see with my own eyes, in this case. I see increased metadata coming out of Apple in the iApps, those apps have been produced *since* Jobs and Avie came back, I see Giampaolo being hired on, with his BeFS work being mentioned as a strong reason, I see rich hierarchical filetype metadata being introduced into the developer tools... these are all empirical evidence. There are no assumptions, no suppositions, no drawn conclusions at this point. Each of these is irrefutable, empirically provable and easily checked.



    The only conclusion that I am making is that Apple is interested in metadata, and that they are extremely likely to be working on it at the filesystem level, given the *filetype* metadata support in the dev tools... only the Finder is going to be using that type of data on a regular basis.</strong><hr></blockquote>



    Clearly you overrate each of these issues. The increased metadata in the iApps isn't that strong an indication that metadata will be moved to the FS. Again, probably 99.9% of the apps in existance have metadata. This has been the case since the dawn of time (the 60's). How many FS's with metadata have we seen since then? The fact that Apple hires a guy famous for journaling doesn't mean that he will work on metadata or have any influence to get metadata in the FS. Some metadata was already in OS 9 (in the dev tools and it was used by the finder). I can't remember stumbling upon an extendable metadata system. It would be far more exciting if the data was in fact not intended for the finder (if you start to support one regular app, then why not another, etc, etc).



    Then I look at the people on this board dancing around the campfire and singing: "We will get FS metadata, rejoice!". Somehow that doesn't seem to mesh with the facts.



    [quote]<strong>I make no assumptions as to what Giampaolo may or may not be working on, whether his tasklist has any bearing on what Apple as a corporation may or may not be in interested in, how his presence affects Apple's larger concerns, etc, etc, etc.</strong><hr></blockquote>



    Then why do you note the hiring of Giampaolo as a indicator that Apple is working on FS metadata? That doesn't make any sense at all if you don't assume he's working on metadata or influencing agenda's. Because if you do, the only conclusion that you can draw is this fact: Giampaolo is working on journaling. And I thought you said that journaling and FS metadata weren't related. <img src="graemlins/hmmm.gif" border="0" alt="[Hmmm]" />



    [quote]<strong>*THAT* is what I'm arguing against here. Much of your argument appears to be based on personal beliefs, assumptions, and chains of conclusions, and I'm attempting to point out the problems with that.



    I have no vague beliefs in what Apple's 'good intentions' are, nor am I drawing false conclusions based on little or no supporting data. My position is clearly given above, in total.



    Of the four items at the end of your previous post, two are drawn conclusions without much foundation other than your personal beliefs. If that is your position, then it's obvious that this discussion *has* been a waste of my time.</strong><hr></blockquote>



    I have made some fairly reasonable assertions concerning Giampaolo's agenda. I think you can only argue about the size of the minus, not whether it exists. The other conclusion is certainly not without hard proof. Unfortunately Apple pulled the incriminating developer note, but the omnigroup mailing lists showed some hefty outrage. Many developers were angry that Apple wanted them to remove some very important features in return for a broken system. This resulted in a half-hearted policy chance (using a confusing mix of creator/type and extensions). I guess that even you didn't miss the fuss about file extensions that resulted (see the arstechnica reviews for some info on the problems with the current system). Do you find it unreasonable that, for someone who feels that (non-free) file extensions are a regression, Apple's actions show either a lack of commitment or a lack of understanding? Is this anymore vague than the assertion that metadata in the iApps and the finder points to almost certain adoption of some advanced metadata system that goes far, far beyond the basic demands of these few apps. You can support many features that the advanced metadata system will bring within the apps themselves. Of course, a universal metadata FS will bring those advantages to most other apps, will create enormous network effects and shall give us some great options, but Apple could continue to expand its use of metadata without implementing an advanced, extensible, abstracted metadata system. There is no reason why Apple couldn't just put a big database in iTunes. There is no reason why they can't use file types within document and application packages (using property lists, a desktop DB and a lot of caching).



    I think that Apple can and should do better, but I fear they will not. Why do you think they shall?



    [quote]<strong>In that context proximity is a variable to a function of shopper mobility.



    Provide for me a description of your function determining a relationship between orthogonal filesystem features, and we can discuss *that*.</strong><hr></blockquote>



    You hire someone for the one, he influences you on the other. How do you want to call it? The package effect (you buy the whole package)?



    [quote]<strong>Variable. Precisely.



    Describe the function that takes that variable, and we'll have a basis for discussion. Put it in context for *this* discussion regarding metadata and journaling implementations in filesystems. Stating that because proximity is used as a variable in other relationship functions means that it is automatically relevant here is a fallacy.</strong><hr></blockquote>



    Did I ever say it was automatically relevant? Assuming it is can only be qualified as nonsensical. I don't see why you even considered it to be a valid conclusion. I never did (and thus didn't see a need to actually state that I was talking about a relationship function instead of an automatically relevant thing).



