This is probably the biggest thing in Tiger, really.
Core Image/Video? Sweet.
Automator? Excellent!
But *THIS*? Revolutionary isn't too far from the truth, really.
Remember the buzz about MS's OneNote, where you can organize various MS documents (Exchange, Word, Excel, PowerPoint) by project? This kicks its sorry ass all *OVER* the damned playground. And back. Twice. And then steals its lunch money.
You can organize *any* file, from *any* app, in *any* way that *you* want to.
In the Finder.
With a simple, already understood and known UI element from iTunes and iPhoto.
This is, quite simply, the biggest leap in user content accessibility since google.
With a simple, already understood and known UI element from iTunes and iPhoto.
And I think that's the best part of this new feature : no need to learn to use it properly, since we've already used it in a less general context. That means more productivity, out of the box!
And that's huge! I'm glad that Steve closed the HI division and dispatched its members in all the software divisions, that's how such revolutionnary features can be introduced without having to learn everything again!
I am apprehensive about it. But maybe it will help me to stop creating a subfolder under my documents directory for every category of data. If that happens, I am a happy customer.
I was wondering if anyone knew how this compares to what microsoft has (had) envisioned for longhorn? is this more or less exactly what they were thinking?... did apple completely blow Microsoft out of the water, or is this not quite up to what (was) (supposedly but won't) be out in longhorn?
I am apprehensive about it. But maybe it will help me to stop creating a subfolder under my documents directory for every category of data. If that happens, I am a happy customer.
That's it... exactly! You'll become a happy customer in less than a year!
I was wondering if anyone knew how this compares to what microsoft has (had) envisioned for longhorn? is this more or less exactly what they were thinking?... did apple completely blow Microsoft out of the water, or is this not quite up to what (was) (supposedly but won't) be out in longhorn?
What I have understood about Longhorn is that this is Microsoft's Copland... and that also works for all the metadata stuff. They wanted to create a revolutionnary metadata manager, and completely blow out the spatial dimension of the filesystem, which would have been pretty cool actually. Only problem : compatibility. Microsoft wants to build new great features and keep the compatibilty with old rotten stuff (and that's a very Coplandish approach...).
So eventually, they decided to build metadata management in a very incremental way, much like Apple. Only problem : with Tiger, Apple is 2 year-early, compared to M$ (okay, 10 year-late compared to BeOS...!), so I guess that when Longhorn is ready, Mac OS X will completely kill it, in the metadata field (well, in the other fields too, actually!).
Only problem : with Tiger, Apple is 2 year-early, compared to M$ (okay, 10 year-late compared to BeOS...!), so I guess that when Longhorn is ready, Mac OS X will completely kill it, in the metadata field (well, in the other fields too, actually!).
Over at Arstechnica they apparently quote Apple, and this might not be metadata after all. Looks like it is a huge, periodic, indexing scheme of everything in the FS. This'd mean the database is not there... Souped Sherlock or souped up FS?
Over at Arstechnica they apparently quote Apple, and this might not be metadata after all. Looks like it is a huge, periodic, indexing scheme of everything in the FS. This'd mean the database is not there... Souped Sherlock or souped up FS?
ZoSo
A searchable index containing data about files and their contents isn't a database?
Over at Arstechnica they apparently quote Apple, and this might not be metadata after all. Looks like it is a huge, periodic, indexing scheme of everything in the FS. This'd mean the database is not there... Souped Sherlock or souped up FS?
ZoSo
well first, i believe it is metadata
and second, what the nerds over at ars dont get, is that if this stuff works and appears to the user the same way, it doesnt matter what the underlying technology is. this stuff works, and we'll have it early next year hopefully. and it works damn well it appears.
and second, what the nerds over at ars dont get, is that if this stuff works and appears to the user the same way, it doesnt matter what the underlying technology is. this stuff works, and we'll have it early next year hopefully. and it works damn well it appears.
very exciting
Well, if there is need for a periodic indexing of a hard drive contents, then obviously for a time-til the next sync, that is-the results to the live-queries won't correspond to the actual condition of your data/docs. Am I right here?
Also, I wonder what kind of a burden updating the index will cause. Quick but the whole system is blocked for 60 seconds? Responsive, but it takes 30 minutes? I don't know, I was just considering the possibilities...
and second, what the nerds over at ars dont get, is that if this stuff works and appears to the user the same way, it doesnt matter what the underlying technology is.
