We have an ~8TB JFS array used for shared storage of databases on our linux boxes, as well as for another smaller arrays used for VM's and general file serving duties. Bitchin' fast even with small files, no extra latency incurred by using ACL's as you get with ext3, very fast to format and fsck (ever tried this on a 2TB ext3 array? Painful!), lowest CPU utilisation of any filesystem we tested, frags up alot less than ext (yay for extent-based filesystems), ridiculously quick deletes, equal to XFS in speed at dealign with large files with less overhead at the same time. Only real disadvantage is that it doesn't support shrinking, but that wasn't considered a big downside for us.
Never tried LVM on anything that big though, ours is just stacked hardware RAID10 stripes across a starburst of cheap SATA drives. Use it on a 3TB LVM at home though, not had any problems with it - but that's comparatively zero load.
It wasn't popular initially since it wasn't supported by Redhat, but we have less issues with the JFS partitions than we've ever had with the ext3 ones (yeah yeah, famous last words;)) - besides, we use CentOS boxes for these anyway (yeah yeah, more famous last words...).
Most underrated filesystem available for Linux IMHO, and those of you who've used it on AIX will already know its awesomeness and how comparatively gimped the linux version really is:) Bottom line: we use JFS on any disc-based IO that's performance critical, and ext3 everywhere else in order to attempt compliance with corporate support;)
I've set up suspend-to-file before, since my laptop has 8GB of RAM and taking a >8GB chunk out of even a 250GB disc is no small feat - as long as your kernel has support for your filesystem of choice built in, you just run a tool to find the address of the file in question (think it's just a mater of setting your file in the suspend config and then looking in/sys for the address). This way, if I desperately need an extra 8GB of space, I can just `rm/swapfile`, instead of going through the palaver of resizing swap and the surrounding partitions. For the record, I keep a 512MB swap file of which I've never seen more than 15MB used. Anyone's who's ever done anything like accidentally rsync 8 billion files [thank god for rsync 3] will know that having gads of swap available is a curse, since disc speeds haven't kept parity with RAM - the system will become almost totally unusable until it hits the memory limit and oom-killer has its way. In such an event, you *want* to run out of swap as soon as possible.
I thought the idea of using an 8GB swap file ridiculous, since my apps rarely use more than 2GB physical for their working data set - but they do lots of random lookups on data so the huge file caches are a massive bonus to me.
What I can't understand is why suspend doesn't seem to flush the file caches from memory when hibernating - writing 8GB of memory to swap, even when using compression, takes an utter age.
Edit: just found this, added in 2.6.16. Will give this a whirl and see if it improves hibernate any... http://linux-mm.org/Drop_Caches
Just some anecdata; we have an IBM 3850 with eight cores (four dual-core Netburst-based 7120N Xeons) running ESX 3, and we can state that for our workloads (bunch of debian VM's running mostly python) it's spanked by a) the four core Core2 Xeon box and b) the four core opteron box (olde style, not barcelona). Profiling shows that the opteron appears to be kept at the highest utilisation because the VM's wait less on memory access.
Granted, it's not a direct comprison of modern CPU's but the opterons hold their own against the supposedly much faster Core2 Xeons. VMMark is a good general indicator for generic VM performance, but as with any benchmark it's only useful if it's close to your workload. If we get a new cluster it's almost certainly going to be a pair of eight-way boxes and we're quite optimistic that opterons could well play quite nicely, given our VM's thirst for low memory latency.
I know they don't offer an eight *socket* server, but even with the Core2 quad Xeons there's still alot of cache snooping going on over the FSB, and this can kill performance in highly concurrent SMP apps.
Another vote for Draytek. Been using them at home and buying them for others for about five years. Never crash, tons of features, most reliable wireless I've ever come across in a compact router, wicked little VPN capabilities for those of you not wishing to shell out a shitload on Cisco but not wanting to set up linux boxes either. Expensive, but will last as long as you can resist the "oooh, shiny!" factor of their next model.
Have been running a 2800VG for about 2 1/2 years now and the only times I've rebooted it are when we've been having problems with the line when the equipment at the exchange goes south. In my last house the thing wasn't rebooted for the whole year it was in there.
Primarily because, in my experience, most users of the word "pro-active" are unaware of it's anti-reactive connotations and use it to describe singularly reactive situations ("I want us to respond to this pro-actively"), or even in just syntax-ruining "I've learnt a cool new word" non-sequiturs ("our new rubber grommets have a 100% pro-active paradigm"). In other words, I'm convinced that alot of people use it because they think it sounds More Important than "active" or lack the vocabulary to better describe it.
It's kinda acceptable in most sysadmin circles as most geeks are aware of things like "pro-active" support (I prefer to call it preventative maintenance myself since it means less fuzzyness for the recipient, which we abbreviate to premaint in conversation) but neologisms are mostly a matter of taste./spot the word-snob;)
The way I've found to restore sane address-bar behaviour was to do the following: Install the oldbar extension about:config browser.urlbar.matchOnlyTyped boolean set to true.
Completely agree that the current situation WRT to anything pornographic is somewhat stupid - as you say, there are various definitions of what constitutes pornography, and there are some very, *very* grey areas. My favourite example of joined up thinking is clause 63 of the UK criminal justive bill here in good ol' blighty that makes owning images of "extreme pornography" illegal and could well end up with you on the sex offenders register. Fair enough some say, it's "extreme porn".
