And this is exactly why OSS people will always be just a bunch of bickering dilettantes. They prefer to own it all, just so they can have the hubris of giving it away.
LOL. I love it. That explains so many things. OSSers love to re-invent the wheel because pointing at an inferior copy of somebody else's application and muttering the magic incantation, ``But this one is FREE,'' gives them a heady rush of moral superiority that far outweighs the feelings of technical inferiority they get from having written it in the first place.
This is my favorite new theory on why OSSers are they way they are.
It reminds me of a great quote from Douglas Adams; it's not exactly the same idea, but I think it expresses the same sentiment, more or less. I'm too lazy to look it up right now. I'll paraphrase: The sense of satisfaction a user gets from getting the machines to work at all blinds the user to the machines' essential uselessness. In other words, their superficial design flaws completely hide their fundamental design flaws.
Mistake? Where did you get the idea that we made a mistake? We've been deploying web applications for key business operations gradually over the past couple of years, and our overall efficiency has skyrocketed. It used to take more than a week to get a sales order filled for a customer. Now that order can be entered in the morning, product ships around lunchtime, and the invoice goes out that afternoon. There's no mistake here.
This is my whole point, really: W3C standards are absolutely irrelevant to us. They're just pieces of paper as far as I'm concerned. What's relevant to my company is our existing base of web applications. If Mozilla (or whatever) can't run those applications well enough that the users don't notice a difference, then Mozilla won't find a home here.
Why get religious about standards? They're important only to the extent that they improve things. Releasing a browser that's 100% compliant with the specification of the standard but that can't do the job of the incumbent browser doesn't help anybody. The Mozilla team should have spent more time implementing compatibility with IE-- or at least an IE compatibility mode that can be turned on and off in the ``advanced'' pane of the preferences dialog-- and less time implementing features that nobody needs. How much time was spent implementing the ``skinnable'' interface? How much time was spent on the mail and news readers? The browser doesn't have a goddamn ``home'' button on the toolbar, but it reads news, and look at all these kewl skinz, d00d!
What a waste. What an unbelievable waste of time and effort. Is it any wonder that IE is the more popular browser? For all their faults, Microsoft implemented features that developers and users actually wanted, while the Mozilla coalition gave us kewl skinz.
If you want to talk about mistakes, I think we can start right there.
I agree. Improvement is a good thing. But Ogg isn't an improvement on MP3, except in some very marginal ways. Best I can discern, Ogg is about equivalent to MP3, except for the lack of software and hardware that supports it.
Wish I had mod points. After a discussion a few days ago, on the occasion of the 1.0 release, I came to the same conclusion. Ogg offers a couple of features that MP3 lacks, but they are useful only in very specific situations. I'm a heavy user of MP3s-- as I mentioned elsewhere, I have ripped my collection of about 400 CDs to MP3, and I play selections from that library practically every day-- and I can't come up with a good reason why I would be interested in Ogg. The software is there, the hardware is there. Unless something changes drastically, I predict that Ogg will pretty much continue to be irrelevant to all but a select few.
All, interestingly enough, of these sources have egos the size of Texas and consciences the size of Guam.
Bad example. While Guam might be small compared to Texas, it's still pretty darned big in the absolute sense. I.e., it's a lot bigger than I am. For future reference, I would choose something like ``consciences the size of a flea.''
On a more serious note, might I ask that you consider being a little bit more careful with your words in the future. Saying this like, ``All of these sources [politicians, CEOs, doctors...] have egos the size of Texas,'' is an overgeneralization, and a rude one. My girlfriend is a doctor, as are most of our friends. While there are egoists out there, I'd have to say that all but one or two of the doctors I know are the most humble people you could hope to meet. We've talked about it before, over dinners and such: they all agree that being a doctor is an overwhelming responsibility. If you let it go to your head, you might get arrogant, but the fact is that simple fear keeps you from thinking too highly of yourself.
I'd just ask that you think twice before generalizing in the future.
Vote? What vote? This is a company, not a cub scout troop. TCO is the only legitimate reason to choose one type of technology over another. If you believe that you can do business by standing on your principles and using Linux just because it's ``free'' (whatever that means), then you're in for an unpleasant surprise.
