If Apple, Google, Mozilla, or anybody else thinks I'm going to upload my Porn fiction and store then as "web docs" online, they can just think again. The last thing I need is to have that crap out in the open where any FBI agent or Employer Detective can find them.
"Take me, Employer Detective Magnum Chandler!" whispered FBI Special Agent Scully Pandora huskily from her C3P0 bedsheets to the unshaven silhouette lounging moodily in the doorway of the cheap New Mexico motel room, backlit by flashes of desert lightning and the headlights of cars - OR WERE THEY??? - passing along the Extraterrestrial Expressway. "Take me like you captured those illegal Martian H-1B immigrants restocking the photocopiers in Accounting! Take me like a cheap 1970s reprint of a 1926 Hugo Gernsback's Amazing Stories featuring 'Off On A Comet' by Jules Verne! Take me like all the contents of the wellhouse at the end of the road in Colossal Cave!"
"Grarrrrough!" replied the Employer Detective, stepping into the light of the swinging single lampshade and the FBI Special Agent screamed as she saw that he was not only a Vampire but now a Zombie Werewolf Pirate Ninja as well! and also possibly a dinosaur.
And yet, the alternatives are all so, so much worse.
I see your Outlook and raise you.... Novell Groupwise.
It couldn't give you a list of pending appointments either, except via the Calendar view. And if you clicked on the default weekly Calendar view, you had no way of moving backwards or forwards in time, because they hadn't thought to put a button there. You had to close the whole Calendar view and reopen in the Monthly view. That window was created around 1996, but they never updated it. User-configurable? Good lord, no. Why give the users anything? They'll just ask for more.
And creating a recurring appointment was a whole new adventure. There was that cute little pseudo-English text box where you could say 'every second Thursday' and hope it understood you... or you could click on every. single. day...
Good times, good times.
I swear, there's some dark cosmic law of the universe that says every groupware package must suck to preserve the Balance.
no one is more pro-Microsoft and anti-user-choice than IT. It requires buy in from everyone so that they don't go and create some automation script that requires others to use an Outlook form.
It's not actually the case that IT departments are by definition pro-Microsoft, but they are very often pro-single-source provider. If you want to know why, your next sentence just answered your own strawman.
The entire purpose of an IT department is to automate rollout and delivery of services, to create a unified corporate 'look and feel', and generally to make life simple and happy and productive for all the actual value-generating staff of the organisation. Anything that gets in the way of that reduces value for the company, and since IT is a cost centre to start with... well, we don't get paid just to play with shiny toys. We play with shiny toys so that other people can work with the shiny toys. Sometimes that means taking some of the shiny off in order to save users from themselves.
And yes, we deploy a lot of automation scripts. Guess what? If your proposed alternative 'solution' to a widely deployed, fully vendor-supported, must-have-it-or-the-enterprise-breaks-instantly product is so different that it doesn't even respond to the same scripts... which means all our infrastructure has to be duplicated, triplicated, tested, multi-tested, with tens to hundreds of man-hours per script... and that's even assuming that it a) provides all the required organisational features and interoperate exactly...
So the organisation instantly suffers breakage of automation, increased user training, more IT hours burned, more support calls filed with vendors, and what, exactly, does the organisation gain by doing this? It would have to be some really big, must-have innovative feature which the current core system doesn't have, and which isn't available as an extension.
Look, on my desk right now I have Internet Explorer and Firefox. It's my job to keep both of them patched and up to date with the latest security patches. How hard to you reckon that ought to be?
Well, with IE8, I just go to the WSUS server, triage the latest Microsoft patches as they come in on Black Tuesday, stage them through our testing groups, and it's done. If there are any security announcements that don't have patches, we can use Group Policy to instantly block access to websites, turn off features, set security levels, etc. One click in Active Directory and the whole site is secure.
