I've played with a N900. Why? Cuz my friend has one. He has one because he works for Nokia.
Here's the problem.
Everything else you just said, means little to me because I'm not even sure why this is a plus.
Repositories? Add them myself? Where do I go find these? How do I know these apps won't screw me? (Granted, Android has this problem too)
It's Fedora/Debian + some UI? This means what for me exactly? I know that Fedora and Debian are Linux distros, and I've installed both of them as well as Gentoo and Ubuntu on a PC of mine at some point in time. But I don't see why I should care. I'm not even sure about what's the difference between Fedora and Debian except that Fedora took up much more space on my PC.
If these are all the N900 has going for it, it's not hard to see why it's not selling. It's because nobody cares about this stuff.
With it's meager market share, I'd say the market says that the N900 doesn't compare favorably. (in the US; I dunno about Europe or elsewhere)
It's huge (though, some people don't mind that) 3G in the US only works on T-mobile. It's probably got as many well-polished apps in Ovi as the Palm Pre.
I want a N900 because it seems like it might be kinda cool to have, but I don't think I'm actually going to buy the N900. I'd probably get an Android phone first because I'd be more likely to be able to do more with it.
I learned C first. I learned Java in college and programmed in that for years. Then I learned Objective-C. Then I tried the Android SDK. Then I tried the iPhone SDK. Then I just tried the Android SDK again.
My conclusions as of last week (when I tried the Android SDK) 1) I prefer Obj-C over Java for the flexibility and readability. It's just simply easier and at least as powerful. I get stuff done faster in Obj-C despite having more years of Java experience. 2) Xcode has bugs. Gee, what a surprise. 3) iPhone SDK is really easy to use for a mobile dev platform. 4) iPhone SDK provisioning (loading a cert to be able to sign and run on a device is a pain in the ass, but they just fixed that recently) 5) The Android SDK / Eclipse feels like being thrown back into the stone age.
All in all, I'd so much rather program in Obj-C than Java. And the Android SDK needs some serious help in order to compete with iPhone. I'd even be inclined to say I'd probably rather even use older conventions to program for OpenStep/GnuStep when it comes to multiplatform GUI apps, but honestly, I haven't tried GnuStep yet. So I can't be sure.
Sure, using Java is nice to have crossplatform compatibility from the start. And I'm definitely inclined to do crossplatform dev because I used to use Windows, OSX, Linux and Solaris. But I've found that getting something built faster and tested is more important than getting something built for a bunch of platforms.
And might I add in that ARM effectively has some of the code compression ability of x86 via Thumb without the need for such high complexity during instruction decode.
UI design is no fun for us because we typically suck at it.
I've seen many people who's idea of UI design is to sketch up a mock window in photoshop with pretty buttons and then stop. That's about the extent of most UI "design" these days.
I've worked with what I would consider my example of an actual professional UI designer. He'd start with the sketch up and a storyboard, then start annotating everything about the workflow. "x button is here, leads to y. defaults prefs to z because between the two most likely user workflow scenerios, the majority would benefit from ijk." "exploring user doing task A. We expect this to happen frequently, so let's make this fast. Note from usability test: this wasn't good enough, users had a hard time understanding what just happened" "this area seems cluttered, revisit for rev2."
All sorts of things. The stuff he did made sense to me and I could easily how he reasoned things out. It wasn't about the pretty. It was about a smooth, annoyance-free experience for the user. If I started implementing a design and had a question or a complaint, it was actually fun to go converse with him about what he thought about it. That he was open to suggestion and actually took suggestions made it all that much worthwhile.
I'd say that Google probably drive away people because of their use of FOSS because they don't advocate that you use FOSS nor even give the average user any obvious indication of their use of FOSS. But then again, there's plenty of other reasons Google could be driving people away that are much more meaningful to the user population.
The response from the author in the comments mention that when they say "free" they meant the common use of "free", not "GNU/Free". This makes sense because their readers are most likely not familiar with "free" as in "GNU/Free".
I hate the "Your real problem is that Gimp doesn't work like Photoshop" answer.
