Earlier today, I had one of those fat finger moments and managed to accidentally delete a whole directory containing hundreds of bookmarks. Obviously, I had to find an easy way to recover them. Thankfully, Firefox makes it pretty easy to restore bookmarks from backup. Since the browser itself keeps backup copies neatly dated, it was quite easy to solve following those directions. {link to this entry}

While checking out Slashdot today, I found a report that when the FBI seizes our messages from Big Tech, we may not even know for years. So, basically, American law enforcement (chances are the situation is not so different elsewhere) can serve Big Tech with subpoenas or search warrants to gain access to a given user's data without her knowledge, and keep it that way for years, if they want. Obviously, this means the person being investigated has no way to challenge it in a court of justice. Hardly what everyone thinks as proper in a liberal democracy. Worse yet, Facebook has acknowldged that in the last six months of 2020, it received a total of 61,262 such requests. We don't have any data for Google or Apple. Mind you, not the any of this should come as a big surprise. However, it may be about time for US legislation to catch up with the reality of new technology in a manner a bit more respectful of personal freedom. The problem, of course, is that when US politicians cannot even agree on basic policies to face the pandemic, ther is little hope they will be able to handle more complicated issues. {link to this entry}

If I remember correctly, I recently wrote a few lines somewhere about the main reason why I still run Firefox. Aside from the fact that I still like it, the perspective of a world where Chrome and Chrome-derived browsers are the only game in town saddens me. As it happens, today, I ran into an old article on how Google was locking down Chrome to restis the rist of Chromium-based browsers. To be fair, I don't totally agree with the author on this particular count. Yet, it seems clear to me that Google does control Chrome and Chromium. There is no real "community" behind Chromium, even if it's open source. Steven Vaughan-Nichols writes about the same topic here. Taking all this into account and checking out the statistics on the usage of web browsers clearly shows the extent of the problem. Chrome is the undisputed leader everywhere in the world, and it currently has a lead that appears to be insurmountable. It all comes down to one thing: browsers are such a central piece of our technology-centered world that we should not trust it to one single vendor. {link to this entry}

Finally, an easy to understand explanation of what blockchain is. Yes, the article has plenty of semi-humorous parts that feel like a wasteful interruption of the train of thought. However, it's still a good read, I think. {link to this entry}

I was recently testing something with Docker, and ran into issues when trying to pull a Docker image from their repo:

sys:~ # docker pull nginx
Using default tag: latest
Error response from daemon: Get https://registry-1.docker.io/v2/: net/http: request canceled while waiting for connection (Client.Timeout exceeded while awaiting headers)

I had the suspicion the problem was caused by the presence of a proxy in my network. So, the next question was how to configure Docker to use a proxy. I found a few different solutions when searching the Web but, at least on my version of Docker (20.10.6), the correct solution is documented on this forum exchange. Basically, follow these steps:

# mkdir -p /etc/systemd/system/docker.service.d

# File: /etc/systemd/docker.service.d/http-proxy.conf
[Service]
Environment="HTTP_PROXY=208.123.76.34:8080"
Environment="NO_PROXY=localhost,127.0.0.1"

# systemctl daemon-reload

# systemctl restart docker

That did the trick. {link to this entry}

A good friend of mine shared a graph displaying the vertigo-inducing list of Linux distribtions and their heritage together with a comment on how it seems as if the proliferation and diversity of Linux is almost organic. Here is my answer:

Allow me to add a couple more thoughts apropos the graph.

First of all, it does appear as if, once you "set something free" (e.g., in a free market type of scenario or, as in our case, in an open source manner; but not in a closed and controlled environment, for obvious reasons), proliferation and diversity will ensue.

But, second, also notice how the proliferation and diversity stays on the more superficial level. There is a very different trend going on at the deeper (i.e., "infrastructure") level. See next point for a clarification on this.

Third, notice how Debian and Red Hat account for most of the derivative distributions. It's almost as if, in a free and open environment, things naturally tend to concentrate around very few solutions, at least at the infrastructure level. I get the feeling this works not only in this field, but also in other fields.

Finally, even though the concentration mentioned in the third point above does appear to occur, there is still more diversity (even at the infrastructure level) than in the case of closed systems.

{link to this entry}

It looks as if ProtonMail was recently asked by the Swiss authorities (on behalf of Europol) to provide information about an IP address accessing one of its email accounts, and all hell broke lose in the web forums. Here and there, I've seen posts from enraged users claiming that they would be ending their ProtonMail accounts, and switching to their own personal mail servers. There are, it seems to me, a few misunderstandings. First of all, while ProtonMail may host people's messages in an encrypted manner, some of the metadata (e.g., the IP addresses of people accessing their servers, as well as the email headers) are not (cannot be) encrypted. That's just the way the email protocols work. It seems a bit silly to blame ProtonMail for that. Likewise, ProtonMail is a legal company involved in perfectly legal activities that never intended otherwise. In that sense, I'm not sure I fully understand those users who claim they were somehow fooled or misled. But, finally, I'm not sure those enraged users who yell that they will just use their own personal mail servers are fully aware of the fact that most web host companies out there (sure, not every company) would behave the same way as ProtonMail if they were to be asked by the authorities to hand in a report with the IP address, or even the identity of their customer. In conclusion, while some of the points being raised are valid and these are definitely issues we all should be aware of, one gets the feeling that this whole crisis is a bit of a storm in a teacup. {link to this entry}

Here is a highly recommended piece: The Descent to C, by Simon Tatham. It provides a great introduction to the C programming language for anyone used to high-level languages, such as Java or Python. Along the way, we learn both the power and the potential pitfalls of C. It's a great piece. {link to this entry}

A few days ago, I wrote a short piece on Firefox Containers and Linux vs. UNIX. The latter topic was in reference to an article by Ruben Schade on the topic. Well, today, while perusing Lobsters, I noticed an article written by Chris Siebenmann on how large UNIX programs were historically not all that portable between UNIXes that is worth a read. Said that, I'm not sure Siebenmann fully understands Schade's point, I think. Yes, since Linux won the Linux vs. UNIX race and it runs everywhere, applications may be more portable now. But portability was never Schade's concern. He was more concerned, I think, with the fact that Linux appears to be moving further and further away from the old UNIX philosophy. That's all. {link to this entry}