Well, not exactly, IMO. Remember Sherlock indexing? It takes forever, consumes a hell of resources and saves huge index files. I use content indexing/searching on my /Developer/Documentation folder and a folder containing pdf and html books and, man, it's ugly, slow and sometimes cannot find things because the index hasn't been updated.
Quote:
Originally posted by applenut
well first, i believe it is metadata
I hope so, too. Let's see:
Quote:
The metadata engine uses special importer technology to open and read heterogeneous file formats. Tiger includes importers for some of the most popular file formats. The metadata engine can be extended to any new file format, automatically adding an application?s information to your search results...
This means that searching through a 1GB psd is going to be fast: we extract only textual and numerical info because users can't search for pixels.
Quote:
Even documents without any metadata are included in searches...
Metadata exists! ***CONF!RMED***
Quote:
By combining the right search technology for each type of searching task, Tiger delivers the best possible results and experience...
Aha, so it's possibly a combination of databases, indexes and BeOS-style metadata.
Quote:
The Macintosh file system is highly optimized for quickly and efficiently reading and writing files. The metadata search engine builds on this core strength by transparently extracting metadata from files in the background and indexing it for optimal search responsiveness...
So HFS+ is going to give us much better metadata support, because without the help from FS... you know how it feels to work with content indexing in the background. However, in the background it may index only metadata, which is a much less resource-intensive process.
Quote:
A separate full-content index optimizes performance for file content searches...
So a full-content index co-exists with metadata. I want to try this search system myself. I suspect, though, that stunning performance is achieved only on 1000 text files on 2x2.5GHz machines with 4GB of RAM, but God only knows how I want to be surprised!
Since this topic has been discussed speculatively for so long, and it's finally here, could a mod possibly change this topic from the kinda anemic "new search" to something a little more descriptive...? "Spotlight: new search and metadata"? "Spotlight and Metadata"?
With Spotlight, I see the Finder becoming just another app...like iTunes. The Finder will become an app to physically move files around and store them in a spatial (semi-spatial) environment or group them using Smart Folders. But I think the majority of people will start using the Spotlight menu in the top-right corner to find their files. After all, it's a global menu. It's there no matter what app you're using.
Apps like the Finder and iTunes will always be there to serve their purpose. iTunes lets you organize your music files and play your music. The Finder will let you install apps and browse your HD in a hierarchical way...but with Spotlight, hierarchy isn't very important anymore, file location isn't important anymore and will only be popular with neat-freaks or people set in their ways and don't want the OS to evolve.
When Spotlight finally arrives, I can just throw everything in the Documents folder (I'll be nice and throw music in the Music folder and video in the Video folder...although file location really just doesn't matter anymore) and let Spotlight find the files I want.
And, while I haven't seen any screenshots of this or any mention of this yet, Open/Save dialogs using Spotlight technology is probably there too...iTunes-like filtering of files from within apps via Open dialogs would make all existing apps infinitely more usable...especially if I'm going to throw all my files into one folder.
The apps could initially and by default filter by recognizable extension or file type. Then all the relevant files would show up, then you could start filtering from there.
Spotlight alone will be worth whatever price tag Apple puts on Tiger. But H.264/AVC and CoreImage is just too fuckin' cool too.
I am apprehensive about it. But maybe it will help me to stop creating a subfolder under my documents directory for every category of data. If that happens, I am a happy customer.
Well, not exactly, IMO. Remember Sherlock indexing? It takes forever, consumes a hell of resources and saves huge index files. I use content indexing/searching on my /Developer/Documentation folder and a folder containing pdf and html books and, man, it's ugly, slow and sometimes cannot find things because the index hasn't been updated.
Ever notice that happening now? No?
Guess what - it's still indexed.
Been fixed a long time now... a one-time long index, then background updates and file notifications (remember, that's now robust in the OS and the Finder (and other apps, say, an indexer) can be written to respond to them) happen on the fly when you're not doing much else.
Comments
Originally posted by a_greer
macrumors:SWEEEEEEET I WANT IT!!!!!!!!!!!
It'll be eight long years of waiting, then, til we finally get those Be OS live queries again! Whoo-hoo!
