Are you using Swift Playgrounds? If so, I’d love to hear about your experience & any tips and tricks you might have, workflow-wise. I’ve been wanting to start tinkering with it for a while.
Ah no, I mean development from the iPad sorry, not developing for the iPad.
Mostly trying to get https://blink.sh/ going with GitHub codespaces, might pull the trigger and setup a remote dev VM on the homelab instead with code-server over SSH (over Tailscale, natch.)
Ah, that’s fair, thanks. :) I’d be interested in hearing from people who’ve used Swift Playgrounds to create non-trivial (as in “marginally useful or interesting”) iPadOS apps.
I am considering this too, for my workstation upgrade. My plan so far is to build up a beefy NixOS desktop, and use an iPad with Blink and SSH. I have some concerns though:
I use LaTeX for letters, papers, invoices etc., and I depend on vim + pdflatex + zathura + synctex for that. I don’t think I can replicate that flow on the iPad.
I sometimes work on web stuff, for which I use a browser and its console. Most mobile browsers do not expose developer tools.
I have heard that the iPadOS multi-tasking situation is in flux.
The rest is just great though: lightweight, great display, great touchscreen experience (as compared GNOME/KDE), great form-factor.
I am looking at making webauthn (server and client side) from scratch without big libraries that do everything. Understanding CBOR is my next goal and I might compare it with ASN.1 DER in the future in writing.
I’m trying something new: inspired by this article I’ve decided to split my time working on my startup and open source projects 50/50 between development and marketing - doing a week of each in turn.
So this week is my first ever attempt at a “marketing week”. I’m going to be mainly trying to book demos with potential users and customers (and do activities to support that).
I’m preparing a presentation for tomorrow, about a paper that I recently submitted to a conference. Afterwards, not exactly sure as the experiments that weren’t done in time did not result in anything interesting. I’ll probably be shifting my attention elsewhere to a new project.
I am a bit annoyed at myself that I want to start writing a blog (mostly summarizing my research in a less formal setting) but I always get hung up on the minor technical details. Of course this probably doesn’t matter as much to others as I think it will, but I somehow am afraid that what I write will not really be future-proof in that the text will not survive a migration to another static site generator. I guess I am just mentally stuck here (and have been for a while).
One thing that helped ease my mind was the realization that HTML is “valid” Markdown, and Markdown is supported by virtually every static site generator in existence.
I’ve migrated parts of particularly stubborn content by just copy/pasting rendered HTML from a WordPress site into a Markdown file. I feel that as long as you’re sticking to plaintext formats, it’ll be as future-proof as anyone could hope for.
I’ve been tinkering with a tool to watch all those agreements you’ve agreed to (terms of service, privacy policy, etc.), for changes. My hope is to shine light on – even if I’m the only one seeing it – the types of changes that are made to all the policies/terms we agree to. I plan on doing this through character-level diffs, rather than the traditional line-level diff, which will help highlight what has changed with precision.
The impetus for this idea was receiving a bevy of those “Our terms have recently changed…” emails, without any indication as to what had changed. There are obvious applications outside of “terms and agreements”, but this is the application that currently interests me.
My initial experiment uses go diff for two reasons:
It supports character-level diffs
It support pretty HTML output for a relatively effortless way to serve results publicly via web page.
As much as I love git as a version control system, I find its interface unpleasant. So naturally, while I thought of using git for this project initially, after asking myself “can git produce character-level diffs instead of line-level?”, I didn’t even bother to find the answer – opting to learn something new rather than go with what I already know, because what I know is git and its unpleasant interface.
Point (2.) above is relevant because I’d like to keep this fun and be able to share policy changes with people by sharing links to diffs I find interesting. I suspect some terms and agreements changes would be quite embarrassing to companies with the diff light shone on them.
Running around Solutionizing. Do we have all the cables we need for this custom device? Oops, we’re missing this one, do we have the hardware in stock to build it or will we have to order one? Ok, now does the thing actually work when we turn it on? Oops, it’s trying to boot from the wrong hard drive, where’s the debug cable so we can fix that? Ok, now does it actually automatically start everything it needs? Ok, do we have a spare operator laptop just in case? Oops, it’s been stolen to be used by another program and is currently in Florida, can we get an identical one out of inventory and put the right software on it? Oops, the customer might change what they want to test, can we set up the cloud server to support two devices at once just in case? Ok the software now handles that fine, can I update the documentation with that info? Ok do we have all the emergency tools, spare parts, spare cables, spare SD cards imaged with the right OS, and so on that we might need for this demo?
