I take it you write your own user docs. Have you considered the possibility that you're not very good at it? Do you know how explain complicated features in a simple manor? Do you have the patience to discuss nit-picky details that you know intuitively but about which your users are totally clueless? Do you know how to organize an immense body of facts so that the reader doesn't go crazy trying to find one small unimportant fact -- unimportant to you, but essential to the user?
Too many engineers look at tech writers as clueless English majors, useful only for cleaning up spelling and grammar. Or arrogant, burned-out former programmers who think they know everything, and really know very little. True, there are a lot of tech writers like that, and their product is not worth reading -- assuming anybody can read it. But there are also serious, motivated tech writers who know a lot about communicating technical subjects.
I like to think I'm one of those. I'm biased of course, but I have been told, more than once, that nobody understood how a product or technology really worked until I sorted a huge pile of random facts into a useful form. (This is especially nice to hear when it comes from the people who designed the system in the first place!) I'm probably not the best in my field, but I think I earn my pay. (Well, no pay right now, industry slump y'know. Oh well.)
Of course, not all documentation really needs a tech writer. You sound like you mostly write little end-user apps. For those, I agree, a good GUI is more important than a good manual.
But consider my own specialty, the API manual. How many of those have you seen that don't make you scream with frustration? Writing good API docs is hard. (Lotta fun though, at least for a compulsive nit-picker like me.)
And writing isn't as hard as maintaining. My last job involved a development framework with more than 10,000 APIs. Which was maintained in RTF. (Please don't laugh, it's not funny.) And which had to be single sourced for four different product targets. (Windows and Linux, two different programming languages.) And, oh yeah, my boss thought that version control was a silly idea.
Why was this documentation base such a mess? Because at this company, the "nobody reads the docs" mentality prevailed. Even the writing team was infected with it. And this self-fulfilling cynicism really hurt the product. The API has a reputation for being obscure and hard to use. Whereas it's really pretty elegant, and even easy to understand, if properly explained. In this case, bad documentation is doing a lot to consign a superior product to undeserved oblivion.
I should end with that pithy comment, but I have to drag the discussion back ontopic. Because of the company's indifference to doc issues, they're only now converting the documents from RTF to markup, something they should have done 10 years ago. Alas, the project is headed up by an intelligent but technologically clueless individual who thinks a little XML transform experience makes him an expert on content management. (Sour grapes? I guess. Then again, I did recommend hiring the guy.) Last I heard, the project was months behind schedule, and was close to being in deathmarch mode.
So I'm deeply interested reading Ament's book. Maybe it'll be useful on my next job. But even if it's well-written, I don't think I'll enjoy the read. Too many lost opportunities.
Well, Sun's worse than most. But I've never worked for any large organization that didn't have this kind of problem. Getting a lot of people to work together without screwing each other up is not easy. Especially when a lot of them have big egos.
jesus! RTFM! what exactly did you used to do at Sun when you worked there? stock the vending machines?
RTF what M? All the docs say is, "run the installer". I run the installer, which is some weird homebrew thing. I end up with a.BAT file that runs a complicated java.exe command. This command fails because it can't find the entry point class. This may have something to do with the installer assuming that my system disk is C: when it's actually F:. Call me lazy, but I can't be bothered to debug this.
I have to wonder why this program doesn't use a commercial installer (Sun has site licenses for InstallShield and ZeroG). And why the program isn't packaged as an executable JAR (rather than a library JAR), which would simplify the installation and avoid potential (or in this case, real) CLASSPATH issues. None of which makes me confident that the program is solid or properly supported.
Anyway, even if there were no problems like this, making the installation so complicated makes the whole thing kind of pointless. The Netscape and Real download managers install themselves when you click on a "give me the product" link. If Sun isn't going to do this, they should just provide a link to a freeware download manager, one that's been around and known to be stable. Politics. Bureaucracy.
