Actually that makes me angry. One of Microsoft's biggest arguments against OOo has been the retraining cost - but then they turned around and forced anyone who wanted to upgrade to retrain anyway! It's a very slimy thing to do, IMO.
Doesn't help if all you need to do is a few basic documents and spreadsheets and you run in Linux most of the time. I actually own Office Home and Student, but although though I find it to be more polished for even basic tasks, I can't be bothered to try to get it to work under Wine, so I stick with OOo.
Usually it's Microsoft Works. I honestly have no idea why Microsoft wastes development time on that particular product; it would be smarter to bundle trimmed-down versions of Office.
Whether a person finds OOo "great" depends in large part on the particular feature set used by that person.
For example, if all you use Calc for is basic spreadsheets with simple formulas, then you'll find Calc to be amazing; on the other hand, if you find yourself needing to use a lot of Office 2003-format Excel documents, perhaps with a sprinkling of VBA, you're going to find Calc lacking.
(A note to the OOo fanboys: I'm not saying that Calc is only good at simple tasks; don't read too much into the specifics of my example. It's just an example using extremes for emphasis.)
Depends heavily on the service in question. If we're talking about Amazon S3, it's fairly trivial to switch to a standard web server for serving files. If we're talking about something complicated like Google Docs, then it gets messy.
But calling all SaaS "evil" because it might go wrong sometimes is ridiculous.
He's not just saying that all currently used solutions have huge issues - he's saying the currently used solutions are inherently evil! That's a far cry from "they have huge issues".
In any case, I don't have issues with his philosophy so much as I have issues with the solutions made necessary by living under his philosophy. It's ridiculous to believe that everyone should run their own home server - it's not financially viable (in many cases) nor is it an appealing solution to the average user.
Yes, it is often the case that providing a service yourself can reduce costs. However, it's not always the case:
The company I just left was hosting their software distribution on a dedicated FTP server. The bandwidth used by the server was costing them $200/month on top of the $195/month they were paying to co-locate the server. (The data center has a much bigger pipe than they had on-site, so we were essentially "doing it ourselves".)
I migrated our software distribution to Amazon S3; they essentially guarantee redundancy and availability (point 1 above). If they were to discontinue service, we could go back to a colocated server or host things on-site if necessary (points 2 and 4). We have local copies of the data (we have to generate the files before posting them for distribution!), and there's no concern about proprietary storage formats or anything, since Amazon simply stores the files we put there (point 3).
Our costs went from $200/month in bandwidth to about $90/month in Amazon S3 charges, but here's the kicker - the number of downloads increased fivefold, and our customers stopped complaining about slow downloads (Amazon's pipes are pretty big). Even if Amazon's prices were to double (unlikely, as they've actually been slowly decreasing over time), we'd still be saving money and utilizing a much superior service.
In otherwords, moving away from self-service not only saved us over 50%, but it vastly improved the service provided to our customers and increased our ability to provide for demand in the process.
It's hard to argue that we'd have saved money doing it ourselves, because our experience was precisely the opposite.
I know, I know, anecdotal evidence is just anecdotal, and YMMV, but that's precisely why RMS's philosophy is not always logistically doable. SaaS providers can amortize the costs of big pipes and data centers over all users of the service, but a small company could easily bankrupt itself trying to match the capabilities of SaaS providers in situations like that of my old employer.
I think this is a problem in the OSS world in general, not just with RMS. People can't agree on specs for new projects, so they just all make competing programs, and in the long run few of them survive, whereas if they'd just work together they'd see some significant progress.
I applaud Google and other companies who pay full-time employees to work on existing OSS projects. I wish more companies would do it - it means that features that are actually needed by the sponsoring company get written, rather than the spiffy new pet features of the hobbyist devs. (I'm not saying all hobbyist devs focus on pet features, I'm just saying it happens all too often.)
Correct me if I'm wrong, but RMS seems to have claimed that letting your own computations be run on someone else's machines is a bad idea:
"If you do your computing on someone else's server, you hand over control of your computing to whoever controls the server."