Glad somebody in Cupertino finally woke up! Oh wait, they had to hire the very same engineer to do it...
ZoSo
Originally posted by ZoSo
It'll be eight long years of waiting, then, til we finally get those Be OS live queries again! Whoo-hoo!
Glad somebody in Cupertino finally woke up! Oh wait, they had to hire the very same engineer to do it...
ZoSo
Yep, thanks Dom!
Core Image/Video? Sweet.
Automator? Excellent!
But *THIS*? Revolutionary isn't too far from the truth, really.
Remember the buzz about MS's OneNote, where you can organize various MS documents (Exchange, Word, Excel, PowerPoint) by project? This kicks its sorry ass all *OVER* the damned playground. And back. Twice. And then steals its lunch money.
You can organize *any* file, from *any* app, in *any* way that *you* want to.
In the Finder.
With a simple, already understood and known UI element from iTunes and iPhoto.
This is, quite simply, the biggest leap in user content accessibility since google.
Originally posted by Kickaha
With a simple, already understood and known UI element from iTunes and iPhoto.
And I think that's the best part of this new feature : no need to learn to use it properly, since we've already used it in a less general context. That means more productivity, out of the box!
And that's huge! I'm glad that Steve closed the HI division and dispatched its members in all the software divisions, that's how such revolutionnary features can be introduced without having to learn everything again!
Again, that's HUGE!!!!!!
Originally posted by talksense101
I am apprehensive about it. But maybe it will help me to stop creating a subfolder under my documents directory for every category of data. If that happens, I am a happy customer.
That's it... exactly! You'll become a happy customer in less than a year!
Originally posted by berrism
I was wondering if anyone knew how this compares to what microsoft has (had) envisioned for longhorn? is this more or less exactly what they were thinking?... did apple completely blow Microsoft out of the water, or is this not quite up to what (was) (supposedly but won't) be out in longhorn?
What I have understood about Longhorn is that this is Microsoft's Copland... and that also works for all the metadata stuff. They wanted to create a revolutionnary metadata manager, and completely blow out the spatial dimension of the filesystem, which would have been pretty cool actually. Only problem : compatibility. Microsoft wants to build new great features and keep the compatibilty with old rotten stuff (and that's a very Coplandish approach...).
So eventually, they decided to build metadata management in a very incremental way, much like Apple. Only problem : with Tiger, Apple is 2 year-early, compared to M$ (okay, 10 year-late compared to BeOS...!), so I guess that when Longhorn is ready, Mac OS X will completely kill it, in the metadata field (well, in the other fields too, actually!).
Originally posted by The One to Rescue
Only problem : with Tiger, Apple is 2 year-early, compared to M$ (okay, 10 year-late compared to BeOS...!), so I guess that when Longhorn is ready, Mac OS X will completely kill it, in the metadata field (well, in the other fields too, actually!).
Over at Arstechnica they apparently quote Apple, and this might not be metadata after all. Looks like it is a huge, periodic, indexing scheme of everything in the FS. This'd mean the database is not there... Souped Sherlock or souped up FS?
ZoSo
Originally posted by ZoSo
Over at Arstechnica they apparently quote Apple, and this might not be metadata after all. Looks like it is a huge, periodic, indexing scheme of everything in the FS. This'd mean the database is not there... Souped Sherlock or souped up FS?
ZoSo
A searchable index containing data about files and their contents isn't a database?
Notice that Apple implemented SQLite in Tiger.
Originally posted by ZoSo
Over at Arstechnica they apparently quote Apple, and this might not be metadata after all. Looks like it is a huge, periodic, indexing scheme of everything in the FS. This'd mean the database is not there... Souped Sherlock or souped up FS?
ZoSo
well first, i believe it is metadata
and second, what the nerds over at ars dont get, is that if this stuff works and appears to the user the same way, it doesnt matter what the underlying technology is. this stuff works, and we'll have it early next year hopefully. and it works damn well it appears.
very exciting
Originally posted by JLL
A searchable index containing data about files and their contents isn't a database?
Notice that Apple implemented SQLite in Tiger.
My reference was to a database(-like) FS.
I just hope this indexing method is not clunky as Sherlock's.
