Slashdot Mirror
Creative Documentation
FuriousCurio writes "Linux kernel hackers appear to be an endlessly creative group of individuals. In response to previous documentation attempts not having been read by many people, KernelTrap is reporting about how the lguest documentation was prepared to be something of an adventure story. Self-proclaimed to turn you into an lguest expert, lguest being one of the new solutions for running a virtual instance of the Linux operating system as a user process within a real instance of the Linux operating system, the documentation mixes humor and wit into puzzles, poetry, and of course source code and a low-level understanding of virtualization. But the questions remains, will making documentation more entertaining actually work to get people to read it?"
If you compile the Anarchists' Cookbook you wind up with Windows 3.11 for Networking.
...until I was eaten by a Grue.
Slashdot Burying Stories About Slashdot Media Owned
Q1. What is lguest?
A. RTFM n00b.
An Indian-American Hindu committed to non-violent thought/speech/action alarmed by the global explosion of radical Islam
Hell no. Just make sure I can search it when I get stuck.
Even the author of TFA thinks this doc is crap (you need "grep" to get off the first page?):
Should be clear, complete and timely.
Every time I've tried to solve a linux problem I've run into docs that miss one, two or all three of those things
Documentation has to be very clear, very unambiguous, and very specific. When you're already up against a problem you don't want to be guessing at what the docs are trying to tell you.
Looking at TFA, my suggestion to not waste everyone's time with cutesy games - hire a real professional to write and edit your docs and get them right the first time.
Three Squirrels
here at slashdot we can't even rtfa, let alone rtfm.
Sometimes, life itself is sarcasm...
As a documentation writer, I've learned from experience that 95% of any typical audience won't read the docs, no matter how many pictures you include, or how entertaining you try to make it. People, in general, just don't like to read, period. They'd rather call support or fumble around and try to figure it out on their own.
The other 5%, though, read the docs so thoroughly that they'll find the tiniest mistakes or oversights. This basically means that the docs have to be perfect, even though only a fraction of the audience will bother to use them.
Having thorough documentation still occasionally helps the other 95% though -- it gives the Tech Support guy something to point to ("see page 108 of the User Guide") when dealing with idiot questions from people who should know better.
There are two purposes to documentation: one is to serve as a reference. That is, when someone who is generally familiar with the system needs to know how to do a particular thing (be it design a cursor, add a command line switch, locate a config file, apply an update, or whatever), there needs to be a document that can be used to easily find that particular answer. This is the goal being striven towards (largely successfully) by man pages.
The other purpose, however, is to make someone who is completely ignorant of the system familiar with it. Most software documentation is terrible at this. Telling me how to do something isn't helpful if I don't know why I'd want to - or, worse, don't even know that such a thing can be done.
Since I haven't used a bad car analogy in a while: having a document that explains how to install a cold-air intake on my car is useless if I've never heard of a cold-air intake.
What the lguest docs are trying to do is solve the latter problem. They're trying to take a system that someone doesn't know anything about (aside from just enough to be interested in it at all) and get that someone up to speed in a general way.
"How" is a good question to answer, but so are "why" and "what." Gimmicky documentation isn't necessary or desirable for the first, but may very well be both for the second and third.
Reality has a conservative bias: it conserves mass, energy, momentum...
Wiki's make the best manuals
..
when properly implemented
clear, concise, to the point, easy to search
and when new problems arise they're easy to add
With this you'd have to add a whole new realm or player class just to tackle the issue and stay with the theme
----------
Trying to fix or change something only guarantees and perpetuates it's existence
There's something to be said for the humor/wit approach I think.
My college physics instructor used the same approach in writing his weekly homework assignments. Essentially, the year's homework detailed the exploits of "Green Sarge" (A real-life version of those plastic soldiers you find at the dollar store) vs the "Beige Chumps" and, later, his arch nemesis -- the Fez-wearing, scimitar-wielding Evil Physics Monkey. Even if the students didn't start the homework immediately, they would always read it to see what Sarge's next exploits would be, and the problem would be in the back of their mind ready to consume any spare brain-cycles. The humorous problems also lead to a lot of impromptu discussion about the problem as well, just talking in the hall or over a lunch table. I think it went a great way towards getting the students to embrace their homework.
[from (vague) memory]
Q) At what velocity must the Evil Physics Monkey fire himself head-on into the Green Army supply train in order to stop it? The Train has a mass of 80,000kg and is traveling at 50km/h. The Evil Physics Monkey has a mass of 100kg. The EPM's scimitar has a mass of 15kg, recalculate the problem assuming that the EPM has carried his scimitar into battle with the Green Army supply train. Assume structural and soft-tissue damages are not a factor.
Will making documentation more entertaining actually work to get people to read it?
s -tank-manuals-unexpectedly.html
The Germans thought so:
http://www.darkroastedblend.com/2007/02/wwii-nazi
"All great things are simple & expressed in a single word: freedom, justice, honor, duty, mercy, hope." --Churchill
the king of off-beat technical documentation is Why's Poignant Guide to ruby.
Aha! So that's why I know so freakin' much about aardvarks, but jack sh!t about zebras.
John
At the risk of getting flamed into hell and back out the other side, I seem to remember reading that the Solitaire game that came with MS Windows originally ('way back when) was designed specifically to get people used to using a mouse, since up until that point, you had DOS at home and LIKED it that way! /a 'mouse'?! //wtf would I do with THAT?