I interpret that to mean that if I don't have physical control of the machine, then it's someone else's machine, and it's a bad idea to put personal information on it. It's all well and good to have access to the patches and such, but if you don't own the machine then someone else can take away control at any point, and then you're hosed. That seems to be what RMS is warning about.
Or have I grossly misinterpreted this sentence from TFA?
I should also mention that paper money has value because the government keeps a stash of gold as backing for the money. Theoretically you can go to the government with cash and ask for the gold backing it. (Obviously the government doesn't have enough to back all the currency in circulation, but you can still ask for gold for your money.)
The government has great control over the value of the dollar. If they wanted it to be worth less (or worthless) for whatever reason, they'd simply have to print out so many bills that the value decreases. We're trusting them not to screw us over like that. (Some inflation is necessary as the population grows and such; we trust them not to overdo it.)
If you prefer, we trust our elected officials (and the people appointed by our elected officials) not to screw us over.
I recall an instance in the Dominican Republic where they elected a president, who promptly stole much of the government's money, fled the country, and was never heard from again: this was an abuse of the kind of trust I'm talking about.
Your average computer illiterate can pay the kid next-doors twenty bucks to customize his software, and your average non-IT company can hire consultants to customize their software.
I think we're using different definitions of the word "can". You're using it in the theoretical sense; I'm using it in the logistical likelihood sense.
First obstacle: Why should Joe Sixpack be required to run his own server just so he can use e-mail? Joe Sixpack will not like having to purchase the extra machine, nor will he like the resulting increased power bill, nor will he like having to call the kid next door every time it stops working. (Let's face it: even Linux machines stop working once in a while.)
But let's suppose Joe does purchase a server and install the various applications on it that he regularly uses. What are the chances that the kid-next-door will a) know how to properly modify the software, and b) be willing to do it for $20?
A thought occurs to me. What about social networks? Even if they were open-source, it would be a nightmare to get everyone's personal installation of the software to interoperate - not to mention that reliability would go down the toilet. Joe's fiber line was cut by a backhoe? Well, too bad, his e-mails bounce and his Facebook page is unavailable until the ISP replaces the cable.
That is the kind of logistical problem inherent to RMS's position. Sure, we'd solve all our privacy problems since the only possessor of my data would be me. But we'd then have to deal with all sorts of other problems, and some existing applications would simply be impractical without centralized servers under the control of a single party. Or do you believe Facebook would function correctly if it were decentralized and everyone had their own custom version of the software? (Disclaimer: I don't have a social networking site profile, Facebook or otherwise. I simply use Facebook as an example.)
I believe this is entirely orthogonal to RMS' endeavours and perhaps not possible anyway. Being evil can presumably be profitable at times so it doesn't seem feasible in the general case to construct a profit-based argument in favour of ethical behaviour.
Indeed you have a point. However, in a world where most users are not ever going to understand the difference between open source and closed source, RMS must be insane to think that telling people "SaaS is evil" will be productive! The average lay person won't know what he's talking about, and as we can see from this discussion here on Slashdot, the nerd populace of the world is far from united on the subject. If RMS wants to convince any average non-technical person, he needs to present reasons why the average person would want Facebook (for example) to be open-source.
Personally, I don't think the average person could care less whether Service X is open source, they just want it to work when they need it - and they certainly don't ever want to buy their own server to run their own services (complete with all its maintenance problems and power bills).
Doesn't he, though? Let's examine what he claims: closed-source software is evil. His solution to the evils supposedly inherent in closed-source software? Use open-source software.
Therefore, he claims that using open-source software will solve our closed-source-software-related problems.
I was perhaps a bit vague when I used the phrase "solve all our problems", however, I supposed it would be self-evident that I meant "solve all our problems related to this topic". Next time I will try to be more verbose.
That's exactly my point - this nebulous (to most people) idea of "freedom" (which, you must admit, the average non-programmer will not be able to exploit) is not a convincing argument against paying someone else to host a service for them (and certainly not against using a free service like Google Apps).
