life.swp 2020-07-03

Life's been tough. To say the least. I've gotten a decent amount done in the past week or so but I just haven't felt like writing about it. So instead of a bunch of blogs I'm just writing the one. Another life.swp.

First, I've been using vimwiki for notetaking and generally anything I just write down, tasks, etc., and was interested in trying to generate a site from the Markdown files. Managed to do it using pandoc and a Makefile. Apparently Netlify has pandoc built in, so now I can just deploy via git and it stays updated. It's currently hosted at vimwiki.standingwater.io. Don't mind the color scheme, it's pretty ugly and sure to change soon. I was experimenting with Nord. Don't like the result in the least.

I recently started using Instapaper, and paying for premium ($3 a month). For the past year or so I have been collecting an absurd amount of bookmarks, almost like a digital scrapbook of events (2020 has been movin' pretty fast, as everyone knows). A ton of the bookmarks have been tweets, but there were still an absurd amount of news articles, so I was interested in a more formal way to archive this stuff, and to clean out my bookmarks. Instapaper doesn't seem to handle tweets very well, but I've ported all of my news article bookmarks over, and am just going through all of the ones I haven't quite read yet, and the "read laters" that have been building up in my feedly for nigh on two years.

It took some work trying to get an Instapaper link in my headers on standingwater.io and blog.standingwater.io. They aren't available as FontAwesome icons, so I had to use SVG's, and the result feels just a little more hacky than I'd like. These sites are so personal and will be with me so long that I have an interest in modularizing them and making them very easy to modify and extend, so I will probably be finding a way to improve on what I have. I actually have been getting a little dissatisfied with standingwater.io and am inclined to code up something myself instead of using the template it's built with.

More interesting, though, is that I want to use the API to automatically (or at least programmatically) download all of my highlighted selections and notations from Instapaper and write them to markdown files in the passages portion of my vimwiki.

More to the point, life has been pretty tough precisely for one reason: the job hunt. It is not going well. I get a response from one in twenty companies, and of the responses I've gotten, only two have led to an interview, one of which I choked on, and the other one of which I completely blew by forgetting about it for the fifteen minutes it was scheduled for. The second one, most ironically, originally said it was just for equity (part of the reason I blew it off) and now says is for $120k. God that feels so shitty.

Sometimes it feels like I'll never get a job. That I am doomed to make $12 an hour for the family company for the rest of my life. It gets hard thinking that. I'm trying to stay positive, but ho boy is it difficult.

I made a comment on Hacker News about anno.wiki and was pleasantly surprised by the feedback I got. It just re-iterated to me that the reason it failed was multi-fold, but that I can still fix it.

I have learned a ton in the past two months. It is surprising how fast and fluent a lot of tech learning feels to me now. This is part of the reason I feel like I should be able to just get a job, but I understand rationally this is not how the world works. My lack of professional experience weighs heavily against me, even if I might be more experienced than a lot of juniors. One of the interviewers with the first company I interviewed with said his best hire ever was a Pizza Hut manager, so it's clear that it is possible. It's just hard as hell to get your foot in the door. And with the pandemic, the future feels so much more bleak.

I am fully confident we'll be entering a serious recession, if not depression. It will rewrite the way we work, and the economy itself. I imagine that things will be more remote, which, on the face of it, means it should be easier to get a remote job, but I'm not confident that's the case. When a position is remote, there is so much more competition because it's not geographically limited. I just don't know.

I think that I need to do two things now.

One, formalize my application process. That is to say, I need to get a cover letter system. It seems they are very helpful. I need to start using Hacker Rank to apply for jobs. I need to start compiling a spreadsheet of all the companies I apply for, what I apply with, when, and I need to start applying at companies that don't have job postings.

Two, I am going to start redesigning anno.wiki. I'm going to go hard on it. I am going to build an entirely new front end in ReactJS and relegate the current one to https://old.anno.wiki. The following list of changes will be incorporated:

Autocomplete omni search. I want you to be able to search for annotations, users, and, most importantly, lines from the omni search bar, with incremental response. This seems essential to the functioning of this app. You need to be able to read Shakespeare in your bed with your own copy, write something in the margin, think "this would make a good annotation," go to your desk, pop open anno.wiki, type the first few words of the first line into the omnibar, and open the passage for annotation in less than ten seconds.
A landing page. I might even pay for an impact video. This is something I've been thinking about for six months, and the comments I got at HN made clear that no one knows what to do with the site.
Analytics.
A React Native mobile app. This may, in fact, be the biggest boost to utility. We will see.
As for more technical information, I want to use graphql in the front end. There is a library called Graphene Python that allows you to easily (I hope) incorporate graphql into python, and there's a specific version called Graphine-SQLAlchemy for SQLA. I am hoping this is easy to graft onto my existing structures. I have become very used to graphql in the past six months in writing Gatsby, and I like it a lot. The idea that I can build an easy data api for Flask without writing an insane amount of routes would be absolutely wonderful.
I am interested in migrating the database a bit. I have been thinking about the possibility of using two different databases, one for data storage (lines) and one for annotations, comments, and general CRUD operations. I do not know. This is very future thinking, after I get the front end running. I'm not sure how that would affect the SQLA, how it would affect the cross-queries, especially the more complex ones that allow me to query all annotations with a particular tag. But we'll see. I like the idea of having a database specific to persistent data, and keeping it separate from all CRUD operations. It just feel safer and cleaner.

This is going to be a big project. Hopefully not as big as building anno.wiki in the first place, but we'll see. I hope I can get it done in 3ish months but that seems extreme. I just don't know. I'm definitely hoping to not make another year long project of it.