I guess I'm switching distros then. Oh well that's what happens with corporate overlords... they take the joy out of the world in the name of shareholder value.

Been doing this for a while and have a few websites handling thousands to millions of requests per day written running on PHP and Python with various levels of code quality. Here's a few takes:

1) Frameworks are garbage for maintainability and performance over the long term. Spaghetti code can beat them, simple, well written code always scales better and requires the least amount of maintenance... Lavarvel, Django, Symphony, React, Angular none of them are guaranteed to lead you to anything but constantly needing to update your code to keep your site running.
2) Python vs PHP vs Java vs Node... PHP is the hands down winner in performance. It's just not funny. A minimal droplet over at Digital Ocean running well written PHP outperforms expensive fancy pants AWS hosting for the others. (Flask ain't bad but PHP still edges it out) If you want more performance than PHP you need to be thinking Go or Rust and at that point maybe you really want a PHP API which queues jobs and Go or Rust code running the queued jobs because 90% of your PHP code won't run any faster than compiled code because the bottleneck not the PHP runtime.
3) People write bad code in any language. Some people can write good, high performance code in a language. Few write good code in many languages. Pick the tool that is best for you and forget the peer pressure.

I work as a developer and am the only person on my team to shun Copilot. In terms of experience I have more years of experience than half the team and less than the other half. Since my coworkers have adopted Copilot Github says that the number of comments I have left on their PRs has roughly doubled. Even the more experienced developers no write code that looks like it was copy pasted from stackoverflow without regard to correctness. This increased burden of review and the increased number of bug fixes I have had to create do make it feel like I am less effective as a developer and have shredded the joy I had in creating as I'm mostly in conflict with coworkers about their O(n^5) and unauthenticated exploitable code.

In the last year I have written two data driven RESTful APIs using PHP, replaced a NodeJS project with PHP, built a RESTful API using Python, and maintained a mess of Matlab/Visual Basic (soon to be replaced with Python). PHP is easy to work with provided you don't have preconceptions of what a language should be. It is both a world class language that a subset of developers love to work with and a vocal sub set of developers love to poo poo. It is in some ways a throwback to the days of C. In other ways it is a hybrid made to get the job done in a way that works for clients. It is not a language that insists on "purity" the way some (looking at you Python) do. What I love about PHP is that I can write beautiful, maintainable code though I recognize bad code can be written in PHP. Arguably, it is possible to write ugly unreadable code in any language... the NodeJS project was so bad it was easier to scrap it and do a ground up rewrite in PHP improving performance in the process. Having worked in C#, C, C++, Objective C, Python, Perl, Matlab, and Fortran; most code is an ugly, unholy mess. The language doesn't matter nearly as much as the developer writing it. Bottom line, if you want to write code that will run on basically any web server and be cost effective for clients to host, it is difficult to do better than PHP. If you want to write an SPA use Javascript. If you want to write a GUI for Linux, chose Python or Vala. If you want to do high performance or embedded computing C/C++ are probably the right choice. Those are the choices I have been happiest about in the past year of my life.

Who needs cell service when you have a good book? Most of my mobile time is spent on a Kindle Paperwhite.

Historically speaking, breaking up companies has been good for innovation and the economy. Breaking up Bell Telephone created competition in the market lowering prices for consumers and rolling out a host of products including the fax machine and really the commercial internet. Breaking up Facebook might stop it from snuffing competitors and would be good for advertisers because really advertising costs a fortune. Breaking up Google or Apple, I don't know they have their cash cows which they plow into research but then again breaking them up might be great. Spun out those research projects are 100 million dollar a year enterprises instead of side projects to the billion dollar endeavor. In fact I will be bold and say I don't think our government has the ability to effectively regulate the industry unless they first create the competition. It turns out it is much easier to regulate 100 small companies competing with one another and ratting each other out than 5 multinationals that are locked in a mutually assured destruction pentagram and can't really rat each other out. So Warren is either flat out correct or she is correct for the wrong reasons but either way FAANG have got to go.

Amazon having trouble rolling out a platform migration does not mean Oracle is a reliable platform. On the contrary, my experience is that due to the high licensing costs, many business forego implementing the replication and redundancy measures needed to make Oracle's db reliable. Amazon having trouble rolling out a platform migration only goes to show that scale makes such migrations difficult and underscore how important planning is in IT.

50 years to boot strap a civilization, because let's face it a city on Mars means bootstrapping civilization there, requires an enormous amount of energy. Energy we just don't currently have.

So you want to build a city, that takes masonry, lumber and steel. On Mars there is no wood so you would have to import all of it. Bamboo is quick growing and would be the first realistic source of lumber on Mars but even bamboo requires good soil, water and time to grow. Which means first enclosing a space, building soil or importing it and waiting. So steel, there is no steel on Mars except what for a few small parts of some small rovers which we have spent and some meteorites. Which means importing it or refining natively which means having infrastructure which would need to be imported. And masonry, actually masonry is pretty easy if you have the machines to mine the raw materials, the machinery to grind the ores and the furnace to make clinker/cure bricks etc. But without steel that means importing it. Basically the amount of raw materials increases geometrically as you shorten your time line. To the point that if you want a city in 50 years you need to think about what would the delta v energy to lift New York city into orbit be? A lot. Global yearly energy usage levels a lot.

