Bloat is load without worthwhile benefit, by definition.
Worthwhile to who?
One should establish POV during the discussion, true. In this case, I'm arguing for software quality, user productivity, and cost savings. I suppose from the POV of the marketing department, flashy but functionless graphics might considered be worthwhile. However, I think that if you studied user experience, they would find software that works well more worthwhile that something with a pretty package, in the long run. Software is a tool, and a tool should be measured by how well it does its job.
You want to spend $50 less on RAM and just use "non bloated software?" You are assuming such hypothetical software would cost no more than $50 extra per machine. As a developer, when you tell me "don't bloat your code," you are telling me several things.
Actually, when I say "don't bloat your code", I'm not so much talking to you as Microsoft and other large corporate software houses, and mainly what I'm saying is: Don't build program that look like a home stereo, or change the icons in every release, or add animations and gradients and transparency or themeable UIs. Stop changing the UI around every year just because some focus group says it's 3% better -- or worse still, because someone in marketing drew a picture. Don't pump out a new version just to keep revenues up. Don't change formats without cleaning up the design. Stop adding checklist features and chasing whatever consumer trend is hot that year.
Bloat is load without worthwhile benefit, by definition. You're talking about design trade-offs, which is a something else. Although, getting in to that, I would point out that well-written code is not only usually more lightweight, but also more stable, more reliable, more secure, and easier to debug. So as long as we're looking at total costs, let's look at time spent by the user dealing with bugs and poor design and data loss and patches and security compromises, or time spent by your employees trying to maintain the ungainly mess you wrote so cheaply. Since you bring it up, I mean.
I find it difficult to believe that people are buying new machines with less than 4 gig of ram.
General objection: Just because something is cheap doesn't mean we should spend it wastefully. I'd still rather that RAM be put to better use than code bloat. I'd rather the PC be faster, or do more things, or do new things, or be cheaper still. Maybe if software wasn't bloated, PCs would be less than $100 these days. Then we could put that money towards better support.
Specific examples:
For a business with 100s or 1000s of PCs, spending an extra $50 on more RAM just because it's cheap means thousands or 10s of thousands of dollars. Code bloat means we have to. We would prefer to save the money.
For a smaller business with 100 PCs (like where I work), upgrading PCs isn't something that happens every 3 years. We've got 10s of older desktops with only 2 GB RAM. (And they've been upgraded to their max.)
Even top-of-the-line compact and subcompact notebooks are often limited to 2, 3 or 4 GB RAM max.
Tablets and handhelds are limited to even less RAM. Microsoft wants Windows running in that space. So do their customers. (I'm no fan of Microsoft, but fact is a lot of stuff is Windows-only. Given that fact, I'd like to be able to use Windows-based software more effectively.)
I'm not sure if that was a proper reply or just a stream of consciousness.:)
My only point with "memory mapped disk I/O" is that one of the uses of VM is to abstract disk I/O into regular memory accesses. Program just accesses memory, but the MMU and OS turn that into disk I/O calls behind the screens. If you don't have a disk, and instead just have one big RAM (or one big NUMA), you don't need to do that because there's no disk I/O calls to be translated to. Of course VM is still useful for other things; indeed that was my point, and I said as much.
Beyond that, I can't speak to Shelob or pop psychology or evolutionary models, but I don't think I can find them on the address bus anyway.:)
So in The Terminator, humanity is destroyed when the power-mad AI "Skynet" launches nuclear missiles. That's been the popular conception of computer-driven destruction ever since.
Here we have computers controlling flying killer robots. Said computers have been compromised by malware. This was detected weeks ago, the malware is still a threat, and they're still flying them.
I'm starting to really believe that WarGames will be the more accurate prediction. Humanity won't be destroyed by machines which try to take our place. Humanity will be destroyed when some punk teenager hacks into a weapons system and pushes the wrong button.
Unlikely. Bus speed and lengths are already an issue. It doesn't if your storage is uniformly fast if the pipe to the processor can't keep up. The von Neumann bottleneck. Massively parallel, distributed systems would fix this, but people are really bad at thinking about problems that way, so I'm not sure how practical that would be.