Firefox? Well now, that's an entirely different story. One, it doesn't support Group Policy configuration, so we have to write a script to read the prefs.js, edit it to set the variables we want, and yet preserve the users' existing settings. And make sure we do this at a time when the user doesn't have the file locked open, but the file is still accessible. And make sure the user profile gets copied correctly in the right order, because it's all just files, but it's cached locally or remotely depending on.... and then, actual updates? Well, again, there's no enterprise deployment mechanism for patch updates, so I have to write a custom script to somehow prevent the user from starting Firefox while we run the install, decode the silly salt directory (a 'security feature' intended to block admins from doing exactly what we need to do), preserve the user's preferences and bookmarks and plugins, then run the install... then figure out which plugins the user has installed and run the installers for those, without letting the installers do what they want to do natively which is hit the Net and do their own installing...
Firefox seems to always think it's running on a home user's machine, and thinks it owns the place. Internet Explorer is aware that it's likely running on a corporate desktop, and plays nice. There's a world of difference between the 'feel' of the two.
And I *like* Firefox, and I miss not having it as my standard browser, and I run Ubuntu at home, but work demands i
And there are other collaborative programs out there. I seriously doubt MS Exchange Server/Outlook is indispensable.
If you seriously doubt that, then you simply haven't worked in business.
I'm truly sorry, but yes, this is exactly the situation. It's nothing to do with being Microsoft; I'm as surprised as you are that MS managed to create a good product in Exchange. But they did, and now they have total enterprise collaboration dominance, and ignoring it won't help.
The sheer length of that list of would-be contenders is part of the problem. There's a million half-assed, incomplete, and mutually incompatible attempts at creating an Outlook/Exchange replacement. The fact that they don't interoperate is key to why they don't work.
What we need is not a product, but a protocol - in the same way that what unseated the 1980s Online Services oligopoly of Compuserve, GEnie, BIX et al wasn't a competitor, but HTTP and SMTP. But sadly, what we have for protocols in the collaboration space we can count on one hand: iCalendar, CalDAV, SyncML. They don't entirely interoperate, and they certainly don't specify all of the solution space.
I'm baffled as to why this is so. But probably, Google not being at all interested in the desktop and preferring to rent web applications, and Apple having their own tidy lock-in game on the iPhone/iPad would explain why there's been no frontal challenge from them. And IBM has been committed to Lotus Notes, which was a good idea in theory but... and Novell had Groupwise which, just, ouch, and now they're toast. And Sun, I mean Oracle, never even tried.
But the open source guys, I was sure they were going to crack it, but nope. Beats me why not.
Apparently not, and not for want of trying. I've been watching the OSS world for, what, about 14 years now? And nothing has come even close to touching Outlook/Exchange. We've had Evolution. We've had Chandler. We've had iCalendar, CalDAV, SyncML. Nothing's filled the whole solution space.
To this day I do not comprehend how come email got world standardised via SMTP in 1982 (okay, with glaring security holes like the Sender: field, but still, interoperable), yet calendaring and contacts is still impossible to interoperate except through one defacto-standard system. Seriously, you have some kind of database, some kind of syncing system, a schema of object types... it's not rocket science, right? It could have been sorted around 1983 at the latest?
And yet... we just migrated from Groupwise to Outlook/Exchange at work, and I can tell you, it's been a huge step up.
Having java installed doesn't magically put you in danger. The only place where java increases your attack profile is when its running as a browser plug in, and that is easy to disable without removing java.
Now, if you just don't know how to do it, and refuse to learn, that's a different issue entirely.
There's 'don't know how', and then there's 'your system will helpfully automagically revert to an insecure configuration every time you apply a purported security upgrade'.
Yes, you can bash your will into many things if you hit them with blunt system configuration editors hard enough, but life is generally easier when the tools you use don't actively work against you.
On top if it, if you read about the main vectors for malware, you'll see java vulnerabilities top the list.
Indeed. Isn't it funny how when Java first launched, its big promise was 'no potential for malware because it runs in a sandbox'?
And that was way before the big malware wave. 1996 was such a kinder, more innocent time. Six years before SQL Slammer and three before even Melissa, and Java's security seemed too paranoid.
