phones, queueing theory and a software project

by pjvm

12-08-2023

log entry #4 (#6 overall)

... in which I actually use this as a log and tell you about what is going on in my life at some length.

smartphones

It seems evident to me that if everybody keeps buying a new phone every three years, there are going to be some problems. Running out of certain metals is one that's on the horizon, and most people reading this will already be familiar with the growing global problem of e-waste.

So I consider it more or less a responsibility to use my devices for as long as possible. In particular: I'm currently still using my first smartphone, an iPhone 5s that I got in 2015.

Unfortunately, I'm seeing some signs that I might not be able to keep that up for much longer. By all accounts, iOS 12 has received its last update and will likely be designated end-of-life later this year, and with it the 5s. That electronic devices have an "end of life" is one of those things... if only it wasn't the norm for software to always become bigger and more demanding.

In less than 2 weeks, Signal will stop supporting my phone. I feel glad that I only introduced 3 people to it, now I only have to tell 3 people they can't reach me through it anymore. I have to ask, though: when a version of your software works on an old device, is it so hard to just keep it working? Especially for such a well-funded organisation? To be honest, it has long been clear that Signal's priorities don't align with my values, with its intrusive pop-ups that you can't turn off and that, upon clicking them away, say "We'll remind you later". Ugh.

When I can't use Whatsapp (hereafter watsep) on this phone anymore, that will really be the nail in the coffin. I dislike and distrust watsep, given facebook's ownership of it and all, but unfortunately it is used by roughly everybody I know, and not just for trivial things but coordination on important matters. For example, I have recently joined the board<1> of my choir ad interim, and of course, e-mail is used when necessary and everything else goes through watsep.

In fact, watsep is one of the only things I even use my phone for, besides alarm clocks - and Microsoft 2-factor authentication (thanks, university ICT).

So now I'm looking into getting a new phone. There are some nice things about this: I can ditch Apple, take a phone that I can easily install a free OS on and do so. The freedom of setting a custom ringtone... I'll probably have a better microphone and certainly much more storage, so I can start making music recordings a lot more. And whoops, there I am reading about DivestOS and the OpenStreetMap editor apps in F-Droid at 2 in the morning...

Still, I don't like it. On top of the sustainability aspect, there are a lot of annoyances: I'll have to sift through all the data that's on there and get anything important off of it; I'll want to delete my Apple account, which still uses my old e-mail address and by the way, do I even have the password to that? And so on.

The question remains what to do with my current phone. Maybe I can find someone to whom it will be useful, but I doubt it.

queueing theory

Something I'm more structurally busy with at the moment is learning about queueing theory, the area of mathematics dedicated to waiting in line. This is in preparation for a challenging course that I'm hoping to follow. I say hoping, because I still have about a book's worth of prerequisite knowledge that I need to come to understand if I want any chance of succeeding in this course. I have 30 days before I need to make my decision on whether to take this PhD course, or opt for a more doable master course.

I'm hoping that after a bunch more preparation, I will be able to take the challenging course with confidence. Doing so was part of my original, neat plan for my curriculum and to be honest there's also some ego involved. I'm trying to figure out how much of my motivation for this course is about proving that I can do it. The plan with the master course also has some problems, but it is much less risky. Perhaps the rational thing to do would be not to take the PhD course even if I'm confident I can do it. Nevertheless, I am going to try and see how far I come with the prerequisites.

I must say I am a bit anxious about this year in terms of workload. Next to my courses I will be doing a research assistantship with a professor, then there are my various tasks within the choir (plus practicing of course), and I hope to continue my semi-volunteer tutoring work.

On the whole, I think I've chosen the right study, but there are definitely some things suboptimal about it. The curriculum puzzle is one; another is the lack of documentation for the programme itself and, well, one specific course really. For this course, no syllabus was provided to us and I was missing some pretty critical information (the course's coordinator is a nice man; he self-descibes as "notoriously bad with reading and responding to e-mails").

So, for the benefit of next year's students, I'm planning to write a syllabus for this course and a document with a bunch of information about the study programme. Information that I and my two fellow students had to extract from the administration across many visits (the number of students in this study probably has something to do with it).

FOSS things

Recently I reported a pretty obvious bug that I was surprised had gone unnoticed for 2 whole years. I speculated I might be the first to try and use the functionality.

the relevant message on the skaware mailing list

The functionality in question? Optional client certificates! So as you might guess I'm working on a gemini thing. It is kind of supposed to be a summer project, but I'm not sure I'll get much work done on it before uni starts back up, with the whole queueing theory business.

As an aside: it looks like there will be a new release with this bug fixed relatively soon, within a couple of months at most - which is very nice. In small projects with low activity, it can happen that user-impacting code gets merged but then sits around in git for ages waiting for the next release. This can be rather frustrating for both contributors and users (I will not name the project, but...). Of course it's extra frustrating if the problem is an "immense oversight" and not having it fixed is a "dealbreaker", but really, any user-impacting fix or addition is important and deserves to be delivered to the users in a timely manner. If I ever have a project that gets used by other people, I hope I will avoid this.<2>

Speaking of things taking far too long: about half a year ago I wanted to try out the NNTP client nn, only to find out that the package for it in the Arch User Repository hadn't been compiling since january 2021. I flagged the package as out-of-date, and it remains so to this day. Recently I took some time to figure out how to write a PKGBUILD, and produced an updated one that seems to work. I expect a functioning nn package will appear in the AUR on monday.

log index

main page (with contact info)

<1> It is one of those moments where English suddenly turns back into a second language. What a strange word "board" seems right now when I hear how it sounds and see how it looks. Is "board" even the right word when it's an association like a choir? Wait, would a choir be called an "association" or something else? Surely a choir wouldn't be called a "club"? I wish English would just import the words "bestuur" and "vereniging" like it has done with all those German words...(nevermind me, I'm still disappointed that "Schadenfreude" was picked over "leedvermaak")

<2> "hope" being again the key word, and to be clear: I'm not in any way saying that people are entitled things from volunteer developers, nor am I trying to put pressure on or ridicule people (hence not naming people or projects here).