@d4ng3r0u5 look at the code. It has no ID. It has no href. It has no class. It is useless.

@cursee it's a deployment webhook in our production environment. It should *always* be master.

@Matthewb it's my boss's commit, lol

There are two hard problems in computing: cache invalidation, naming things, and off-by-one errors.

What godawful encoding is this?!

@bittersweet That is the best explanation of Dutch Santa I've seen.

> post-meritocracy

Implying there was ever a meritocracy...

Even double-insulated devices can have ground faults, it's just very unusual. I had the same problem with a laptop 20 years ago.

After many years as a electronics hobbyist, I now know that the solutions to this problem are: (1) get a new charger, or (2) add an isolation transformer between the wall socket and your charger.

Pro tip: A new device charger is cheaper and lighter than a transformer.

Anyone who uses Octopus Deploy should be seriously considering another continuous delivery product right now. With a Brisbane-based development team, they're a high-value target for both state actors and third parties who would exploit any mandated backdoors.

Just think of all the half-assed firmware which vendors will ship on their "Australia versions"... and the testing effort that will go into them!

So, a coding co-op? I'd be down for sure.

I've been seeing a LOT of casual sexism recently as well. It's disappointing, but I guess it's indicative of the industry at large. Still sucks though.

@AlmondSauce you should totally name names. Bad security deserves public shaming.

Firefox works great without signing in. I hear Pocket and Sync are nice, but I just won't.

@Stebner55 it's what the brand "Thermos" is derived from!

@terimaaki I'm always curious about what people don't like about .NET. The only thing that really bugs me about it is how verbose MS is with their namespacing. Other than that, IMO, C# is a rather friendly OO language, which is packed full of features like async, and tuple assignment/deconstruction.

@FrodoSwaggins fortunately, this one is easy to test, requiring only momentary access to a Windows machine

@FrodoSwaggins It probably shouldn't match, but it does. The screenshot doesn't lie.

@irene @rads23594: FYI, this is called "proxy bidding"

AUX, COM[0-9], LPT[0-9] and others are "globals" in DOS (and therefore Windows). You can pipe to them from anywhere, without specifying a path like /dev/ttyUSB 0.

Don't forget, if all else fails, RTFM: https://wiki.archlinux.org/

@irene Oh, could you imagine how much slower RAM would be if your MMU had to bounds-check on n arbitrary values, instead of a handful of bitmasks?

Yuck.

The competing JEDEC convention of "1MB = 1048576 bytes" doesn't help at all.

I guess RAM manufacturers figured underselling was not worth the engineering pain, compared to storage manufactures.

Let's not forget that this whole thing is the fault of storage manufacturers' marketing departments!

inb4 the punch line is "btw i use arch"

Gods, I hate the clickable touchpad on my new laptop. It needs a light touch to move the cursor but a heavy push to click. Having to do both with the same finger slows me down by at least 50%. And they never seem to work _quite_ right when you leave your thumb on the bottom and pretend it's not clickable.

Never mind the fact that holding my finger still while clicking it is nigh-impossible because of the difference in force.

Nobody uses IIS anymore. Kestrel is the new hotness. And ASP.NET Core 3 is a bunch faster because of allocation-free memory addressing...

Ah yes, the elusive "everything".

@theKarlisK you know about the DRI_PRIME=1 environment variable to force an application to use the dGPU, right?

Gets around a lot of fuckery on my [Raven Ridge + RX560] hybrid graphics.

y tho

Activision Blizzard (ATVI) is an awful company.