oxplot
I'm very surprised when people voice this concern. These keys cost a mere $20 or so. There's no limit to how many website you can set them up with. So $40 or even $60 is all you need to invest in.
How do you do that?
Passing Passwords (2015)
Galera Cluster for MySQL: The True Multi-Master
I expected email notifications but all I managed to do was to add the saved queries on the home page. Am I missing something?
Faces of Open Source
Any provider has the right to terminate your account for variety of reasons even if you're a paid customer.
Like I've said in the past here [1], if something is so important to you, you need to treat it as such. Continuing to use a service that offers no guarantees, availability or even continuing access to your account is a sign of ignorance, when you rely on that service so much.
Now ignorance is not the fault of the user. It's primarily the failure of education. There are multiple deficiencies at play here:
1. Lack of knowledge of general population about significance of terms of service, and literacy to read and understand them.
2. Lack of regulation to enforce service providers to provide concise and less technical terms of service.
3. Lack of clear options provided by the service provider for users, so that if a paid option with guarantees is available, it's easy to sign up for and use.
In case of Google, you can, for $60 a year, have an email account under your own domain (which by the way decouples you from Google if you wish to move to another provider in the future), a 24/7 phone support and other goodies. The issue is the number of hoops you need to jump through to set that up, making it inaccessible for the average user. This is Google's fault.
1. Names can be faked. I don't see any long term and practical way for OKC to enforce real names (as FB and G+ showed in the past). So this is really a non-issue.
2. Photos are much more problematic because it's assumed that one cannot search based on a photo. A 5 second Google Image search on profile pictures from OKC however shows that a lot of people re-use photos they post on various social sites. It takes one of those places to have your real name next to your photo and you're identified! Google Image search is just the tip of the iceberg. It's getting easier and faster every day to compile database of photos and run a AI algo on it to recognize faces.
Any company who boasts protecting privacy of its users ought to find a way to protect against facial recognition.
How about your photos? Those are as PII as it gets. If you're not careful (and majority of people aren't), and upload the same photo you use for your linkedin or twitter account, a 5 second Google Image search can easily reveal your true identity.
In fact, I have, in the past, emailed OKC and others and asked them to put measures in place so when a user uploads a photo, it's checked against various sources online and if it's a match, the user would be warned that it's potentially revealing their identity. I'm not sure if they took note of that.
HTTP URL Path Parameter Syntax
To infinity and beyond, with fractals
try/except in python doesn't fit that goal: looking at a line of code in python, one cannot with any reasonable level of confidence determine what types of exception could be raised, where they will be raised or if any, unless they read the documentation of the every method call that could happen by running that line.
Rock climbing when I can be bothered going to gym. Not lately.
Breathing Retraining
As a personal example, I've searched for answers related to golang far less than I have for python because I tend to find my answers by simply reading golang docs but in case of python, docs do not help me as much so I turn to google and consequently to SO.
Currently, the only print ready HTML to PDF processor that I know is Prince [1] and to a lesser extent Firefox.