Yes, there are legitimate privacy concerns inherent to storing your data on other people's machines. However, unilaterally calling SaaS "evil" is not going to convince anyone that his arguments have merit. He needs to present more concrete arguments that the average businessman will understand - and, in most cases, that means talking about money. Unfortunately, it will often be the case that bringing up money will be counter-productive to RMS's argument.
You can't hire someone to fix software if the source is not available. Logistics mean nothing if the choice is not available in the first place.
You misunderstand. I'm not arguing that things shouldn't be open source. On the contrary, I'm arguing that, assuming the source code is available, the logistics of using that source code can make the situation just as difficult to resolve as if it were closed-source software.
In other words, claiming that making software open source will solve all our problems is, at best, overly simplistic.
Again, my argument is not against making software open source - my argument is that RMS's philosophy is flawed and, in many cases, logistically impossible in the real world.
The validity of the check is not as relevant as the accurate record of your identity. You see, a store will trust you as long as they have your ID on file, i.e. they have you by the balls in case if your check bounces. This is not what I would call a trust-based transaction.
I've never had a store make a copy of my ID when I hand them a check. What makes you think they have my ID on file? All they have is a check, and until they cash it they're simply trusting that it's valid.
In other words, you're completely ignoring the inherent trust in the system. Even if they did make a photocopy of your ID, the store is trusting you that the ID you have provided is not fake. If it's not fake, the store trusts the issuing authority that the ID was issued to the correct person.
You can't claim there's no trust simply by ignoring the trust involved.
All I know is that dollars are legal tender and people do accept them as payment for anything I ever care to obtain.
And why is a $10 bill legal tender? Because we trust the government to uphold its value. The selfishness of individuals does not remove the trust inherent to the system.
Credit card companies trust you as much as a hold-em player trusts you when you make a large bet.
I wasn't talking about credit card companies trusting the credit card holders, I was talking about merchants trusting the credit card issuers.
In any case, loans (credit cards or not) are also based in part on trust. The loaner trusts that you'll repay the loan with interest, rather than simply taking the money and disappearing. They trust that the documentation of your income (which you provided) is legitimate (this is especially the case if you're self-employed). There is a lot of trust built in to the system, and claiming it's not there is laughable.
Again, ignoring the trust inherent to a system doesn't prove there is no trust involved; it just shows you don't want to believe there's trust inherent to the system.
Oh, no, you misunderstand. I'm not arguing that anything should be more strict than what RMS wants. (Quite the opposite; I think RMS is ridiculously purist in his views.)
I was simply pointing at a flaw in my parent post's argument (he claimed that simply releasing the source code to the SaaS would be sufficient to make that SaaS "free" under RMS's definition). This is why I put "free" in quotes - I'm using the word as defined by RMS, but I don't agree with that definition.
You've simply shifted the trust from money to other places.
For example:
When you write a check, you have to show your ID, because (drum rolls) the store does not want to give you a big screen TV in exchange for a piece of paper.
This does not verify the validity of the check. It simply verifies that the name on your ID is the same as the name on the check. They still have to simply trust you that the check won't bounce, and that it's not a fake check. Furthermore, they're trusting you that the ID itself isn't fake - and they're trusting the issuing authority to do their job correctly when deciding whether to issue you the ID in the first place.
When it comes right down to it, if the store hands you the TV without making sure the check clears first (and this takes a few days, you know), they are, in fact, handing you a TV in exchange for a piece of paper.
I don't know why it is such a shock to you that money is designed to be functional even for parties which distrust each other completely.
Any money system inherently trusts the issuing authority - whenever I use or accept a $20 bill as payment, I am trusting that the US government is vouching for the value of that bill. Yes, stores do a quick check to see whether large bills are counterfeit - but have you ever seen them check a $10? A $5? Certainly you haven't seen anyone try to determine whether a $1 bill is counterfeit. (This is a function of cost - trusting $1 bills to be real costs a lot less than trusting $20 bills to be real, in the event that a bill is fake.) But even so, the store trusts that when they deposit the bills in the bank, the bank will credit their bank account correctly, the bank trusts the government to keep the bills' value relatively intact, and so on. There's no escaping trust.