First off, there's currently no cut-and-dried definition of "extreme porn" either in the law statutes or the public eye - extreme porn generally just means "something that I, the beholder, consider pornographic and extreme" - too vague for my liking. Secondly, and far worse IMHO, is that the act itself isn't illegal - the law is worded as such that, for example, adults are still allowed to engage in whatever extreme fetish they like as long as it's not against any of the participants wills and the usual H&S/due diligence stuff, but if they want to record it for posterity all of a sudden it's illegal. Makes no sense to me at all.
Which brings me back to complete agreement with your point, or I think so at any rate. As you say, laws concerning things like paedophilia are meant to be there to protect people who can't neccesarily protect themselves. Looking at, for want of a better word, "cartoon" CGI images not based on pre-existing illegal material might still make you and everyone else involved in its creation and distribution a sick and twisted fuck, but what *real* harm has it done? Just being a sick and twisted fuck in itself isn't illegal (yet), and going down the road of "this non-harmful/legal activity MIGHT lead to this harmful/illegal activity" takes us stright back into the "gateway drug" quagmire that you can apply to practically any human endeavour. Alcohol, violent movies/computer games/music, cigarettes, bird-watching (hey, it won't hurt if I just collect *one* egg)... can't have it both ways people. Unless you're cool with the idea of thoughtcrime:)
Both of these bring us back to the forensic aspect though - at the moment, it's still relatively easy to tell when a digital image has been doctored, especially when "big name" software is used with minimal post-processing that would otherwise remove any obvious artifacts (not trying to sound like an expert, just providing anecdata).
Of course, I don't envy the first people to be convicted of having real images doctored to look like fake ones, or fake images doctored to look like real ones. Just expect the law to follow the path of least resistance and outlaw anything that looks like realistic illegal porn on the grounds that the owner can't reliably prove whether it's lawful (perhaps it was a CGI take on an unlawful image in the first place) or not. We're already in the arguably slightly ridiculous situation where someone 1hr under the AoC is still legally a child, and last time I checked porn downloads didn't come with digitally signed Authentic VeriSign Porn Star Birth Certificates (it becomes even more confusing when you realise that the AoC for sex and/or porn is different across much of even just the western world), so when the situation does arise expect people to undergo trial by mudslinging and "guilty until proven innocent/no smoke without fire" hysteria before the law catches up with technology.
M'lud, I present exhibit A - a 1024-node renderfarm accompanied by a small army of animators and artists that we believe was used to fabricate exhibit B;)
At the moment (as far as I'm aware - I have a friend who works in forensic IT who has a colleague that specialises in detecting doctored images and video), even the top-end CGI is relatively easy to distinguish from the real thing, especially where humans are involved (even more so for video). Whilst I agree there's a possibility that tech and skills capable of making realistic human animations and the like may only be a few years away, I still think it'll be a long time before such fare becomes indistinguishable from the real thing, and even if it was there'd be an inevitable paper trail (or lack of it) concerning the origin of the pics/vids.
...if it comes into law because it means that anti-Scientology protestors will be legally obliged to obfuscate their placards so as not to say anything offensive.
SCIENTOLOGY - IT'S A CU*T WHAT A BUNCH OF CU*TISTS COME ON YOU CU*TS
N.B. to americans not used to judicious use of the "C" word perjorative, please don't take this post as some sort of sweary troll. It's nowhere near as offensive here in the UK as it is in the states. It's not the sort of thing you'd say at dinner with the in-laws, but it's not the sort of thing that gets you blacklisted as some sort of social outcast. If this is the case in the states as well I apologise for being patonising, since it's the general consensus that saying "cunt" gets you universally reviled by everybody over there.[/derail]
The best thing about these new bits of Sony software is that you don't even need to type in your name, DoB, SSN or any of the rest of it - it finds it all out for you! Nonsensical Bullshit Paradigm Shift 2.0 is SO cool!
Yup, I'm looking forward to being banned from the internet after my flatmate downloads a bunch of music from the net. Seems like the only way to have a shared internet account not being terminated in this manner is for me to police every machine in the house, which means invading the privacy of my flatmates.
As you point out, an internet connection is not a person, or a company, or a family, or anything. It is an internet connection that may be being used by all of these.
When I was a penniless student, I participated in a study very similar to this, also at a London university, using CAVE. If this test was anything like the one I did, it was purposefully engineered to try and generate a feeling of discomfort - you "walked" around a room a bit like a library, seeing people mooching about and sitting at desks and reading.
As you walked about, you'd sometimes hear an unintelligible mutter from behind, or in front, as you approached people and you gauged you reaction as to whether a particular person or particular mutter made you feel uncomfortable, or glance at you in someting you perceived as a disdainful expression (I remember the facial animations being pretty convincing for circa 2000) - they fitted you up with a pulse thingy and something that was probably like a galvanic skin response dooferwhatsit as well as relying on your description. I never found out what the results of the study were, or even what the experiment was meant to be proving, but I imagine it was either something akin to judging peoples reaction to VR people, a la Snow Crash's Juanita.