Calling it the "experience" is just a pretentious way of saying that Macs are easy to use. (Maybe it also covers paying premium prices for hardware just because the case looks cool. Yawn.) I can't believe it's that hard to duplicate a Mac's ease-of-use features. But even if it is, so what? People who build their own computers are sophisticated enough to live with that. And lots of us would be very happy to see a low-budget way to play with software that only runs on the Mac platform.
I saw a piece (Charles Kuralt I think) on the building of the National Cathedral some time ago. One way they pay for the building is by allowing people to design gargoyles, for a fee. One woman had a bunch of gargoyles made that are caricatures of her grandchildren. Unsuprising that a Star Wars fan took advantage of this. The only strange thing is that it didn't happen years ago. Anyway, basing a gargoyle on a fantastic or imaginary creature has a long tradition.
Sorry, wrong cliche. But you get my drift. Right now it's possible to access the internet directly, even if your provider wants you to go through their "portal". (Although, as you point out, it can be a pain in the ass, especially with AOL.) Maybe most users are too tech-illiterate to do this, but that's changing rapidly.
Which is why some providers want to control how you access the internet. Want to surf the web? Gotta go through the portal, no generic IP applications. Want to untether your access? No do-it-yourself WiFi, gotta lease their web tablet. Want to plug in multiple computers? Gotta lease a router from the provider.
The obvious implication is cost, of course. But what worries me is the free flow of information. Big media companies have a bad track record for providing equal access to all sources of information. It's not so much censorship as a desire to avoid offending kneejerk pressure groups, and thus losing business. Still, the effect is the same.
There's a certain irony to seeing Disney and Microsoft complaining about this trend. Disney owns go.com, which started out as Infoseek, a leading search engine; Disney spent billions in an unsuccessful attempt to create a click-monopoly. Microsoft owns MSN, which was originally an unsuccessful AOL clone, then became a struggling ISP and portal. If either company had had more success with their own attempts to monopolize clicks, they'd be singing a different tune!
You're right, nobody told me. So, just out of curiousity, I installed it. Just a few problems.
It only accesses the "Sun download center", not the JDK files on java.sun.com.
There's a JDK on the SDC. Not the latest beta, but newbies want a production release anyway. Except that I can't find a link to the Windows version. It's supposed to be there, but whatever I do I always end up with the Solaris/86 version!
There's a link to the SDC on java.sun.com, but it isn't very conspicious. Someone browsing java.sun.com could be excused for thinking that the SDC didn't include any Java software.
I couldn't get the stupid download manager to work, anyway.
All this weirdness would seem to be the result of the long rivalry between two of Sun's operating companies, JavaSoft and SunSoft. Of course, they're called something else now (I can't be bothered to keep up), but the rivalry it the same as ever. Did I mention that there's a lot of politics and bureaucracy at Sun?
Sure, it'll open the MS equivalent of a RTF but throw complicated formatting at it and it bombs. My company uses Excel *a lot* and we have some pretty complicated formulas which we *depend on* to get what we do done. Open Office isn't an alternative. Period.
More to the point, it never will be. By which I mean it will never be practical to run a mixed Open/Microsoft shop. For that to happen, Open Office would have to be 100% compatible. Not 90%, not 99%, not even 99.9%. 'Cause even if the incompatibility is trivial, Murphy's Law dictates that it will come into play when you can least tolerate it.
Which is why I find Open Office's emphasis on MS compatibility depressing. They even try to support Visual Basic macros!
It's not suprising. Open Office development is driven by Sun. A company that seems incapable of admitting that Microsoft won the desktop wars long, long ago. But it's too bad. I'm starting to get impressed with OO's maturity, feature set, and API. These might form the basis for solid competition with Microsoft -- if Sun ever stops fighting battles that are already lost.
You seem to be in love with the word "castrated". Getting your libido suppressed by drugs is a long way from having your genitals mutilated.
The fact is, most older men go through similar hormonal changes (I'm going through them myself). They're unpleasant and undignified and embaressing. And in Turing's case they must have been particularly bad. But they're still quite different from becomming a eunuch.
You're making inferences about Turing's state of mind based on what you think he went through and how you think he must have responded. That's not totally unreasonable, but it simply isn't consistent with what Turing said and wrote during that period. Yes, he was persecuted and abused. But not to the point of fatal depression and suicide.