Many of us here acknowledge that Photoshop and GIMP have different UIs. Why? That's uncontestable. They're different apps. They're made by different teams. They're different.
We acknowledge that both apps have good parts about their UIs and bad parts about UIs. It's not that Photoshop is the golden standard or GIMP's just a different way of doing things. It's that for a given task, there's a lot of things that make sense for a larger majority of users.
In my usage of Photoshop, it DOES do exactly what you said: namely "use the last action with the tool in question as it's current setting". If using the last set options as default is not what you meant, then perhaps a clarification is necessary. I've left unusual options on and forgotten about them until they bit me and realized, "oh yeah, I haven't used this tool since xyz and I used some weird ass feature that I left on". So if you checked a box for aspect ratio locking and left it that way, it'll stay that way until next time. I agree it'd be a retarded default to have on, but nevertheless, I too can guarantee you that it is not an application default to have that on. Simply, the user last set it that way. Only God knows how many countless times I've ended up drawing with a ridiculously wrong brush because I forgot I set it to something different.
As for the transparency example in GIMP. Maybe there's something that I'm missing, but to me, if I set something to be transparent. It's pretty logical that I want it transparent. So it should stay transparent. It's nice to have an option to make it untransparent, but that doesn't change the fact that if I set something transparent, it's pretty safe to assume I want it transparent.
This isn't supposed to be a "The Photoshop Way" versus "The GIMP Way" argument. There's been plenty of UI fuckups in Photoshop itself, and we all know that. The point is, for every UI complaint, there is an acceptable solution: think it through. It should never "well, photoshop does it that way, we should do it that way." It should always be, "okay, what do I want, what am I doing, and what does that imply, and how can I make this easier for the user?" For the dragging problem about, this would be "well, I have a transparency value, and I have a preference for opaque-during-drag" which is by default off." Why? Because, it is more likely that a user would want the transparent image dragged transparently, as opposed to opaquely, since being transparent makes it easier to see what you're dragging across and see how it'll low when you let the mouse button go. That makes sense. It's not the right answer for the moment because that's what Photoshop is doing, it's the right answer for the moment because it make sense and makes it less likely that the majority of users will need to hunt for some odd preference.
Point is: Every complaint of "unintuitive" should never be answered with "well, that's cuz you're accustomed to using photoshop". It's an opportunity for improvement or for educating a user why it's the better way. The GIMP developers should take the time to listen to the complaints and revise those workflows, do stuff to make things easier, make the application help the user more. It should never be dismissed as wanting it to be like Photoshop. Because copying Photoshop doesn't help the user unless for a given task, it really is the best way. And we know that Photoshop is far from the best way.
Having a guy who sounds like he enjoys making power supplies blow up and understands how to use an oscilloscope is definetely a good reviewer. Having the guy equipped with a instrumented load is even better.
Wow, it just occurred to me it's an AT (not ATX) power supply that you're pointing at. Either way, it seems the reviews for some of their newer ATX offerings seem good, so I'll keep them in mind for future home-built PCs.
ZFS. As in Sun's ZFS file system. As I see more and more drives dying at work, I've been considering building a file server a home with redundancy and checksumming.
Whoa, and how does your post not make you sound like an arrogant prick?
Yeah, there's plenty of fans in both the Android and Apple camps that should keep their mouths shut. That's obvious. But it's not helping your case to respond like that to a poster who actually reasons out his side of the argument.
Because while open source allows people to inspect the source and understand what it does, that assumes somebody actually is trying to inspect and understand the source. Open source isn't a magical solution to make the community care about actually doing the work to improve the project. But as we see, eventually somebody noticed.
So it seems most of the community would rather use the phone than read the code. Not that surprising, isn't it?
What I find most interesting is that the security researcher's test app was a benign test for a rootkit bootstrap. To me, his results seems to imply: 1) At least 200 people don't read, understand, or care about the permissions warnings, in favor of "ooh, Twilight! *swoon*" 2) If the rootkit bootstrap worked, the app removal process depends on Google asking a rootkit-compromised phone to be trustworthy enough to remove the app. Isn't that kinda silly?