... so there would be no need for a VM system... VM was considered important not just for simulating a larger memory system...
It's a common misconception (reinforced by several GUIs) that "virtual memory" means "using disk as virtual RAM". That's not accurate.
Virtual memory means the address a program sees is not the real address in storage. It's translated by the MMU (Memory Management Unit). This lets you do any number of things, only some of which have to do with paging/swapping to disk. In particular, it lets different programs have different views and access of the same storage, which is important for separation of privileges between processes (useful for even single-user systems, and required for practical multi-user systems).
It would be the case that some of the uses for VM would go away, though. No need for disk swapping/paging. No need for memory mapped disk I/O if you have no disk, either.
When you change channels, you will occasionally get an hourglass for a few seconds, for no apparent reason.
Every time there's a change to the channel line-up, you'll have to download a 300 MB update, wait ten minutes for it to install, and then restart your TV.
Instead of the NTSC color bars, during "technical difficulties" they will show a blue screen with white text.
Every year they'll change what all the buttons on the remote do. Next year volume up/down will be mapped to numbers 2 and 7.
"Hi there! It looks like you're trying to find something good to watch. Would you like me to suggest something?"
Since Giant MR it's been very hard to read overwritten blocks without probe microscopy ($$$)...
Cost (money, manhours, etc.) are often cited when it comes to why the NSA's requirements are silly, but people forget that the NSA is not working in the same problem space most are. The NSA and it's foreign counterparts *do* have the resources to spend on that kind of thing. They may even have technologies not publicly known. But they're also protecting things like nuclear launch codes. Chances are, your corporate payroll isn't capable of ending civilization as we know it.:)
I can't stand "security" people in business in general with this impulsive urge to physically destroy hard drives because of the data they stored.
I suspect it's because of NSA/DOD, which has very specific rules for how to destroy data. NSA did studies 30+ years ago and found a simple overwrite wasn't always sufficient. (Keep in mind back then you could prolly read the data with iron filings and a magnifying glass, so magnetic remanence was rather more plausible than today.) And since 2007, overwrite hasn't been acceptable at all. It's degaussing or physical destruction only. Whether this is due to some special recovery technique, or just because hard disks are smart and people are dumb, only the NSA knows, and they aren't telling.
But if you're not protecting classified national security information, it does seem rather likely to be overkill.
One of the PalmOS browsers worked this way, doing pre-rendering at the other end to help compensate for a slow connection and a small device. If the ISP's server farm went down, so did your web browser.
I've used a satellite Internet provider that did similar as well, parsing the HTML at the provider ground station so it could fetch all the needed objects and send them in a single stream to the sky. This eliminated a lot of repeated fetch requests from the client over sat, which made a big difference since RTT was around one second.
And given that it was satellite, wouldn't that also qualify as "in the cloud"?;-)
Plus, I don't have advertisements constantly flashing before my eyes (which you pay to watch on cable; you "get them" for free on non-cable) telling me "consume, consume, CONSUME."
I believe a more apt term would be megalomaniacal; believing oneself to have absolute moral superiority -- in this case, over a craven race of incipient terrorists, pedophiles, and copyright infringers.
Hey now! Do you have any evidence at all that any copyright infringement is going on?
But don't worry; someone will tell you it can be done, and you'll pay them a lot of money, only to realize they lied.
The NSA employs more mathematicians than any other organization in the world. I don't know you from Adam, but it's still a near-certainty that they have people much smarter than either of us working for them. They often fab their own silicon, build their own hardware, write their own software -- all from the ground up.
Whether or not this particular project will be a success is an open question -- the NSA is hardly immune to the Dilbert-style failings of any large bureaucracy, and "National Stupidity Agency" is a common-enough expansion -- but don't assume they'll fail just because you disagree with their mission and/or policies.