I suppose one could just as easily ask, who launches icons from the windowing GUI using the keyboard, but not using Spotlight? Who does it so often that a key combination rather than a keypress slows them down?
Interesting you should say that.
It's only my anecdotal observation, but whenever I compare the Windows to Mac users I know, I always get the impression that the Mac users are very keyboard-shy compared to the Windows ones, and less efficient.
The Windows users will use a very fluid mix of mouse and keyboard gestures, but the Mac users tend to use mouse-only gestures, and generally take a much longer time to get anything done. In fact it's painful for me to sit behind a Mac user and watch them deliberately left-click, drag, move, for about ten seconds when it would take about two accelerator keys. I see their hands all bunched up in claws from the intensity of extended delicate mousing and the muscle concentration it requires and think 'that can't be healthy'.
Maybe I just have a poor selection of non-proficient Mac users to observe? But they seem to be the 'power user' type to me.
mac hasn't been "for design" since os9, maybe os8, now it's for college students, aspiring writers, and people who worry more about what other people think about their computer then how useful the computer is.
Cough. Tell that to all the designers and design schools I know. Macs are only for design in the IT world I'm aware of. How's life in yours?
Don't kid yourself about OSX. You may like it, but it has it's own share of UI disasters.... Some, like having all of the menus at the top of the screen made sense when we were on low resolution single screen systems, but are detriments in multi-monitor high resolutions system.
I used to dislike the Apple menubar at the top of the screen, but since seeing Office 2011 on OSX just the other day, I had a sudden epiphany:
Taking the existence of the menu bar away from the control of applications means application designers can't remove it just because, like a cat, they decided to chase a shiny ribbon instead that day.
(I suppose Microsoft could have just pulled all the entries off the Office 2011 menubar and left it with just the 'Apple', but as it happened, they chose not to do that because it would have pointed out just how annoying they were.)
I so wish we could enforce something like this on Windows.
I despise programs that rearrange your files for you, make ridiculous subdirs w/out permissions, etc.
++this, though I don't actually mind iTunes (actually I have Rhythmbox on Ubuntu for my main music store, and use iTunes on Windows 7 just for purchasing new singles). I'm sorta okay with the 'Artist/Album' structure, but it sometimes makes organising albums of songs by multiple artists annoying.
But for photos... I had a very bad experience with F-Spot wanting to take over the process of importing - and then losing - all my photos. Shotwell seems better behaved. Every few weeks I sync my camera - I do it manually, creating a folder under 'photos' in the form yyyy-mm-dd, copy all the photos since last time, then import that folder into Shotwell. Never ever do I let a dedicated photo management application try to put stuff anywhere. That's what the filesystem is for, a reader application's job is to read it, not to write it.
I don't like the proliferation of all these 'organiser' apps. It seems to be fundamentally wrongheaded to me. If I want to organise my files, that should be a function of the filesystem, not an app.
I wish an operating system had a concept of 'query folders' where I could have honest-to-goodness filesystem folders, indistinguishable from real ones, where membership and even name was the result of a script. Then I could create as many such folders as I wanted to organise my stuff, and let my apps get on with editing and displaying and NOT organising. That way I could use whatever app I wanted to do whatever I wanted and not be in the position of 'darn, I renamed my external hard drive and now Rhythmbox can't find anything'.
You can link a file into as many directories as you like with the 'ln' command.
Well, sort of. You can create a hardlink or symlink in the Posix model easily enough, certainly. But the link is only one way - you can't easily find, given a file, where all its links are. So they can tend to get caught up in the bit-rot. And there's enough of a stigma around symlinking - let alone hard-linking - that very few tools can be relied on to support it in all cases.
A true tagged or non-exclusive-directory filesystem would, I assume, have proper two-way linking between a file and and its links, so you could query a file and get a list of its tags/locations. And all the tools, without exception, would fully support it. This would include things like copying a 'folder' to removable media - you would need to standardise what it means. You can't just copy the links and you can't just turn the links into unlinked files.
