You know the drill, comrades. ← Copied from the thread last week by @JordiGH. I think threads like these should not be automated because it feels better to converse on posts by someone than on posts by something. ;-)
So what are you going to work on this week?
Feel free to ask for help or advice or just talk about whatever it is you’re doing, work-related or not.
Studying Haskell Programming From First Principles, I’m midway through ch6 on typeclasses. Also, getting married. Not in that order of priority.
Aren’t you supposed to be offline? ;-)
Was trying to figure that out as well lol.
I’m at the chapter 7 on functionnal patterns right now, while the exercices are often a little simple, they do a great job of making sure you understand the theory!
Howdy!
As usual, I’ve kept myself busy with work on Dylan …
I’ve written documentation for my statistics library (which is still in-progress). This was fun as I finally got a chance to try embedding math equations into the documentation, which looks great, I think. Example
I’m still working on my new low level posix sockets binding. I’ve taken the opportunity to remove a lot of things and stick to only the newer APIs.
I finally fixed a long-standing (years) bug in our bindings generator so that it correctly maps
const char*to our<C-string>type without needing any boiler plate in the binding configuration file. I’ve no idea why I didn’t do this 3 years ago. There are some other small bugs in the bindings generator that I’m working on fixing. These are more paper-cut type things.I’d started updating our
libgit2bindings, but they were never all that far along in the first place. It might be best to just treat it as a new project. We currently handle library dependencies as git submodules. It would be nice if we could uselibgit2to build this into our build tools to provide better / more direct support.Now that the extension SDK for Visual Studio Code is public, I published my Dylan language plugin officially. I’ve started extending it a bit more locally and will see how that goes. I’ve also been looking at re-doing our editor syntax support using some of the infrastructure from Magic Python, so that we can more readily share our support between Visual Studio Code, Atom, Sublime Text and Textmate.
I took most of this last weekend off and spent it in the mountains with my family on a nature education event for young children. Our daughter isn’t 5 yet, but was pretty excited by going on a trek through some jungle and seeing new things.
I started work on an Emacs
dmenureplacement fordwma few weeks ago and I hope to release something this week. It’s now morphed into something that doesn’t use Emacs (it useszshinstead). Funny how these things go.As for the “why did I get into this” category, I’m going to write a new backend for Gnus to manage my paper database. I realized that they’re just “articles” and I want to mark them as read as well as make notes on them. I use Gnus and it seems like a decent fit for my needs (either that or it’s my hammer of late).
I actually wrote a
dmenureplacement myself, originally about 7 years ago. I wanted a GTK window for it instead of the minimalistdmenudesign. It was also an excuse to write an application in Haskell: gtkmenuI’d like to add fuzzy searching to it and clean up the code a bit, but it’s something that I use every day without fuss.
Cool! I’ll check that out. Thanks for the pointer.
Well, I was gonna do some other stuff, but instead I’m having my imposter syndrome triggered by this thread.
Here, let me help you out:
I’m doing nothing interesting this week, which is why I haven’t responded to the top-level comment.
edit: Huh, “troll”? I think it’s entirely reasonable to let other people know that not everyone is doing something interesting. I’m not being sarcastic or mean-spirited. Sometimes we are just boring and that’s ok.
Does sound rather passive-aggressive.
I really don’t understand. In fact, I’m thoroughly confused. I have not responded to the top-level request for what am I doing this week, because I am not doing anything. Nothing to say, so I stay silent. What are people thinking that I mean? How are people thinking I am being aggressive?
I think the assumption is that you were implying that corbyhaas shouldn’t have replied to the top-level comment either.
Ohhhhhhh, crap. That’s not what I meant at all. I was meaning to say, “it’s ok, lots of people aren’t doing interesting things, you are not alone, but you won’t hear about people not doing interesting things because they won’t say anything.” I was trying to indicate that only the loudest people would be talking in this thread, and that there are lots of silent people who just were not as exciting right now. It is perfectly ok to be merely adequate once in a while, or all the time.
Last week was pretty productive, tested out a couple of ideas and got a couple of pull requests for https://github.com/alexflint/go-arg done. Also tried out Electron a bit, got two test apps done https://github.com/walle/lights-out and https://github.com/walle/gobenchui-gui.
This week I’ll continue to experiment with Electron and try to write some more Go. Hopefully this week will be as productive!
Working slowly through the Matasano Crypto Challenges in Clojure. Yesterday (Sunday morning) I woke early and had a couple hours before my wife and son woke, in which I got the pieces together to nail set 1 challenge 6. I am really enjoying these challenges. It really is a nice way to explore features of a language you may not use so often. (I’ve never used byte arrays in Clojure before, for example.) Also it’s really driving home how good the Clojure experience in Emacs, Cider and CLJ-Refactor is. REPL-driven development is so interactive and fun! Early exploration in the REPL, adding a few simple
deftestcases once the interface is starting to emerge, and then adding somedefspecentries using test.check to find more edge cases.So far, because this is for my own amusement rather than for a paying employer, I’ve only added happy path tests. I have been thinking about extending to test failure scenarios with
test.check, but not really sure where to draw the line. If I expect a byte array, but the implementation will actually work if it’s fed a list or a vector of ints/bytes, should my function blow up on that? I don’t really feel that is in the spirit of dynamic languages.At work, doing RPM upgrades and rebooting machines with embarrassingly high uptimes. Among other things :-)
Have been fiddling with implementing Hutton’s parser combinators in F#. There are other F# PCs, but I like Hutton’s Haskell and I wanted to compare them to understand design tradeoffs. This is all to improve sample code for my CodeMash presentation on building purely functional compilers from scratch.
