SSDs and disk speed have nothing to do with this. None of these writes are hitting disk. All they've shown is that when you cache a write to disk, the operating system might add data to it more efficiently than the slow Python and Java string code can expand a string.

There's minimal evidence that healthy eating alone lowers your long-term body weight. That's part of the point really. Any diet change, be it fad dieting or more sustainable health eating choices, they are all capable of short-term weight loss. Keeping that weight off on the long-term is is a so much harder problem, it's barely related to what works for losing a large amount of weight in the first place.

If you read studies about people who lose and keep weight off, like Long-term weight loss maintenance, the common factors that always show up are both very low calorie counts and constant feedback. Basically, chart your weight all the time, and cut your calories if it ever goes up. That is brutally difficult to sustain for years at a time. If you follow any sort of hunger-driven diet, with healthy foods or not, you will probably go back to whatever weight your body likes over time. That's how hunger works.

That grass-roots FLOSS development only happened after the GPL does not mean it was necessary to create it, nor even caused directly by it. Giving away free software to promote consulting and support revenue can be a profit center independently of other motives. I can easily imagine an alternate 2015 where there was no Stallman, so instead consulting companies shared boring infrastructure code to split its development costs.

Good old dictionary.com says paranoia is baseless or excessive suspicion of the motives of others. There are a few examples where I think Stallman is excessively paranoid. I personally like using the web only over e-mail to avoid "survellance". Wander that deep down the rabbit hole, and the all powerful three letter agencies out to get you will also have secret exploits for Lynx. Seriously, it's all in the Snowden documents! And I totally did remember to take my medication!

However, there are way less examples that seem extreme like that today then there used to be. Re-writing your hard drive firmware with secret monitoring tools? In 2015 evidence that might be happening is reasonable news, not paranoia.

I've seen plenty of examples of companies who do not want to share code unless compelled to. There are software compliance tools for lawyers whose main purpose is checking corporate source code repos to make sure there's no GPL code. But the number of corporate contributors to all the BSD distributions says the GPL is not mandatory to develop open code. Did it help? Sure. I think open source software as a way to share overhead on boring infrastructure code was inevitable though, even if there was no "free software" (tm).

The world could have collaborated and built the modern Internet just fine on BSD licensed software, which is itself a variation of public domain. What Stallman deserves credit for is inventing the Copyleft license as a way to compel source code sharing. He's stayed relevant beyond that as source for paranoia about software being used against people, a stance that looks more prescient each year.

The problem with the PHP community is that it's filled with the kind of people who feel it's good idea to work on PHP.

And recently all over again with the EDM kids and their Javascript. My lawn, etc.

The SQL gap is not what let MySQL take off. PostgreSQL didn't get a good Windows port until version 8.0 in 2005. That delay is what let MySQL get such a major installed base ahead of PostgreSQL being usable for the typical desktop app.

VividCortex is a commercial tool similar to JetProfiler. The company is run by guys who are also MySQL experts, so I expect them to assimilate all the interesting JetProfiler features into there eventually. And the core Postgres profiling keeps getting better each release too.

Postgres stopped using OIDs for regular tables in version 8.1, many years ago. You can force the old behavior with default_with_oids, but no one does that anymore. OIDs haven't been needed for referential integrity in quite a while. Only the system tables still use them to connect tables on a typical server, which is mainly because those need to be read during bootstrapping before all the SQL features are available.

The pgAdmin developers agree, and have already started on a full rewrite in Python. The goal of having a GUI app that's cross-platform used to take tools like C++ and wxWidgets, and that's what the current pgAdmin III is written in. The new one doesn't need to be so low-level in its library use.

Oracle has features like Edition-Based Redefinition for this job. They are a lot more work for the simple job of transactional DDL.

There's an old but accurate at the time page comparing this feature across databases at Transactional DDL in PostgreSQL: A Competitive Analysis.

I live very close to the Baltimore distribution center. What they've done there is position it right in the middle of where all the major highways here intersect. You really can get to any other part of Baltimore in 25 minutes from there. I suspect they're going to limit this service to popular items in cities where the layout makes things feasible.

And I can't get 4 quarter horses to run as fast as a whole one, no matter how much duct tape I use.

That's not how Chrome works. The only answer to all Chrome issues is "get the latest Chrome version where that's fixed". The concept of fixing a problem in an older version just doesn't exist in that mindset. It's rather fundamentally at odds with how Debian manages releases due to that, which is why I'm not surprised at their lack of fucks here.