October 07 2018, 19:29

By the way, about online learning. I just bought a course that costs $500. I plan to get my money back since I’m participating in a program that refunds it upon successful passing of exams. But here’s what’s interesting. The course was funded by the US Department of Defense; in essence, the entire Solution Engineering Institute’s activities are sponsored by this $411 million contract. The course is intriguing, yet it is absolutely non-interactive. Meaning, the content created with DoD funds is also being sold. This actually raises a question about how profitable it is to work in the online education market now, if you create quality content and find the right investors.

Just a side note, they have a very interesting eLearning engine in terms of information security. Downloading videos and slides is no problem, but the lecture’s speech transcript is transmitted to the browser as encrypted XML, encoding curves assembled into words by the browser, and in this XML, the commands are also encrypted. So, for those who want to extract these transcripts (though who needs them), there’s no simple solution. If you decipher the logic of obfuscated javascript, at most you can extract the coordinates of these curves, but not the actual text.

Dmitry Voloshin

October 07 2018, 19:07

Currently, I’m listening to the Software Architecture course from Carnegie Mellon University, and in the third module they discuss the architecture of WWW and CGI. It’s amusing to listen to, but the subject is different. There, the guy explains that clickable images are one of the CGI features. And I just can’t understand it – it’s the map tag, to which you give a dozen different pages and the coordinates of the areas you click on. What does this have to do with CGI? It turns out that there is an older alternative to maps – the ISMAP attribute of the IMG tag. If it is present (value does not matter), and the img is surrounded by an A tag, the coordinates of the point clicked on are automatically added. If there are two images inside the A tag, and both have ISMAP, the one you click on will have its coordinates transmitted. And as part of CGI, there was a module in the web server that read these coordinates and navigated somewhere depending on what they were. I thought that ISMAP was initially introduced, and then the MAP tag was introduced later. But both are described in the HTML 3.2 specification.

October 06 2018, 00:16

It turned out that many companies decided to rename the architectural pattern master/slave in their documentation due to its negative connotation. “Primary” and “Replica” are used in the documentation of IBM, Microsoft, Amazon Web Services, ACM, Python, Django, Drupal, CouchDB, Redis, and MediaWiki (the latter decided not to abandon “master”).

Update: I think it’s because in a true master-slave setup, only the slaves should work, while something like a zookeeper would be on the master.