Yep, it's mostly used for front-end duties like connection pooling, load balancing, SSL offloading, gzip, that type of thing. If you're running PHP stuff, it's still debatable whether you want to go FCGI or PDM instead of Apache's built-in module. There are ups and downs in both cases and you'll have to see what works best for your site. At my company we use Nginx up front (with server type obfuscated) for SSL offloading and gzip and connection pooling. From there it goes into a varnishd cache on the same server (stored in 100% RAM) which handles the static stuff. Varnishd then forwards remaining requests to an L7 load balancer appliance type thing which then drops requests to each of 10 web "application" servers which are a combination of Apache with mod_php, Tomcat and Jetty Java servers. We've also used Nginx as an IMAP proxy and cache and it works quite well for that.

Apache has a good architecture but it's horrible at handling a lot of simultaneous connections and recycling them (that will change in 2.4 but it's not out yet). Also, if you're using mod_php, over time each Apache process will take the total maximum amount of RAM your php process uses, and many of our PHP applications use 128-256M of RAM or more (data management type stuff). So you can run a server out of RAM if you're trying to maximize connections.

Nginx can handle 10K connections on a little box with very little RAM due to the way it threads stuff. It's basically a copy engine and it's very fast. Varnishd can also handle a lot of connections and can serve up content straight from RAM in less time than apache takes to build a connection. That being said, Apache is reliable, and has I feel better logging at the moment and just more of everything. It's a reference implementation. It's actually fine for most purposes but if you're handling 1000 users simultaneously and they are making 10-20 connections each with various service calls and static downloads, you gotta have something that can pool the conenctions on the front end and handle static content or you're going to spend a lot of money on RAM. And if you're serving up static content with Tomcat, Tomcat is absolutely garbage. I think it has to boot the whole JVM to serve up your one file. If not that bad, it's still awfully slow, and it REALLY benefits from caching up front. BTW, Nginx does caching as well but varnishd seemed more mature and elegant.

Now lastly, you can just go out and buy an F5 BigIP and it does all this stuff on specialized hardware (Ok, special board, intel chip) and it's out of the box. But even the little ones are $20K which is a lot of software dev hours and/or web server/database/storage hardware. Would be nice and fun to have but if you can't spend the money on hardware (and training!) the nginx/varnishd frontend is pretty much the best setup in my book at the moment. A little complex but once it's set up you just let it run. I made an internal nginx cache for all our internal sites, including some Java apps (e.g. Jira) and with requests going through the cache everything just flies. If you use sharepoint on IIS, you would be prettty stupid to not try a cache server up front, it's amazing. If nginx fixed mod_rewrite stuff to be the same as apache, it would probably be possible to make it into an application server, and we're going to get a test environment set up with php-fpm and see how it fares. We'll see how managable it is though.

No one's made the Steve Jobs Borg Icon for Apple yet.. just give it time for this glorious torch passing. Meanwhile Gates gives another $2B away this year to eradicate malaria worldwide and saves 200M lives.

By faaaar and away the 9.0 Japanese earthquake and the subsequent Fukushima Daiichi nuclear disaster was the most important story of the year. I'm not in Japan but Japan, the 3rd largest economy, was Stunned by it. We are still seeing economic and emotional effects. Not to mention just being sick for all those people who had to move or were hurt. And the fact that the land is now closed off for decades. I don't know how you can put Steve Jobs in the same category. Sure, he did a lot at Apple... I think it's weird to worship the guy though, especially when he was a well-known dick and basically his legacy trying to put an end to general purpose computing. I don't think it'll happen but they definitely tried. Ritchie is definitely far more worthy of worship, being the humble person he was. But when 100+ million people are affected by a natural disaster and then a big man-made disaster, that pretty much takes the cake over any single person dying, no matter where they were in the tree of the economy.

Anyway, there were hundreds of stories on Slashdot about Fukushima, so many that I had to turn my back on them after a while because the situation was so depressing. I think we can start looking back at it now though for some closure. As far as I know it's still pretty fucked there, and yet Kim Kardashian marraige/divorce is the main story of the year on the pop news sites. Oh to be that uninformed and ignorant. I wish I could go back to before Fukushima and become blind and deaf for 9 months so I never had to hear about it..