And then there's credit cards - stores trust that Visa will provide the charged funds.
So you see, rather than go into detail about where exactly the trust lies, I simplified the explanation in saying that the money system is based on trust, as I have just shown.
Oh, I understand RMS's position - I just think it's fundamentally flawed in certain respects.
Besides, having the source code doesn't mean you can do something about it - you have to find someone with the skills necessary to understand the source code and make the appropriate modifications without breaking anything else, and you have to be able to afford to hire that person.
This is much easier said than done, of course; how is your average small business owner supposed to be able to determine whether programmer X is better than programmer Y, or even whether either programmer actually has the skills necessary?
Most of my disagreement with RMS's position comes from the actual logistics of what he wants people to do - he ignores the real-world with his fanatic purism, and IMHO the entire open source community is suffering because of it.
Vocal OSS purists like RMS are, IMHO, driving the average person away from open source software. A good many people shy away from zealots and their goods.
I decided to access my e-mail (incidentally run on my own server) via IMAP, knowing that if my internet connection goes out I wouldn't be able to read my e-mail. Why did I choose to do this? Because I wanted to be able to access my e-mail from other computers.
For me, the advantages of having access to my e-mail from anywhere outweigh the risk of losing my internet access and thus not having access to my e-mail. Yes, this has inconvenienced me on occasion, but I'm the one that took the risk, and I find it acceptable.
My point is, we often do consider what happens when we lose our internet connection, and we often decide it's worth the risk.
Certain aspects of it do operate without trust. Any time you see 1000000.00 written on a check, you can (and should) forget about trusting anyone.
Except the check system itself is based on trust. Your local grocery store trusts that your check won't bounce, and they trust that when they take the check to the store it won't be fake, and that the bank will process it correctly and deposit the appropriate amount in their bank account. You trust that your bank will debit the appropriate amount from your checking account.
You are thinking of what we would do if we did not have a money-based economy. This has nothing to do with trust.
The money system itself is based on trust. We trust that the government isn't printing too much and making it worthless. We trust that they're actually keeping enough gold to back some of it. We trust other people not to give us counterfeit bills (or do you examine the $20 the grocery store gave you as change for your $50?).
It is laughable to claim that the money system has nothing to do with trust.
Sure, there are instances of ridiculousness that outweigh the trust inherent in the system - when an 18 year old college student writes a $1 million check, for instance.. But who is RMS to decide that SaaS is inherently untrustworthy?
The way I see it, when RMS claims that SaaS is evil because someone else has control of your data, he's claiming that other people are not to be trusted. This view, when carried out to its logical conclusion, ends up just where your parent post put it - with everyone making their own milk with their own cow in their own backyard.
RMS may be nuts, but he is looking after your interest, unlike MS, Apple, Google.
One could argue that by asking us to provide services ourselves, and thus spending more money, RMS is actually harming our well-being, while Google, who gives us Google Apps at no charge, is improving our well-being. This isn't a flawless argument, of course, but it's just as valid as RMS's (IMNSHO flawed) philosophy.
Then he's failing in his argument - business decide most things by cost. If using SaaS costs $X/year, and running it themselves costs $(X + 5000)/year, why on earth would they run it themselves?
Not to mention the initial setup cost - purchasing servers, hiring someone to set up those servers and software, and so on.
It's all well and good to say "This is evil!" but if he really wants people to stop using SaaS, he needs to provide concrete reasons that a business should take the more expensive route.
If you're about to tell me that running it themselves isn't really more expensive... go ahead and show me how I can run a mail/calendar/document server on site for less than what Google charges for Google Apps. (Hint: Google Apps is free, and you can pay a small fee to get uptime guarantees.)
Slashdot Mirror
Actually that makes me angry. One of Microsoft's biggest arguments against OOo has been the retraining cost - but then they turned around and forced anyone who wanted to upgrade to retrain anyway! It's a very slimy thing to do, IMO.