But even six years later, I don't think animation has yet progressed to the state where we can no longer differentiate between reality and CGI, even if we're attempting to suspend disbelief, and as such I don't think any meaningful conclusions can be drawn as to whether it correlates with our real world experiences. Although I will say that spending any amount of time on the tube, particularly at rush hour or after the arseholes stagger out of the pubs, is enough to make anyone wish for their own jet pack:)
In response to a couple of other comments (slashdot has gone and buggered up my comment view, and it makes viewing and/or responding to comments a complete PITA, as well as hiding virtually everything unless I browse at threshold 0 - I don't get any "X comments below your threshold" linsk any more) - sorry, but appearance is one of the quickest way of judging people, and an evolutionary tool that's served us very well in the past. Clothes, the "default" facial expression, the way someone walks, the way someone holds themself, the way they talk (visually and aurally), the way their eyes move - put them all together and it can prove a very useful metric for a quick assesment of someone's likely reaction. As to the black guys who mentioned they regularly suffer from verbal abuse - let's just say I'm shocked that such things still happen in civilised (and assuming western) society. Must have been living in my own bubble of liberals for too long.
...which brings me back to my original point: KDE no longer give a crap about the power users, they just want their DE to look the blingiest and to hell with usability.
...the new OSX interface has shown us that we don't need so many colours. Colours in a computer eat up the memory bits and distract us from our reverence. Personally, I'm going to take Steve's advice and go get my eyes chromed.
I've been struggling to remain sufficiently Zen amongst all this as well. I don't recommend you read some of the comments on that page if you value your blood pressure, there seems to be a colossal mentality of "Novell spent loads of money on it, and it's scientific, therefore is MUST be true!".
The search function seems retarded to me as well, for the precisely the reasons you state - we already have a method for launching applications where you're able to type in the start of the name of a program, press a button to initiate a search and then press return to launch the application, it's called a bash shell with tab complete in my book. Woo fucking hoo for your innovative search function. As you say, pretty much the whole point of a GUI, and the reason they're easier to pick up by n00bs, is that they expose you to more than you're currently using. Kickoff is the complete antihesis of this, yet it still manages to be complex enough to be confusing.
The most disturbing thing is the attitudes of some of the key developers who seem to think that a) the current users (especially the power users) don't have valid opinions and b) that it's better to cater to the lowest common denominator. Sorry, that's why I dumped GNOME. I don't want to dump KDE, but if you make everything as ball-gratingly annoying to use as kickoff and amarok I'm not going to have a choice.
Here's to hoping that the "It's finished, honest!" 4.1 will either a) put off a load of users and spur the devs into using some common sense instead of putting faith in design by committee or b) actually be halfway usable. Until then, I'm going to continue hugging my beloved KDE 3.5 - hands down the best UI I have ever used, imagined, experienced, witnessed, praised or wiggled.
So, why the change at all? Again, one part of the answer is the code base of the old solution: it was, according to the developers, ugly. Additionally, Novell spent quite some money and time on the question how a menu actually should look like usability wise. They checked back with users (like, real people, not computer people ) and tried to figure out what actually is usable. Most people like the classical menu because they are used to - that has nothing to do how well this type of menu is actually suitable to the task. And usability wise it is totally nuts to navigate such menus with a mouse! (Do I have to add that I dont use menus at all as long as I can avoid it? Long live Alt+F2!) So they started developing Kickoff - with average or new users in their mind.
In the end the situation was that usability wise Kickoff was a strong improvement to the old menu. It doesnt matter if you disagree with that personally, because that has nothing to do with the figures gathered in a scientific test. Also, there was developer power there to port Kickoff. And they did. Originally, there was even another menu planed, but that was delayed.
For one thing, you describe the way you work your way around cities in the exact same way I do (I similarly have terrible linguistic memory but highly developed spatial and visual), and my experiences with the new kicker exactly mirror yours. Likewise, I rarely remember the name of an app, I either remember the shape of the letters or the shape of the icon (I generally read a word at a time, rather than letter by letter if you catch my drift). The new K menu is so visually inconsistent to me that I find myself literally having to read each and every option nearly every single time, and it also took me minutes to find applications that had taken me seconds before.
The fact that you can't just hover your mouse to see if $app is listed under "tools" or "configuration" or whatever is the biggest failing in my book, it's now click, move, click, maybe scroll up and down a bit, move back, click, maybe scroll up and down a bit. Having to literally click on every single category is so maddening I can't begin to describe it.
Net result of your rather excellent blog post? I'm now of the conclusion that HCI testing is complete bullshit, especially when the reasons SuSE give are "we didn't like it any more" (change for changes sake, anyone?). Where are the options to return some sanity to that clusterfuck of a start menu? It makes Vista's shutdown menu look like a paragon of simplicity./disgust
KDE 4 currently contains so few tweaking options (as in the features are missing from the code, not that you have to edit some obscure XML file or what have you) that it isn't an option at the moment. Since I know exactly what I want to do, I can go from a vanilla KDE install to a desktop that's almost fully tweaked to my needs in about half an hour, after 90 minutes of fiddling with KDE 4 I gave up.
Progress on adding features in preparation for the "real" KDE4 has been glacial enough that I can say it won't have the functionality I desire. But it'll look Shiny, so that's an adequate subtitute.
If I hadn't already posted I'd give you all my mod points.
But then I guess you're preaching to the choir - your setup sounds like it mirrors mine in a great many ways (although I keep my "here's where I start apps and look at the systray" panel on autohide at the top and have one long taskbar panel that spans the screen at the bottom. I too require contrastyness, multiple desktops and a window manager I can configure the hell out of to keep all my apps nicely segregated (although I do run two 1920x1200 desktops on my work Linux box, I find the real estate invaluable).