Primaries are just polls. They are useful to the parties to determine the public tone of the upcoming election. Most likely the national GOP already knows who they are going to run, and they are using the primaries as a way to find the relevant messages they will focus on in 2012. This new system here is of course endorsed by many prominent politicians. This is because such a system, controlled by the powers that be, would enable more complete and more real time polling. Polling serves the purpose of determining not only what people think of individuals but it actually closes the loop on a complex quality control process. If a candidate can know how different sections of the electorate react to each thing they say, they can hone that message to reach the most people. If people think that logging in and "voting" in this system will somehow allow them to have influence, they will do it. However, time and time again you see the politicians' ACTS are very different than their SPEECH. A real revolution in democracy would be a version control system for the U.S. Code so we can see every debate and comment on each piece of code and generate summary reports of the changes in the code and how it relates to the "politics", that is the sampling and influence of public opinion. Of course they will find ways to ruin that too so they can make money and get power (spamming it or something) but at least it's a technical step in the right direction that will maybe open the doors to more of the younger generation to really get involved in making laws in this country. Because that's what this is all about--what are the laws and why.

$2.75 earnings per share and a 3.1% dividend yield on 23 billion in net income, I think they're doing fine. They are a money machine, they aren't hipsters. Why invest all the time and money in huge new projects and technology when you have stuff making tons of money and more importantly profit. MSFT isn't even in the same class as Apple or whatever people are comparing them to nowadays (Apple, I'm guessing). Geeze, look at IBM, same story. No one balls out IBM for having tons of patents and no real products.

Alfresco implements the sharepoint protocol so you can save and load straight from MS Office. It does a lot of what Sharepoint does. It's a few years behind, though, missing stuff like faceted navigation, a really good search/webcrawl, and all the single sign on stuff. But it is standards based, uses a JCR repository, implements CMIS, they have workflow and records management and web content management.

Couldn't you jam GPS and pretend you were the GPS satellites and then send singals to the drone in such a way that it thought it was going home but wasn't? Hopefully they kept some inertial guidance on board to at least provide error detection.

Workers unionize and professionals unite under a cabal. IT should not be relegated to the realm of the auto worker or grocery store clerk (ok, maybe tech support). But high level IT people are professionals just like MD, lawyers and accountants. There needs to be a real professional organization like the AMA or ABA and professional licensing like the CPA to set some walls about who can do what position in this business. Otherwise the undercutting will continue as any kid from college can walk across town and "be the IT guy" and outsource it to a cloud run by the same kids. Stuff like this bill and the enevitable data disasters--which are coming, just wait--will cause a requirement for a certain educated elite to decide the direction of the management of the countries' tech so we all keep each other safe. Because in a networked world we are all somewhat dependent on each other. We need coordination. We need professionalism. We need barriers to entry. Tech is too valuable, it's going to be the new human life soon, invaluable, lots of growth potential. Don't sell out to the other professionals.

I think there's good things happening around security, I think the certs are a step in the right direction, but every stupid company has certs. I do like stuff like SAGE and LOPSA and I think those organizations could be remodeled after AMA and really build something that will stabilize the industry. I hope this bill is the final news you need to make a decision that you are now needing to make a choice--will you be a mechanic, a dock worker, a factory worker? Or will you be a doctor, lawyer, or accountant.

Yes, IE is by far the most common browser for our sites, with about 50% IE, 30% FF and the rest chrome. Geeks use Chrome because it seems faster (especially when using their beloved Google Apps), but check out your process list. It uses a pretty meaty process for each tab and eats RAM like crazy. For traditional web browsing, Mozilla is the best. And it does have some deployable settings for a corporate environment. It supports Kerberos. Etc. Etc. Chrome is for like javascript stuff only. All the kids might say, "Who uses Kerberos instead of a web sso solution?" but the answer is millions of people at their jobs do. Finally, I agree with the parent. StatCounter is a geek counter.. not a everyday joe and grandma counter. Look at your own browser analytics and determine for yourself what audience is using your sites.

This is actually the perfect place to incubate distributed object stores (e.g. Hadoop on one end, something like Zimbra on the other). One namespace .gov, with sub-namespaces. With a CMIS interface. Anyone see VMWare Project Octopus yet? Well, take that times 10,000 and you have a pretty nice records management system, platform independent. There's also Alfresco which is using the JCR spec which I believe can be moved to some type of distributed backend. But it implements CMIS, has a DoD spec records management system.. So the general spec would be a CMIS framework, each department/branch/whatever makes available a service for document retrieval, central .gov listing of the services, basically what Amazon does for literally everything it does. Do not compromise, executive order Jeff Bezos style, everything is a service with a public interface. I think it is possible, but it would take a lot of just plain buying in and our government (the bureaucratic, non-political side) has gotten really really good at dragging their feet and doing nothing. The cuts are coming though, and they will have to improve efficiency just like we all have in the private sector. Of course Defense is the worst, but education can use some work as well.

