245 karma · joined February 21, 2013
Check what? A cookie (expires), a session (doesn't traverse devices) or an IP address (traverses entire networks)? Bookmark the login page, every one of those sites have one.
If I had to scale for tens of millions of users, 80% of which log in at least once per day, I would be focusing my effort on the most EFFICIENT technologies (watts per request). If there's capital, I'd look at what Facebook has done already and give it my own twist. I'd purchase the property and hardware up front and get to work building the platform, forking existing code where needed.
If we're talking about 10,000 users, then I'm going to pick a technology that gives me a lot more flexibility. In fact, I would probably just extend an off-the-shelf PHP/MySQL package with an nginx EC2 instance to save myself the effort of building from scratch, because I wouldn't have to.
I used Bootstrap 3 to take away the load of developing a native app for each platform and as of this weekend I've been working on a replication scheme which should get our command delays down to within a few seconds. The next process will hopefully be to eliminate PLCs and get Arduinos or similar hardware involved.
A similar face in this thread is dangrossman, who created the awesome Bootstrap date range picker that's plastered all over our graphs and historical reports.
Unfortunately it's still very much in beta and I have contractual obligations so I can't you a full tour but the marketing page (WIP) can be found here: http://concar.co.nz/services/rms/
Backend: nginx, php5-fpm, MariaDB 10 on Amazon EC2 (2 instances, no caching). Homemade framework & API.
Development: OS X 10.9, Sublime Text 2, Transmit, git
I know I'm old fashioned, but it works quite well even with ~300 requests per second. Currently experimenting with node.js & Meteor.
It seems like you have to live in Silicon Valley, or some Californian suburb, to even attempt to answer these types of requests. I love UX design and development, I'm working on some insanely cool stuff involving automation and industrial wireless, but the audience around here don't even know what they want, let alone wanting to spend a pretty penny on something they haven't had 10 years to "prove it works".
+1 for anyone living on the other side of the railroad tracks. Train for the hardest situations, so when the real ones pop up you can manage them with confidence.
Just don't live in Auckland.
Chromium workers die after you've closed their respective tabs. Once the page finishes loading the worker is no longer needed, another one can pick up the next request. I've never come across the line 0 error, I don't task a single PHP script with a respectably large operation, it's not designed for that - even the default script execution timeout offers a hint.
Obviously session information and state can be stored in a cache, database or file. If you need to trigger a maintenance script create an XHR or cron job. 1x1 transparent PNGs is a horrible way to do some back-end dirty work that the client need not be associated with. If you're doing this to run DB optimization, purges or bulk emailers, you're doing it wrong.
An obscure limitation? Yes, but there are bigger issues in PHP than death-by-design.
Write drafts, stock up on lots of them. Go back the next day and revise. Research your points. Feel free to take excerpts from other websites to drive those points, but don't copy entire paragraphs.
Write about stuff that interests you or enrages you, stuff you're passionate about. The rest should come naturally. If you're having trouble finding something to write about, don't force it. Find inspiration.
Keep your blog design relevant. If you want to start a tech blog, go for a design that supports that. You don't have to go all out on an amazing design, but it definitely helps presentation. Some opt for a minimal, fixed, black on white design. Remember that this is your blog and it represents you.
Typography is more important than you think.
Lastly, don't write an article about something because you were asked to, and don't publish an article you aren't sure about. Readers see really quickly the difference between writing from the heart and marketing.
That's all I can think of right now.
Not everybody has to agree, it is not productive to try and change others' beliefs in a hateful manner. The Internet is written in permanent marker, it's just going to add to the hate out there already, clogging up our mind tubes.
What a shame, for humanity and technological advancement. Aaron would have advocated for his right to post, even if he believed differently. So glad this made the front page.