To me, the clean, grey lines and (more or less) subdued icons of a default KDE are a thing of simple, elegant beauty (default GNOME was too until they decided users hated choice) and the default KDE4 has been such a radical departure from the KDE3 philosophy that I'm not sure I'd call them the same environment at the moment
FTR I'm a sysadmin, whereas I imagine you're a developer but I typically spend about half my life in a konsole window. As a KDE "power user" (in my book that's someone who actually uses the extensive window class placement functionality), KDE 4.0 is such a colossal step backwards in missing features, pointless gloss, huge, flashy icons and basic usability horror (don't make me rant about kicker's replacement again, I grind my teeth whenever I have to use it) that I really, really can't believe they had the gall to release it without an alpha label, and the.0x releases have done precious little to redress the balance, to the extent that I don't think 4.1 is going to come close to approaching KDE 3.5.
Some of us are acutely sensitive to colour, and randomly coloured crap and needless animation is enough to send me round the bend. I use colour inside applications to draw attention to things, I don't want colour clustered around the application drawing attention away from it. I personally find the default theme in KDE4 ugly beyond words.
Posted from a bleak grey desktop, and proud of it.
For one thing, I don't want to have to be continually defining new playlists (I have about twenty that I use for when I want a specific subset of music but that's about it), and secondly how do you script "play random stuff from this folder, this folder, this folder, this folder and this folder until I decide I want to listen to some Tori Amos"? I'm used to just switching to amarok, typing in "tori" and leaving it. I'm not aware of any method in amarok 2 that allows me to do this quickly. Like I said, I don't spend much time looking at my music player - that's what movie players are for.
For some reason I hadn't come across plastique - I've just checked again and it is available in ubtuntu whereas I'm pretty certain it wasn't at first. At last some sanity from all the low-contrast roundiness! Cheers for the prod.
The whole release management thing for KDE4 has been done to death, but I'll restate my position on this - if they call it a 4.0 instead of a 3.99, then I'm going to treat it like a 4.0 and bitch about bugs and missing features. I didn't cut Vista any slack when it was released in an only semi-usable state, and I don't really see how KDE 4.0 is much different. To say I'm looking forward to KDE4 becoming usable is true, to say that I'm a bit narked with the progress and methodology so far is also true. Time will tell I guess.
I know it's not a "true2 part of KDE and I hate to make another rant about this, but I've been unable to see anything but a horrific downturn in amarok, both in terms of usability and basic visual appeal and the developers are convinced their "content-centric" way of doing things is The Right Way. In my experience this translates to: Making it look like superficiallyiTunes whilst continuting to ignore the ability to have the user decide where to put things like the playback controls and position slider Seemingly ephemeral "content window" taking up greater than a third of the main app real estate so I can repeatedly read the wikipedia articles on my bands, or something. Why would I want to do this all the time? Oh right, because it's plasma Aforementioned content window gets in the way of dragging things from the tree browser on the left to the playlist on the right Playlist has been severely gimped compared to amarok 1.4 IMHO. Devs have been telling users like me that keep several thousand items in their playlist are stupid (the only valid reason I ever saw was because it increases startup time - something I'm not particularly worried about with my current amarok uptime being about fifteen days) whilst failing to provide me with a convenient way to listen to my music in the way I liked (generally on random/semi-random unless I want to listen to a particular album or artist, in which case I use the boolean filter)
Maybe I'm horrifically sad and very much music 1.0 or some such crap, but I use amarok because it makes managing and quickly picking out music from a massive collection really, really easy. Amarok 2 just seems to me to be a catalogue of style over usability and change for changes sake. Pretty much every criticism I've seen of the new UI on the blog from the very first mockups has been shouted down with either "these aren't even alpha yet, shut up, the final design will look nothing like this!", "you're wrong, this way is prettier", "we think it's more usable even if you don't, no we won't provide that as an option, it goes against our philosophy" or "can't change it now, we're too close to release". Seriously, how much time to/.'ers spend staring at their music application (not counting pretty visuals like projectm)?
On a more KDE-centric level, I'm not enjoying the low-contrast Qt themes with the insistence of rounding every possible corner, and I've yet to come across any themes that give be the beautific simplicity of Plastik The new XP-style kicker replacement is an absolute abomination to use. Too many clicks, practically impossible to browse the program hierarchies quickly. Everyone says "use the search!" - sorry, I shouldn't have to use the search function because you neglected basic functionality Still doesn't like working across multiple monitors Panel and window configuration options are still severely lacking Seeming enforcement of "the desktop is the application!" metaphor with the proliferation of widgets replacing apps. The desktop, in my way of working at least, is visible for about three seconds after login until an app or five autostarts and covers it. Thanks to KDE's fantastic setup of multiple individually configurable panels and/or kb shortcuts I was able to do away with all of that tiresome minimising of windows. If you're going to make us use widgets, at least give us the option to make them use the window manager so they get an entry in the taskbar, please. The lets-have-windows-without-taskbar-entries philosophy is annoying enough on windows, as anyone who's spent time trying to find that security dialog box that took a minute or two to appear will testify Speaking of the taskbar, the icons are still huge and it still doesn't play very nicely with having lots of windows open Last time I checked, those somewhat confus
Slashdot Mirror
We have an ~8TB JFS array used for shared storage of databases on our linux boxes, as well as for another smaller arrays used for VM's and general file serving duties. Bitchin' fast even with small files, no extra latency incurred by using ACL's as you get with ext3, very fast to format and fsck (ever tried this on a 2TB ext3 array? Painful!), lowest CPU utilisation of any filesystem we tested, frags up alot less than ext (yay for extent-based filesystems), ridiculously quick deletes, equal to XFS in speed at dealign with large files with less overhead at the same time. Only real disadvantage is that it doesn't support shrinking, but that wasn't considered a big downside for us.
