This reminds me of SF short story, where people came up with idea of robotic doves (birds) acting as police and paralysing people who wanted to commit murder. But they had to adapt to do the job properly - to detect intent even in most ruthless killers. Soon they started to prevent people killing insects. After that, it was not possible to switch off TV set. And solution for that was to create self-evolving robotic killer hawks to catch the doves... anybody knows what was the name of the story, cannot find it now?

You're looking for Robert Sheckley's 1953 short story Watchbird , via Project Gutenberg. There was a TV adaptation in 2007's Season 1, Episode 6, Masters of Science Fiction.

Great read.

I've posted most of this on the "blog" site where it's likely to be read instead of buried in a 1000-post thread, but this seems the right place to follow up with your well-articulated, broad-based global objections (with which I agree 110%), and outline the nits.

Upon re-reading this list, it's depressing just how many things about the 3.0 redesign that I'm already thinking of blocking/hacking out client-side via greasemonkey or local CSS overrides. The depressing part isn't that I'm willing to do it; I love the site enough to go through the trouble. The depressing part is that the only reaction I can have to all this effort is to start thinking about how I can disable it.

1) Images: Meh, I can take 'em or leave 'em. I can understand users' frustration, but they're trivial to block client-side.

2) Whitespace:

Narrow the spacing between lines.

It's like reading in doublespaced/triplespaced form.

3) Whitespace. I think people have
told you the fixed-width column
was too narrow. But just in case,
here's another reminder.

4) Content and presentation of article summaries:
(From the click-to-expand department)

All that whitespace, and you can't even display the full article summary? Because some web designer said all summaries had to fit within a maximum number of vertical pixels before requiring a mouse click? And you(...rest of this objection after the jump ... *click*)
believed him? Really? :)

5) Comments. User numbers (UIDs) need to be displayed. They're a useful indicator age of account and therefore useful for helping mentally filter trolls/shills. (Umm, sorry, noobs, but if your UID indicates an account created in the past day or so, it takes me a while to accept you as a regular ;)

6) Comments. Timestamps need to be timestamps. Sometimes it's critical to know who was the first to make a joke or link to a reference. "A few minutes ago" or "An hour ago" isn't enough. Going further out, "Two years ago" is meaningless if you're talking about things like whether someone called a corporate takeover or tech development before or after the news actually came out. To illustrate the problem by way of example, "1 year ago" could mean at any time during 2012, 2013, or 2014, for any time period from 8 months ago to 18 months from now, and is no longer useful for gauging whether someone successfully predicted the eventual fate of Blackbrry. Slashdot is an easily-googlable source of record, and it's *vital* to know on what day it reported on something.

P.S. Just because you read it on a blog doesn't mean it's true. http://graysky.org/2013/09/blog-timestamp/ And even this author notes that for some publishing, the timing is highly relevant. If you want to be the blog of record, your content is such content.

7) Comments. Needs filtering or a one-click-load-all-comments button.

D1, its bugs notwithstanding, could do this with three middle clicks into new tabs of about 100 comments per tab.

D2 could do this with two drags over the slider and a load-all-comments. (or a load-500-comments and then a load-all-comments).

D3 doesn't seem to be able to do this as far as I can tell.

8) Black-on-grey is less readable than black-on-white.

Sorry, OS X people, this is fail. I can tolerate this only because I can manually override it client-side. It's horrible and makes the site unreadable, but, well, it's something even an idiot like me can forcibly override client-side in 5 minutes. It's hardly the worst defect of the redesign.

9) Floating DIVs. Really? *REALLY?!?!* Some of us use something other than mice or greasy fingers on touchscreens to scroll.

10) Auto-refresh. There's a preference to disable this, right? Right?

11) Will D1 be preserved? I felt that D2 was something I could adapt to, and on occasion, I prefer its presentation to that of D1. This is unusable, and I will leave if it goes through as presented.

12) Like most UX redesigns, I know that the overwhelming flood of negative feedback will be ignored. We're just the users. We don't know a thing about design, and it's the designer's attitude that matters, not whether it's usable or not.

This means I'm likely to be leaving for other places soon. I'm not sure where I'll go yet, but I'll find a community somewhere. Fark's fun but nontechnical. Digg's dead, and good riddance. Reddit requires too many mouse clicks to do anything. HN is clean, elegant, technical, informative and so bone-dry sterile that I can only go there once a day.

Thank you, /., for 15 years of providing a place for funsightformative coments. There was truly no place like this. I respect that the Dice sale was as good an exit as you could have made under the circumstances (I thought SlashBI might have actually gotten some traction given some time), but failing to prevent their UX people from killing Slashdot was a pretty ignoble end to what was once a proud website.

Good luck in your future endeavors.