That's another thing. People have these stereotypical notions of depression. Somebody goes through a bad patch, and then dies violently, people find it easy to assume that they were depressed and suicidal. And if you're going to have somebody killed, you count on that stereotype.
But in the real world, everybody's different. Some people will kill themselves over minor embaressments. Others can undergo the most extreme embaressment and humiliation and still maintain their sense of self-worth. Reading about Turing's life convinces me he was closer to the later than the former.
I agree, this wouldn't be easier if the user was working with a single version. But it might be easier than downloading all of those versions of the JDK, if the user wants to try each one.
Strictly speaking, there is only one version. Or if you want to get really picky, there's four versions, three of which are unsupported. There are an ungodly number of releases, which contain things like bug fixes and minor new features. But these are supposed to be backward compatible with the old versions. Anything that's not is a bug.
Of course, in the real world, bugs happen, and people need to use old versions of the JDK to work around them. But anybody who's working at that level is a sophisticated Java programmer, not a newbie. Such a person does need the whole JDK!
In theory you're right. But that single file is huge. When I worked for JavaSoft, there was a lot of negative feedback from people who had slow and/or unreliable internet connections, and couldn't keep an ftp or http connection open long enough.
The obvious solution is to provide some kind of download manager (as Netscape and Real now do). Or organize the Java SDK into smaller independent packages. Or allow outsiders to start up mirror sites overseas. Or... But life at Sun is full of politics and bureaucracy (which is why I no longer work there), and getting even the simplest solution in place is like pulling teeth.
Real-time programs, especially games, tend to have a lot of hand-optimized assembly code. Lots of programmers trust their own optimizations better than those of a compiler -- though I suspect most are wrong to do so. Anyway, porting assembly language between two Pentium-based platforms is easy. Whereas porting to PPC essentially means starting from scratch.
There's also the testing issue. I'm suprised they were willing to spend the bucks to do QA on two different platforms. Forget about three. Especially when the third one requires special hardware on which to do the testing.
Jeesh, nobody has a sense of irony any more. In fact, fans of MySQL consider it a relational dbms. It has relational features, but they don't work all that well.
I find it easy to understand why the Slashdot editors (and probably most readers) are uninterested in Codd's life or death. He did come up with the basic ideas that make modern databases work. But nowadays people take these ideas for granted.
You see the same thing in math. Everybody knows how to prove that there's no largest natural number, but who's ever heard of Georg Cantor? The problem is that his theories suffuse modern math, and his important ideas are extremely easy to understand, so nobody feels obliged to give him credit.
Turing had just been forcibly castrated by court order for the heinous crime of being homosexual,
You're misinformed. (Internet rumor? Read Hodges's book.) Turing had been forced to undergo hormone therapy that was supposed to "cure" his sexual orientation. Barbaric, but not enough to destroy him, either psychologically or sexually. He was well past that stuff, and even sexually active again, when he died.
Hm, Von Neumann's role as a computer science pioneer is controversial. He did forsee the coming dominance of "stored-programming computers" -- but contrary to popular belief, he didn't invent the model.
Now, Turing has some claim to being the very first computer scientist. His attempts to explain mathematical logic in terms of mechanistic abstractions predates the computer revolution by a couple of decades. And when actual computers did start to appear, he was one of the first to deal seriously with what we now call software.
But here's another bummer. It seems pretty obvious that the security bozos murdered Turing because they couldn't stand the idea of faggot with all those state secrets in his head. I should emphasize that this is my paranoid theory, not Hodges's -- as a UK citizen, he could get in trouble just for suggesting the possibility.
Who needs relational databases?
on
MIT Gnome Invasion
·
· Score: 1, Offtopic
I hate environmentalists who don't do their homework. They play into the hands of anti-environmentalists who think that making fun of the evironmental movement answers all arguments.