Not using credit cards is commendable. Not using a debit card is moronic.
Just curious, in all seriousness, I thought it was the other way around? Given the rewards points that credit cards give, it seems wiser to always use the credit card to get rewards points and pay it off in full every month? This way you benefit from: 1) a 1 month interest-free buffer for managing your savings (high interest) to checking account (low interest) flow. 2) rewards from your credit card 3) being able to file a claim if somebody screws with your card
I'm not sure in what situation I would actually want to use a debit card?
But the reason is that a normal iPhone app can't. Call history is not available to apps. Ever. So there's no warning. SMS history and ability to send/recieve/edit messages are also not available. Existing mail and mail accounts are not available.
Whether you think any of these should be accessible is a different topic entirely.
As far as I know, what is available is the following: 1) Address book access. 2) Location (a permission dialog appears when this info is requested, so even if you don't want to allow the app to know, you're supposed to still be able to use the rest of the app) 3) a sandboxed folder, essentially a per-app home directory. No app can access outside that folder except... 4)... the camera roll folder.
Actually... they were quite a formidable semiconductor company.
For all those PowerPC chips, Apple designed northbridges. Uni-North (Uni-N) is family featured in the G4s. U3 and U4 is used in G5s (http://en.wikipedia.org/wiki/PowerPC_970).
"Not to their idiot pandering gormless shareholders, but to their audience."
Um... it seems like their audience freakin' LOVES them.
Yea, there's plenty of evidence that there are apps some people want on the app store and it got rejected for some reason. The news gets everywhere every time it happens, like all other Apple news. Even Verizon's commercials make mention of this. I'm fairly certain most people know about the curated computing and limited app possibilities on the platform.
But, I find it hard to believe that if their business model is as bad as/. makes it out to be, they couldn't possibly be selling an iPad every 3 seconds.
You're correct that on the iPhone, files do have names. After all, it's a unix box underneath. But well-designed apps are supposed to abstract that out so that while we as programmers know there's a filename, the user doesn't need to care.
Take the simple notes app for example; I don't know how it's built, but if I were to guess, it's probably a bunch of files. Instead of creating a note and giving it a file name, you hit the plus button in the corner, and a note sheet pops up. The title of the note is the first line of the note, not the file name.
If you look at the photo browser, you don't see the file names either even though they're coming from the DCIM directory like any other camera. Instead, you see small thumbnail or the picture. Like a camera, you're going to be interacting with it as "next photo", "previous photo", or "pick one from this sheet". The file name doesn't come into play for the user because, well, it's not like "IMG0010.JPG" is going to actually give meaning.
If it one day does mean anything, then fine, publish a photo browser app that displays the file name in the title, and I guess you'll make money.
-----
So, yeah, backing up while traveling is hard to do. The best option is to find somebody else's PC or Mac, and back it up and take the backup and put it on your thumbdrive.
Off the bat, 3G and WiFi sound like great ideas, but if you actually tried it, will hit upon more problems.
1) You're travelling. Therefore 3G data is expensive. Really expensive. Or just not available. (yes, I thought about this once too when I was in Japan with a hacked iPhone from AT&T US)
2) WiFi takes energy. As it is, I think smartphones have ass battery life as it is. Afterall, I'm used to my old Motorola which I charged once a week. An iPhone 3G gets 6 hours on WiFi according to some web page. My iPhone backup is apparently 64MB on my computer. That's not too bad of a size I guess. I dunno how big others are. Googling for GoogleWifi's free wifi speeds, some guy at PC Magazine tried measuring it. The highest upload speed mentioned is 198kb/s. So that translates to roughly 25 kilobytes/sec. Transferring 64MB across GoogleWiFi on a good day will take 43.7 minutes. Nevermind that it just burned a little over 10% of my battery, it also just burnt over half an hour of my time.:(
-------
No, it's not like seat belts. People understand what seat belts do. And if you really don't want to wear them, it's not like it's a pain in the ass to avoid. You plop down in your seat and drive, and well, it's not like the belt's going to get in your way every time you sit in the car.