The current STE (Secure Terminal Equipment, the rename of the STU (Secure Telephone Unit) series) costs around $3500 for the basic model. The technology in it is rather inferior my contemporary geek standards. One of the big reasons it costs so much is all the critical technology is sourced within the US from trusted sources. (Well, that's the theory, anyway.)
The NSA goes to considerable lengths and expense to protect their supply chain. (It's easy to spare no expense when you're spending others' money.)
But can the researchers detect the government mind control rays transmitted from cell phone towers? If it could do that, I wouldn't have to wear this tin-foil hat all the time!
For me it is a big deal as I deal with customers and ephemeral contacts, not just a set of constant contacts.
If they really are ephemeral, why are you taking the time to memorize them? Write them down (or type them in) once. Read your notes to put them into short-term memory, dial them once, then forget about them.
If they're worth memorizing, chances are, a smartphone (or computer dialing program for a desk phone) can do a better job remembering them than you can.
I probably wasn't clear: the muscle-memory that I mentioned was for specific phone numbers, not the keypad in general.
I think the parent poster understood that. Point is, you can still train yourself to that. It's a bit harder, but the left-right orientation is the same, which helps a lot. You just have to learn how to select the top/middle/bottom row for a digit for the session.
In school, I knew someone who had taken the time to practice writing upside down. They got to the point where they could do it almost as fast as they wrote normally. (A mostly useless skill, to be sure, but it's the kind of thing you do when you're a kid.) Point being, you can train yourself to this.
You'll have a lot more luck with that then you will with finding odd keyboards, and then you'll still be able to function on other people's equipment, too.
That's a name from Slashdot's user-relationship system.
You can tell the system that another user is a Friend or a Foe of yourself. This shows up as a little colored dot next to their username. You can also optionally have it score comments of Friends/Foes higher/lower.
Fan = someone who has marked *you* as a Friend Freak = someone who has marked *you* as a Foe
And even GNU has put out some good man pages. My intro to shell programming was reading the man page for bash. It's remarkably comprehensive. I don't know how that slipped through GNU quality control.;-)
Slashdot Mirror
Since you're not saying anything Runaway1956 didn't say in his original post, I'll just refer you to my original rejection of that assertion.
One should establish POV during the discussion, true. In this case, I'm arguing for software quality, user productivity, and cost savings. I suppose from the POV of the marketing department, flashy but functionless graphics might considered be worthwhile. However, I think that if you studied user experience, they would find software that works well more worthwhile that something with a pretty package, in the long run. Software is a tool, and a tool should be measured by how well it does its job.
You want to spend $50 less on RAM and just use "non bloated software?" You are assuming such hypothetical software would cost no more than $50 extra per machine. As a developer, when you tell me "don't bloat your code," you are telling me several things.
Actually, when I say "don't bloat your code", I'm not so much talking to you as Microsoft and other large corporate software houses, and mainly what I'm saying is: Don't build program that look like a home stereo, or change the icons in every release, or add animations and gradients and transparency or themeable UIs. Stop changing the UI around every year just because some focus group says it's 3% better -- or worse still, because someone in marketing drew a picture. Don't pump out a new version just to keep revenues up. Don't change formats without cleaning up the design. Stop adding checklist features and chasing whatever consumer trend is hot that year.
Bloat is load without worthwhile benefit, by definition. You're talking about design trade-offs, which is a something else. Although, getting in to that, I would point out that well-written code is not only usually more lightweight, but also more stable, more reliable, more secure, and easier to debug. So as long as we're looking at total costs, let's look at time spent by the user dealing with bugs and poor design and data loss and patches and security compromises, or time spent by your employees trying to maintain the ungainly mess you wrote so cheaply. Since you bring it up, I mean.
I find it difficult to believe that people are buying new machines with less than 4 gig of ram.
General objection: Just because something is cheap doesn't mean we should spend it wastefully. I'd still rather that RAM be put to better use than code bloat. I'd rather the PC be faster, or do more things, or do new things, or be cheaper still. Maybe if software wasn't bloated, PCs would be less than $100 these days. Then we could put that money towards better support.