Agriculture has a huge environmental impact. There's toxic runoff (even with "organic" methods), erosion, replacement of natural growth with an artificial monoculture. Worst of all, you have to burn a lot of fuel (and create a lot of greenhouse gases) to grow a crop. Nobody who cares about the environment should advocate more agriculture than is necessary to feed people or produce other necessary products.
Not that non-fossil diesels are necessarily a bad idea. These guys burn vegetable oil in their diesel boats so they won't spread particulate all over the marine sanctuary they cruise over. But that's a limited, specific goal. As a general replacement for fossil fuels, "crop" fuels are just plain stupid.
When you post a story like this, say "pf (a packet filter package)" or something like that. Otherwise people have to click a half dozen times just to find out if they're interested in the story!
OK, I'm not up on this issue, but I don't see how you can call ethanol "envirofriendly". Yes, burning ethanol produces less pollution than methanol, to say nothing of diesel fuel. But you have to figure in the environmental cost of producing the ethanol. Don't you have to grow sugar beets or something similar, then ferment and refine the result? These are all energy-intensive and unenvironmental processes!
The purpose of things like zoning laws and homeowner agreements is to protect people from inconsiderate neighbors. In places were "zoning" is a dirty word, you can end up with a sweatshop or a nightclub or even a factory next door, and be unable to do anything about it.
On the other hand, lots of people break these rules right and left, and nobody cares, because they're considered good neighbors. I have a friend who's operating a business out of her condo in total violation of her housing association rules. All the other condo owners know about it, but she's such a valued member of the community (networks a lot, goes out of her way to make friends and help people) that nobody's inclined to make an issue of it.
Of course, if she ever does make a enemy who want to shut her down, she's screwed! Worth bearing in mind before you build that illegal granny flat.
I would think a Goldtouch would work well for you -- unless you have extremely strange arms! The whole point of the hinge design is to allow you to stick your elbows out in a natural manner.
But I'm no ergonomics expert. You should send your question to their support people.
Slashdot Mirror
Too many engineers look at tech writers as clueless English majors, useful only for cleaning up spelling and grammar. Or arrogant, burned-out former programmers who think they know everything, and really know very little. True, there are a lot of tech writers like that, and their product is not worth reading -- assuming anybody can read it. But there are also serious, motivated tech writers who know a lot about communicating technical subjects.
I like to think I'm one of those. I'm biased of course, but I have been told, more than once, that nobody understood how a product or technology really worked until I sorted a huge pile of random facts into a useful form. (This is especially nice to hear when it comes from the people who designed the system in the first place!) I'm probably not the best in my field, but I think I earn my pay. (Well, no pay right now, industry slump y'know. Oh well.)
Of course, not all documentation really needs a tech writer. You sound like you mostly write little end-user apps. For those, I agree, a good GUI is more important than a good manual.
But consider my own specialty, the API manual. How many of those have you seen that don't make you scream with frustration? Writing good API docs is hard. (Lotta fun though, at least for a compulsive nit-picker like me.)
And writing isn't as hard as maintaining. My last job involved a development framework with more than 10,000 APIs. Which was maintained in RTF. (Please don't laugh, it's not funny.) And which had to be single sourced for four different product targets. (Windows and Linux, two different programming languages.) And, oh yeah, my boss thought that version control was a silly idea.
Why was this documentation base such a mess? Because at this company, the "nobody reads the docs" mentality prevailed. Even the writing team was infected with it. And this self-fulfilling cynicism really hurt the product. The API has a reputation for being obscure and hard to use. Whereas it's really pretty elegant, and even easy to understand, if properly explained. In this case, bad documentation is doing a lot to consign a superior product to undeserved oblivion.
I should end with that pithy comment, but I have to drag the discussion back ontopic. Because of the company's indifference to doc issues, they're only now converting the documents from RTF to markup, something they should have done 10 years ago. Alas, the project is headed up by an intelligent but technologically clueless individual who thinks a little XML transform experience makes him an expert on content management. (Sour grapes? I guess. Then again, I did recommend hiring the guy.) Last I heard, the project was months behind schedule, and was close to being in deathmarch mode.