And as for FUD, as I've said before, FUD stands for fear, uncertainty, and doubt. I'm not talking about any of those things. I'm stating cold, hard facts: it costs more to re-train your staff to use Linux than it does to keep them on Windows, because the training costs associated with keeping them on Windows is zero. Your staff already knows how to use Windows, because they're already using it. It's simple math, not FUD.
Be that as it may, simply saying, ``it ain't necessarily so'' doesn't make for a very productive or interesting argument.
And your example is a terrible one. The idea the current flows from the positive pole to the negative pole springs-- when it does spring-- from the arbitrary names attached to the charges. If they were called ``right'' and ``left'' or ``square'' and ``round,'' your ``common sense'' would tell you something entirely different.
We're talking about drawing conclusions from facts: people know how to use Windows, but they don't know how to use Linux. Therefore, people will have to be taught how to use Linux, at a nonzero but undetermined cost. Training costs to maintain the status quo are fixed at zero. Voila, it costs more in training to move people to Linux than it costs to keep them on Windows. If the cost of the retraining (as yet undetermined) exceeds the cost of staying on Windows (also undetermined), it makes more sense to keep using Windows.
I'd love to hear somebody explain to me how that line of reasoning is flawed. Because I haven't heard it yet.
I'd love to hear what, exactly, you consider to be ``power user'' features in, say, a word processor. Printing, maybe? Spell-checking? Style sheets? Table of contents and index generation? Change control?
At my company, every employee, from the interns to the CEO, uses those features regularly to produce our documents. And these aren't complex documents; they're contracts, and proposals, and whitepapers. Standard stuff. To us, these aren't ``power user'' features. They're basic tools.
Maybe the problem, here, is that the developers (or at least advocates) of open source office software don't actually use office software themselves.
De facto standards are all well and good, but why should your application require to use anything outside the feature set supported by all browsers?
Because when we contracted to have this app, and others, written, ``all browsers'' consisted of Netscape 4 (which was long obsolete even then), IE 5 for Windows and Mac, and various second-tier products that were of no relevance to us. Mozilla didn't even exist yet.* It's unreasonable to expect that any software vendor should be held to a standard of forwards-compatibility. We would not expect them to guarantee that the apps would work with browsers that hadn't even been released yet. That's just absurd.
In any case, this job was done and paid for a long time ago, so we're in no position-- nor have we any particular desire-- to argue over it.
Also, just FYI, all our web apps are J2EE, not CGI. They have a surprising amount of presentation-layer logic, more perhaps than you might have guessed.
* Please don't talk to me, as so many others have, about the various 0.n versions. Neither I nor my company has time to mess with software that is of less-than-1.0 quality. As far as we're concerned, Mozilla didn't even exist until just a few weeks ago.
Did you not read my post in which I talked about de facto standards? All the paper standards in the world can't help you if you decide to release a web browser that doesn't render pages in the way that the user needs it to. Consider my situation. We spent a good chunk of change-- not millions, but several thousand bucks-- getting our trouble ticket application written and debugged. We have the option of either contracting with the vendor to open up that application and make the changes necessary to make it work properly with Mozilla 1.0, for a few thousand more bucks, or use IE instead of Mozilla, at no cost at all. Which do you think we chose?
With open source software it's not really an important question, since there's no new license to pay for.
Absolutely it's an important question. Whether the software costs money or not, it still has to be acquired, installed, and (possibly) configured. This is a trade-off. Windows software has to be bought, but installing and configuring it is (usually) very simple. Linux software, on the other hand, doesn't have to be bought, but is (usually) quite tedious to install and configure. So if you told me that the commercial product costs $100 and has a three-click installer, but that the free alternative consists of seventeen packages that must all be installed individually, I think the smart choice is to just buy the commercial product and be done with it. But that's my perspective; I'm in the position of having more money than time, at least when we're talking about $100.
Your post is well-written and well-thought-out. Unfortunately, I disagree with most of it. Rather than just saying ``nuh-uh'' about thirty times, I'll let you have the last word on this issue.