Let's instead think about getting a toehold on Mars? Still a lot of energy. Radiation shielding for transit, energy to to shift solar orbits, energy to descend to Martian surface, raw materials to begin building infrastructure. It is still a lot of energy. Industrialized nation-state yearly energy usage levels. That's a lot of energy. Saving that much energy (hyrdocarbons to turn into rocket fuel) is something of a 7year project and remember that's a foot hold on Mars which will take a generation or more to scale to a city even with a regular supply of immigrants and raw materials. Think of landing a comet on Mars type resource challenges.

Oh and that's just the Earth side of things because remember there are not likely to be hydrocarbons on Mars. Solar is great for powering everyday living to a point but Solar as we currently know it and consume it does not allow for the infrastructural building needed on Mars. Fusion reactors? Still years away, would need exported to Mars and still we just aren't that good at using electricity to make industrial goods. Cement, glass and steel production as we know them are all heavy based on hydrocarbons; petroleum and coal which again are not on Mars.

Could we in 50yrs solve the energy problem? Perhaps. I hope so because that would allow us here on Earth to curb Global Warming. Do we have the spine to make the sacrifices needed? I doubt it.

So recently (gave on on the idea last week) I looked into starting a business offering offsite backups for NAS owners. After surveying the market I focused on a price of $5/TB/mo comparable to BlackBlaze. Turns out that price point is pretty hard to beat. The minimum scale needed to break even over the expected lifetime of the hardware at that price point for me was 50TB at 80% utilization of storage*. Now I won't go into all of the numbers and there are some economies of scale but for a company to offer $5/TB/mo to businesses (which tend to have lower support costs though higher liability) is cutting it pretty close. To make a living at it, and employ the sorts of people needed to get to that scale I figured out I'd need to scale into the 10,000TB (10PB) range. That's possible I suppose but if you are selling primarily to consumers with their high average support costs on already thin margins, well it is not a recipe for financial stability or success.

*It can be done cheaper of course but I was looking to do things right™; sufficient UPSs (30min) and redundant power supplies, drive and enclosure level fail safes, redundant internet connectivity, reasonable insurance coverage... I also thought about having offline snapshots but did not in the end include that in my cost estimations.

The are making it easy to kinda sorta mitigate a problem they created by not thinking through a design decision?. No a snooze button is not the answer. Reboot only upon user interaction would be making it easy to avoid unwanted reboots. Snooze just kicks the unwanted reboot down the road.

Dear Braxton Carter,

I'm afraid you do not understand what industry you are in. You are a public utility. As a public utility consumers do not want you to "innovate". The Public, yes with a capital P synonymous with People in "We the People...", have a vested interest in you serving as a public utility and that means not "innovating". Innovation that is simply creative ways of charging people more money for less utility is not in the Public interest. What is good for T-Mobile is not necessarily good for the Public. Net Neutrality even in it's current watered down version was a huge step forward for consumers. No, it did not benefit T-Mobile, but it was one small step forward for the Public and your industry.

Aluminum:~ redacted$ ps -ef | grep Firefox
    502 290 1 0 Wed09AM ?? 85:15.56 /Applications/Firefox.app/Contents/MacOS/firefox -psn_0_36873
    502 2036 290 0 7:54PM ?? 0:11.86 /Applications/Firefox.app/Contents/MacOS/plugin-container.app/Contents/MacOS/plugin-container /Library/Internet Plug-Ins/Flash Player.plugin -greomni /Applications/Firefox.app/Contents/Resources/omni.ja -appomni /Applications/Firefox.app/Contents/Resources/browser/omni.ja -appdir /Applications/Firefox.app/Contents/Resources/browser 290 gecko-crash-server-pipe.290 org.mozilla.machname.1962407656 plugin
    502 2747 1905 0 4:16PM ttys000 0:00.00 grep Firefox

Does not look to me like plugins are running in the main process.

It is fairly obvious that it is in Google's best interest to provide these API's and to an extent they have. Apple makes a killing in the phone world not because their phones are locked down but because they are free of bloat. Heck if I could install a better browser and music player I'd be quite happy with iOS but I can't because it is locked down. Similarly my experience with Android is that I can't remove the Verizon crap and the HTC/Samsung crap so while Android is nice in an emulator (if really depressingly dark) it does not work in the real world without root because I need to substitute stock vanilla android to get the simple easier (Google you have a lot of work to do here...) to use interface that lets me just use my phone.

Okay, you may think you only have time for 200 pages. And you may have some students putting in only the minimum effort but you really ought to have more than 200 pages. One of the best teachers I had assigned 100 pages a week for 11 grade history. I haven't read Blown to Bits yet (downloading) but it looks good. I would stay away from fiction even near reality works like Little Brother and 1984 as the primary source but they are important if only in how they have changed how we look at technology. Put them on an additional reading list and have them handy for the student that resonates with the material. Also consider having at least an excerpt of Lessig or watch one of his presentations in class. Have it ready for a substitute, he is a great speaker and I use him as an example when teaching presentation skills. You might also consider Bruce Schneier's blog as a source. Bruce has essays about this material and links to scholarly and popular works in this areas all the time.

8% may be low, just right or high depending on what else is in your "engineering budget". The 50% number is management BS. A number pulled out of someone's arse that is completely meaningless. You could just as easily spend more on engineering salary and grow the "engineering budget" to the magical 4% spending target as slash costs to the point you are out of business. Let me suggest the better approach to your manager. Look at where you are spending money, evaluate alternatives and if you see something offering a significant cost saving like say dropping expensive Microsoft Office licenses for Libre Office (especially in the engineering dept.) then make the decision to save some money.