Specific examples:
I'm not sure if that was a proper reply or just a stream of consciousness. :)
My only point with "memory mapped disk I/O" is that one of the uses of VM is to abstract disk I/O into regular memory accesses. Program just accesses memory, but the MMU and OS turn that into disk I/O calls behind the screens. If you don't have a disk, and instead just have one big RAM (or one big NUMA), you don't need to do that because there's no disk I/O calls to be translated to. Of course VM is still useful for other things; indeed that was my point, and I said as much.
Beyond that, I can't speak to Shelob or pop psychology or evolutionary models, but I don't think I can find them on the address bus anyway. :)
It's a weapons platform that's been compromised by mainstream malware. From that alone, the pooch is jolly well being gang-banged.
So in The Terminator, humanity is destroyed when the power-mad AI "Skynet" launches nuclear missiles. That's been the popular conception of computer-driven destruction ever since.
Here we have computers controlling flying killer robots. Said computers have been compromised by malware. This was detected weeks ago, the malware is still a threat, and they're still flying them .
I'm starting to really believe that WarGames will be the more accurate prediction. Humanity won't be destroyed by machines which try to take our place. Humanity will be destroyed when some punk teenager hacks into a weapons system and pushes the wrong button.
$DEITY help us.
Unlikely. Bus speed and lengths are already an issue. It doesn't if your storage is uniformly fast if the pipe to the processor can't keep up. The von Neumann bottleneck. Massively parallel, distributed systems would fix this, but people are really bad at thinking about problems that way, so I'm not sure how practical that would be.
It's a common misconception (reinforced by several GUIs) that "virtual memory" means "using disk as virtual RAM". That's not accurate.
Virtual memory means the address a program sees is not the real address in storage. It's translated by the MMU (Memory Management Unit). This lets you do any number of things, only some of which have to do with paging/swapping to disk. In particular, it lets different programs have different views and access of the same storage, which is important for separation of privileges between processes (useful for even single-user systems, and required for practical multi-user systems).
It would be the case that some of the uses for VM would go away, though. No need for disk swapping/paging. No need for memory mapped disk I/O if you have no disk, either.
I have mod points but you're already at 5. They should remove the cap on just this one comment so it can go to a million.
http://dictionary.reference.com/browse/joke
http://dictionary.reference.com/browse/prediction
Oh, boy, this is comedy gold:
If Microsoft made TV...
When you change channels, you will occasionally get an hourglass for a few seconds, for no apparent reason.
Every time there's a change to the channel line-up, you'll have to download a 300 MB update, wait ten minutes for it to install, and then restart your TV.
Instead of the NTSC color bars, during "technical difficulties" they will show a blue screen with white text.
Every year they'll change what all the buttons on the remote do. Next year volume up/down will be mapped to numbers 2 and 7.
"Hi there! It looks like you're trying to find something good to watch. Would you like me to suggest something?"
Since Giant MR it's been very hard to read overwritten blocks without probe microscopy ($$$) ...
Cost (money, manhours, etc.) are often cited when it comes to why the NSA's requirements are silly, but people forget that the NSA is not working in the same problem space most are. The NSA and it's foreign counterparts *do* have the resources to spend on that kind of thing. They may even have technologies not publicly known. But they're also protecting things like nuclear launch codes. Chances are, your corporate payroll isn't capable of ending civilization as we know it. :)
I can't stand "security" people in business in general with this impulsive urge to physically destroy hard drives because of the data they stored.
I suspect it's because of NSA/DOD, which has very specific rules for how to destroy data. NSA did studies 30+ years ago and found a simple overwrite wasn't always sufficient. (Keep in mind back then you could prolly read the data with iron filings and a magnifying glass, so magnetic remanence was rather more plausible than today.) And since 2007, overwrite hasn't been acceptable at all. It's degaussing or physical destruction only. Whether this is due to some special recovery technique, or just because hard disks are smart and people are dumb, only the NSA knows, and they aren't telling.