Doesn't help if all you need to do is a few basic documents and spreadsheets and you run in Linux most of the time. I actually own Office Home and Student, but although though I find it to be more polished for even basic tasks, I can't be bothered to try to get it to work under Wine, so I stick with OOo.
You know, you could look it up yourself. It's not very hard.
Usually it's Microsoft Works. I honestly have no idea why Microsoft wastes development time on that particular product; it would be smarter to bundle trimmed-down versions of Office.
Whether a person finds OOo "great" depends in large part on the particular feature set used by that person.
For example, if all you use Calc for is basic spreadsheets with simple formulas, then you'll find Calc to be amazing; on the other hand, if you find yourself needing to use a lot of Office 2003-format Excel documents, perhaps with a sprinkling of VBA, you're going to find Calc lacking.
(A note to the OOo fanboys: I'm not saying that Calc is only good at simple tasks; don't read too much into the specifics of my example. It's just an example using extremes for emphasis.)
Someone mod parent up :) I hadn't thought of RMS's statements in that light.
Depends heavily on the service in question. If we're talking about Amazon S3, it's fairly trivial to switch to a standard web server for serving files. If we're talking about something complicated like Google Docs, then it gets messy.
But calling all SaaS "evil" because it might go wrong sometimes is ridiculous.
He's not just saying that all currently used solutions have huge issues - he's saying the currently used solutions are inherently evil! That's a far cry from "they have huge issues".
In any case, I don't have issues with his philosophy so much as I have issues with the solutions made necessary by living under his philosophy. It's ridiculous to believe that everyone should run their own home server - it's not financially viable (in many cases) nor is it an appealing solution to the average user.
Yes, it is often the case that providing a service yourself can reduce costs. However, it's not always the case:
The company I just left was hosting their software distribution on a dedicated FTP server. The bandwidth used by the server was costing them $200/month on top of the $195/month they were paying to co-locate the server. (The data center has a much bigger pipe than they had on-site, so we were essentially "doing it ourselves".)
I migrated our software distribution to Amazon S3; they essentially guarantee redundancy and availability (point 1 above). If they were to discontinue service, we could go back to a colocated server or host things on-site if necessary (points 2 and 4). We have local copies of the data (we have to generate the files before posting them for distribution!), and there's no concern about proprietary storage formats or anything, since Amazon simply stores the files we put there (point 3).
Our costs went from $200/month in bandwidth to about $90/month in Amazon S3 charges, but here's the kicker - the number of downloads increased fivefold, and our customers stopped complaining about slow downloads (Amazon's pipes are pretty big). Even if Amazon's prices were to double (unlikely, as they've actually been slowly decreasing over time), we'd still be saving money and utilizing a much superior service.
In otherwords, moving away from self-service not only saved us over 50%, but it vastly improved the service provided to our customers and increased our ability to provide for demand in the process.
It's hard to argue that we'd have saved money doing it ourselves, because our experience was precisely the opposite.
I know, I know, anecdotal evidence is just anecdotal, and YMMV, but that's precisely why RMS's philosophy is not always logistically doable. SaaS providers can amortize the costs of big pipes and data centers over all users of the service, but a small company could easily bankrupt itself trying to match the capabilities of SaaS providers in situations like that of my old employer.
I think this is a problem in the OSS world in general, not just with RMS. People can't agree on specs for new projects, so they just all make competing programs, and in the long run few of them survive, whereas if they'd just work together they'd see some significant progress.
I applaud Google and other companies who pay full-time employees to work on existing OSS projects. I wish more companies would do it - it means that features that are actually needed by the sponsoring company get written, rather than the spiffy new pet features of the hobbyist devs. (I'm not saying all hobbyist devs focus on pet features, I'm just saying it happens all too often.)
Correct me if I'm wrong, but RMS seems to have claimed that letting your own computations be run on someone else's machines is a bad idea:
"If you do your computing on someone else's server, you hand over control of your computing to whoever controls the server."