Your experience and mine appears to differ greatly, as we probably interact with different people. When it comes to programs, it's different. I've seen so many dialogs just come and go, and within seconds, the user asks me a question that was exactly answered by the dialog. Yeah, it's a facepalm moment. I think to myself, "Here I am, sitting with a PhD in physics (or bioengineering). A person who's technical education has surpassed mine, and he/she won't bother to read a dialog in the center of their screen?" How many people have you seen click "allow" when Norton Firewall or ZoneAlarm or McAfee or Windows Firewall asks if a program should be allowed to access the internet? Even when the subtext says something like, "This program may have your computer." or even if I'm visiting them cuz I just fixed their computer a few days ago. You'd think that'd be fresh in their mind. At least, I thought it would be fresh in their mind. Thing is, we joke about the "allow or deny" ads and stuff like that. We'd like to think that requiring a user enter their root password to access system settings would encourage good behavior. I won't deny that it probably has made some users better or more knowledgable. But for every user who has become better computer users, there's many more who will mindless click through everything in the blind hope that the machine does what they wan
Slashdot Mirror
Can we please stop using "free" when we mean "libre"? You know, like normal people? "free" is too ambiguous.
I've played with a N900. Why? Cuz my friend has one. He has one because he works for Nokia.
Here's the problem.
Everything else you just said, means little to me because I'm not even sure why this is a plus.
Repositories? Add them myself? Where do I go find these? How do I know these apps won't screw me? (Granted, Android has this problem too)
It's Fedora/Debian + some UI? This means what for me exactly? I know that Fedora and Debian are Linux distros, and I've installed both of them as well as Gentoo and Ubuntu on a PC of mine at some point in time. But I don't see why I should care. I'm not even sure about what's the difference between Fedora and Debian except that Fedora took up much more space on my PC.
If these are all the N900 has going for it, it's not hard to see why it's not selling. It's because nobody cares about this stuff.
Cuz they're be flogged in the media for looking like they're trying to copy not just Apple's products, but Apple's strategy? :P
With it's meager market share, I'd say the market says that the N900 doesn't compare favorably. (in the US; I dunno about Europe or elsewhere)
It's huge (though, some people don't mind that)
3G in the US only works on T-mobile.
It's probably got as many well-polished apps in Ovi as the Palm Pre.
I want a N900 because it seems like it might be kinda cool to have, but I don't think I'm actually going to buy the N900. I'd probably get an Android phone first because I'd be more likely to be able to do more with it.
Plus, I'm not on T-mobile.
I learned C first. I learned Java in college and programmed in that for years. Then I learned Objective-C. Then I tried the Android SDK. Then I tried the iPhone SDK. Then I just tried the Android SDK again.
My conclusions as of last week (when I tried the Android SDK)
1) I prefer Obj-C over Java for the flexibility and readability. It's just simply easier and at least as powerful. I get stuff done faster in Obj-C despite having more years of Java experience.
2) Xcode has bugs. Gee, what a surprise.
3) iPhone SDK is really easy to use for a mobile dev platform.
4) iPhone SDK provisioning (loading a cert to be able to sign and run on a device is a pain in the ass, but they just fixed that recently)
5) The Android SDK / Eclipse feels like being thrown back into the stone age.
All in all, I'd so much rather program in Obj-C than Java. And the Android SDK needs some serious help in order to compete with iPhone.
I'd even be inclined to say I'd probably rather even use older conventions to program for OpenStep/GnuStep when it comes to multiplatform GUI apps, but honestly, I haven't tried GnuStep yet. So I can't be sure.
Sure, using Java is nice to have crossplatform compatibility from the start. And I'm definitely inclined to do crossplatform dev because I used to use Windows, OSX, Linux and Solaris. But I've found that getting something built faster and tested is more important than getting something built for a bunch of platforms.
And might I add in that ARM effectively has some of the code compression ability of x86 via Thumb without the need for such high complexity during instruction decode.
UI design is no fun for us because we typically suck at it.