What you could do, perhaps, is store all the originals (including folders) in a single universal folder as a globally-unique identifier (it can't be just system-unique, because what if you copy a file to someone else's machine?), then make the other folders on a system contain only hardlinks, and have the file-and-folder copy algorithm copy both a subset of the originals folder and all the appropriate tag folders...
It gets messy, is what happens, because things like disk drives fundamentally have a notion of containment (my file is either on this disk or it's not, it doesn't help if it's 'virtually somewhere out there in the cloud' once I've pulled the network plug) while tags don't. I'm sure we could solve these problems, but they need to be solved correctly and with mathematical rigor at the lowest layer of the filesystem. I don't see any serious attempts to do that in any of the tagged filesystem approaches I've seen yet.
Sources in the underworld tell me - and this is just an unsubstantiated rumour, so please don't freak out, but I think it's worth repeating as an advisory - that the Conspiracy may now have not just an agenda but minutes.
Please, try to remain calm. They don't yet have Daytimers. For now.
To over-stretch the military analogy, they're much more the light infantry than the intelligence corps.
I would have thought 'the crazed LSD freak tossing live frag grenades into the barracks at random between shooting own toes with shotgun' would have been a more appropriate analogy. Has Anonymous ever actually done *anything* constructive and useful, or do they just create mayhem?
By the way, would you like to buy some quadruply-hedged derivatives? I'm told by the best authorities that they contain only wafer thin slices of subprime mortgage, you'd barely notice it.
An addendum. By 'the human mind is infinite', I literally mean that. One of the implications of this, and I think is the intuition that Searle and I share, is that the human mind is capable of almost arbitrary meta-cognition: we can choose to be conscious not just of our thoughts, but of our thoughts about our thoughts, and so on to what appears to be an unlimited degree. This is a feature that algorithms (at least ones implemented in Newtonian physics; quantum algorithms might be capable of some infinite regresses) don't seem to share; infinite regresses take infinite time or space to compute and eventually fail. Yet the human mind can grasp some very nastily recursive ideas almost instantaneously, and without falling prey to logic bombs - we don't consume infinite resources like a Turing machine would.
This capability for unlimited introspection, I think, is what Searle is getting at, though as a committed materialist he can't come out and just say it; he has to believe that the mind is a function of the brain; I don't. But the reason the Chinese Room 'works' is that it plays to this intuitive sense we all have of an infinitely unfolding 'interior life' of the mind which, no matter how closely we approximate it in algorithms, never is exactly the same as the approximation, because it just keeps going on and on indefinitely.
This is also why we respond intuitively negatively to machine-images of society like the Borg- we sense at some level that as far as we are minds we are not machines, because machines are a crude approximation of the vast mental spaces within us. We sense, if we can't measure, that we are big inside, that our thoughts-about-thoughts-about-thoughts, and in fact the much deeper unconscious realms of our psyche, are not at all reducible to language and symbols, and no matter how much we communicate of ourselves, there's always more inside that remains unsaid.
There may be ways to approximate this capability at varying levels of detail, but I'll go out on a limb and predict that what we'll find is that the approximations all tail off very quickly long before they could meaningfully pass the Turing test with a determined interviewer. We could at some level approximate some elements of the human emotional matrix - but a real human always has the ability to bring more to the table than is already there, while an algorithmic AI will only be able to remix what's there in an obviously mechanical way. It will never quite feel correct, will always have something left out. Even in text chat, it will quickly become obvious that the interviewer is dealing with a simulation and not the full - infinitely large - human mind.
Slashdot Mirror
If Apple, Google, Mozilla, or anybody else thinks I'm going to upload my Porn fiction and store then as "web docs" online, they can just think again. The last thing I need is to have that crap out in the open where any FBI agent or Employer Detective can find them.
"Take me, Employer Detective Magnum Chandler!" whispered FBI Special Agent Scully Pandora huskily from her C3P0 bedsheets to the unshaven silhouette lounging moodily in the doorway of the cheap New Mexico motel room, backlit by flashes of desert lightning and the headlights of cars - OR WERE THEY??? - passing along the Extraterrestrial Expressway. "Take me like you captured those illegal Martian H-1B immigrants restocking the photocopiers in Accounting! Take me like a cheap 1970s reprint of a 1926 Hugo Gernsback's Amazing Stories featuring 'Off On A Comet' by Jules Verne! Take me like all the contents of the wellhouse at the end of the road in Colossal Cave!"