    [quote]<strong>Wow. Stunning bit of artificial suppositions slapped together into a single framework, I'll grant you that... but you didn't exactly show how this supports your assertion that Giampaolo's presence and workload (as affected by his presumed work on Elvis) affect the decision to pursue metadata at the filesystem level. I mean c'mon, I can come up with imagined conversations to support any position I care to come up with.</strong><hr></blockquote>



    It doesn't seem entirely unlikely for something like this to happen. Don't you ever get idea's from fellow employee's? I guess not. In the real world however, companies might actually use the knowledge of their employees or there might be a water cooler or email system that is used to communicate between employees. Perhaps you work for the pointy heared boss and he makes every decision himself, without consulting those with experience? And perhaps you are also forbidden to talk to other employees unless the allows you to?



    [quote]<strong>I understand the point you are attempting to show, that Giampaolo's employment at Apple is likely to mean that they are pursuing metadata.</strong><hr></blockquote>



    No, the point I'm trying to make is that Giampaolo may influence Apple to pursue advanced, extensible metadata in the FS. Apple is already pursuing metadata, but not in the way I like (mostly inside apps and a weak type/creator/extensions abomination). I hope someone may entice Avie with a system so advanced and customizable that he will forget the nonsense about giving every file an extension (bla.basestation, WTF!) and removing all metadata from the FS. Giampaolo may be that guy (probably not, but I'm an optimist).



    [quote]<strong>What I don't agree with is your apparent assumption that he is the sole repository or source for metadata design and implementation</strong><hr></blockquote>



    Before he came, Apple caused quite a mess in OS X (why do we need X-Ray to understand what app opens your file and why?). Why do some apps set the type/creator and others don't? Why is the creator missing in some Apple apps? I hope you don't blame me for pinning my last hopes on him, his predecessors messed things up pretty badly.



    [quote]<strong>Giampaolo is not the only person to have ever thought of rich metadata. He's not the only person to have *implemented* filesystem level metadata. He is the only person I know of that has implemented such a system essentially by himself, which shows real talent... but not uniqueness in the concepts involved.</strong><hr></blockquote>



    True, he is the only person with that track record within Apple, making him the most likely candidate to create an advanced metadata system or to oversee or advise it's creation. Of course, such a job will certainly mean that he will influence the system. I don't think you'll get him to work on a system that he doesn't agree with very easily. Suppose that you build a system that succesfully uses method A after you have evaluated and turned down method B. Now you move to a different employer and are asked to create something similar. Will you advise the use of B? Will you insist?



    [quote]<strong>His presence at Apple is much more likely to mean that Apple is interested in his talents (metadata) than he is spontaneously going to infuse Apple with his interests.</strong><hr></blockquote>



    Would it not be likely that Apple hired Giampaolo for his journaling experience? It seems that he is some kind of authority on the matter. In that case, it would not be unlikely that he might advance his view of metadata somewhat spontaneously, instead of being asked to create/organize/advise an advanced metadata system.



    [quote]<strong>Besides, I have to admit I'm getting confused here... we were discussing proximity of metadata and journaling in BeFS, and how that proximity, you claim, means that the two orthogonal concepts are somehow tied at a conceptual level... and now we're discussing Giampaolo's proximity within Apple as a seed for meme distribution?</strong><hr></blockquote>



    Suppose that Giampaolo was only hired for journaling and perhaps other non-metadata issues. Suppose that advanced metadata comes up and someone decides to ask G. Voila. We go from journaling to metadata. The other possibility that G was hired for his knowledge of metadata is far less interesting in this regard. Then we don't need journaling as a stepping stone.



    [quote]<strong>So, restating again that we know nothing about his workloads, etc... you're saying that because he (presumably - eWeek is the only source on this) worked on Elvis, that he couldn't even be an advisor on a metadata project? After all, that's the only way to support the following statements:</strong><hr></blockquote>



    I pointed out already that I picked the wrong words. Can you please stop pounding on that time and again?



    But you said...

    -I took that back

    But you said...

    -I took that back

    But you said...

    AAAAAAaaargh



    [quote]<strong>'I don't think'... 'We might get'....



    Obviously a strong factual foundation there from which to argue your position.</strong><hr></blockquote>



    The normal way to debate is by asking someone to substantiate his claims. In a debate one does not have to provide proof for all his statements because one would have to write a thick paper to express even a few opinions. As such, it is common to simply state one's position and only defend the statements that are challenged.



    Challenge me if you wish, I can make provide an extensive argument to support my position.



    [quote]<strong>You're correct, it doesn't. What it does illustrate is a newfound interest in metadata as an organizational tool for files... and one that has only shown up in Apple *after* the return of Jobs and the introduction of Tevanian and the ex-NeXT engineers. The *filetype* information in the dev tools would only really be useful to the Finder, in a general case, which would benefit most from filesystem level support.</strong><hr></blockquote>



    True, but this doesn't support your conclusion that iApp metadata and a bit of file type info will be a big + towards an extensible, advanced metadata FS. I've seen many metadata debates and precious few debaters made the link between application-specific metadata and a global system. Why wouldn't the iTunes team just keep adding music specific metadata, the finder team a bit of new file info, the iPhoto team some photo info, etc, etc? After all, didn't developers do that for dozens of years (except on BeOS and a few mainframes).



