140 karma · joined May 27, 2020
Earthly lets you abstract anything you do inside a container into an Earthfile, that runs locally on a dev's machine, but also in any CI, making CI scripts more portable. However, testing other CI-specific things before committing, such as GitLab rules dictating when jobs actually run, remain unsolved. But they also remain unsolved with this glci solution, right?
In the end, my practical recommendation is to just reduce the number of services - not necessarily get down to 0. You'd be at a loss. There is no good GMaps alternative (not in Germany anyways, the level of detail of the maps data and/or search results of any competitor is poor in comparison, including HERE WeGo or OSM). Same for web search / google.com. I have experienced many times that DDG/Startpage/Bing would not find what I was looking for, then I switched to Google, and found something relevant right on page 1.
Other services, especially those with standardized exchange-import/export formats are easier to migrate, though. For instance, GMail, including calendar and contacts sync, or using another provider for your files (GDrive).
What I did wrong (and sometimes still do) is that you cannot just add new things without removing other things. It helps to be fully aware of what kinds of things you'll have to remove from your life. What is the habit you're replacing(!) with a new habit? Food for thought ... ;).
I use Anki (and the classic version of "SuperMemo") a lot, too. However, I mostly use it for learning languages (vocabulary), or for more complex material in case a test/exam is coming up (to avoid "cramming").
However, I see two problems with such SRS systems, now that I've been using them for over 10 years:
1) I can often not recall the knowledge in real life, simply because I study _both_ the question and answer side of the card - in other words, if I'm asked the question in a rather different way, I might not recall that I know that fact at all. The most extreme example of this was that certain flashcards broke after a while, e.g. showing a gray (instead of a white) background in the question card, for some reason. Whenever that gray card comes up I know the answer immediately, without even having fully read the question (which I forget over time). So in the end, what my brain remembers is: <question card with a medium-length sentence and gray background> comes up -> answer = "propensity". (I'm learning English as a foreign language).
2) A lot of knowledge in books that I read (let's take the "Pragmatic programmer" as an example) is not suited for flash card system learning. It's knowledge of the form "if this happens, do that", e.g. "if you are about to write a new method in your source code, think of how you would test that method first, to improve the method's quality". Sure, I could create a flash card that has the question: "what should you do when you write a new method?", but it's pointless. These kinds of things require deliberate practice for the specific situations, when you are in the IDE, not in the SRS software. It's methodological knowledge where, in the end, it doesn't matter if you could answer the question when someone asked you about it on the street, but it matters that you remember when you're in the actual situation.
There's ample evidence that such kinds of speeds are not generally possible with every kind of book. See e.g. https://www.scotthyoung.com/blog/2015/01/19/speed-reading-re...
I did take speed reading courses myself. Sure, it helps you speed-read, skimming over sections that seem unfamiliar, and detect when you need to slow down, because the subject matter becomes too unfamiliar/new. I've also had the impression that retention improves a bit after the couple of training days - at least that's what the workshop instructor makes you believe (it's their job to sell success and good feelings so that you recommend the course to your peers).
In the end, I highly doubt that anyone who claims to have such high speeds is sincere. The simple fact is that marking text with a highlighter, making notes (with a pen) and thinking about the content (do you?) takes time. If you accounted for that and took the average, you'd get lower numbers. And that is fine.
You don't need an account. Just open a private browsing window - at least this works for me.