"Grarrrrough!" replied the Employer Detective, stepping into the light of the swinging single lampshade and the FBI Special Agent screamed as she saw that he was not only a Vampire but now a Zombie Werewolf Pirate Ninja as well! and also possibly a dinosaur.
yes, I can see the problem.
Outlook is terrible.
And yet, the alternatives are all so, so much worse.
I see your Outlook and raise you.... Novell Groupwise.
It couldn't give you a list of pending appointments either, except via the Calendar view. And if you clicked on the default weekly Calendar view, you had no way of moving backwards or forwards in time, because they hadn't thought to put a button there. You had to close the whole Calendar view and reopen in the Monthly view. That window was created around 1996, but they never updated it. User-configurable? Good lord, no. Why give the users anything? They'll just ask for more.
And creating a recurring appointment was a whole new adventure. There was that cute little pseudo-English text box where you could say 'every second Thursday' and hope it understood you... or you could click on every. single. day...
Good times, good times.
I swear, there's some dark cosmic law of the universe that says every groupware package must suck to preserve the Balance.
And why would an office suite have an email reader?
Why wouldn't it? Does your office get by without email?
no one is more pro-Microsoft and anti-user-choice than IT. It requires buy in from everyone so that they don't go and create some automation script that requires others to use an Outlook form.
It's not actually the case that IT departments are by definition pro-Microsoft, but they are very often pro-single-source provider. If you want to know why, your next sentence just answered your own strawman.
The entire purpose of an IT department is to automate rollout and delivery of services, to create a unified corporate 'look and feel', and generally to make life simple and happy and productive for all the actual value-generating staff of the organisation. Anything that gets in the way of that reduces value for the company, and since IT is a cost centre to start with... well, we don't get paid just to play with shiny toys. We play with shiny toys so that other people can work with the shiny toys. Sometimes that means taking some of the shiny off in order to save users from themselves.
And yes, we deploy a lot of automation scripts. Guess what? If your proposed alternative 'solution' to a widely deployed, fully vendor-supported, must-have-it-or-the-enterprise-breaks-instantly product is so different that it doesn't even respond to the same scripts... which means all our infrastructure has to be duplicated, triplicated, tested, multi-tested, with tens to hundreds of man-hours per script... and that's even assuming that it a) provides all the required organisational features and interoperate exactly...
So the organisation instantly suffers breakage of automation, increased user training, more IT hours burned, more support calls filed with vendors, and what, exactly, does the organisation gain by doing this? It would have to be some really big, must-have innovative feature which the current core system doesn't have, and which isn't available as an extension.
Look, on my desk right now I have Internet Explorer and Firefox. It's my job to keep both of them patched and up to date with the latest security patches. How hard to you reckon that ought to be?
Well, with IE8, I just go to the WSUS server, triage the latest Microsoft patches as they come in on Black Tuesday, stage them through our testing groups, and it's done. If there are any security announcements that don't have patches, we can use Group Policy to instantly block access to websites, turn off features, set security levels, etc. One click in Active Directory and the whole site is secure.
Firefox? Well now, that's an entirely different story. One, it doesn't support Group Policy configuration, so we have to write a script to read the prefs.js, edit it to set the variables we want, and yet preserve the users' existing settings. And make sure we do this at a time when the user doesn't have the file locked open, but the file is still accessible. And make sure the user profile gets copied correctly in the right order, because it's all just files, but it's cached locally or remotely depending on.... and then, actual updates? Well, again, there's no enterprise deployment mechanism for patch updates, so I have to write a custom script to somehow prevent the user from starting Firefox while we run the install, decode the silly salt directory (a 'security feature' intended to block admins from doing exactly what we need to do), preserve the user's preferences and bookmarks and plugins, then run the install... then figure out which plugins the user has installed and run the installers for those, without letting the installers do what they want to do natively which is hit the Net and do their own installing...