Never tried LVM on anything that big though, ours is just stacked hardware RAID10 stripes across a starburst of cheap SATA drives. Use it on a 3TB LVM at home though, not had any problems with it - but that's comparatively zero load.
It wasn't popular initially since it wasn't supported by Redhat, but we have less issues with the JFS partitions than we've ever had with the ext3 ones (yeah yeah, famous last words ;)) - besides, we use CentOS boxes for these anyway (yeah yeah, more famous last words...).
Most underrated filesystem available for Linux IMHO, and those of you who've used it on AIX will already know its awesomeness and how comparatively gimped the linux version really is :) Bottom line: we use JFS on any disc-based IO that's performance critical, and ext3 everywhere else in order to attempt compliance with corporate support ;)
I've set up suspend-to-file before, since my laptop has 8GB of RAM and taking a >8GB chunk out of even a 250GB disc is no small feat - as long as your kernel has support for your filesystem of choice built in, you just run a tool to find the address of the file in question (think it's just a mater of setting your file in the suspend config and then looking in /sys for the address). This way, if I desperately need an extra 8GB of space, I can just `rm /swapfile`, instead of going through the palaver of resizing swap and the surrounding partitions. For the record, I keep a 512MB swap file of which I've never seen more than 15MB used. Anyone's who's ever done anything like accidentally rsync 8 billion files [thank god for rsync 3] will know that having gads of swap available is a curse, since disc speeds haven't kept parity with RAM - the system will become almost totally unusable until it hits the memory limit and oom-killer has its way. In such an event, you *want* to run out of swap as soon as possible.
I thought the idea of using an 8GB swap file ridiculous, since my apps rarely use more than 2GB physical for their working data set - but they do lots of random lookups on data so the huge file caches are a massive bonus to me.
What I can't understand is why suspend doesn't seem to flush the file caches from memory when hibernating - writing 8GB of memory to swap, even when using compression, takes an utter age.
Edit: just found this, added in 2.6.16. Will give this a whirl and see if it improves hibernate any... http://linux-mm.org/Drop_Caches
Just some anecdata; we have an IBM 3850 with eight cores (four dual-core Netburst-based 7120N Xeons) running ESX 3, and we can state that for our workloads (bunch of debian VM's running mostly python) it's spanked by a) the four core Core2 Xeon box and b) the four core opteron box (olde style, not barcelona). Profiling shows that the opteron appears to be kept at the highest utilisation because the VM's wait less on memory access.
Granted, it's not a direct comprison of modern CPU's but the opterons hold their own against the supposedly much faster Core2 Xeons. VMMark is a good general indicator for generic VM performance, but as with any benchmark it's only useful if it's close to your workload. If we get a new cluster it's almost certainly going to be a pair of eight-way boxes and we're quite optimistic that opterons could well play quite nicely, given our VM's thirst for low memory latency.
I know they don't offer an eight *socket* server, but even with the Core2 quad Xeons there's still alot of cache snooping going on over the FSB, and this can kill performance in highly concurrent SMP apps.
£0.02
Another vote for Draytek. Been using them at home and buying them for others for about five years. Never crash, tons of features, most reliable wireless I've ever come across in a compact router, wicked little VPN capabilities for those of you not wishing to shell out a shitload on Cisco but not wanting to set up linux boxes either. Expensive, but will last as long as you can resist the "oooh, shiny!" factor of their next model.
Have been running a 2800VG for about 2 1/2 years now and the only times I've rebooted it are when we've been having problems with the line when the equipment at the exchange goes south. In my last house the thing wasn't rebooted for the whole year it was in there.
Primarily because, in my experience, most users of the word "pro-active" are unaware of it's anti-reactive connotations and use it to describe singularly reactive situations ("I want us to respond to this pro-actively"), or even in just syntax-ruining "I've learnt a cool new word" non-sequiturs ("our new rubber grommets have a 100% pro-active paradigm"). In other words, I'm convinced that alot of people use it because they think it sounds More Important than "active" or lack the vocabulary to better describe it.
It's kinda acceptable in most sysadmin circles as most geeks are aware of things like "pro-active" support (I prefer to call it preventative maintenance myself since it means less fuzzyness for the recipient, which we abbreviate to premaint in conversation) but neologisms are mostly a matter of taste. /spot the word-snob ;)
Damnit, can't use mod points in this thread now.
The way I've found to restore sane address-bar behaviour was to do the following:
Install the oldbar extension
about:config browser.urlbar.matchOnlyTyped boolean set to true.
Bingo! No more hyper-annoying address bar.
Completely agree that the current situation WRT to anything pornographic is somewhat stupid - as you say, there are various definitions of what constitutes pornography, and there are some very, *very* grey areas. My favourite example of joined up thinking is clause 63 of the UK criminal justive bill here in good ol' blighty that makes owning images of "extreme pornography" illegal and could well end up with you on the sex offenders register. Fair enough some say, it's "extreme porn".
:)
First off, there's currently no cut-and-dried definition of "extreme porn" either in the law statutes or the public eye - extreme porn generally just means "something that I, the beholder, consider pornographic and extreme" - too vague for my liking. Secondly, and far worse IMHO, is that the act itself isn't illegal - the law is worded as such that, for example, adults are still allowed to engage in whatever extreme fetish they like as long as it's not against any of the participants wills and the usual H&S/due diligence stuff, but if they want to record it for posterity all of a sudden it's illegal. Makes no sense to me at all.