Signing off, Tackhead (#54550)
5-digit-club, with 43 achievements, 2^9 +5 comments, 2^8 consecutive daily reads, embarassingly low 2^2 metamod score; I suppose I'd have metamoderated more often if the UI for that hadn't been broken in the upgrade to Slashdot 2.0, (I still don't know if +/- means that the comment was good/bad, or if the moderation done to the comment was fair/unfair, and yes, that distinction is important in the case of "+1 Funny" vs "-1 Flamebait" because the mod missed the joke) and maybe it's fitting that Diana Moon Glampers: UX Designer was my last +5.

(P.S.: Does anyone know how I can tell how many comments I've posted in total? I'd like to know before I go.)

So what's the plan going forward? I've had a couple of hours to cool down and formulate my objections more objectively.

1) Images: Meh, I can take 'em or leave 'em. I can understand users' frustration, but they're trivial to block client-side.

2) Whitespace:

Narrow the spacing between lines.

It's like reading in doublespaced/triplespaced form.

3) Whitespace. I think people have
told you the fixed-width column
was too narrow. But just in case,
here's another reminder.

4) Content and presentation of article summaries:
(From the click-to-expand department)

All that whitespace, and you can't even display the full article summary? Because some web designer said all summaries had to fit within a maximum number of vertical pixels before requiring a mouse click? And you(...rest of this objection after the jump ... *click*)
believed him? Really? :)

5) Comments. User numbers (UIDs) need to be displayed. They're a useful indicator age of account and therefore useful for helping mentally filter trolls/shills. (Umm, sorry, noobs, but if your UID indicates an account created in the past day or so, it takes me a while to accept you as a regular ;)

6) Comments. Timestamps need to be timestamps. Sometimes it's critical to know who was the first to make a joke or link to a reference. "A few minutes ago" or "An hour ago" isn't enough. Going further out, "Two years ago" is meaningless if you're talking about things like whether someone called a corporate takeover or tech development before or after the news actually came out. To illustrate the problem by way of example, "1 year ago" could mean at any time during 2012, 2013, or 2014, for any time period from 8 months ago to 18 months from now, and is no longer useful for gauging whether someone successfully predicted the eventual fate of Blackbrry. Slashdot is an easily-googlable source of record, and it's *vital* to know on what day it reported on something.

P.S. Just because you read it on a blog doesn't mean it's true. http://graysky.org/2013/09/blog-timestamp/ And even this author notes that for some publishing, the timing is highly relevant. If you want to be the blog of record, your content is such content.

7) Comments. Needs filtering or a one-click-load-all-comments button.

D1, its bugs notwithstanding, could do this with three middle clicks into new tabs of about 100 comments per tab.

D2 could do this with two drags over the slider and a load-all-comments. (or a load-500-comments and then a load-all-comments).

D3 doesn't seem to be able to do this as far as I can tell.

8) Black-on-grey is less readable than black-on-white.

Sorry, OS X people, this is fail. I can tolerate this only because I can manually override it client-side. It's horrible and makes the site unreadable, but, well, it's something even an idiot like me can forcibly override client-side in 5 minutes. It's hardly the worst defect of the redesign.

9) Floating DIVs. Really? *REALLY?!?!* Some of us use something other than mice or greasy fingers on touchscreens to scroll.

10) Auto-refresh. There's a preference to disable this, right? Right?

11) Will D1 be preserved? I felt that D2 was something I could adapt to, and on occasion, I prefer its presentation to that of D1. This is unusable, and I will leave if it goes through as presented.

12) Like most UX redesigns, I know that the overwhelming flood of negative feedback will be ignored. We're just the users. We don't know a thing about design, and it's the designer's attitude that matters, not whether it's usable or not.

This means I'm likely to be leaving for other places soon. I'm not sure where I'll go yet, but I'll find a community somewhere. Fark's fun but nontechnical. Digg's dead, and good riddance. Reddit requires too many mouse clicks to do anything. HN is clean, elegant, technical, informative and so bone-dry sterile that I can only go there once a day.

Thank you for 15 years of providing a place for funsightformative coments. There was truly no place like this. I respect that the Dice sale was as good an exit as you could have made under the circumstances (I thought SlashBI might have actually gotten some traction given some time), but failing to prevent their UX people from killing Slashdot was a pretty ignoble end to what was once a proud website. Good luck in your future endeavors, but if there's no D1/D2 preservation, I'm afraid I won't be riding this train to wherever it's going.

Signing off,
5-digit-club, with 43 achievements, 2^9 +5 comments, 2^8 consecutive daily reads, embarassingly low 2^2 metamod score; I suppose I'd have metamoderated more often if the UI for that hadn't been broken in the upgrade to Slashdot 2.0, (I still don't know if +/- means that the comment was good/bad, or if the moderation done to the comment was fair/unfair, and yes, that distinction is important in the case of "+1 Funny" vs "-1 Flamebait" because the mod missed the joke) and maybe it's fitting that Diana Moon Glampers: UX Designer was my last +5.