I've seen many people who's idea of UI design is to sketch up a mock window in photoshop with pretty buttons and then stop. That's about the extent of most UI "design" these days.
I've worked with what I would consider my example of an actual professional UI designer. He'd start with the sketch up and a storyboard, then start annotating everything about the workflow.
"x button is here, leads to y. defaults prefs to z because between the two most likely user workflow scenerios, the majority would benefit from ijk."
"exploring user doing task A. We expect this to happen frequently, so let's make this fast. Note from usability test: this wasn't good enough, users had a hard time understanding what just happened"
"this area seems cluttered, revisit for rev2."
All sorts of things. The stuff he did made sense to me and I could easily how he reasoned things out. It wasn't about the pretty. It was about a smooth, annoyance-free experience for the user. If I started implementing a design and had a question or a complaint, it was actually fun to go converse with him about what he thought about it. That he was open to suggestion and actually took suggestions made it all that much worthwhile.
No. And few.
I'd say that Google probably drive away people because of their use of FOSS because they don't advocate that you use FOSS nor even give the average user any obvious indication of their use of FOSS. But then again, there's plenty of other reasons Google could be driving people away that are much more meaningful to the user population.
I pretty much agree with everything you just said, except the hammer part. :)
http://www.epinions.com/review/Stanley_Antivibe_16_Ounce_Rip_Claw_Hammer_51_942/content_133678534276
While I don't own one (since I don't use a hammer all that often), the first time I saw an ad for the Stanley FatMax Xtreme Antivibe Hammer, I was like, "whoa, that looks pretty darn cool for a hammer."
My guess is that the common person will be much more excited about a hammer than over FOSS.
The response from the author in the comments mention that when they say "free" they meant the common use of "free", not "GNU/Free".
This makes sense because their readers are most likely not familiar with "free" as in "GNU/Free".
I hate the "Your real problem is that Gimp doesn't work like Photoshop" answer.
Many of us here acknowledge that Photoshop and GIMP have different UIs. Why? That's uncontestable. They're different apps. They're made by different teams. They're different.
We acknowledge that both apps have good parts about their UIs and bad parts about UIs. It's not that Photoshop is the golden standard or GIMP's just a different way of doing things. It's that for a given task, there's a lot of things that make sense for a larger majority of users.
In my usage of Photoshop, it DOES do exactly what you said: namely "use the last action with the tool in question as it's current setting". If using the last set options as default is not what you meant, then perhaps a clarification is necessary. I've left unusual options on and forgotten about them until they bit me and realized, "oh yeah, I haven't used this tool since xyz and I used some weird ass feature that I left on". So if you checked a box for aspect ratio locking and left it that way, it'll stay that way until next time.
I agree it'd be a retarded default to have on, but nevertheless, I too can guarantee you that it is not an application default to have that on. Simply, the user last set it that way. Only God knows how many countless times I've ended up drawing with a ridiculously wrong brush because I forgot I set it to something different.
As for the transparency example in GIMP. Maybe there's something that I'm missing, but to me, if I set something to be transparent. It's pretty logical that I want it transparent. So it should stay transparent. It's nice to have an option to make it untransparent, but that doesn't change the fact that if I set something transparent, it's pretty safe to assume I want it transparent.
This isn't supposed to be a "The Photoshop Way" versus "The GIMP Way" argument. There's been plenty of UI fuckups in Photoshop itself, and we all know that. The point is, for every UI complaint, there is an acceptable solution: think it through.
It should never "well, photoshop does it that way, we should do it that way."
It should always be, "okay, what do I want, what am I doing, and what does that imply, and how can I make this easier for the user?"
For the dragging problem about, this would be "well, I have a transparency value, and I have a preference for opaque-during-drag" which is by default off." Why? Because, it is more likely that a user would want the transparent image dragged transparently, as opposed to opaquely, since being transparent makes it easier to see what you're dragging across and see how it'll low when you let the mouse button go. That makes sense. It's not the right answer for the moment because that's what Photoshop is doing, it's the right answer for the moment because it make sense and makes it less likely that the majority of users will need to hunt for some odd preference.