Which brings me back to complete agreement with your point, or I think so at any rate. As you say, laws concerning things like paedophilia are meant to be there to protect people who can't neccesarily protect themselves. Looking at, for want of a better word, "cartoon" CGI images not based on pre-existing illegal material might still make you and everyone else involved in its creation and distribution a sick and twisted fuck, but what *real* harm has it done? Just being a sick and twisted fuck in itself isn't illegal (yet), and going down the road of "this non-harmful/legal activity MIGHT lead to this harmful/illegal activity" takes us stright back into the "gateway drug" quagmire that you can apply to practically any human endeavour. Alcohol, violent movies/computer games/music, cigarettes, bird-watching (hey, it won't hurt if I just collect *one* egg)... can't have it both ways people. Unless you're cool with the idea of thoughtcrime
Both of these bring us back to the forensic aspect though - at the moment, it's still relatively easy to tell when a digital image has been doctored, especially when "big name" software is used with minimal post-processing that would otherwise remove any obvious artifacts (not trying to sound like an expert, just providing anecdata).
Of course, I don't envy the first people to be convicted of having real images doctored to look like fake ones, or fake images doctored to look like real ones. Just expect the law to follow the path of least resistance and outlaw anything that looks like realistic illegal porn on the grounds that the owner can't reliably prove whether it's lawful (perhaps it was a CGI take on an unlawful image in the first place) or not. We're already in the arguably slightly ridiculous situation where someone 1hr under the AoC is still legally a child, and last time I checked porn downloads didn't come with digitally signed Authentic VeriSign Porn Star Birth Certificates (it becomes even more confusing when you realise that the AoC for sex and/or porn is different across much of even just the western world), so when the situation does arise expect people to undergo trial by mudslinging and "guilty until proven innocent/no smoke without fire" hysteria before the law catches up with technology.
M'lud, I present exhibit A - a 1024-node renderfarm accompanied by a small army of animators and artists that we believe was used to fabricate exhibit B ;)
At the moment (as far as I'm aware - I have a friend who works in forensic IT who has a colleague that specialises in detecting doctored images and video), even the top-end CGI is relatively easy to distinguish from the real thing, especially where humans are involved (even more so for video). Whilst I agree there's a possibility that tech and skills capable of making realistic human animations and the like may only be a few years away, I still think it'll be a long time before such fare becomes indistinguishable from the real thing, and even if it was there'd be an inevitable paper trail (or lack of it) concerning the origin of the pics/vids.
Ever since my patch went through, the Debian GPG uses either double or quadruple ROT13
;)
I kid, I kid
...if it comes into law because it means that anti-Scientology protestors will be legally obliged to obfuscate their placards so as not to say anything offensive.
SCIENTOLOGY - IT'S A CU*T
WHAT A BUNCH OF CU*TISTS
COME ON YOU CU*TS
N.B. to americans not used to judicious use of the "C" word perjorative, please don't take this post as some sort of sweary troll. It's nowhere near as offensive here in the UK as it is in the states. It's not the sort of thing you'd say at dinner with the in-laws, but it's not the sort of thing that gets you blacklisted as some sort of social outcast. If this is the case in the states as well I apologise for being patonising, since it's the general consensus that saying "cunt" gets you universally reviled by everybody over there.[/derail]
The best thing about these new bits of Sony software is that you don't even need to type in your name, DoB, SSN or any of the rest of it - it finds it all out for you! Nonsensical Bullshit Paradigm Shift 2.0 is SO cool!
Yup, I'm looking forward to being banned from the internet after my flatmate downloads a bunch of music from the net. Seems like the only way to have a shared internet account not being terminated in this manner is for me to police every machine in the house, which means invading the privacy of my flatmates.
As you point out, an internet connection is not a person, or a company, or a family, or anything. It is an internet connection that may be being used by all of these.
When I was a penniless student, I participated in a study very similar to this, also at a London university, using CAVE. If this test was anything like the one I did, it was purposefully engineered to try and generate a feeling of discomfort - you "walked" around a room a bit like a library, seeing people mooching about and sitting at desks and reading.
:)
As you walked about, you'd sometimes hear an unintelligible mutter from behind, or in front, as you approached people and you gauged you reaction as to whether a particular person or particular mutter made you feel uncomfortable, or glance at you in someting you perceived as a disdainful expression (I remember the facial animations being pretty convincing for circa 2000) - they fitted you up with a pulse thingy and something that was probably like a galvanic skin response dooferwhatsit as well as relying on your description. I never found out what the results of the study were, or even what the experiment was meant to be proving, but I imagine it was either something akin to judging peoples reaction to VR people, a la Snow Crash's Juanita.
But even six years later, I don't think animation has yet progressed to the state where we can no longer differentiate between reality and CGI, even if we're attempting to suspend disbelief, and as such I don't think any meaningful conclusions can be drawn as to whether it correlates with our real world experiences. Although I will say that spending any amount of time on the tube, particularly at rush hour or after the arseholes stagger out of the pubs, is enough to make anyone wish for their own jet pack
In response to a couple of other comments (slashdot has gone and buggered up my comment view, and it makes viewing and/or responding to comments a complete PITA, as well as hiding virtually everything unless I browse at threshold 0 - I don't get any "X comments below your threshold" linsk any more) - sorry, but appearance is one of the quickest way of judging people, and an evolutionary tool that's served us very well in the past. Clothes, the "default" facial expression, the way someone walks, the way someone holds themself, the way they talk (visually and aurally), the way their eyes move - put them all together and it can prove a very useful metric for a quick assesment of someone's likely reaction. As to the black guys who mentioned they regularly suffer from verbal abuse - let's just say I'm shocked that such things still happen in civilised (and assuming western) society. Must have been living in my own bubble of liberals for too long.
