I think you missed the point in your rush to object. What's the legal difference (IANAL) between optimizing HTML and inserting ads? In both cases X leaves the source, Y arrives at the destination. Opera does something like this for their Opera Mini browser: the content that is delivered to the browser isn't even HTML, it's some proprietary format, although the browser usually correctly renders it to what the HTML would look like. However, in case of Opera Mini, I explicitly agree to such manipulations and to accompanying technical solutions.
Once again, this may be a good move on Cotendo's part that will lower their costs and improve end user experience, but it is a dangerous one, because if ISPs and CDNs automatically receive the right to manipulate transmitted content however they please, it will certainly lead to abuse in some cases.

It doesn't really matter if it's a technically good move; if this sticks, we might be getting lots of ISP-inserted ads, iframed toolbars and other "value-added" stuff in non-encrypted HTTP traffic pretty soon.

I personally find Instant Search an awesome feature. However, it seems to conflict with an experimental search feature I love (namely, the keyboard shortcuts), so until Google introduces a version that supports both Instant and keyboard shortcuts, the latter feature wins.

Don't know it this is a slashvertisement or not, but the original post on the Ksplice blog sounds much more adequate and calm to me. I wouldn't bet that the AC who posted this works for Ksplice.

Oh. I see. They're going to lock down fully capable CPUs. Now that's a shame.

Please keep in mind that usually such downgraded hardware is actually a top of the line CPU that has some defects and was unable to pass all the QA tests imposed on the flagship product. So while this stuff may be cracked, it is quite possible that people who do so will experience lots of bugs due to failing cache blocks and hyper-threading modules...
If that's the case, I don't see a problem here: Intel just charges more for a CPU that has all of its components working properly, and less for units that had high failure rates of the said components and had them disabled.

Wrong. Twitter doesn't treat the username as metadata - it's still limited by that SMS size. From the help center:

While theoretically it's definitely possible, I'm not sure if storing BLOBs in a database is a good solution; at least none of our colleagues from other universities and agencies like ESA or NOAA that I personally know (or at least have some knowledge of how their systems are built) use this approach. All of them do exactly as GP said: store metadata and e.g. file paths in the database, and store tables, BLOBs etc in the file system. Since some of these projects have a huge amount of resources, I think that this approach may have more to it than it may seem.
And as far as backups and mirroring go, there are projects that use popular protocols like BitTorrent for mirroring and load balancing, and there are some that have built non-trivial custom data distribution systems (like the SDO guys).

Actually, these 85k rows represent only a day's worth of data from a single satellite instrument. We often operate on data sets tens of millions of rows large, and the general performance trend is nearly the same.
About scaling: as far as I understand, PyTables' main bottleneck is raw disk I/O. I think that applying a couple of primitive tricks like using memcached and a large amount of RAM will tip the scales in its favor even further, but that needs to be verified.

Hello, I'm a space research guy.
I've recently made a comparison of MySQL 5.0, Oracle 10i and HDF5 file based data storage for our space data. The results are amusing (the linked page contains charts and explanations; pay attention to the conclusive chart, it is the most informative one). In short (for those who don't want to look at the charts): using relational databases for pretty common scientific tasks sucks badly performance-wise.

Disclaimer: while I'm positively not a guru DBA and thus admit that both of the databases tested could be configured and optimized better, but the thing is that I am not supposed to. Neither is the OP. While we may do lots of programming work to accomplish our scientific tasks, being a qualified DBA is a completely separate challenge - an unwanted one, as well.

So far, PyTables/HDF5 FTW. Which brings us back to the OP's question about organizing these files...

Hell, it's about time!

In the US, maybe. In most other countries, not quite. E.g. in Russia you can get as much as 15% off the retail price, so most expensive and cutting-edge hardware is bought online. Last time I checked, Nexus one was both expensive and cutting-edge :)

FTFY :)

I did follow the TFA to the origin of the story (MyDroidWorld forum), and I still don't see any code, captured data or even a photo of the said eFuse chip inside the DroidX. I understand that the original poster appears to be a reputable hacker, but come on, what kind of real reporting is this? Can anyone else verify these claims? More information needed, thank you very much whoever posts it, because if true, this is an outrage.

You can't have a lot of kids knowing how to program tomorrow if you don't spark their interest with such a tool today. And IMO this will be great not only for attracting and educating future software engineers, but also to tap into the pool of active talented kids who are not going to be software engineers, ever. The kids who will be nuclear physicists, radio geeks, astronomy fans, journalists will also acquire basic programming abilities without distracting from their main specialty to learn a programming language or two, dive into a complex SDK and constantly work to keep these skills up to date.
In short, I think that App Inventor is pretty awesome.