Point is: Every complaint of "unintuitive" should never be answered with "well, that's cuz you're accustomed to using photoshop". It's an opportunity for improvement or for educating a user why it's the better way. The GIMP developers should take the time to listen to the complaints and revise those workflows, do stuff to make things easier, make the application help the user more. It should never be dismissed as wanting it to be like Photoshop. Because copying Photoshop doesn't help the user unless for a given task, it really is the best way. And we know that Photoshop is far from the best way.
Yeah, the weight of the PSU is a pretty reliable indicator too. Although a year ago I found this site: http://jonnyguru.com/modules.php?name=NDReviews&op=Review_Cat&recatnum=13
Having a guy who sounds like he enjoys making power supplies blow up and understands how to use an oscilloscope is definetely a good reviewer. Having the guy equipped with a instrumented load is even better.
Wow, it just occurred to me it's an AT (not ATX) power supply that you're pointing at. Either way, it seems the reviews for some of their newer ATX offerings seem good, so I'll keep them in mind for future home-built PCs.
ZFS. As in Sun's ZFS file system. As I see more and more drives dying at work, I've been considering building a file server a home with redundancy and checksumming.
Mind if I ask for the model of that power supply? Sounds like something I'd be interested in if I need to build me a ZFS server.
Whoa, and how does your post not make you sound like an arrogant prick?
Yeah, there's plenty of fans in both the Android and Apple camps that should keep their mouths shut. That's obvious. But it's not helping your case to respond like that to a poster who actually reasons out his side of the argument.
This is why I enjoy visiting Japan more than Taiwan.
While in Taiwan: "Why can't you speak Mandarin properly? Why can't you understand Taiwanese at all?"
While in Japan: "Oh, how cute! It's a Chinese guy trying to speak our language! Pretty good!"
The N900 doesn't support rSAP either....
That's not surprising though. It seems only european cars support rSAP in the first place. It doesn't seem to be popular elsewhere.
Hurray for security by obscurity?
Because while open source allows people to inspect the source and understand what it does, that assumes somebody actually is trying to inspect and understand the source.
Open source isn't a magical solution to make the community care about actually doing the work to improve the project. But as we see, eventually somebody noticed.
So it seems most of the community would rather use the phone than read the code. Not that surprising, isn't it?
What I find most interesting is that the security researcher's test app was a benign test for a rootkit bootstrap. To me, his results seems to imply:
1) At least 200 people don't read, understand, or care about the permissions warnings, in favor of "ooh, Twilight! *swoon*"
2) If the rootkit bootstrap worked, the app removal process depends on Google asking a rootkit-compromised phone to be trustworthy enough to remove the app. Isn't that kinda silly?
Not using credit cards is commendable. Not using a debit card is moronic.
Just curious, in all seriousness, I thought it was the other way around?
Given the rewards points that credit cards give, it seems wiser to always use the credit card to get rewards points and pay it off in full every month? This way you benefit from:
1) a 1 month interest-free buffer for managing your savings (high interest) to checking account (low interest) flow.
2) rewards from your credit card
3) being able to file a claim if somebody screws with your card
I'm not sure in what situation I would actually want to use a debit card?
You're correct that the iPhone doesn't warn you.
But the reason is that a normal iPhone app can't.
Call history is not available to apps. Ever. So there's no warning.
SMS history and ability to send/recieve/edit messages are also not available.
Existing mail and mail accounts are not available.
Whether you think any of these should be accessible is a different topic entirely.
As far as I know, what is available is the following: ... the camera roll folder.
1) Address book access.
2) Location (a permission dialog appears when this info is requested, so even if you don't want to allow the app to know, you're supposed to still be able to use the rest of the app)
3) a sandboxed folder, essentially a per-app home directory. No app can access outside that folder except...
4)
Altivec is for PowerPC. It's still used in PowerPC chips from both IBM and Freescale/Motorola.
But since Apple switched to Intel, they'll probably never mention it again. Especially since it still spanks SSE4.
Actually... they were quite a formidable semiconductor company.