...which brings me back to my original point: KDE no longer give a crap about the power users, they just want their DE to look the blingiest and to hell with usability.
If you think catering for people like me means pissing all over everyone else, you're mistaken. Isn't KDE meant to be configurable?
...the new OSX interface has shown us that we don't need so many colours. Colours in a computer eat up the memory bits and distract us from our reverence. Personally, I'm going to take Steve's advice and go get my eyes chromed.
I've been struggling to remain sufficiently Zen amongst all this as well. I don't recommend you read some of the comments on that page if you value your blood pressure, there seems to be a colossal mentality of "Novell spent loads of money on it, and it's scientific, therefore is MUST be true!".
The search function seems retarded to me as well, for the precisely the reasons you state - we already have a method for launching applications where you're able to type in the start of the name of a program, press a button to initiate a search and then press return to launch the application, it's called a bash shell with tab complete in my book. Woo fucking hoo for your innovative search function. As you say, pretty much the whole point of a GUI, and the reason they're easier to pick up by n00bs, is that they expose you to more than you're currently using. Kickoff is the complete antihesis of this, yet it still manages to be complex enough to be confusing.
The most disturbing thing is the attitudes of some of the key developers who seem to think that a) the current users (especially the power users) don't have valid opinions and b) that it's better to cater to the lowest common denominator. Sorry, that's why I dumped GNOME. I don't want to dump KDE, but if you make everything as ball-gratingly annoying to use as kickoff and amarok I'm not going to have a choice.
Here's to hoping that the "It's finished, honest!" 4.1 will either a) put off a load of users and spur the devs into using some common sense instead of putting faith in design by committee or b) actually be halfway usable. Until then, I'm going to continue hugging my beloved KDE 3.5 - hands down the best UI I have ever used, imagined, experienced, witnessed, praised or wiggled.
As metamatic linked to in his excellent post;
http://liquidat.wordpress.com/2007/11/30/kde-4-some-reasons-for-design-decisions/
So, why the change at all? Again, one part of the answer is the code base of the old solution: it was, according to the developers, ugly. Additionally, Novell spent quite some money and time on the question how a menu actually should look like usability wise. They checked back with users (like, real people, not computer people ) and tried to figure out what actually is usable. Most people like the classical menu because they are used to - that has nothing to do how well this type of menu is actually suitable to the task. And usability wise it is totally nuts to navigate such menus with a mouse! (Do I have to add that I dont use menus at all as long as I can avoid it? Long live Alt+F2!) So they started developing Kickoff - with average or new users in their mind.
In the end the situation was that usability wise Kickoff was a strong improvement to the old menu. It doesnt matter if you disagree with that personally, because that has nothing to do with the figures gathered in a scientific test. Also, there was developer power there to port Kickoff. And they did. Originally, there was even another menu planed, but that was delayed.
Good blog post :)
/disgust
For one thing, you describe the way you work your way around cities in the exact same way I do (I similarly have terrible linguistic memory but highly developed spatial and visual), and my experiences with the new kicker exactly mirror yours. Likewise, I rarely remember the name of an app, I either remember the shape of the letters or the shape of the icon (I generally read a word at a time, rather than letter by letter if you catch my drift). The new K menu is so visually inconsistent to me that I find myself literally having to read each and every option nearly every single time, and it also took me minutes to find applications that had taken me seconds before.
The fact that you can't just hover your mouse to see if $app is listed under "tools" or "configuration" or whatever is the biggest failing in my book, it's now click, move, click, maybe scroll up and down a bit, move back, click, maybe scroll up and down a bit. Having to literally click on every single category is so maddening I can't begin to describe it.
Net result of your rather excellent blog post? I'm now of the conclusion that HCI testing is complete bullshit, especially when the reasons SuSE give are "we didn't like it any more" (change for changes sake, anyone?). Where are the options to return some sanity to that clusterfuck of a start menu? It makes Vista's shutdown menu look like a paragon of simplicity.
KDE 4 currently contains so few tweaking options (as in the features are missing from the code, not that you have to edit some obscure XML file or what have you) that it isn't an option at the moment. Since I know exactly what I want to do, I can go from a vanilla KDE install to a desktop that's almost fully tweaked to my needs in about half an hour, after 90 minutes of fiddling with KDE 4 I gave up.
Progress on adding features in preparation for the "real" KDE4 has been glacial enough that I can say it won't have the functionality I desire. But it'll look Shiny, so that's an adequate subtitute.
If I hadn't already posted I'd give you all my mod points.
.0x releases have done precious little to redress the balance, to the extent that I don't think 4.1 is going to come close to approaching KDE 3.5.
But then I guess you're preaching to the choir - your setup sounds like it mirrors mine in a great many ways (although I keep my "here's where I start apps and look at the systray" panel on autohide at the top and have one long taskbar panel that spans the screen at the bottom. I too require contrastyness, multiple desktops and a window manager I can configure the hell out of to keep all my apps nicely segregated (although I do run two 1920x1200 desktops on my work Linux box, I find the real estate invaluable).