I interpret that to mean that if I don't have physical control of the machine, then it's someone else's machine, and it's a bad idea to put personal information on it. It's all well and good to have access to the patches and such, but if you don't own the machine then someone else can take away control at any point, and then you're hosed. That seems to be what RMS is warning about.
Or have I grossly misinterpreted this sentence from TFA?
I should also mention that paper money has value because the government keeps a stash of gold as backing for the money. Theoretically you can go to the government with cash and ask for the gold backing it. (Obviously the government doesn't have enough to back all the currency in circulation, but you can still ask for gold for your money.)
The government has great control over the value of the dollar. If they wanted it to be worth less (or worthless) for whatever reason, they'd simply have to print out so many bills that the value decreases. We're trusting them not to screw us over like that. (Some inflation is necessary as the population grows and such; we trust them not to overdo it.)
If you prefer, we trust our elected officials (and the people appointed by our elected officials) not to screw us over.
I recall an instance in the Dominican Republic where they elected a president, who promptly stole much of the government's money, fled the country, and was never heard from again: this was an abuse of the kind of trust I'm talking about.
*sigh*
The stores are trusting that the courts will enforce the law, then. The courts trust that the cops will track down the fugitive.
However you break it up, it comes down to someone trusting someone else to do something properly.
Your average computer illiterate can pay the kid next-doors twenty bucks to customize his software, and your average non-IT company can hire consultants to customize their software.
I think we're using different definitions of the word "can". You're using it in the theoretical sense; I'm using it in the logistical likelihood sense.
First obstacle: Why should Joe Sixpack be required to run his own server just so he can use e-mail? Joe Sixpack will not like having to purchase the extra machine, nor will he like the resulting increased power bill, nor will he like having to call the kid next door every time it stops working. (Let's face it: even Linux machines stop working once in a while.)
But let's suppose Joe does purchase a server and install the various applications on it that he regularly uses. What are the chances that the kid-next-door will a) know how to properly modify the software, and b) be willing to do it for $20?
A thought occurs to me. What about social networks? Even if they were open-source, it would be a nightmare to get everyone's personal installation of the software to interoperate - not to mention that reliability would go down the toilet. Joe's fiber line was cut by a backhoe? Well, too bad, his e-mails bounce and his Facebook page is unavailable until the ISP replaces the cable.
That is the kind of logistical problem inherent to RMS's position. Sure, we'd solve all our privacy problems since the only possessor of my data would be me. But we'd then have to deal with all sorts of other problems, and some existing applications would simply be impractical without centralized servers under the control of a single party. Or do you believe Facebook would function correctly if it were decentralized and everyone had their own custom version of the software? (Disclaimer: I don't have a social networking site profile, Facebook or otherwise. I simply use Facebook as an example.)
I believe this is entirely orthogonal to RMS' endeavours and perhaps not possible anyway. Being evil can presumably be profitable at times so it doesn't seem feasible in the general case to construct a profit-based argument in favour of ethical behaviour.
Indeed you have a point. However, in a world where most users are not ever going to understand the difference between open source and closed source, RMS must be insane to think that telling people "SaaS is evil" will be productive! The average lay person won't know what he's talking about, and as we can see from this discussion here on Slashdot, the nerd populace of the world is far from united on the subject. If RMS wants to convince any average non-technical person, he needs to present reasons why the average person would want Facebook (for example) to be open-source.
Personally, I don't think the average person could care less whether Service X is open source, they just want it to work when they need it - and they certainly don't ever want to buy their own server to run their own services (complete with all its maintenance problems and power bills).
Doesn't he, though? Let's examine what he claims: closed-source software is evil. His solution to the evils supposedly inherent in closed-source software? Use open-source software.
Therefore, he claims that using open-source software will solve our closed-source-software-related problems.
I was perhaps a bit vague when I used the phrase "solve all our problems", however, I supposed it would be self-evident that I meant "solve all our problems related to this topic". Next time I will try to be more verbose.