Driving a car? No. The gas is on the right and the brake is in the middle. The one on the left is the clutch. The big round thing in the middle is for turning. In front of you is a gauge that tells you how fast you're going.
Every car I've ever seen has those characteristics. Right-hand drive, left-hand drive, sports car, truck, whatever. The same basic user interface is found everywhere.
A more apt analogy would be to compare a car to a boat. You can do the same basic things-- forward, left, right, faster, slower-- but some of them have to be done in entirely different ways.
Jee-zus, talk about your knee-jerk reactions. If you read a little more closely, I was asking if OpenOffice has PDF generation built in or if it uses an external application. That's a very important question, because I need to know whether I can generate PDFs within OpenOffice by itself, or whether I'll have to make sure some external software is installed first.
I've gotta tell you, issues of cost and re-training notwithstanding, it's people like you that will put the open source software movement into an early grave.
1. There are literally millions of people out there who have experience using Windows and Windows applications for basic business tasks: email, browsing the web, document handling, and so on. Let's say that 5,000 of them work for MyCorp. If MyCorp wants to deploy 5,000 Linux desktops instead of 5,000 Windows desktops (savings: 5,000 times whatever MyCorp pays per head for Windows), they're going to have to re-train 5,000 people to use different (albeit similar) tools. Whether they have to be re-trained a little or a lot, it still adds up fast. It's not clear, because of a lot of things, whether it would cost more to deploy Linux or to stick with Windows. There is no clear, universally applicable cost-savings case associated with Linux. Blame it on inertia or the entrenchment of Windows or what-have-you, but it's the case nonetheless.
2. The idea of a de facto standard is not new, and it shouldn't be controversial. At my business, we use lots of web apps, for everything from order processing to HR. When one of my coworkers complained last month that the trouble ticket app wouldn't work properly with Mozilla, there was a sudden and final answer: use IE. You can wave all the paper standards you want; if the software isn't useful because of the way it renders-- or fails to render-- pages, then it gets replaced by the better-behaved incumbent. I'm sure this scenario is being played out over and over again all over the world, even as we speak.
Both of these points speak to the same basic idea: if the open source folks want to build software that can be seriously considered as an alternative to the tools people already use, they must not force the user base to compromise. I will change to another operating system if, and only if, it benefits me to do so. Giving me software that's equivalent to, but different from, what I use now will not motivate me to change. Giving me software that isn't as useful to me as what I have now sure as hell won't motivate me to change.
Let's get back to the subject at hand: we're talking about Shake here. If you're using Shake, you send jobs to the render farm to render; you don't use your desktop for that. This is true no matter what your desktop system is. So the only performance question revolves around interactivity: is the computer fast enough for you to interact creatively with the software? I played with Shake running on a dual-processor G4 at NAB, and it was very interactive. So the answer to that question is a qualified yes.
I could put a dozen 2 GHz Pentium-whatevers on my desk, but they'd just spend most of their time waiting on me. The heavy lifting is being done by the render farm in the basement.
So yup, I'm suggesting that a 50% speed difference between a Mac and a PC doesn't mean jack shit, as long as both PC and Mac are fast enough for the artist.
Thanks for the info. I'm glad to see that somebody is willing to discuss this subject with me.
When you use OpenOffice's ``send document as email'' feature, what exactly happens? Does OpenOffice have its own email functionality built in, or does it fire off another email program?
I have basically the same question about PDF generation. Does OpenOffice use its own PDF generator, or does it use Adobe's, or what?
Okay, sorry I jumped the gun on you. But even in your revised post, you still didn't give any specifics.
I'm not going to set up a Linux PC just so I can evaluate OpenOffice. I imagine that's probably the typical position that corporate IT folks will take as well. That's why I asked specific questions; I was expecting (rather, hoping for) specific answers.
Your comment was no help at all. If you've used both MS Office and Linux office applications, as you say, then you should have no trouble at all telling me specifically how one would do the things I asked about. Without some kind of specific information, I'm afraid your post really carries no weight at all.
Man, somebody really has a hard-on for me today. That's three ``-1, Flamebait'' moderations on two posts in just the past hour or so. One was applied to a post in which I said that the TCO of Linux on the desktop compares negatively to the TCO of Windows on the desktop, and two were applied to the parent, in which I said that browsers that fail to render pages the same way IE does should be considered broken. I'm glad to see that reasoned, critical opinions are treated with such respect from the Slashdot crowd.