My job title says “software engineer” but when someone asks me what I actually do for a living I sometimes have no idea how to describe it.
Working on launching a new backend+frontend for Berkeley Graphics website. Completely rewrote it in Django, took 3 months (part time) which also included learning Django itself.
Stack: Django + HTMX + jQuery + PostgreSQL. I’ll migrate the data to PostgreSQL 15 if it releases on time on Oct 13.
Personal: Building a package manager for AutoHotKey. (Paging @hwayne.) That sounds like a daunting task, but a minimal version of ahkpm won’t actually be that difficult because I don’t have to resolve transitive dependencies until AHK library authors start using it.
I’m enjoying learning Go in both the work and personal projects, and despite my annoyances, it might be my new favorite programming language.
Set up my work laptop for occasional personal use. I decided to trade in my personal laptop (early 2015 MB Pro) for a gift card that I plan to use as a downpayment on an Apple Watch.
I barely use a computer outside work nowadays, and my interests are sufficiently in the SFW category, so I’m comfortable using my work laptop.
That saves me spending so much time syncing setup between the two machines (something made harder by the new one I was just issued being an M1 Mac).
Aggressively job (or maybe part-time contract) hunting after a recent layoff, probably getting back to designing (and maybe finally returning to implementing) https://github.com/klardotsh/gluumy, and some volunteer work over the upcoming weekend with FIRST robotics. Should be a solid week!
Same as in the weekend plan. Probably building a bats test suite for one of two shell libraries I want to get tested/refactored/released. I figured out how to clear a few important hurdles, but progress was blunted by breaking a plate and cutting a finger pretty good.
Last week of the season sailing in Greece. Making the most of 20°C weather and sea warm enough to swim in. Chilly in the evenings now though.
Tinkering with iPad development again, mostly as I haven’t brought the laptop with me but have some itches I want to scratch dev-wise.
Are you using Swift Playgrounds? If so, I’d love to hear about your experience & any tips and tricks you might have, workflow-wise. I’ve been wanting to start tinkering with it for a while.
Ah no, I mean development from the iPad sorry, not developing for the iPad.
Mostly trying to get https://blink.sh/ going with GitHub codespaces, might pull the trigger and setup a remote dev VM on the homelab instead with code-server over SSH (over Tailscale, natch.)
Ah, that’s fair, thanks. :) I’d be interested in hearing from people who’ve used Swift Playgrounds to create non-trivial (as in “marginally useful or interesting”) iPadOS apps.
I am considering this too, for my workstation upgrade. My plan so far is to build up a beefy NixOS desktop, and use an iPad with Blink and SSH. I have some concerns though:
The rest is just great though: lightweight, great display, great touchscreen experience (as compared GNOME/KDE), great form-factor.
I am looking at making webauthn (server and client side) from scratch without big libraries that do everything. Understanding CBOR is my next goal and I might compare it with ASN.1 DER in the future in writing.
I’m trying something new: inspired by this article I’ve decided to split my time working on my startup and open source projects 50/50 between development and marketing - doing a week of each in turn.
So this week is my first ever attempt at a “marketing week”. I’m going to be mainly trying to book demos with potential users and customers (and do activities to support that).
What’s your startup?
I’m building the SaaS version of my open source project https://datasette.io - the preview is up at https://datasette.cloud
I’m preparing a presentation for tomorrow, about a paper that I recently submitted to a conference. Afterwards, not exactly sure as the experiments that weren’t done in time did not result in anything interesting. I’ll probably be shifting my attention elsewhere to a new project.
I am a bit annoyed at myself that I want to start writing a blog (mostly summarizing my research in a less formal setting) but I always get hung up on the minor technical details. Of course this probably doesn’t matter as much to others as I think it will, but I somehow am afraid that what I write will not really be future-proof in that the text will not survive a migration to another static site generator. I guess I am just mentally stuck here (and have been for a while).
One thing that helped ease my mind was the realization that HTML is “valid” Markdown, and Markdown is supported by virtually every static site generator in existence.
I’ve migrated parts of particularly stubborn content by just copy/pasting rendered HTML from a WordPress site into a Markdown file. I feel that as long as you’re sticking to plaintext formats, it’ll be as future-proof as anyone could hope for.