That's exactly my point - this nebulous (to most people) idea of "freedom" (which, you must admit, the average non-programmer will not be able to exploit) is not a convincing argument against paying someone else to host a service for them (and certainly not against using a free service like Google Apps).
Yes, there are legitimate privacy concerns inherent to storing your data on other people's machines. However, unilaterally calling SaaS "evil" is not going to convince anyone that his arguments have merit. He needs to present more concrete arguments that the average businessman will understand - and, in most cases, that means talking about money. Unfortunately, it will often be the case that bringing up money will be counter-productive to RMS's argument.
You can't hire someone to fix software if the source is not available. Logistics mean nothing if the choice is not available in the first place.
You misunderstand. I'm not arguing that things shouldn't be open source. On the contrary, I'm arguing that, assuming the source code is available, the logistics of using that source code can make the situation just as difficult to resolve as if it were closed-source software.
In other words, claiming that making software open source will solve all our problems is, at best, overly simplistic.
Again, my argument is not against making software open source - my argument is that RMS's philosophy is flawed and, in many cases, logistically impossible in the real world.
The validity of the check is not as relevant as the accurate record of your identity. You see, a store will trust you as long as they have your ID on file, i.e. they have you by the balls in case if your check bounces. This is not what I would call a trust-based transaction.
I've never had a store make a copy of my ID when I hand them a check. What makes you think they have my ID on file? All they have is a check, and until they cash it they're simply trusting that it's valid.
In other words, you're completely ignoring the inherent trust in the system. Even if they did make a photocopy of your ID, the store is trusting you that the ID you have provided is not fake. If it's not fake, the store trusts the issuing authority that the ID was issued to the correct person.
You can't claim there's no trust simply by ignoring the trust involved.
All I know is that dollars are legal tender and people do accept them as payment for anything I ever care to obtain.
And why is a $10 bill legal tender? Because we trust the government to uphold its value. The selfishness of individuals does not remove the trust inherent to the system.
Credit card companies trust you as much as a hold-em player trusts you when you make a large bet.
I wasn't talking about credit card companies trusting the credit card holders, I was talking about merchants trusting the credit card issuers.
In any case, loans (credit cards or not) are also based in part on trust. The loaner trusts that you'll repay the loan with interest, rather than simply taking the money and disappearing. They trust that the documentation of your income (which you provided) is legitimate (this is especially the case if you're self-employed). There is a lot of trust built in to the system, and claiming it's not there is laughable.
Again, ignoring the trust inherent to a system doesn't prove there is no trust involved; it just shows you don't want to believe there's trust inherent to the system.
Oh, no, you misunderstand. I'm not arguing that anything should be more strict than what RMS wants. (Quite the opposite; I think RMS is ridiculously purist in his views.)
I was simply pointing at a flaw in my parent post's argument (he claimed that simply releasing the source code to the SaaS would be sufficient to make that SaaS "free" under RMS's definition). This is why I put "free" in quotes - I'm using the word as defined by RMS, but I don't agree with that definition.
Very well, I'll play your game.
You've simply shifted the trust from money to other places.
For example:
When you write a check, you have to show your ID, because (drum rolls) the store does not want to give you a big screen TV in exchange for a piece of paper.
This does not verify the validity of the check. It simply verifies that the name on your ID is the same as the name on the check. They still have to simply trust you that the check won't bounce, and that it's not a fake check. Furthermore, they're trusting you that the ID itself isn't fake - and they're trusting the issuing authority to do their job correctly when deciding whether to issue you the ID in the first place.
When it comes right down to it, if the store hands you the TV without making sure the check clears first (and this takes a few days, you know), they are, in fact, handing you a TV in exchange for a piece of paper.
I don't know why it is such a shock to you that money is designed to be functional even for parties which distrust each other completely.