...a point hit home by the success of the "For Dummies/Idiots" series. It's one of the selling points of the books
;-)
I think the very readable For Dummies series of books hasn't reached the seemingly untapped potential of its target audience.
Maybe someone with a better knowledge of history or who has studied technical writing can elaborate on this, but I believe it was the O'Reilly series of books that broke ground on changing the manner in which technical books were written from textbook-ish style to something more informal and entertaining.
I'd guess there's more than a few books in the O'Reilly catalogue, for example, on everybody's favourite list, but the increasing focus on appealing to readers often leads to compromising on actual content. More people educating themselves by buying or reading more books (or on-line documentation) is A Good Thing, of course, but my preference has been for the (apparently dated) textbook-ish approach. Compare, for example, something like Internetworking With Tcp/Ip: Principles, Protocols, and Architecture (Internetworking with TCP/IP Vol. 1) published in 1991 with anything published today on the subject of networking. One is as comprehensive and as well written as it is boring to read, while the others are more accessible and topical and shorter. No surprise which sells more copies.
What I've never got my head around is that people increasingly don't want to read anything. I wonder how somehow making their living as a writer feels knowing that most of us are guilty of relying on a Google search for a quick intro or how-to when the READMEs, man pages, source code, etc. is sitting on their hard drive.
I wonder how somehow making their living as a writer feels knowing that most of us are guilty of relying on a Google search for a quick intro or how-to when the READMEs, man pages, source code, etc. is sitting on their hard drive.
I am a technical writer, and I assure you that it's absolutely no secret that this is the case, and we're OK with it. I can't deny that there are times where I feel a little down after sweating blood over a documentation project that I know 16% of our customer base will ever read (that's an actual statistic, incidentally, from a firm I once worked for), but, in the end, my paycheck still arrives. I will say, though, that both companies for which I've worked as a writer are constantly working to improve documentation content and style in hopes that you'll use it instead of Google. Tech writing, despite how it probably appears, evolves like anything else. Whether its an effective evolution is up to you, I guess. I have my own opinions in that area.
A much, much bigger frustration is the lack of respect given to tech writers by developers and hardware engineers. I couldn't count the number of times I've been handed a pile of "documentation" written by an barely literate ESL developer somewhere with the expectation that I can magically turn it into a public doc in "what, a day or two?" In their eyes, we're typists, and in the two years I've been doing this, one of my greatest professional joys has become the look on the face of some snarky developer when I say, "No, this is more like three weeks. Will that hold up your release?" As joyful as I am, though, there are times where I simply have to produce something in a quarter of the time it actually needs, and that invariably results in garbage. In my opinion, many of the problems with technological documentation could be solved by just keeping me in the loop throughout the project, but that seems to be too much to ask. On the rare occasion where this happens, I've produced award-winning manuals (yes, there really are awards for this) that receive a surprising amount of kudos from customers.
But, most of the time, I'm handed junk information at the last minute and nobody's willing to answer the phone as I try to distill anything meaningful from the whole thing. Then, I either unapologetically delay the project or produce crap. The sun goes up, the sun goes down.
-
Inventor of the term 'pardon my French'.
Bill Stewart
New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
The concept is not new. It is called engaging the reader.
For most of us on slashdot, we are already engaged by the technology. We have no other need to read the documentation. We want to know how to make this work just to know how to make it work. But, the average user could care less about how a thing works, so long as it does what they want it to without any need to learn if possible. Why do you think tutors and techs have so many jobs? Why do you think so many people have diseased computers? Because they are not engaged in learning about how or why it works.
For those old enough to remember, the old TRS80 manuals were good examples of how to write engaging documentation in their day. We can do much better today, but few have done as well since then. People need an emotional tie when learning to truly remember. Think about those things you actually do remember from decades past. They almost all have an emotional anchor, whether it be tears or laughter or something else (excitement of learning?).
So, creating a set of documentation that meets needs of people who do not get the same excitement/enjoyment out of just learning the tech will go far for helping the others out their learn the tech. And we need them to learn the tech. Or linux and OSS will die on the vine.
You can always claim that as long as people can write software, there will be open source. I counter that until the general public has a vested interest that they are aware of and care about, OSS is always at the mercy of government and business. All it takes is a few laws to be passed and OSS goes away. Some are on the books now and some are talked about often enough here.
The best way to fight for the future of anything is marketing. That includes *good*, solid, easy and friendly documentation. That may be the biggest selling point to the home user in the end. "It just works" is not just a slogan, but an expectation of most people. If whatever it is does not live up to that, then whatever claims to be next will steal their attention.
It boils down to loud words mean nothing. Claims of ours is better means nothing. All that means anything is the average parent/sibling/child can sit down and just use it. If the docs are not fun and easy, then that is very unlikely to happen for most people.
InnerWeb
Freud might say that Intelligent Design is religion's ID.
An example from the cc:Mail section: The TFS post office can not be used for addressing. Mail sent directly to the TFS (gatelink) post office
without having been addressed to a routing post office will go to e-mail heaven immediately. It would
not be delivered if you put 40,000 volts through it. This was a big problem. People constantly addressed e-mail to the gatelink PO and they went in the bit-bucket. When I added that snippet to the manual, these problems went way down for new installs. I worked in support as well as doing the docs, so I knew the incidence rates.
Money for nothing, pix for free