So I'm deeply interested reading Ament's book. Maybe it'll be useful on my next job. But even if it's well-written, I don't think I'll enjoy the read. Too many lost opportunities.
Well, Sun's worse than most. But I've never worked for any large organization that didn't have this kind of problem. Getting a lot of people to work together without screwing each other up is not easy. Especially when a lot of them have big egos.
Gawd, I can just imagine trying to convince the Sun bureaucracy to deploy something like Bit Torrent. Can you say, "snowflake in hell"?
Bored now.
I have to wonder why this program doesn't use a commercial installer (Sun has site licenses for InstallShield and ZeroG). And why the program isn't packaged as an executable JAR (rather than a library JAR), which would simplify the installation and avoid potential (or in this case, real) CLASSPATH issues. None of which makes me confident that the program is solid or properly supported.
Anyway, even if there were no problems like this, making the installation so complicated makes the whole thing kind of pointless. The Netscape and Real download managers install themselves when you click on a "give me the product" link. If Sun isn't going to do this, they should just provide a link to a freeware download manager, one that's been around and known to be stable. Politics. Bureaucracy.
Calling it the "experience" is just a pretentious way of saying that Macs are easy to use. (Maybe it also covers paying premium prices for hardware just because the case looks cool. Yawn.) I can't believe it's that hard to duplicate a Mac's ease-of-use features. But even if it is, so what? People who build their own computers are sophisticated enough to live with that. And lots of us would be very happy to see a low-budget way to play with software that only runs on the Mac platform.
I saw a piece (Charles Kuralt I think) on the building of the National Cathedral some time ago. One way they pay for the building is by allowing people to design gargoyles, for a fee. One woman had a bunch of gargoyles made that are caricatures of her grandchildren. Unsuprising that a Star Wars fan took advantage of this. The only strange thing is that it didn't happen years ago. Anyway, basing a gargoyle on a fantastic or imaginary creature has a long tradition.
Which is why some providers want to control how you access the internet. Want to surf the web? Gotta go through the portal, no generic IP applications. Want to untether your access? No do-it-yourself WiFi, gotta lease their web tablet. Want to plug in multiple computers? Gotta lease a router from the provider.
The obvious implication is cost, of course. But what worries me is the free flow of information. Big media companies have a bad track record for providing equal access to all sources of information. It's not so much censorship as a desire to avoid offending kneejerk pressure groups, and thus losing business. Still, the effect is the same.
There's a certain irony to seeing Disney and Microsoft complaining about this trend. Disney owns go.com, which started out as Infoseek, a leading search engine; Disney spent billions in an unsuccessful attempt to create a click-monopoly. Microsoft owns MSN, which was originally an unsuccessful AOL clone, then became a struggling ISP and portal. If either company had had more success with their own attempts to monopolize clicks, they'd be singing a different tune!
- It only accesses the "Sun download center", not the JDK files on java.sun.com.
- There's a JDK on the SDC. Not the latest beta, but newbies want a production release anyway. Except that I can't find a link to the Windows version. It's supposed to be there, but whatever I do I always end up with the Solaris/86 version!
- There's a link to the SDC on java.sun.com, but it isn't very conspicious. Someone browsing java.sun.com could be excused for thinking that the SDC didn't include any Java software.
- I couldn't get the stupid download manager to work, anyway.
All this weirdness would seem to be the result of the long rivalry between two of Sun's operating companies, JavaSoft and SunSoft. Of course, they're called something else now (I can't be bothered to keep up), but the rivalry it the same as ever. Did I mention that there's a lot of politics and bureaucracy at Sun?Which is why I find Open Office's emphasis on MS compatibility depressing. They even try to support Visual Basic macros!
It's not suprising. Open Office development is driven by Sun. A company that seems incapable of admitting that Microsoft won the desktop wars long, long ago. But it's too bad. I'm starting to get impressed with OO's maturity, feature set, and API. These might form the basis for solid competition with Microsoft -- if Sun ever stops fighting battles that are already lost.