For all those PowerPC chips, Apple designed northbridges. Uni-North (Uni-N) is family featured in the G4s.
U3 and U4 is used in G5s (http://en.wikipedia.org/wiki/PowerPC_970).
"Not to their idiot pandering gormless shareholders, but to their audience."
Um... it seems like their audience freakin' LOVES them.
Yea, there's plenty of evidence that there are apps some people want on the app store and it got rejected for some reason. The news gets everywhere every time it happens, like all other Apple news. Even Verizon's commercials make mention of this. I'm fairly certain most people know about the curated computing and limited app possibilities on the platform.
But, I find it hard to believe that if their business model is as bad as /. makes it out to be, they couldn't possibly be selling an iPad every 3 seconds.
You're correct that on the iPhone, files do have names. After all, it's a unix box underneath. But well-designed apps are supposed to abstract that out so that while we as programmers know there's a filename, the user doesn't need to care.
Take the simple notes app for example; I don't know how it's built, but if I were to guess, it's probably a bunch of files. Instead of creating a note and giving it a file name, you hit the plus button in the corner, and a note sheet pops up. The title of the note is the first line of the note, not the file name.
If you look at the photo browser, you don't see the file names either even though they're coming from the DCIM directory like any other camera. Instead, you see small thumbnail or the picture. Like a camera, you're going to be interacting with it as "next photo", "previous photo", or "pick one from this sheet". The file name doesn't come into play for the user because, well, it's not like "IMG0010.JPG" is going to actually give meaning.
If it one day does mean anything, then fine, publish a photo browser app that displays the file name in the title, and I guess you'll make money.
-----
So, yeah, backing up while traveling is hard to do. The best option is to find somebody else's PC or Mac, and back it up and take the backup and put it on your thumbdrive.
Off the bat, 3G and WiFi sound like great ideas, but if you actually tried it, will hit upon more problems.
1) You're travelling. Therefore 3G data is expensive. Really expensive. Or just not available. (yes, I thought about this once too when I was in Japan with a hacked iPhone from AT&T US)
2) WiFi takes energy. As it is, I think smartphones have ass battery life as it is. Afterall, I'm used to my old Motorola which I charged once a week. :(
An iPhone 3G gets 6 hours on WiFi according to some web page. My iPhone backup is apparently 64MB on my computer. That's not too bad of a size I guess. I dunno how big others are.
Googling for GoogleWifi's free wifi speeds, some guy at PC Magazine tried measuring it. The highest upload speed mentioned is 198kb/s. So that translates to roughly 25 kilobytes/sec.
Transferring 64MB across GoogleWiFi on a good day will take 43.7 minutes. Nevermind that it just burned a little over 10% of my battery, it also just burnt over half an hour of my time.
-------
No, it's not like seat belts. People understand what seat belts do. And if you really don't want to wear them, it's not like it's a pain in the ass to avoid. You plop down in your seat and drive, and well, it's not like the belt's going to get in your way every time you sit in the car.
Your experience and mine appears to differ greatly, as we probably interact with different people. When it comes to programs, it's different. I've seen so many dialogs just come and go, and within seconds, the user asks me a question that was exactly answered by the dialog. Yeah, it's a facepalm moment. I think to myself, "Here I am, sitting with a PhD in physics (or bioengineering). A person who's technical education has surpassed mine, and he/she won't bother to read a dialog in the center of their screen?" How many people have you seen click "allow" when Norton Firewall or ZoneAlarm or McAfee or Windows Firewall asks if a program should be allowed to access the internet? Even when the subtext says something like, "This program may have your computer." or even if I'm visiting them cuz I just fixed their computer a few days ago. You'd think that'd be fresh in their mind. At least, I thought it would be fresh in their mind.
Thing is, we joke about the "allow or deny" ads and stuff like that. We'd like to think that requiring a user enter their root password to access system settings would encourage good behavior. I won't deny that it probably has made some users better or more knowledgable. But for every user who has become better computer users, there's many more who will mindless click through everything in the blind hope that the machine does what they wan