Any money system inherently trusts the issuing authority - whenever I use or accept a $20 bill as payment, I am trusting that the US government is vouching for the value of that bill. Yes, stores do a quick check to see whether large bills are counterfeit - but have you ever seen them check a $10? A $5? Certainly you haven't seen anyone try to determine whether a $1 bill is counterfeit. (This is a function of cost - trusting $1 bills to be real costs a lot less than trusting $20 bills to be real, in the event that a bill is fake.) But even so, the store trusts that when they deposit the bills in the bank, the bank will credit their bank account correctly, the bank trusts the government to keep the bills' value relatively intact, and so on. There's no escaping trust.
And then there's credit cards - stores trust that Visa will provide the charged funds.
So you see, rather than go into detail about where exactly the trust lies, I simplified the explanation in saying that the money system is based on trust, as I have just shown.
Oh, I understand RMS's position - I just think it's fundamentally flawed in certain respects.
Besides, having the source code doesn't mean you can do something about it - you have to find someone with the skills necessary to understand the source code and make the appropriate modifications without breaking anything else, and you have to be able to afford to hire that person.
This is much easier said than done, of course; how is your average small business owner supposed to be able to determine whether programmer X is better than programmer Y, or even whether either programmer actually has the skills necessary?
Most of my disagreement with RMS's position comes from the actual logistics of what he wants people to do - he ignores the real-world with his fanatic purism, and IMHO the entire open source community is suffering because of it.
Vocal OSS purists like RMS are, IMHO, driving the average person away from open source software. A good many people shy away from zealots and their goods.
I decided to access my e-mail (incidentally run on my own server) via IMAP, knowing that if my internet connection goes out I wouldn't be able to read my e-mail. Why did I choose to do this? Because I wanted to be able to access my e-mail from other computers.
For me, the advantages of having access to my e-mail from anywhere outweigh the risk of losing my internet access and thus not having access to my e-mail. Yes, this has inconvenienced me on occasion, but I'm the one that took the risk, and I find it acceptable.
My point is, we often do consider what happens when we lose our internet connection, and we often decide it's worth the risk.
Certain aspects of it do operate without trust. Any time you see 1000000.00 written on a check, you can (and should) forget about trusting anyone.
Except the check system itself is based on trust. Your local grocery store trusts that your check won't bounce, and they trust that when they take the check to the store it won't be fake, and that the bank will process it correctly and deposit the appropriate amount in their bank account. You trust that your bank will debit the appropriate amount from your checking account.
You are thinking of what we would do if we did not have a money-based economy. This has nothing to do with trust.
The money system itself is based on trust. We trust that the government isn't printing too much and making it worthless. We trust that they're actually keeping enough gold to back some of it. We trust other people not to give us counterfeit bills (or do you examine the $20 the grocery store gave you as change for your $50?).
It is laughable to claim that the money system has nothing to do with trust.
Sure, there are instances of ridiculousness that outweigh the trust inherent in the system - when an 18 year old college student writes a $1 million check, for instance.. But who is RMS to decide that SaaS is inherently untrustworthy?
The way I see it, when RMS claims that SaaS is evil because someone else has control of your data, he's claiming that other people are not to be trusted. This view, when carried out to its logical conclusion, ends up just where your parent post put it - with everyone making their own milk with their own cow in their own backyard.
RMS may be nuts, but he is looking after your interest, unlike MS, Apple, Google.
One could argue that by asking us to provide services ourselves, and thus spending more money, RMS is actually harming our well-being, while Google, who gives us Google Apps at no charge, is improving our well-being. This isn't a flawless argument, of course, but it's just as valid as RMS's (IMNSHO flawed) philosophy.
Then he's failing in his argument - business decide most things by cost. If using SaaS costs $X/year, and running it themselves costs $(X + 5000)/year, why on earth would they run it themselves?
Not to mention the initial setup cost - purchasing servers, hiring someone to set up those servers and software, and so on.
It's all well and good to say "This is evil!" but if he really wants people to stop using SaaS, he needs to provide concrete reasons that a business should take the more expensive route.
If you're about to tell me that running it themselves isn't really more expensive... go ahead and show me how I can run a mail/calendar/document server on site for less than what Google charges for Google Apps. (Hint: Google Apps is free, and you can pay a small fee to get uptime guarantees.)