@matigo Speaking of lame development jokes, I have a pretty lame alias in my .gitconfig now: https://social.10centuries.org/#208480
// @jws
@matigo Yeah, but there should be some sort of a mention that you need to do this certain thing in this certain way so that you don't end up having problems. That is worth writing some documentation. ?
// @jws
@jws Better documentation would have helped. I think I need to write that thing down somewhere so that in the future we don’t end up in this situation again.
Yeah, I don’t like blaming people for making mistakes either, but there are some people in this project who like to play the blame game. Luckily they are being phased out from the project.
Okay, time for some form of hibernation. Tomorrow will be a long day, I am all alone with the kids while my wife is on a cruise.
In other words, the original problem was a programming mistake. I could blame the junior developer for introducing that bug in the small refactoring she made, but because I did review that change the blame is all on me (as it should). Even the best of us (as if I would be even be in that group) make mistakes every now and then.
Sigh… The problem with those AsyncTasks was caused by a bug elsewhere™. I had added a lock mechanism for the database access (multithreading, yay!) and the lock was not released in one crucial network call that updates the data oftentimes. Now it is fixed and all those AsyncTasks work just fine.
But, as I was debugging this I decided I should refactor a couple of the bigger tasks to be based on IntentService instead. That then prompted me to refactor one continuously running service to be based on the plain Service instead. There can be only one IntentService running at any given time. Once the previous one completes, the next one will be started. And if that continuously run service would be based on IntentService, that would effectively block all other IntentService work, making all that other work pointless.
@jws Well, I tend to disagree. While RxJava is cool, I find the simplicity of AsyncTasks just great for simple tasks. For more complex things, of course, there are better alternatives. The thread jumping is pretty straightforward once you understand in which thread which methods are executed (onPreExecute and onPostExecute on the main thread, doInBackground in, well, a background thread). You just need to pay attention to where you update the UI.
// @matigo
@tewha Pnut seem to have a bunch Japanese and German active users, I have observed. But then again, I'm not there that often (maybe once a month), and I have barely any time to pop in here either.
// @larand
@larand Same here. When I started using Twatter more earlier that year I had high hopes. But that summer, when they announced the first big change and people got angry, I jumped to the ADN bandwagon and never looked back. I still have an account there, mostly for claiming my user name, but I don't post there on purpose1.
Except occasionally when we need to post some work related stuff to social media to show that we actually do some work for a thing we got for free from work.