You're making inferences about Turing's state of mind based on what you think he went through and how you think he must have responded. That's not totally unreasonable, but it simply isn't consistent with what Turing said and wrote during that period. Yes, he was persecuted and abused. But not to the point of fatal depression and suicide.
That's another thing. People have these stereotypical notions of depression. Somebody goes through a bad patch, and then dies violently, people find it easy to assume that they were depressed and suicidal. And if you're going to have somebody killed, you count on that stereotype.
But in the real world, everybody's different. Some people will kill themselves over minor embaressments. Others can undergo the most extreme embaressment and humiliation and still maintain their sense of self-worth. Reading about Turing's life convinces me he was closer to the later than the former.
Of course, in the real world, bugs happen, and people need to use old versions of the JDK to work around them. But anybody who's working at that level is a sophisticated Java programmer, not a newbie. Such a person does need the whole JDK!
The obvious solution is to provide some kind of download manager (as Netscape and Real now do). Or organize the Java SDK into smaller independent packages. Or allow outsiders to start up mirror sites overseas. Or... But life at Sun is full of politics and bureaucracy (which is why I no longer work there), and getting even the simplest solution in place is like pulling teeth.
There's also the testing issue. I'm suprised they were willing to spend the bucks to do QA on two different platforms. Forget about three. Especially when the third one requires special hardware on which to do the testing.
I find it easy to understand why the Slashdot editors (and probably most readers) are uninterested in Codd's life or death. He did come up with the basic ideas that make modern databases work. But nowadays people take these ideas for granted.
You see the same thing in math. Everybody knows how to prove that there's no largest natural number, but who's ever heard of Georg Cantor? The problem is that his theories suffuse modern math, and his important ideas are extremely easy to understand, so nobody feels obliged to give him credit.
Now, Turing has some claim to being the very first computer scientist. His attempts to explain mathematical logic in terms of mechanistic abstractions predates the computer revolution by a couple of decades. And when actual computers did start to appear, he was one of the first to deal seriously with what we now call software.
For details of Turing's life, I heartily recommend Andrew Hodges's biography.
But here's another bummer. It seems pretty obvious that the security bozos murdered Turing because they couldn't stand the idea of faggot with all those state secrets in his head. I should emphasize that this is my paranoid theory, not Hodges's -- as a UK citizen, he could get in trouble just for suggesting the possibility.
Certainly not Slashdot, which is based on MySQL.
Agriculture has a huge environmental impact. There's toxic runoff (even with "organic" methods), erosion, replacement of natural growth with an artificial monoculture. Worst of all, you have to burn a lot of fuel (and create a lot of greenhouse gases) to grow a crop. Nobody who cares about the environment should advocate more agriculture than is necessary to feed people or produce other necessary products.
Not that non-fossil diesels are necessarily a bad idea. These guys burn vegetable oil in their diesel boats so they won't spread particulate all over the marine sanctuary they cruise over. But that's a limited, specific goal. As a general replacement for fossil fuels, "crop" fuels are just plain stupid.
Gee, I knew there was a reason I should have taken Prescience 101. Then I could read Slashdot posts before they're written!!
Are you sure it's not "pudge facker"?
When you post a story like this, say "pf (a packet filter package)" or something like that. Otherwise people have to click a half dozen times just to find out if they're interested in the story!
OK, I'm not up on this issue, but I don't see how you can call ethanol "envirofriendly". Yes, burning ethanol produces less pollution than methanol, to say nothing of diesel fuel. But you have to figure in the environmental cost of producing the ethanol. Don't you have to grow sugar beets or something similar, then ferment and refine the result? These are all energy-intensive and unenvironmental processes!
On the other hand, lots of people break these rules right and left, and nobody cares, because they're considered good neighbors. I have a friend who's operating a business out of her condo in total violation of her housing association rules. All the other condo owners know about it, but she's such a valued member of the community (networks a lot, goes out of her way to make friends and help people) that nobody's inclined to make an issue of it.
Of course, if she ever does make a enemy who want to shut her down, she's screwed! Worth bearing in mind before you build that illegal granny flat.
But I'm no ergonomics expert. You should send your question to their support people.