"No the year dammit ... TELL ME THE YEAR!"
...
"Oh God no ... I've gone back too far, I'll never stop it now! We're doomed I tell you all of us DOOMED!"
Then I'd grab fistfuls of hair (if I had enough hair to form fistfuls) and run off madly into the street. I've always wanted to run up to someone and do that. Maybe not at a police station or in a psych ward, but y'know, somewhere.
I mention this deranged fantasy to say that the date is important! A friend recently sent me a link to an article for which he wanted an opinion as to the quality of the technical information it contained. It seemed pretty good to me, but I wanted to check the date to make sure it was current as well. However as I've seen on a disturbingly increasing number of articles there was no publishing date, not even in the URL. All the comments on the posting were deserving of datestamps, but the article itself apparently hadn't yet reached this exalted level of worthiness.
"Silicon doesn't wear out; microchips were effectively immortal. The Wig took notice of the fact. Like every other child of his age, however, he knew that silicon became obsolete, which was worse than wearing out; this fact was a grim and accepted constant for the Wig, like death or taxes ..."
The very same is true of information, probably more so because it's one of the few things that changes faster than silicon. So writers out there I implore you, if your ideas are worth reading then they're worth putting a date on! Besides, maybe you're the first person to dream up whatever you're putting on the page. If you can't point to a date and say "I thought this up first", how will you ever sue someone for a ho-jillion dollars and retire to Fiji?
(Tuesday Sept 15, 2009)
(1:58pm)
3 comments:
A bit of artistic license, methinks. Silicon doesn't "wear out" in the traditional sense, but it is quite fickle and degrades easily. See this for starters, as well as this and so on.
Anyways, I think you hit on the crux of the issue in your last paragraph. If your opinions have some sort of validity or merit, then by all means, timestamp it. On the other hand, I would argue that 99% of web content fails this basic test, with a +/-1% margin of error. Web 2.0 just makes it a lot easier for these people to agree with each other and subsequently convey implicit acceptance. This is one of the reasons things like Wikipedia will never match things like Intellipedia in terms of quality, relevance, and veracity triumphing over bias.
This sort of issue, and your friend's predicament are a very big problem with uncorroborated technical opinion/fluff pieces that so rampantly infest the tubes. It's not really a surprise that professional societies segregate their access to information completely for this reason, as well as cases like citeseer having fairly smart algorithms for filtering out superfluous citations as a way to boost popularity. Some stuff does still get through, but certainly not to the same extent.
I basically don't do any technical reading online outside of my inbox. And while I read a lot of news online, I spend most mornings with the dead tree variety much for the same reason. At least in the latter case if someone is way off base it is easy to call them on it and have them face up to it, as opposed to the blogosphere where comments are either moderated, ignored, or lost in the abysmal signal/noise ratio that seems to follow these things around.
These to me tend to be the far more critical issues than the date of publication. In technology especially, whats old is new again. This has been the case with everything from 64-bit computing to transactional memory to dataflow (or any non Von Neumann architecture) to offload engines and everything in between. Incidentally, most of the technical papers I have been reading lately and making subsequent use of were written in the late 80s or early 90s. If a piece of technical writing has withstood the test of time, it almost certainly has done so for reasons of merit and general acceptable, not merely implicit indifference. Consider this before putting too much stock in something as superfluous as a timestamp.
I agree that quote takes a bit of articstic license, though those are Bill's words not mine. With regard to the quality of technical info online, true it varies wildly but that could be said of the quality of *anything* online from prose to porn. Admittedly I could have pointed my friend in the direction of a much more complete and thorough explanation of what he was looking for (how to connect to a wireless network from the command line in linux) but for the topic at hand a simple few paragraph tutorial sufficed.
I must call shenanigans on your last comment however, if something has stood the test of time no doubt it has done so on merit. But if I've never heard of it or it's merits how the hell will I know if it's stood the test of time if no one bothered to put a date on it?
I thought that was obvious. Useful information that has withstood the test of time is oft cited and frequently returned to. In terms of research papers and the like, citations of old papers by new ones happen as a matter of course, and as a result constantly bump up the old one. If an old paper is constantly returned to, there must be some reason for it. This is the sort of metric that is frequently employed by systems (like citeseer) that attempt to balance the signal to noise ratio, despite being susceptible to the same sort of systematic circumvention as pagerank.
My point regarding timestamps is that they are ultimately irrelevant. If something is cited by hundreds of different papers it is worth looking at regardless of whether it was written a month ago or 20 years ago. One can also extrapolate the relative age by looking back at early references and citations, although again, this has very little to do with relevance. Some of the best technical material I've read and that I often refer to hasn't changed in 30 years. If it had been written yesterday, it would ultimately not change anything.
Did you never stop to wonder why people that are heavily involved in one particular area tend to have all effectively read the same things one way or another despite all having totally different backgrounds? Useful information has a way of shaking itself out and separating itself from the noise, and unfortunately there just isn't very much of it (whether online, in print, etc.)
Post a Comment