This is already a solved problem:

www.crooklynclan.net

The first thing you see on the page is the most recent entries, no matter what they are. Genre pages are available, with each genre getting their own page of recent entries. Completely separate to that are 'charts', which show the top tracks from this month, past six months, and 'all time', with both site-wide and per-genre charts available.

The site's search feature needs work, but that's a different problem altogether. The point is that there is room for both a 'recently added' and a 'most popular' set of rankings, and the way to do it is already in place, and in service. Crooklyn Clan has plenty of the same issues that cause the problems shown in the App Store (lots and lots of contributions, poor SnR, frequent turnover, fickle customers), and they've been doing it right for years. All Apple needs to do is forego their adherence to "not invented here".

So, a major company uses open source software and Slashdot complains about it? There's just no winning.....

It's only a good idea if they can negotiate a peering agreement with Verizon so that they don't end up getting the slow internet anyway...but then Verizon will be mad at them and try to get the internet on their side by writing a public nastygram, which might actually be a good thing because Verizon will find itself on the wrong end of the "Think of the Children!" argument.

That will only take you so far. With so many programs defaulting to the My Documents folder, it'd be annoying at best to have to point to c:\realdocs "because viruses". The user could point the "My Documents" folder to c:\realdocs, but now we're in the same boat again. Even if a user decided it was worth the hassle to deprecate the use of the system variable, c:\realdocs would still be accessible by the same user. From Windows' security standpoint, there's no difference between the user being attacked by ransomware, and the user adding a password to an Excel sheet. Thus, ransomware doesn't need root privileges to mess up a user's files.

Even beyond that, the next generation of ransomware wouldn't exactly need a foundational rewrite to go to %user%\recent and see where those files point to and encrypt all the .docx, .xlsx, and .qif files there. I'm sure that somewhere in userland, there's some indication as to where the Dropbox/OneDrive/Gdrive folders are, and encrypting all that stuff. Even less complicated would be to search all available hard drives for user generated file types. .dll files wouldn't be worth it, but .qbw files very much would be. Ultimately, trying to thwart an attack of this nature would be of limited success, because from the most literal of standpoints, the virus is doing nothing different than what a user would be doing.

Amongst the things that makes this kind of attack so successful is that very problem: if you're trying to prevent outbound traffic at the firewall, you've already lost, basically. How does security software distinguish. technically, between a cryptovirus taking a file hostage, and a user passwording a file with WinRAR and uploading it to SpiderOak? That, good friends, is a question that I pay ESET a nontrivial sum to discuss and determine.

See. what I thought would be a useful stat to show would be "the average amount that those who spend, spend". In other words, if Google showed how much was spent on a given "freemium" app by those who spent >$0. This would give users a meaningful metric with which to decide whether it's worth it to attempt to use the app, because they could, on average, expect to spend that amount. If an app has a spending average of precisely $4.99, and the pro version costs $4.99, then it's fair to assume that users only pay for the 'pro' key within the app, and it won't nickel-and-dime all day. If $25 is the going rate, it's clear that the game is a skinner box and isn't worth it.

Of course, the bleeding obvious issue is that developers wouldn't be too fond of that number getting too high, which people would be less inclined to do once they have the feeling of going 'above average'.

You're negating the value of the "plus expenses" part when paired with a little creativity...

"I'd never make it there in time to help you if I didn't rent that Aston Martin!"
"The only place to eat between my location and your office was Ruth's Chris Steakhouse. I certainly didn't want to drive out of my way and delay this further!"
"The only laptop capable of handling that kind of process was the top-tier Macbook Pro...but I negotiated a discount on the iPad that I gave to the CEO in your name to ensure that you get full credit for leveraging the synergies!"

Three reasons I personally can think of. First, NetBSD specifically is a fork intended to run on basically anything with a microprocessor. CentOS will run on x86 hardware, and in the form of Pidora and similar, runs on ARM. Try it on an Itanium or SPARC or PowerPC Mac, and things get a smidge more interesting.

Second, ZFS. Now cue those who believe that file system nirvana is found in btrfs or ReiserFS or HFS+, but I'm a huge fan of ZFS as a file system. If you're like me, you'll be using BSD in the form of one of its descendants, like FreeNAS or NAS4Free, where ZFS makes lots of other things much easier.

Finally, the license. I'm neither a programmer nor a recompiler so my use of BSD licensed software is essentially identical to my use of GPL software ('free as in beer', with the occasional bug report). For purists and programmers, there is a difference in what is and isn't allowed under the respective licenses.

This video explains the situation very effectively.
http://www.youtube.com/watch?v...

Oh c'mon, you're trying to tell me that *nobody* at NASA had the common sense to call a few Libyan nationalists and order some used pinball machine parts off of Amazon?

Now admittedly, I'm a bit bitter about a problem that's not really Creative's fault. I bought an Audigy 2 ZS for my laptop using PC Card...and then the next wave of laptops only came with an Expresscard slot. So, I ponied up again for an X-Fi card that fit the Expresscard slot...and then laptops stopped coming with those. Now I fully admit that Creative isn't to blame for that, but it is sad just the same. However, I digress.