This has nothing to do with advanced users. This has to do with my secretary, who knows how to email Word documents and web pages to me by going to the ``File'' menu and selecting ``Send To Mail Recipient'' (or whatever it is). Up pops an email message with the file attached, and she just types my name and clicks ``Send.'' She's not dumb, by any stretch of the imagination, but she's not a computer hobbyist, either. She doesn't mess with 'em for fun. She uses her PC as a tool to get her job done.
When you're using (insert name of word processor here) under Linux, can you go to the ``File'' menu (or any menu), select ``Sent To Mail Recipient'' (or something like that), type in an email address, and click ``Send?'' If you can't, then my secretary will have to be re-trained to use Linux. Can you create a PDF by printing to the Acrobat Distiller printer? If you can't, then my secretary will have to be re-trained to use Linux.
I'm not speaking from a position of experience here; I have no idea how to do those basic, ten-times-every-day tasks with any combination of applications under Linux. But no matter how easy it may be, it's different, and that means I'll have to re-train my secretary. That'll cost me money and time. If the cost of re-training my secretary exceeds the cost of a copy of Windows plus the necessary software, then Linux costs me more than Windows.
You said, ``the functionality that they do make use of is already replicated quite well under Linux.'' Back that up. Explain to me, please, how to do the following task with any combination of applications for Linux.
1. Open the Excel file containing such-and-such vendor's price list. (We get these from our vendors, so we don't get a say in their format. They come as Excel files.)
2. Open the word processing document that I just emailed you and find the table cell marked, ``paste description here.'' (I create that document, so it can be in any format, as long as it supports in-line tables.) Paste the description of part so-n-so from the price list into the table cell.
3. Turn the word processing document into a PDF, and mail it back to me for approval.
In real life, the job would be a lot more complex, of course, but it would basically consist of that list of steps, permuted in various ways. She can carry that job out very easily with Windows, using Excel, Word, Outlook, and Acrobat Distiller. How would you do that job with Linux and Linux applications? If you can do it at all, is the process close to or very different from the process my secretary currently uses? (The more different, the more it will cost me to re-train.)
I'm going to stand behind my call of ``blindingly obvious'' on this one.
You're under the mistaken impression that people buy Macs because Macs are fast, therefore the fact that Macs aren't as fast as PCs means more people will buy PCs. That's wrong.
People buy Macs so that they (the people, not the computers) will be productive. I have a 750 MHz PIII on my desk (using it now) and a 500 MHz iBook. Right now, I'm using the PC to surf and goof off, but in a minute I'm going to go back to using Illustrator and InDesign on my iBook. It's not that my iBook is faster, and it's not that I'm using software that's not available on Windows. It's that I'm more productive when I'm using a Mac. Lots and lots of people feel the same way; even if it's only 5-10% of the desktop computer market, it's still millions of people.
Bottom line: fast is nice and all, but there are things that are more important than fast. That's where Apple's market share comes from.
I knew it! I knew it! I was even thinking, as I posted my comment, ``-1, Flamebait.'' And there it is!
I shouldn't be surprised, of course, but it always catches me off-guard when the Slashdot community (for lack of a better word) operates so predictably. Any comment critical of Microsoft gets +1, Insighful, regardless of merit. Any comment critical of Linux gets -1, Flamebait, regardless of merit.
I understand your point, but I think you're being kinda silly. What research is necessary to show that people who are trained to use MS Office under Windows have to be re-trained to use some other application under Linux? This is blindingly obvious. And while the learning curve from one platform to the other may be shallow in some specific areas-- like typing in a word processor, for instance-- it gets steeper when you move into other common tasks, like emailing a word processor document to a colleague. You do it one way with Word and Outlook, and an entirely different way with whatever-you-please on Linux. Voila, retraining.
You don't need a trade study to know how zero retraining on the left compares to non-zero retraining on the right.
Slashdot Mirror
And this is exactly why OSS people will always be just a bunch of bickering dilettantes. They prefer to own it all, just so they can have the hubris of giving it away.