Firefox seems to always think it's running on a home user's machine, and thinks it owns the place. Internet Explorer is aware that it's likely running on a corporate desktop, and plays nice. There's a world of difference between the 'feel' of the two.
And I *like* Firefox, and I miss not having it as my standard browser, and I run Ubuntu at home, but work demands i
And there are other collaborative programs out there. I seriously doubt MS Exchange Server/Outlook is indispensable.
If you seriously doubt that, then you simply haven't worked in business.
I'm truly sorry, but yes, this is exactly the situation. It's nothing to do with being Microsoft; I'm as surprised as you are that MS managed to create a good product in Exchange. But they did, and now they have total enterprise collaboration dominance, and ignoring it won't help.
The sheer length of that list of would-be contenders is part of the problem. There's a million half-assed, incomplete, and mutually incompatible attempts at creating an Outlook/Exchange replacement. The fact that they don't interoperate is key to why they don't work.
What we need is not a product, but a protocol - in the same way that what unseated the 1980s Online Services oligopoly of Compuserve, GEnie, BIX et al wasn't a competitor, but HTTP and SMTP. But sadly, what we have for protocols in the collaboration space we can count on one hand: iCalendar, CalDAV, SyncML. They don't entirely interoperate, and they certainly don't specify all of the solution space.
I'm baffled as to why this is so. But probably, Google not being at all interested in the desktop and preferring to rent web applications, and Apple having their own tidy lock-in game on the iPhone/iPad would explain why there's been no frontal challenge from them. And IBM has been committed to Lotus Notes, which was a good idea in theory but... and Novell had Groupwise which, just, ouch, and now they're toast. And Sun, I mean Oracle, never even tried.
But the open source guys, I was sure they were going to crack it, but nope. Beats me why not.
And no other software does what Outlook does?
Apparently not, and not for want of trying. I've been watching the OSS world for, what, about 14 years now? And nothing has come even close to touching Outlook/Exchange. We've had Evolution. We've had Chandler. We've had iCalendar, CalDAV, SyncML. Nothing's filled the whole solution space.
To this day I do not comprehend how come email got world standardised via SMTP in 1982 (okay, with glaring security holes like the Sender: field, but still, interoperable), yet calendaring and contacts is still impossible to interoperate except through one defacto-standard system. Seriously, you have some kind of database, some kind of syncing system, a schema of object types... it's not rocket science, right? It could have been sorted around 1983 at the latest?
And yet... we just migrated from Groupwise to Outlook/Exchange at work, and I can tell you, it's been a huge step up.
Having java installed doesn't magically put you in danger. The only place where java increases your attack profile is when its running as a browser plug in, and that is easy to disable without removing java.
Now, if you just don't know how to do it, and refuse to learn, that's a different issue entirely.
There's 'don't know how', and then there's 'your system will helpfully automagically revert to an insecure configuration every time you apply a purported security upgrade'.
Yes, you can bash your will into many things if you hit them with blunt system configuration editors hard enough, but life is generally easier when the tools you use don't actively work against you.
On top if it, if you read about the main vectors for malware, you'll see java vulnerabilities top the list.
Indeed. Isn't it funny how when Java first launched, its big promise was 'no potential for malware because it runs in a sandbox'?
And that was way before the big malware wave. 1996 was such a kinder, more innocent time. Six years before SQL Slammer and three before even Melissa, and Java's security seemed too paranoid.
Sigh. I don't really miss those years, but...
or using the same UI target for the trashcan and eject.
I have no idea what you're talking about.
Aww, so you're so new to Macs you don't even remember that one? So cute!
I suppose one could just as easily ask, who launches icons from the windowing GUI using the keyboard, but not using Spotlight? Who does it so often that a key combination rather than a keypress slows them down?
Interesting you should say that.
It's only my anecdotal observation, but whenever I compare the Windows to Mac users I know, I always get the impression that the Mac users are very keyboard-shy compared to the Windows ones, and less efficient.