But if you're not protecting classified national security information, it does seem rather likely to be overkill.
Opera Mini isn't the only one.
One of the PalmOS browsers worked this way, doing pre-rendering at the other end to help compensate for a slow connection and a small device. If the ISP's server farm went down, so did your web browser.
I've used a satellite Internet provider that did similar as well, parsing the HTML at the provider ground station so it could fetch all the needed objects and send them in a single stream to the sky. This eliminated a lot of repeated fetch requests from the client over sat, which made a big difference since RTT was around one second.
And given that it was satellite, wouldn't that also qualify as "in the cloud"? ;-)
"I can imagine quite a bit." -- Han Solo
Plus, I don't have advertisements constantly flashing before my eyes (which you pay to watch on cable; you "get them" for free on non-cable) telling me "consume, consume, CONSUME."
Just take the special sunglasses back off.
I believe a more apt term would be megalomaniacal; believing oneself to have absolute moral superiority -- in this case, over a craven race of incipient terrorists, pedophiles, and copyright infringers.
Hey now! Do you have any evidence at all that any copyright infringement is going on?
But don't worry; someone will tell you it can be done, and you'll pay them a lot of money, only to realize they lied.
The NSA employs more mathematicians than any other organization in the world. I don't know you from Adam, but it's still a near-certainty that they have people much smarter than either of us working for them. They often fab their own silicon, build their own hardware, write their own software -- all from the ground up.
Whether or not this particular project will be a success is an open question -- the NSA is hardly immune to the Dilbert-style failings of any large bureaucracy, and "National Stupidity Agency" is a common-enough expansion -- but don't assume they'll fail just because you disagree with their mission and/or policies.
The current STE (Secure Terminal Equipment, the rename of the STU (Secure Telephone Unit) series) costs around $3500 for the basic model. The technology in it is rather inferior my contemporary geek standards. One of the big reasons it costs so much is all the critical technology is sourced within the US from trusted sources. (Well, that's the theory, anyway.)
The NSA goes to considerable lengths and expense to protect their supply chain. (It's easy to spare no expense when you're spending others' money.)
But can the researchers detect the government mind control rays transmitted from cell phone towers? If it could do that, I wouldn't have to wear this tin-foil hat all the time!
So I'm your Foe for wanting a uniform interface on like devices?
Nope, I'm just some guy passing by who saw your comment and replied.
You might want to pay a bit more attention to who you're talking to. That will help when you're making all those calls to customers.
For me it is a big deal as I deal with customers and ephemeral contacts, not just a set of constant contacts.
If they really are ephemeral, why are you taking the time to memorize them? Write them down (or type them in) once. Read your notes to put them into short-term memory, dial them once, then forget about them.
If they're worth memorizing, chances are, a smartphone (or computer dialing program for a desk phone) can do a better job remembering them than you can.
I probably wasn't clear: the muscle-memory that I mentioned was for specific phone numbers, not the keypad in general.
I think the parent poster understood that. Point is, you can still train yourself to that. It's a bit harder, but the left-right orientation is the same, which helps a lot. You just have to learn how to select the top/middle/bottom row for a digit for the session.
In school, I knew someone who had taken the time to practice writing upside down. They got to the point where they could do it almost as fast as they wrote normally. (A mostly useless skill, to be sure, but it's the kind of thing you do when you're a kid.) Point being, you can train yourself to this.
You'll have a lot more luck with that then you will with finding odd keyboards, and then you'll still be able to function on other people's equipment, too.
I've been called worse that a freak!
That's a name from Slashdot's user-relationship system.
You can tell the system that another user is a Friend or a Foe of yourself. This shows up as a little colored dot next to their username. You can also optionally have it score comments of Friends/Foes higher/lower.
Fan = someone who has marked *you* as a Friend
Freak = someone who has marked *you* as a Foe
And even GNU has put out some good man pages. My intro to shell programming was reading the man page for bash. It's remarkably comprehensive. I don't know how that slipped through GNU quality control. ;-)