    [quote]<strong>As amply shown by the complete lack of interest in organizational metadata anywhere in Apple's shipping products... *sigh*</strong><hr></blockquote>



    There is an enormous difference between application metadata and the advanced FS metadata I'm talking about. I've seen developers who wanted to have the first, but not the second (almost over their dead bodies). They are the type that have some terrible fear of interoperability. Avie seems to fit the mold. Witness his clear disgust of type/creator in the FS (OS X doesn't use a file creator for 'open with' for instance, instead stuffing it in the resource fork, something we know he hates a lot), even though he seems to want lots of metadata in every iApp. The logical system for such a person to propose would be the one I described earlier in this post (DB in iTunes, stuffing everything in packages). Personally, I'd rather lose some interoperability to get the many advantages that a advanced metadata FS will bring.



    [quote]<strong>Alright, this is obviously a useless argument to be having, since your beliefs are based on more assumption than fact. That's fine, but it certainly makes your position less arguable, which of course makes debate a waste of time. Perhaps we should move this to Fireside Chat? Seems at this point that it would fit in better there.



    Basically your arguments reduce down to a lot of beliefs, assumptions, and drawn conclusions without much empirical evidence to back any of them up. If that's the position you choose to take, that's fine, but I'm afraid I really don't care to continue this, in that case.</strong><hr></blockquote>



    I think your conclusions are certainly not less far-fetched. You ignore the many possibilities other than a metadata FS as if this was a black/white world. The evidence you have that Apple is actually considering something that advanced and useful is sorely lacking. It seems to boil down to Giampaolo's presence which you believe to have some enormous influence and the iApp and finder metadata which you assume Apple wants to extend to third party apps and abstracted instead of being extended like it has before. In the end you draw a conclusion that goes far beyond the facts that you started with and seems to clash with the facts that I pointed out. IMHO, my story fits the facts better. It accounts for both the increased metadata in the iApps and the reduced use of metadata in the FS.



    [ 10-20-2002: Message edited by: wfzelle ]</p>
  • Reply 52 of 55
    jlljll Posts: 2,713member
    [quote]Originally posted by A@ron:

    <strong>This info is non first hand:



    Apple is indeed working on this and recent seeds of 10.2.2 are up to &gt;6F12 so my sources say. We should see this implimented in builds within the next week and a half. Just what I hear.



    A@ron</strong><hr></blockquote>



    Journaling is already in Mac OS X Server 10.2.2 6F12 (which is the newest seeded build).
  • Reply 53 of 55
    frawgzfrawgz Posts: 547member
    [quote]Originally posted by wfzelle:

    <strong>Furthermore, I do believe that Avie is a threat to the cause of advanced metadata, not in the least because he likes to overengineer things (Aqua) and has a very strong bias towards the command line.</strong><hr></blockquote>



    :confused: :confused:



    My two cents: I tend to agree with Buonrotto and others on the Clean Slate Theory of Apple's metadata strategy. Let's agree on one thing first: advanced, pervasive metadata provides enormous potential to be the next paradigm of information access and manipulation. Whether Apple is on its way there or not is still in question, although I like to think there are signs they are interested.



    Personally, while I liked the type/creator model, types and creators were relatively hidden from me as an end user. There was no easy or elegant interface provided for me to access or change types and creators. Also, let's face it: we live in a Windows world. I liked to spout my arrogance and say extensions were esoteric and ugly on my machine, but when it came time for me to share a file with a friend, I had to eat humble pie and tack that extension on for her benefit. I know this is a touchy issue for a lot of us, so let's leave it at that and not let this thread devolve into an extensions vs. type/creator argument.



    Assume for a second that Apple had planned to ditch type/creator for a robust, advanced metadata model to take us into the next decade. Would it have built on the crusty 80's type/creator model? If it were up to me, no. Others have pointed out that the new Extensions/Open With model works in the interim to rid us of the old type/creator dependencies, while putting us squarely within the realm of the de facto standard in file naming schemes. What this does for us when Apple extends this model into the world of advanced metadata is allow portability at the basic level of extensions. What this means for me: I can have the greatest way to access data in the world, and when it comes time for me to share that data, crucial file types won't be lost. This is all speculation, of course.



    What I know for certain now: creator codes allowed us to differentiate within file types so that certain files would be opened by certain applications. This functionality is exposed to me in a simple manner in Mac OS X. I simply select my files and Get Info, then choose the application I want them to open in. No creator codes, just application names. No complaints for now.
  • Reply 54 of 55
    mrmistermrmister Posts: 1,095member
    Kickaha--



    Reductio ad absurdum much?
  • Reply 55 of 55
    Enough's enough, folks.



    These drawn out arguments are wearing thin and are leading nowhere. We're really not making any headway here.



    This whole thread was started based on what I (and probably others) consider to be a logical fallacy. Does the existence of one feature directly result in the non-existence of another? I don't think so. Regardles, it has lead to nit-picking and bickering over speculation of an issue that NO ONE here has all the facts on.



    It ends now.



    [ 10-21-2002: Message edited by: Brad ]</p>
Sign In or Register to comment.