The Windows users will use a very fluid mix of mouse and keyboard gestures, but the Mac users tend to use mouse-only gestures, and generally take a much longer time to get anything done. In fact it's painful for me to sit behind a Mac user and watch them deliberately left-click, drag, move, for about ten seconds when it would take about two accelerator keys. I see their hands all bunched up in claws from the intensity of extended delicate mousing and the muscle concentration it requires and think 'that can't be healthy'.
Maybe I just have a poor selection of non-proficient Mac users to observe? But they seem to be the 'power user' type to me.
mac hasn't been "for design" since os9, maybe os8, now it's for college students, aspiring writers, and people who worry more about what other people think about their computer then how useful the computer is.
Cough. Tell that to all the designers and design schools I know. Macs are only for design in the IT world I'm aware of. How's life in yours?
Don't kid yourself about OSX. You may like it, but it has it's own share of UI disasters. ... Some, like having all of the menus at the top of the screen made sense when we were on low resolution single screen systems, but are detriments in multi-monitor high resolutions system.
I used to dislike the Apple menubar at the top of the screen, but since seeing Office 2011 on OSX just the other day, I had a sudden epiphany:
Taking the existence of the menu bar away from the control of applications means application designers can't remove it just because, like a cat, they decided to chase a shiny ribbon instead that day.
(I suppose Microsoft could have just pulled all the entries off the Office 2011 menubar and left it with just the 'Apple', but as it happened, they chose not to do that because it would have pointed out just how annoying they were.)
I so wish we could enforce something like this on Windows.
Its cool, dude. I have no friends here.
Only foes you have already met?
I downloaded a copy of that as a kid
I feel old now.
Nah, he probably pulled it off Community Memory.
I despise programs that rearrange your files for you, make ridiculous subdirs w/out permissions, etc.
++this, though I don't actually mind iTunes (actually I have Rhythmbox on Ubuntu for my main music store, and use iTunes on Windows 7 just for purchasing new singles). I'm sorta okay with the 'Artist/Album' structure, but it sometimes makes organising albums of songs by multiple artists annoying.
But for photos... I had a very bad experience with F-Spot wanting to take over the process of importing - and then losing - all my photos. Shotwell seems better behaved. Every few weeks I sync my camera - I do it manually, creating a folder under 'photos' in the form yyyy-mm-dd, copy all the photos since last time, then import that folder into Shotwell. Never ever do I let a dedicated photo management application try to put stuff anywhere. That's what the filesystem is for, a reader application's job is to read it, not to write it.
I don't like the proliferation of all these 'organiser' apps. It seems to be fundamentally wrongheaded to me. If I want to organise my files, that should be a function of the filesystem, not an app.
I wish an operating system had a concept of 'query folders' where I could have honest-to-goodness filesystem folders, indistinguishable from real ones, where membership and even name was the result of a script. Then I could create as many such folders as I wanted to organise my stuff, and let my apps get on with editing and displaying and NOT organising. That way I could use whatever app I wanted to do whatever I wanted and not be in the position of 'darn, I renamed my external hard drive and now Rhythmbox can't find anything'.
You can link a file into as many directories as you like with the 'ln' command.
Well, sort of. You can create a hardlink or symlink in the Posix model easily enough, certainly. But the link is only one way - you can't easily find, given a file, where all its links are. So they can tend to get caught up in the bit-rot. And there's enough of a stigma around symlinking - let alone hard-linking - that very few tools can be relied on to support it in all cases.
A true tagged or non-exclusive-directory filesystem would, I assume, have proper two-way linking between a file and and its links, so you could query a file and get a list of its tags/locations. And all the tools, without exception, would fully support it. This would include things like copying a 'folder' to removable media - you would need to standardise what it means. You can't just copy the links and you can't just turn the links into unlinked files.
What you could do, perhaps, is store all the originals (including folders) in a single universal folder as a globally-unique identifier (it can't be just system-unique, because what if you copy a file to someone else's machine?), then make the other folders on a system contain only hardlinks, and have the file-and-folder copy algorithm copy both a subset of the originals folder and all the appropriate tag folders...