LOL. I love it. That explains so many things. OSSers love to re-invent the wheel because pointing at an inferior copy of somebody else's application and muttering the magic incantation, ``But this one is FREE,'' gives them a heady rush of moral superiority that far outweighs the feelings of technical inferiority they get from having written it in the first place.
This is my favorite new theory on why OSSers are they way they are.
It reminds me of a great quote from Douglas Adams; it's not exactly the same idea, but I think it expresses the same sentiment, more or less. I'm too lazy to look it up right now. I'll paraphrase: The sense of satisfaction a user gets from getting the machines to work at all blinds the user to the machines' essential uselessness. In other words, their superficial design flaws completely hide their fundamental design flaws.
Mistake? Where did you get the idea that we made a mistake? We've been deploying web applications for key business operations gradually over the past couple of years, and our overall efficiency has skyrocketed. It used to take more than a week to get a sales order filled for a customer. Now that order can be entered in the morning, product ships around lunchtime, and the invoice goes out that afternoon. There's no mistake here.
This is my whole point, really: W3C standards are absolutely irrelevant to us. They're just pieces of paper as far as I'm concerned. What's relevant to my company is our existing base of web applications. If Mozilla (or whatever) can't run those applications well enough that the users don't notice a difference, then Mozilla won't find a home here.
Why get religious about standards? They're important only to the extent that they improve things. Releasing a browser that's 100% compliant with the specification of the standard but that can't do the job of the incumbent browser doesn't help anybody. The Mozilla team should have spent more time implementing compatibility with IE-- or at least an IE compatibility mode that can be turned on and off in the ``advanced'' pane of the preferences dialog-- and less time implementing features that nobody needs. How much time was spent implementing the ``skinnable'' interface? How much time was spent on the mail and news readers? The browser doesn't have a goddamn ``home'' button on the toolbar, but it reads news, and look at all these kewl skinz, d00d!
What a waste. What an unbelievable waste of time and effort. Is it any wonder that IE is the more popular browser? For all their faults, Microsoft implemented features that developers and users actually wanted, while the Mozilla coalition gave us kewl skinz.
If you want to talk about mistakes, I think we can start right there.
I agree. Improvement is a good thing. But Ogg isn't an improvement on MP3, except in some very marginal ways. Best I can discern, Ogg is about equivalent to MP3, except for the lack of software and hardware that supports it.
Wish I had mod points. After a discussion a few days ago, on the occasion of the 1.0 release, I came to the same conclusion. Ogg offers a couple of features that MP3 lacks, but they are useful only in very specific situations. I'm a heavy user of MP3s-- as I mentioned elsewhere, I have ripped my collection of about 400 CDs to MP3, and I play selections from that library practically every day-- and I can't come up with a good reason why I would be interested in Ogg. The software is there, the hardware is there. Unless something changes drastically, I predict that Ogg will pretty much continue to be irrelevant to all but a select few.
All, interestingly enough, of these sources have egos the size of Texas and consciences the size of Guam.
Bad example. While Guam might be small compared to Texas, it's still pretty darned big in the absolute sense. I.e., it's a lot bigger than I am. For future reference, I would choose something like ``consciences the size of a flea.''
On a more serious note, might I ask that you consider being a little bit more careful with your words in the future. Saying this like, ``All of these sources [politicians, CEOs, doctors...] have egos the size of Texas,'' is an overgeneralization, and a rude one. My girlfriend is a doctor, as are most of our friends. While there are egoists out there, I'd have to say that all but one or two of the doctors I know are the most humble people you could hope to meet. We've talked about it before, over dinners and such: they all agree that being a doctor is an overwhelming responsibility. If you let it go to your head, you might get arrogant, but the fact is that simple fear keeps you from thinking too highly of yourself.
I'd just ask that you think twice before generalizing in the future.
Vote? What vote? This is a company, not a cub scout troop. TCO is the only legitimate reason to choose one type of technology over another. If you believe that you can do business by standing on your principles and using Linux just because it's ``free'' (whatever that means), then you're in for an unpleasant surprise.