(P.S.: Does anyone know how I can tell how many comments I've posted in total? I'd like to know before I go.)

10 Mb/s upstream is not that unusual these days. And many botnets are way bigger than 10000 bots. Even if each bot has just 2 Mb/s upstream, you only need 50000 Bots. And botnets are not limited to infecting computers on home user connections. Infecting 100 servers on 1 Gbit/s connections is also enough.

As much as i applaud Apple for finding homes for physically challenged mice, that doesn't mean the rest of the mice should have to wear sandbags.

Diana Moon Glampers as a UX designer. That explains a lot, actually.

I miss the days when it was UI - the user's interface with the computer. An interface. The thing that makes it possible to make the computer do what you want it to do. Design it for maximum functionality with minimal interference.

Somewhere along the line it became UX - the experience. The fluff. The marketing. Doesn't matter if it's functional or not as long as it feels good. You're not allowed to learn anything, you're not allowed to even know how it works. There's nothing to master. Just one button that says "Make it look like whatever the other UX people think is fashionable this year."

In Windows-land, we lost (unless you hack the registry) focus-follows-mouse from XP to 7, and the ability to resize an arbitrary number of windows when we went from 7 to Metro. In Web-land, we lost Firefox. In GNOME-land, we're about to lose middle-click-to-paste. (I probably shouldn't have mentioned focus-follows-mouse, or they'll take that too.)

First they hide the feature. They they claim telemetry says nobody uses it. Then they take it away. (Never mind the fact that the sort of user who does use the feature either delays the upgrade, hacks around the limitation, and is likely to pre-emptively disable telemetry as a matter of course.)

We used to be Emperors and Empresses over our machines. Now that any fool can design a UX, we have UIs designed by fools for fools. It's all kind of mixed up in my mind, but the past five years of change for change's sake have been a doozy.

The Council has declared a day of rejoicing, relaxation and release as intelligence reports from the blue world confirm that the latest invader from the blue world has failed to detect appreciable quantities of quadrohydrocarbon.

K'Breel, Speaker for the Council, addressed a tightly-clenched world: "Our collective tightening effort over the past year has not gone in vain. Long and hard have we clenched, and now it is time for all right-thinking citizens to reap the rewards. Our symbol must no more be the clenched fist, but the unfolded flower! REJOICE with your podmates, RELAX your cloacae, and RELEASE upon our impoverished atmosphere a deluge of accumulated flatulence so great that the very canyon walls shall shake, enveloping the invaders in dust and cutting off their vital power!"

When a junior reporter reminded the Speaker that the latest invader was powered by something other than mere radiant stellar energy, K'breel, in his mercy, had both of the junior reporter's cloacae sealed until the pressure of accumulated quadrohydrocarbon was released through the second-weakest point of structural failure: the gelsacs.

"O, they ruled the solar system
Near ten thousand years before
'Til one brave advent'rous spirit
Brought that mighty ship to shore."

As you finish the last verse, Floyd smiles with contentment, and then his eyes close as his head rolls to one side. You sit in silence for a moment, in memory of a brave friend who gave his life so that you might live.

would have been awesome had this happened a year or two ago when the company was on the upswing, not shortly-to-be-passé.

Part of the problem with tech industry is that company lifecycles are shortening even faster than product lifecycles.

If it takes two or three years between the decision to go public and the actual IPO (plus another 6-12 month lockout period afterwards), and a company can only exist for 5-10 years before it becomes obsolete, the time between the decision to exit and the actual exit takes up a huge portion of the company's arc.

The VCs who backed GRPN and ZNGA made bank. Almost nobody else did. Sometimes you get lucky, like GOOG and LNKD, who raised enough capital to wall themselves off from competition. Sometimes a company can re-invent itself, like AAPL, AMZN and NFLX. Sometimes, you can find a sucker willing to pay top dollar for a worthless asset - MySpace had a good exit and left someone else holding the bag.

But those are the rare successes. Most of the time, the founder rides the rocket all the way up and all the way down to the ground, and even he ends up getting a fraction of what could have been made if he'd only shopped the company out earlier.

It's incredibly difficult to build a sustainable business in this industry. If you catch the big industry cycles: mainframe to micro, micro to client-server, client-server to cloud (mainframe), you can do so. Get out of sync with the industry fads - whether you're 1-2 years early or 1-2 years late - and your company will not outlast the pendulum swing. The optimum strategy is to find a fad, slap something together for $X, and try your damndest to get acqui-hired for $2X in a 1-2 year timeframe. Lather, rinse, repeat. If it's so important to be a CEO, sell the damn company anyways, and use the money you made off the last one to start the next one.