It gets messy, is what happens, because things like disk drives fundamentally have a notion of containment (my file is either on this disk or it's not, it doesn't help if it's 'virtually somewhere out there in the cloud' once I've pulled the network plug) while tags don't. I'm sure we could solve these problems, but they need to be solved correctly and with mathematical rigor at the lowest layer of the filesystem. I don't see any serious attempts to do that in any of the tagged filesystem approaches I've seen yet.
Last night, a member of cat macro group Anonymous
Better?
a shadowy cabal of leaders with an agenda.
Sources in the underworld tell me - and this is just an unsubstantiated rumour, so please don't freak out, but I think it's worth repeating as an advisory - that the Conspiracy may now have not just an agenda but minutes.
Please, try to remain calm. They don't yet have Daytimers. For now.
To over-stretch the military analogy, they're much more the light infantry than the intelligence corps.
I would have thought 'the crazed LSD freak tossing live frag grenades into the barracks at random between shooting own toes with shotgun' would have been a more appropriate analogy. Has Anonymous ever actually done *anything* constructive and useful, or do they just create mayhem?
trivial to hedge with a derivitive these days.
Good joke, sir!
By the way, would you like to buy some quadruply-hedged derivatives? I'm told by the best authorities that they contain only wafer thin slices of subprime mortgage, you'd barely notice it.
your desire to use your kitteh to login to your computer.
Shoot, the FBI is knocking on my door. Welp, time to napalm the security authentication token... uh oh...
I CAN NOT HAZ DESTRUCTABLE!
Yes, but it will be very -unstable. Like, it won't even start booting.
And if it does, the universe may be destroyed in every way possible (and a few which are essentially impossible).
Kirk, climbing a mountain.
Likewise, we don't need all of the functions of natural brains in an artificial brain.
But do we even know which functions of a natural brain we need?
I think we don't.
An addendum. By 'the human mind is infinite', I literally mean that. One of the implications of this, and I think is the intuition that Searle and I share, is that the human mind is capable of almost arbitrary meta-cognition: we can choose to be conscious not just of our thoughts, but of our thoughts about our thoughts, and so on to what appears to be an unlimited degree. This is a feature that algorithms (at least ones implemented in Newtonian physics; quantum algorithms might be capable of some infinite regresses) don't seem to share; infinite regresses take infinite time or space to compute and eventually fail. Yet the human mind can grasp some very nastily recursive ideas almost instantaneously, and without falling prey to logic bombs - we don't consume infinite resources like a Turing machine would.
This capability for unlimited introspection, I think, is what Searle is getting at, though as a committed materialist he can't come out and just say it; he has to believe that the mind is a function of the brain; I don't. But the reason the Chinese Room 'works' is that it plays to this intuitive sense we all have of an infinitely unfolding 'interior life' of the mind which, no matter how closely we approximate it in algorithms, never is exactly the same as the approximation, because it just keeps going on and on indefinitely.
This is also why we respond intuitively negatively to machine-images of society like the Borg- we sense at some level that as far as we are minds we are not machines, because machines are a crude approximation of the vast mental spaces within us. We sense, if we can't measure, that we are big inside, that our thoughts-about-thoughts-about-thoughts, and in fact the much deeper unconscious realms of our psyche, are not at all reducible to language and symbols, and no matter how much we communicate of ourselves, there's always more inside that remains unsaid.
There may be ways to approximate this capability at varying levels of detail, but I'll go out on a limb and predict that what we'll find is that the approximations all tail off very quickly long before they could meaningfully pass the Turing test with a determined interviewer. We could at some level approximate some elements of the human emotional matrix - but a real human always has the ability to bring more to the table than is already there, while an algorithmic AI will only be able to remix what's there in an obviously mechanical way. It will never quite feel correct, will always have something left out. Even in text chat, it will quickly become obvious that the interviewer is dealing with a simulation and not the full - infinitely large - human mind.
That's my prediction, anyway.