And as for FUD, as I've said before, FUD stands for fear, uncertainty, and doubt. I'm not talking about any of those things. I'm stating cold, hard facts: it costs more to re-train your staff to use Linux than it does to keep them on Windows, because the training costs associated with keeping them on Windows is zero. Your staff already knows how to use Windows, because they're already using it. It's simple math, not FUD.
Be that as it may, simply saying, ``it ain't necessarily so'' doesn't make for a very productive or interesting argument.
And your example is a terrible one. The idea the current flows from the positive pole to the negative pole springs-- when it does spring-- from the arbitrary names attached to the charges. If they were called ``right'' and ``left'' or ``square'' and ``round,'' your ``common sense'' would tell you something entirely different.
We're talking about drawing conclusions from facts: people know how to use Windows, but they don't know how to use Linux. Therefore, people will have to be taught how to use Linux, at a nonzero but undetermined cost. Training costs to maintain the status quo are fixed at zero. Voila, it costs more in training to move people to Linux than it costs to keep them on Windows. If the cost of the retraining (as yet undetermined) exceeds the cost of staying on Windows (also undetermined), it makes more sense to keep using Windows.
I'd love to hear somebody explain to me how that line of reasoning is flawed. Because I haven't heard it yet.
I'd love to hear what, exactly, you consider to be ``power user'' features in, say, a word processor. Printing, maybe? Spell-checking? Style sheets? Table of contents and index generation? Change control?
At my company, every employee, from the interns to the CEO, uses those features regularly to produce our documents. And these aren't complex documents; they're contracts, and proposals, and whitepapers. Standard stuff. To us, these aren't ``power user'' features. They're basic tools.
Maybe the problem, here, is that the developers (or at least advocates) of open source office software don't actually use office software themselves.
De facto standards are all well and good, but why should your application require to use anything outside the feature set supported by all browsers?
Because when we contracted to have this app, and others, written, ``all browsers'' consisted of Netscape 4 (which was long obsolete even then), IE 5 for Windows and Mac, and various second-tier products that were of no relevance to us. Mozilla didn't even exist yet.* It's unreasonable to expect that any software vendor should be held to a standard of forwards-compatibility. We would not expect them to guarantee that the apps would work with browsers that hadn't even been released yet. That's just absurd.
In any case, this job was done and paid for a long time ago, so we're in no position-- nor have we any particular desire-- to argue over it.
Also, just FYI, all our web apps are J2EE, not CGI. They have a surprising amount of presentation-layer logic, more perhaps than you might have guessed.
* Please don't talk to me, as so many others have, about the various 0.n versions. Neither I nor my company has time to mess with software that is of less-than-1.0 quality. As far as we're concerned, Mozilla didn't even exist until just a few weeks ago.
Did you not read my post in which I talked about de facto standards? All the paper standards in the world can't help you if you decide to release a web browser that doesn't render pages in the way that the user needs it to. Consider my situation. We spent a good chunk of change-- not millions, but several thousand bucks-- getting our trouble ticket application written and debugged. We have the option of either contracting with the vendor to open up that application and make the changes necessary to make it work properly with Mozilla 1.0, for a few thousand more bucks, or use IE instead of Mozilla, at no cost at all. Which do you think we chose?
With open source software it's not really an important question, since there's no new license to pay for.
Absolutely it's an important question. Whether the software costs money or not, it still has to be acquired, installed, and (possibly) configured. This is a trade-off. Windows software has to be bought, but installing and configuring it is (usually) very simple. Linux software, on the other hand, doesn't have to be bought, but is (usually) quite tedious to install and configure. So if you told me that the commercial product costs $100 and has a three-click installer, but that the free alternative consists of seventeen packages that must all be installed individually, I think the smart choice is to just buy the commercial product and be done with it. But that's my perspective; I'm in the position of having more money than time, at least when we're talking about $100.
Your post is well-written and well-thought-out. Unfortunately, I disagree with most of it. Rather than just saying ``nuh-uh'' about thirty times, I'll let you have the last word on this issue.
Driving a car? No. The gas is on the right and the brake is in the middle. The one on the left is the clutch. The big round thing in the middle is for turning. In front of you is a gauge that tells you how fast you're going.