ZoSo
Originally posted by applenut
well first, i believe it is metadata
and second, what the nerds over at ars dont get, is that if this stuff works and appears to the user the same way, it doesnt matter what the underlying technology is. this stuff works, and we'll have it early next year hopefully. and it works damn well it appears.
very exciting
Well, if there is need for a periodic indexing of a hard drive contents, then obviously for a time-til the next sync, that is-the results to the live-queries won't correspond to the actual condition of your data/docs. Am I right here?
Also, I wonder what kind of a burden updating the index will cause. Quick but the whole system is blocked for 60 seconds? Responsive, but it takes 30 minutes? I don't know, I was just considering the possibilities...
ZoSo
Originally posted by applenut
and second, what the nerds over at ars dont get, is that if this stuff works and appears to the user the same way, it doesnt matter what the underlying technology is.
Well, not exactly, IMO. Remember Sherlock indexing? It takes forever, consumes a hell of resources and saves huge index files. I use content indexing/searching on my /Developer/Documentation folder and a folder containing pdf and html books and, man, it's ugly, slow and sometimes cannot find things because the index hasn't been updated.
Originally posted by applenut
well first, i believe it is metadata
I hope so, too. Let's see:
The metadata engine uses special importer technology to open and read heterogeneous file formats. Tiger includes importers for some of the most popular file formats. The metadata engine can be extended to any new file format, automatically adding an application?s information to your search results...
This means that searching through a 1GB psd is going to be fast: we extract only textual and numerical info because users can't search for pixels.
Even documents without any metadata are included in searches...
Metadata exists! ***CONF!RMED***
By combining the right search technology for each type of searching task, Tiger delivers the best possible results and experience...
Aha, so it's possibly a combination of databases, indexes and BeOS-style metadata.
The Macintosh file system is highly optimized for quickly and efficiently reading and writing files. The metadata search engine builds on this core strength by transparently extracting metadata from files in the background and indexing it for optimal search responsiveness...
So HFS+ is going to give us much better metadata support, because without the help from FS... you know how it feels to work with content indexing in the background. However, in the background it may index only metadata, which is a much less resource-intensive process.
A separate full-content index optimizes performance for file content searches...
So a full-content index co-exists with metadata. I want to try this search system myself. I suspect, though, that stunning performance is achieved only on 1000 text files on 2x2.5GHz machines with 4GB of RAM, but God only knows how I want to be surprised!
Even just "Spotlight" would be much better...
Lemon Bon Bon
Apps like the Finder and iTunes will always be there to serve their purpose. iTunes lets you organize your music files and play your music. The Finder will let you install apps and browse your HD in a hierarchical way...but with Spotlight, hierarchy isn't very important anymore, file location isn't important anymore and will only be popular with neat-freaks or people set in their ways and don't want the OS to evolve.
When Spotlight finally arrives, I can just throw everything in the Documents folder (I'll be nice and throw music in the Music folder and video in the Video folder...although file location really just doesn't matter anymore) and let Spotlight find the files I want.
And, while I haven't seen any screenshots of this or any mention of this yet, Open/Save dialogs using Spotlight technology is probably there too...iTunes-like filtering of files from within apps via Open dialogs would make all existing apps infinitely more usable...especially if I'm going to throw all my files into one folder.
The apps could initially and by default filter by recognizable extension or file type. Then all the relevant files would show up, then you could start filtering from there.
Spotlight alone will be worth whatever price tag Apple puts on Tiger. But H.264/AVC and CoreImage is just too fuckin' cool too.
Originally posted by talksense101
I am apprehensive about it. But maybe it will help me to stop creating a subfolder under my documents directory for every category of data. If that happens, I am a happy customer.
this is just one purpose i can think of. gooood.
Originally posted by costique
Well, not exactly, IMO. Remember Sherlock indexing? It takes forever, consumes a hell of resources and saves huge index files. I use content indexing/searching on my /Developer/Documentation folder and a folder containing pdf and html books and, man, it's ugly, slow and sometimes cannot find things because the index hasn't been updated.
Ever notice that happening now? No?
Guess what - it's still indexed.
Been fixed a long time now... a one-time long index, then background updates and file notifications (remember, that's now robust in the OS and the Finder (and other apps, say, an indexer) can be written to respond to them) happen on the fly when you're not doing much else.
It'll be fast, trust me.