Wrote a Markov chain ingester and generator for the Lita chat bot (in essence it listens to everyone in public channels and builds a Markov chain for each user, you can then query it to get a random generated “sentence” based on what the bot has seen that user say): https://github.com/dirk/lita-markov (and a blog post explaining it in a bit more detail)
This week I hope to work on:
More math this week, while I figure out how to unroll summary data to get wider summary estimates for some metrics. We have 10, and 1 minute rollups, but would like to be able to go arbitrarily to X minute for display purposes. The biggest problem is percentiles, which I plan to estimate by assuming “some” distribution. Currently, the only percentile data needed is response time, and log-normal looks like it might apply easily, and is described by data that is easily reconstructed (stdev, and mean). Interested in thoughts around this, though.
At home, I’m starting to look for a project to contribute to that benefits a non-profit, such that I can claim volunteering hours my company’s charitable org. Obviously, this shouldn’t be difficult as there are literally 10s of thousands of projects out there that probably apply, but finding projects which I can make good contributions to is tricky, as it involves a bunch of different factors.
For skew data, like often happens with things like counts, I’ve really liked what adjusted boxplots can do, via the medcouple. Best part, it’s nonparametric, so no need to try to fit a distribution to it. Will that help you?
Maybe you can hand-roll your own adjusted violin plots. That would be quite interesting.
Honestly, it might make sense to revisit our visualizations, but for right now we are constrained to pretty basic line graphs.
Thanks for the suggestions, though!
Gosh, you know when you have plans and they just don’t work at all? Well last week was a bust for me on doing anything with Firestr. Maybe this week something will happen? I don’t know, there is this thanksgiving thing and all…..
P.S. Anyone here do anything with GNUnet?
Starting work on an upcoming conference talk, which means collecting data, drafting slides, falling into despair, finding the joy of a well-flowing deck, and of course, almost certainly uncovering some near show stopping bugs :)
Ansible + ec2 deployment stuff.
G'day everyone,
We sent out a 4 question survey to our 20 Peergos alpha testers last week, though not many replies yet. Finished a merkle-b-tree implementation which we need to interface with IPFS. I’m planning a fun Christmas trip back to Melbourne via Argentina.
This is a quiet, nose-to-the-grindstone-before-the-holidays type week. I will put some time into addressing the latest round of feedback on the Rust FAQ (feel free to add some more feedback if you have any), but my main concern is getting extra work hours done before Thanksgiving. My girlfriend is coming out for the holiday (long distance is the worst), and I don’t want to work once she’s here (I get Thanksgiving itself off, but not the days around it).
Midnight today is the submission deadline for ISCA. I’m co-authoring a paper with other students in my lab.
After that, it’s back to working on our computer architecture class project, which involves comparisons of various prefetching and DMA schemes.
Made improvements and relaunched my Rails Rumble game last week, this week I want to see if I can make it more fun and rewarding for players. Trying not to get mired in architecture questions as I do it, but there’s a challenging dearth of available info on how to structure this sort of thing.
Sinatra 2.0 alpha work continues, and I’m not sure we can merge the branch just yet but most of the roadblocks were cleared by the work of a collaborator so we can actually make progress toward that now.
[Comment removed by author]
For work, back on augmenting image collections for recognition tasks; this week, simulating, and removing, motion blur from images. I am in a little over my head here, but it’s still really fun. My eBay scraper did good work, although the eBay API is pretty damn crappy. I got the images that we needed, but I’m not sure that I didn’t miss a whole bunch. Still, results count, and damn if it isn’t nice to be using Python libraries again. Arrow, for instance, really makes dealing with its problem domain pleasurable.
In non-work, I have committed to living on Windows and using Google’s services (in place of my existing Apple/Fastmail/Dropbox/Smugmug melange) for a month. I tried to use Siri in the car the other day and got so angry at how bad the results and interface were that I am considering ditching Apple’s ecosystem full stop. There’s no rage like the rage of an apostate, I guess.
Life is to some degree on hold because our daughter is sick; I’d like to find something computer-ish I can use to distract myself, but I just end up falling back into staring into space and worrying.
I’m four weeks into a new job, and getting my head around moodle - a virtual learning environment (VLE). It’s interesting working in an education environment, where the IT shop is all windows except the VLE which I’m running on a LAMP stack.
Finally solved my bicycle gremlins. Ended up fitting some Jagwire gear cable outers (& inners) that allow a much tighter bend radius round the handlebars than the SRAM outers the bike came with. (This is after having replaced the handlebars with some that had a larger radius as well.) The upshot is—I can shift the rear gears smoothly every time! So now I need to go ride the thing this week (and have my watch stop yelling at me for not moving enough to boot.)
Technology-wise, I’ve recently discovered the amazement of middleman over Nanoc, so I’m currently porting my blog over from the latter to the former.
i’m spending a lot of my free time figuring out how to deploy a yesod thing to some docker image; it’s proving more annoying than i anticipated, but at least i might get an article about how to do it out, after i’m done!
I’ve been using Om to build web(site|app)s in ClojureScript for a while now, but I started trying to learn the newly-reimagined version of Om, Om.Next. To that end, I used om.next to build a birthday card/website for my sister.
I also competed at the Ontario powerlifting provincials and had kind of a disappointing meet.