Every car I've ever seen has those characteristics. Right-hand drive, left-hand drive, sports car, truck, whatever. The same basic user interface is found everywhere.
A more apt analogy would be to compare a car to a boat. You can do the same basic things-- forward, left, right, faster, slower-- but some of them have to be done in entirely different ways.
Jee-zus, talk about your knee-jerk reactions. If you read a little more closely, I was asking if OpenOffice has PDF generation built in or if it uses an external application. That's a very important question, because I need to know whether I can generate PDFs within OpenOffice by itself, or whether I'll have to make sure some external software is installed first.
I've gotta tell you, issues of cost and re-training notwithstanding, it's people like you that will put the open source software movement into an early grave.
Look, this really isn't that complicated.
1. There are literally millions of people out there who have experience using Windows and Windows applications for basic business tasks: email, browsing the web, document handling, and so on. Let's say that 5,000 of them work for MyCorp. If MyCorp wants to deploy 5,000 Linux desktops instead of 5,000 Windows desktops (savings: 5,000 times whatever MyCorp pays per head for Windows), they're going to have to re-train 5,000 people to use different (albeit similar) tools. Whether they have to be re-trained a little or a lot, it still adds up fast. It's not clear, because of a lot of things, whether it would cost more to deploy Linux or to stick with Windows. There is no clear, universally applicable cost-savings case associated with Linux. Blame it on inertia or the entrenchment of Windows or what-have-you, but it's the case nonetheless.
2. The idea of a de facto standard is not new, and it shouldn't be controversial. At my business, we use lots of web apps, for everything from order processing to HR. When one of my coworkers complained last month that the trouble ticket app wouldn't work properly with Mozilla, there was a sudden and final answer: use IE. You can wave all the paper standards you want; if the software isn't useful because of the way it renders-- or fails to render-- pages, then it gets replaced by the better-behaved incumbent. I'm sure this scenario is being played out over and over again all over the world, even as we speak.
Both of these points speak to the same basic idea: if the open source folks want to build software that can be seriously considered as an alternative to the tools people already use, they must not force the user base to compromise. I will change to another operating system if, and only if, it benefits me to do so. Giving me software that's equivalent to, but different from, what I use now will not motivate me to change. Giving me software that isn't as useful to me as what I have now sure as hell won't motivate me to change.
Let's get back to the subject at hand: we're talking about Shake here. If you're using Shake, you send jobs to the render farm to render; you don't use your desktop for that. This is true no matter what your desktop system is. So the only performance question revolves around interactivity: is the computer fast enough for you to interact creatively with the software? I played with Shake running on a dual-processor G4 at NAB, and it was very interactive. So the answer to that question is a qualified yes.
I could put a dozen 2 GHz Pentium-whatevers on my desk, but they'd just spend most of their time waiting on me. The heavy lifting is being done by the render farm in the basement.
So yup, I'm suggesting that a 50% speed difference between a Mac and a PC doesn't mean jack shit, as long as both PC and Mac are fast enough for the artist.
Thanks for the info. I'm glad to see that somebody is willing to discuss this subject with me.
When you use OpenOffice's ``send document as email'' feature, what exactly happens? Does OpenOffice have its own email functionality built in, or does it fire off another email program?
I have basically the same question about PDF generation. Does OpenOffice use its own PDF generator, or does it use Adobe's, or what?
Thanks again.
Good thing we're not talking about motorcycles, then.
Okay, sorry I jumped the gun on you. But even in your revised post, you still didn't give any specifics.
I'm not going to set up a Linux PC just so I can evaluate OpenOffice. I imagine that's probably the typical position that corporate IT folks will take as well. That's why I asked specific questions; I was expecting (rather, hoping for) specific answers.
Your comment was no help at all. If you've used both MS Office and Linux office applications, as you say, then you should have no trouble at all telling me specifically how one would do the things I asked about. Without some kind of specific information, I'm afraid your post really carries no weight at all.