To me, the clean, grey lines and (more or less) subdued icons of a default KDE are a thing of simple, elegant beauty (default GNOME was too until they decided users hated choice) and the default KDE4 has been such a radical departure from the KDE3 philosophy that I'm not sure I'd call them the same environment at the moment
FTR I'm a sysadmin, whereas I imagine you're a developer but I typically spend about half my life in a konsole window. As a KDE "power user" (in my book that's someone who actually uses the extensive window class placement functionality), KDE 4.0 is such a colossal step backwards in missing features, pointless gloss, huge, flashy icons and basic usability horror (don't make me rant about kicker's replacement again, I grind my teeth whenever I have to use it) that I really, really can't believe they had the gall to release it without an alpha label, and the
Where's my KDE "blackjack and hookers" edition?
Some of us are acutely sensitive to colour, and randomly coloured crap and needless animation is enough to send me round the bend. I use colour inside applications to draw attention to things, I don't want colour clustered around the application drawing attention away from it. I personally find the default theme in KDE4 ugly beyond words.
Posted from a bleak grey desktop, and proud of it.
For one thing, I don't want to have to be continually defining new playlists (I have about twenty that I use for when I want a specific subset of music but that's about it), and secondly how do you script "play random stuff from this folder, this folder, this folder, this folder and this folder until I decide I want to listen to some Tori Amos"? I'm used to just switching to amarok, typing in "tori" and leaving it. I'm not aware of any method in amarok 2 that allows me to do this quickly. Like I said, I don't spend much time looking at my music player - that's what movie players are for.
For some reason I hadn't come across plastique - I've just checked again and it is available in ubtuntu whereas I'm pretty certain it wasn't at first. At last some sanity from all the low-contrast roundiness! Cheers for the prod.
The whole release management thing for KDE4 has been done to death, but I'll restate my position on this - if they call it a 4.0 instead of a 3.99, then I'm going to treat it like a 4.0 and bitch about bugs and missing features. I didn't cut Vista any slack when it was released in an only semi-usable state, and I don't really see how KDE 4.0 is much different. To say I'm looking forward to KDE4 becoming usable is true, to say that I'm a bit narked with the progress and methodology so far is also true. Time will tell I guess.
I know it's not a "true2 part of KDE and I hate to make another rant about this, but I've been unable to see anything but a horrific downturn in amarok, both in terms of usability and basic visual appeal and the developers are convinced their "content-centric" way of doing things is The Right Way. In my experience this translates to:
/.'ers spend staring at their music application (not counting pretty visuals like projectm)?
Making it look like superficiallyiTunes whilst continuting to ignore the ability to have the user decide where to put things like the playback controls and position slider
Seemingly ephemeral "content window" taking up greater than a third of the main app real estate so I can repeatedly read the wikipedia articles on my bands, or something. Why would I want to do this all the time? Oh right, because it's plasma
Aforementioned content window gets in the way of dragging things from the tree browser on the left to the playlist on the right
Playlist has been severely gimped compared to amarok 1.4 IMHO. Devs have been telling users like me that keep several thousand items in their playlist are stupid (the only valid reason I ever saw was because it increases startup time - something I'm not particularly worried about with my current amarok uptime being about fifteen days) whilst failing to provide me with a convenient way to listen to my music in the way I liked (generally on random/semi-random unless I want to listen to a particular album or artist, in which case I use the boolean filter)
Maybe I'm horrifically sad and very much music 1.0 or some such crap, but I use amarok because it makes managing and quickly picking out music from a massive collection really, really easy. Amarok 2 just seems to me to be a catalogue of style over usability and change for changes sake. Pretty much every criticism I've seen of the new UI on the blog from the very first mockups has been shouted down with either "these aren't even alpha yet, shut up, the final design will look nothing like this!", "you're wrong, this way is prettier", "we think it's more usable even if you don't, no we won't provide that as an option, it goes against our philosophy" or "can't change it now, we're too close to release". Seriously, how much time to
Since the site seems slashdotted, here's the latest dev image posted to the amarok blog: http://amarok.kde.org/blog/uploads/Newtheme.png
On a more KDE-centric level, I'm not enjoying the low-contrast Qt themes with the insistence of rounding every possible corner, and I've yet to come across any themes that give be the beautific simplicity of Plastik
The new XP-style kicker replacement is an absolute abomination to use. Too many clicks, practically impossible to browse the program hierarchies quickly. Everyone says "use the search!" - sorry, I shouldn't have to use the search function because you neglected basic functionality
Still doesn't like working across multiple monitors
Panel and window configuration options are still severely lacking
Seeming enforcement of "the desktop is the application!" metaphor with the proliferation of widgets replacing apps. The desktop, in my way of working at least, is visible for about three seconds after login until an app or five autostarts and covers it. Thanks to KDE's fantastic setup of multiple individually configurable panels and/or kb shortcuts I was able to do away with all of that tiresome minimising of windows. If you're going to make us use widgets, at least give us the option to make them use the window manager so they get an entry in the taskbar, please. The lets-have-windows-without-taskbar-entries philosophy is annoying enough on windows, as anyone who's spent time trying to find that security dialog box that took a minute or two to appear will testify
Speaking of the taskbar, the icons are still huge and it still doesn't play very nicely with having lots of windows open
Last time I checked, those somewhat confus