Well, there's that. There's definitely that. But what about the replies from Apple. Supposedly you tell the phone something, and then it decodes it and then it can actually change settings on your phone based on what you told it (e.g. send a message, create a calendar, etc.). Now I knew it was cloudy to begin with but I just started thinking, "what about the replies". Because that's the real "back door" here. First of all, Apple can send arbitrary commands to your phone. Not a big deal, but certainly something that might not raise as much suspicion now that there's a constant stream of binary between the two (soon to be encrypted further, I'd guess). Secondly, the requests and responses are travelling through the Internet and of course through the airwaves, which means someone could easily alter those replies with the aforementioned valid keys (pretty easy). And if you keep it quiet, you could even pretend like you're Apple's CA and not have to install a new cert on the phone. So, what is the API for the responses? What calls can a response make on the phone?

That's not to mention, of course, how easy it would be to just speex encode everything you say in a room and send to arbitrary internet hosts. No sir, I don't like it. Of course, I think it's stupid to ask your phone stuff anyway, but that's just me. I would like something speech oriented in a 55" LED format for my TV room, however.

The worst part of all is that Microsoft is now the only company that will make local speech recognition that's not a cloud service. Google, no, Apple, no, because both of these are relying on ads a LOT (and the general decline of traditional ad industries, thus flowing to them), and Microsoft relies on profits from publishing software. Microsoft is actually the ones looking the least evil, or at least the lowest enabler of evil. Sigh, what's the world coming to.

It says 180 when I clicked the link.

The so called repayment is no such thing. Let the Fed show the actual money, because the moment they try to sell the assets that they own from those banks they will find out their actual worth, and it's nowhere near what they tell you, there is no repayment.

Uh, the Fed prints money. They can show whatever they want. The Fed responded to a massive deleveraging (leveraging is when money is spent multiple times). In the U.S. economy, we rely on money being spent around 8 times at once. It's hard to envision but there's a sort of chain reaction effect when money is spent and economists call this the money multiplier. What happened was the housing market was over-represented in the leveraging and when it tanked it took a major multiple of itself out of play. What the Fed did was re-leverage all of that itself by basically printing up a portion of the lost multiplied leverage and just giving it to some big companies. They in turn, spent it and the regular money multiplier took effect enough to lift the credit markets from a standstill. It wasn't fair, no, but we are all better off today that we would be.

Where we stand right now is that the private capital leverage is starting to pick back up again as banks loan money. As that happens, people are going to be willing to take on risk and at that point the Fed will unload the stuff it bought for cash it printed, thus getting that cash back. The Fed can then just burn the money or basically buy back debt from china or something. The beauty is that we make the rules and the rules call for, above all, certainty of value, even if that value is dropping at a certain rate (e.g. inflation). What they did was brilliant, actually made the taxpayers money and, although we will have to suffer some inflation (and it's coming), the vast majority of prices were already inflated due to the over-leveraging (which functions as monetary inflation does on prices), so food isn't going to shoot up and if you're well placed in the market now you will see some very nice gains over the coming decade. Plus, baby boomers are going to go from saving to spending mode starting in 2016 which will just add some more bull market momentum. Assuming we don't overheat again, we are primed for a good 10-15 years of steady growth, and they can keep us from overheating by doing some taxes, or raising rates or a combination of both. I have a feeling it'll be taxes first, so we can buy back those treasuries.

I second LTO tapes if you really want an "archive". Also, storage is a concern, so you'll want to look at iron mountain or access co for offsite storage in a tape vault with Halon. Usually around $25 a month plus tapes and transport.. That being said, you can get a highly probable recoverability by having lots of copies of the data. So lots of drives in lots of locations work. Rsync.net has a really good online service (and it's cheap). I wouldn't recommend any other place since I've worked with rsync and know them to be a good company. But still, unless you have a physical copy on tape somewhere, there's no guarantee. It's all about making it improbable that you'll lose the data. After that, there's insurance ;)

Also, I want to add that I'm not against Scientific Linux. I think it's a great product. I think they tend to rush to the newer releases because they are more likely to be running HPC and other projects needing stuff. I feel like as far as QA, CentOS emphasizes what I need it for, which is stable application servers and database servers. I need a stable OS that I can build my applications and servers on. I don't really depend on Redhat or CentOS for application packages, just the core OS. I find that when I purchase an application for RHEL, it works on CentOS. SL, on the other hand, has much better QA around the HPC clustering, experimental filesystems, math stuff (duh) and some of the graphical environments. So, if I was a scientist doing research, I would definitely want to be in a community with lots of scientists doing QA. Instead, I'm in the cheap businessman community (Non-profit) who can't afford RHEL, thus CentOS. I know they both try to be as close to RHEL as possible, and I'm sure in the vast majority of cases they are identical, but I feel the communities differ and that's what I have based my choice on.