Man, somebody really has a hard-on for me today. That's three ``-1, Flamebait'' moderations on two posts in just the past hour or so. One was applied to a post in which I said that the TCO of Linux on the desktop compares negatively to the TCO of Windows on the desktop, and two were applied to the parent, in which I said that browsers that fail to render pages the same way IE does should be considered broken. I'm glad to see that reasoned, critical opinions are treated with such respect from the Slashdot crowd.
Oh, well. Just blowin' karma today, after all.
This has nothing to do with advanced users. This has to do with my secretary, who knows how to email Word documents and web pages to me by going to the ``File'' menu and selecting ``Send To Mail Recipient'' (or whatever it is). Up pops an email message with the file attached, and she just types my name and clicks ``Send.'' She's not dumb, by any stretch of the imagination, but she's not a computer hobbyist, either. She doesn't mess with 'em for fun. She uses her PC as a tool to get her job done.
When you're using (insert name of word processor here) under Linux, can you go to the ``File'' menu (or any menu), select ``Sent To Mail Recipient'' (or something like that), type in an email address, and click ``Send?'' If you can't, then my secretary will have to be re-trained to use Linux. Can you create a PDF by printing to the Acrobat Distiller printer? If you can't, then my secretary will have to be re-trained to use Linux.
I'm not speaking from a position of experience here; I have no idea how to do those basic, ten-times-every-day tasks with any combination of applications under Linux. But no matter how easy it may be, it's different, and that means I'll have to re-train my secretary. That'll cost me money and time. If the cost of re-training my secretary exceeds the cost of a copy of Windows plus the necessary software, then Linux costs me more than Windows.
You said, ``the functionality that they do make use of is already replicated quite well under Linux.'' Back that up. Explain to me, please, how to do the following task with any combination of applications for Linux.
1. Open the Excel file containing such-and-such vendor's price list. (We get these from our vendors, so we don't get a say in their format. They come as Excel files.)
2. Open the word processing document that I just emailed you and find the table cell marked, ``paste description here.'' (I create that document, so it can be in any format, as long as it supports in-line tables.) Paste the description of part so-n-so from the price list into the table cell.
3. Turn the word processing document into a PDF, and mail it back to me for approval.
In real life, the job would be a lot more complex, of course, but it would basically consist of that list of steps, permuted in various ways. She can carry that job out very easily with Windows, using Excel, Word, Outlook, and Acrobat Distiller. How would you do that job with Linux and Linux applications? If you can do it at all, is the process close to or very different from the process my secretary currently uses? (The more different, the more it will cost me to re-train.)
I'm going to stand behind my call of ``blindingly obvious'' on this one.
You're under the mistaken impression that people buy Macs because Macs are fast, therefore the fact that Macs aren't as fast as PCs means more people will buy PCs. That's wrong.
People buy Macs so that they (the people, not the computers) will be productive. I have a 750 MHz PIII on my desk (using it now) and a 500 MHz iBook. Right now, I'm using the PC to surf and goof off, but in a minute I'm going to go back to using Illustrator and InDesign on my iBook. It's not that my iBook is faster, and it's not that I'm using software that's not available on Windows. It's that I'm more productive when I'm using a Mac. Lots and lots of people feel the same way; even if it's only 5-10% of the desktop computer market, it's still millions of people.
Bottom line: fast is nice and all, but there are things that are more important than fast. That's where Apple's market share comes from.
I knew it! I knew it! I was even thinking, as I posted my comment, ``-1, Flamebait.'' And there it is!
I shouldn't be surprised, of course, but it always catches me off-guard when the Slashdot community (for lack of a better word) operates so predictably. Any comment critical of Microsoft gets +1, Insighful, regardless of merit. Any comment critical of Linux gets -1, Flamebait, regardless of merit.
Good thing none of this matters.
I understand your point, but I think you're being kinda silly. What research is necessary to show that people who are trained to use MS Office under Windows have to be re-trained to use some other application under Linux? This is blindingly obvious. And while the learning curve from one platform to the other may be shallow in some specific areas-- like typing in a word processor, for instance-- it gets steeper when you move into other common tasks, like emailing a word processor document to a colleague. You do it one way with Word and Outlook, and an entirely different way with whatever-you-please on Linux. Voila, retraining.
You don't need a trade study to know how zero retraining on the left compares to non-zero retraining on the right.