I use my onboard audio for nearly all of my listening needs. My internal speakers are utter crap (I think one is blown, actually), and thus, even if Creative added all the super-duper offboard processing in the world, it wouldn't sound any better than what those speakers can pump. Adding a nice set of Sennheiser or Denon headphones, I can start to hear some of the MP3 sizzle in the 128kbps MP3s, and a handful of 192's, depending on the song and the encoder and settings used. Even playing video games, the difference between 'Good Enough' and 'X-Fi Good' never comes into play, because it's the nuts-and-bolts of the big picture that will make or break it in either direction - if the sound effects and musical score is good, the miniscule difference an audio chipset will make has nothing to do with it. If they're crap, a ZxR processor isn't going to change anything.

That being said, I still use offboard audio hardware on a regular basis. I use my Rane SL3 to DJ with Serato. Even if it wasn't a de facto hardware dongle to unlock the Serato software, there's no motherboard chipset that supports 2ms latency from end-to-end of the audio path. In other words, my SL3 can reliably take an audio signal from my turntable, translate it into speed and directional data, and send MP3 audio back out, in 2ms. Creative doesn't make hardware like that. The story is pretty similar for my Audio6 (which I use for Traktor) and my Connectiv (which I used to use for Torq and Deckadance, though it required closer to 5ms latency to be stable). I have a MobilePre USB that I use occasionally for XLR and 1/4" recording. These are niche products for niche purposes, but the fact that your local Guitar Center sells a range of these kinds of interfaces demonstrates that there's indeed a market for discrete audio hardware. Creative just doesn't make it.

Your ideas are intriguing and I wish to subscribe to your newsletter. No seriously, I'm with you on the 3mbps/sec DSL situation and am wondering what software/hardware you use for this. I see this as being quite handy on Patch Tuesday and similar. I have half-ideas as to how to make it work, but I'm interested to hear about your tried-and-true setup.

Well, duh. MapPoint and S&T was a plastic-disc software title, intended for end users to do stuff without an internet connection. See kids, in the days between the joys of attempting to re-fold a paper map and always-on, always-connected internet streamed maps, companies got all the street information together and sold a software release in a perpetual licensing format. People could then take their laptops and a serial (later USB and/or Bluetooth) GPS add-on and navigate with a laptop, without worrying about data plans, cellular outages, or getting stuck on a necessary phone call that brought into question one's allegiance to accurate navigation.

In the case of MapPoint, routes and distances were mass queried and used in tandem with Access and Excel to make geographical and topological data useful in a business context.

Websites are going to use Google maps (or yahoo/mapquest/bing, to a much lesser extent) because their APIs allow embedded maps nice and easily. For folks who need offline information, Google Maps was never intended to fill that space. Now, it seems, Delorme is the sole holdout for plastic disc mapping software.

Annoyingly, it's not just Encarta. It's seemingly any offline reference title. Grolier's is paywalled to oblivion, Britannica gives the first two paragraphs, Simon & Schuster haven't sold a reference app in years, and Wikipedia is, well, Wikipedia.

Now yes, the internet is how we get data around fastest, and even CDs were a de facto subscription since you'd buy a copy every year or two to stay current. I get that. Where plastic disc media had some usefulness to it was that, for K-12 schooling, it was easier to cite them as one would cite a traditional printed volume. Additionally, even if not the most bleeding edge information, most information contained therein would remain relatively consistent from year to year (especially ones on historical matters; technological matters, less so for obvious reasons). It also provided a baseline with which to compare other sources. If Encarta and Wikipedia disagreed, it'd pose the question of 'why'. Was there some sort of major breakthrough that allows Wikipedia to show its strengths as being an up-to-the-minute, crowdsourced reference, or is the Wikipedia article amidst an edit war? At least with Encarta, there's some semblance of "information freeze" where it's accurate to the point where the disc was pressed, and can be relied upon as such.

Sending reference works "to the cloud" makes sense, until companies paywall the whole thing, you don't know what you're really getting when you fork over your Mastercard, and it causes people like me to wax nostalgic for the plastic disc for well-written, relatively unbiased descriptions of WWII battles.

Using a Hotmail account.

I'm more of the persuasion that the reason why Java is still in widespread use in the enterprise is because it predates most other solutions and no one wants to pay between five and nine figures to replace the existing system.

Java is getting particularly annoying in that they're try to make the runtime environment more secure...and in doing so, have a tendency to break things to the point where it's a requirement to undo all the new security defaults in order to make the Java stuff actually load. Oracle has indicated that it will soon remove the ability to allow things to run by clicking 'yes/allow/run' to half a dozen warning error messages, which means that the amount of time and effort to make the JRE security requirements happy may eclipse the time saved in using it in the first place. Java is also a nonstarter on mobile devices. Finally, I've had major issues reminiscent of IE6 hell - $SOME_APPLET is only compatible with a particular version of the JRE and it's impossible to upgrade without breaking it, so people are stuck on that particular variant of Java.

Disclaimer: I haven't written a line of code since college. I have, however, had to support Java applets and, without exception, they cause these kinds of problems. I don't care if you use PHP, Perl, Python, Ruby, or .NET...just please...PLEASE spare the support staff the hell of dealing with end user Java sites.