Great content doesn’t need catchy presentation to capture people’s interest. Case in point: https://danluu.com/
I’ve been tinkering with a tool to watch all those agreements you’ve agreed to (terms of service, privacy policy, etc.), for changes. My hope is to shine light on – even if I’m the only one seeing it – the types of changes that are made to all the policies/terms we agree to. I plan on doing this through character-level diffs, rather than the traditional line-level diff, which will help highlight what has changed with precision.
The impetus for this idea was receiving a bevy of those “Our terms have recently changed…” emails, without any indication as to what had changed. There are obvious applications outside of “terms and agreements”, but this is the application that currently interests me.
Have you considered tracking those using a git repository?
I’ve been contemplating doing something like that using git scraping: https://simonwillison.net/2020/Oct/9/git-scraping/
That’s a cool idea; thanks for the link.
My initial experiment uses go diff for two reasons:
As much as I love
gitas a version control system, I find its interface unpleasant. So naturally, while I thought of using git for this project initially, after asking myself “can git produce character-level diffs instead of line-level?”, I didn’t even bother to find the answer – opting to learn something new rather than go with what I already know, because what I know isgitand its unpleasant interface.Point (2.) above is relevant because I’d like to keep this fun and be able to share policy changes with people by sharing links to diffs I find interesting. I suspect some terms and agreements changes would be quite embarrassing to companies with the diff light shone on them.
Iirc git has character-level diffs, and it probably also has some html-output plugin.
What about an alternative VCS like mercurial?
Running around Solutionizing. Do we have all the cables we need for this custom device? Oops, we’re missing this one, do we have the hardware in stock to build it or will we have to order one? Ok, now does the thing actually work when we turn it on? Oops, it’s trying to boot from the wrong hard drive, where’s the debug cable so we can fix that? Ok, now does it actually automatically start everything it needs? Ok, do we have a spare operator laptop just in case? Oops, it’s been stolen to be used by another program and is currently in Florida, can we get an identical one out of inventory and put the right software on it? Oops, the customer might change what they want to test, can we set up the cloud server to support two devices at once just in case? Ok the software now handles that fine, can I update the documentation with that info? Ok do we have all the emergency tools, spare parts, spare cables, spare SD cards imaged with the right OS, and so on that we might need for this demo?
My job title says “software engineer” but when someone asks me what I actually do for a living I sometimes have no idea how to describe it.
Site reliability engineer? Solutions architect?
Working on launching a new backend+frontend for Berkeley Graphics website. Completely rewrote it in Django, took 3 months (part time) which also included learning Django itself.
Stack: Django + HTMX + jQuery + PostgreSQL. I’ll migrate the data to PostgreSQL 15 if it releases on time on Oct 13.
Here are some screenshots:
Account Management System: https://twitter.com/berkeleygfx/status/1578275905255841792
PDF invoices: https://twitter.com/berkeleygfx/status/1578259210579558400
Work: Building out a scale testing tool.
Personal: Building a package manager for AutoHotKey. (Paging @hwayne.) That sounds like a daunting task, but a minimal version of ahkpm won’t actually be that difficult because I don’t have to resolve transitive dependencies until AHK library authors start using it.
I’m enjoying learning Go in both the work and personal projects, and despite my annoyances, it might be my new favorite programming language.
Really cool idea! Hope it is successful
Set up my work laptop for occasional personal use. I decided to trade in my personal laptop (early 2015 MB Pro) for a gift card that I plan to use as a downpayment on an Apple Watch.
I barely use a computer outside work nowadays, and my interests are sufficiently in the SFW category, so I’m comfortable using my work laptop.
That saves me spending so much time syncing setup between the two machines (something made harder by the new one I was just issued being an M1 Mac).
Learning nodejs, unfortunately
Aggressively job (or maybe part-time contract) hunting after a recent layoff, probably getting back to designing (and maybe finally returning to implementing) https://github.com/klardotsh/gluumy, and some volunteer work over the upcoming weekend with FIRST robotics. Should be a solid week!
Same as in the weekend plan. Probably building a bats test suite for one of two shell libraries I want to get tested/refactored/released. I figured out how to clear a few important hurdles, but progress was blunted by breaking a plate and cutting a finger pretty good.