Powered by TagBoard
![[Powered by Blogger]](http://buttons.blogger.com/bloggerbutton1.gif){kind=small}
Wednesday, October 16, 2002
edit your blog:
While two reasons for MacOS X' glacial release progress - especially until the beginning of this year - have been the unexpected difficulty in marrying the Mac GUI and the NeXT/*nix underpinnings, and the parallel clashes between the "old" and the "new" program development culture, a third and less conspicious factor has added at least an equal amount of inertia to it: structuring the new OS in such a way that it was ready for all the new - "insanely great", if you will - capabilities that the inspired minds of Apple could dream up.
The Unix underpinnings of OSX have received a lot of attention for their stability and maturity, but less has been written about the flexibility inherent in its design. Unix was originally geex-country, the playfield of binary bravadoros, but became stale and standardized country when the desktops took over as the area of expansion and invention. Linux is where it's at nowadays, but the OSX mixture of glamour surface, accepted standard programs, and modularized libertarianism seems to be luring more and more of the post-nerds into the fold. (how's THAT for mixing metaphors?!) It was a major revolution when Apple released the HFS+ file system for their OS, but in OSX it is but one of the systems available to the system. Anyone who wants to (and has the skills to), can write a driver for a new file system and dump it into the relevant folder (/System/Library/Filesystems/), and then any volume using that standard of organization is transparent to the system.
This is a flexibility built into Unix as standard, but there was also an incredible number of hooks and nooks added to the original APIs, and we are only beginning now to see why they were put there. It was no PR-phrase when Mr. Jobs said about OSX that "We designed this OS to last at least 15 years". One of the initial phases of developing OSX was gathering together all the projects, ideas and suggestions that had accumulated in Apple's and (equally much) NeXT's years of turmoil. Some of them were were extensions of the basic ideas already found in the OS'es, such as the development of RendezVous - a TCP/IP superset - to give the user-friendliness that Appletalk was all about. Others were totally new ideas, dreamed up by laterally fertile minds but never implemented in any practical fashion; from the Quartz graphical layer to UI details such as Dialog Sheets and Drawers. Others again were new developments gathered from the "outside".
OSX 10.2 was, as I said in a previous BloQ, "strutting hooks and possibilities in all directions, like spines on a hedgehog". We have only seen the tiny start of program interoperability, as demonstrated in the interplay between iDisk and iPhoto, iChat and Mail, iSync and Address Book, iTunes and RendezVous.... Like Mr. Jobs likes to say, ".... there's one more thing" - Xcept there are many more than one.
I was finishing the second part of the "Building the Next OSX" series when the story broke about a new update of OSX being released within the next couple of weeks, incorporating the long awaited Journaling File System for Mac, something that I had been told would not happen until sometime next year, as part of the next major update, 10.3. That news, released by two of the most highly estimated Mac journalists, Matthew Rothenberg and Nick dePlume, will not only stir up a hornets' nest and steal most of the limelight that was expected to fall on the roll-out of IBM's plans for the future main PowerPC processor in Big Macs, it will also provide users with hitherto unthought possibilities.
For one, the ability to Undo on a Finder level - that is, cancel the most recent actions, even things like Move or Delete -; but more interestingly, the ability to assign almost any value to almost any file. By value, I mean meta-data. By meta-data, I mean things like the Labels provided in MacOS 7.5 onwards for color marking folders for urgency, or mood, or genre. Or, if you have tried the Gnutella-based program LimeWire, adding data such as Author, Title, Publication data, Length, Feel-Good value, etc to an outline text of a book. The meta-data could also contain details about which addresses it had been saved to, providing you with a quick overview of who in your company had seen it (or at least had had it in possession). As Asterix would say, the possibilities are legio... no wait, he would say, "All good things ..."
I will discuss meta-data further in my revised suggestions for Building the Next OSX, hopefully available tomorrow.
"... comes to those that wait ... and wait ..."
While two reasons for MacOS X' glacial release progress - especially until the beginning of this year - have been the unexpected difficulty in marrying the Mac GUI and the NeXT/*nix underpinnings, and the parallel clashes between the "old" and the "new" program development culture, a third and less conspicious factor has added at least an equal amount of inertia to it: structuring the new OS in such a way that it was ready for all the new - "insanely great", if you will - capabilities that the inspired minds of Apple could dream up.
The Unix underpinnings of OSX have received a lot of attention for their stability and maturity, but less has been written about the flexibility inherent in its design. Unix was originally geex-country, the playfield of binary bravadoros, but became stale and standardized country when the desktops took over as the area of expansion and invention. Linux is where it's at nowadays, but the OSX mixture of glamour surface, accepted standard programs, and modularized libertarianism seems to be luring more and more of the post-nerds into the fold. (how's THAT for mixing metaphors?!) It was a major revolution when Apple released the HFS+ file system for their OS, but in OSX it is but one of the systems available to the system. Anyone who wants to (and has the skills to), can write a driver for a new file system and dump it into the relevant folder (/System/Library/Filesystems/), and then any volume using that standard of organization is transparent to the system.
This is a flexibility built into Unix as standard, but there was also an incredible number of hooks and nooks added to the original APIs, and we are only beginning now to see why they were put there. It was no PR-phrase when Mr. Jobs said about OSX that "We designed this OS to last at least 15 years". One of the initial phases of developing OSX was gathering together all the projects, ideas and suggestions that had accumulated in Apple's and (equally much) NeXT's years of turmoil. Some of them were were extensions of the basic ideas already found in the OS'es, such as the development of RendezVous - a TCP/IP superset - to give the user-friendliness that Appletalk was all about. Others were totally new ideas, dreamed up by laterally fertile minds but never implemented in any practical fashion; from the Quartz graphical layer to UI details such as Dialog Sheets and Drawers. Others again were new developments gathered from the "outside".
OSX 10.2 was, as I said in a previous BloQ, "strutting hooks and possibilities in all directions, like spines on a hedgehog". We have only seen the tiny start of program interoperability, as demonstrated in the interplay between iDisk and iPhoto, iChat and Mail, iSync and Address Book, iTunes and RendezVous.... Like Mr. Jobs likes to say, ".... there's one more thing" - Xcept there are many more than one.
I was finishing the second part of the "Building the Next OSX" series when the story broke about a new update of OSX being released within the next couple of weeks, incorporating the long awaited Journaling File System for Mac, something that I had been told would not happen until sometime next year, as part of the next major update, 10.3. That news, released by two of the most highly estimated Mac journalists, Matthew Rothenberg and Nick dePlume, will not only stir up a hornets' nest and steal most of the limelight that was expected to fall on the roll-out of IBM's plans for the future main PowerPC processor in Big Macs, it will also provide users with hitherto unthought possibilities.
For one, the ability to Undo on a Finder level - that is, cancel the most recent actions, even things like Move or Delete -; but more interestingly, the ability to assign almost any value to almost any file. By value, I mean meta-data. By meta-data, I mean things like the Labels provided in MacOS 7.5 onwards for color marking folders for urgency, or mood, or genre. Or, if you have tried the Gnutella-based program LimeWire, adding data such as Author, Title, Publication data, Length, Feel-Good value, etc to an outline text of a book. The meta-data could also contain details about which addresses it had been saved to, providing you with a quick overview of who in your company had seen it (or at least had had it in possession). As Asterix would say, the possibilities are legio... no wait, he would say, "All good things ..."
I will discuss meta-data further in my revised suggestions for Building the Next OSX, hopefully available tomorrow.
