wtracy

I'm surprised that Overstock is simultaneously still running at a loss and is seeing rising share values.

Every time I've tried to shop there, I've found a terrible selection, middling prices, and a mediocre UI. (To be fair, every e-commerce giant has a mediocre UI.) I see them casually mentioned in lists of "Amazon alternatives", but I've never once had someone recommend Overstock for a specific purchase. I don't think I've even talked to anyone who actually buys there.

Are they growing gangbusters in some niche I'm not aware of?

The possibility that a lot of that investment might be coming from pension funds or from government-insured banks is concerning.

Folks also get so worked up about highly efficient new renewable power sources that they forget about transit and storage.

I don't know about the situation today, but for years the massive wind farms on Tehachapi weren't even connected to the grid. Even after PG&E bought the electricity to meet their renewables quota, it was cheaper to waste it than to build the transmission infrastructure.

Renewables have the inverse of the problem you described: They are too distributed, and it's hard to get enough to power a community without running transmission lines over hundreds of miles.

Yes, the centralization of infrastructure required for efficient nuclear energy production is a problem. But don't forget that renewables have equivalent problems.

I think there's a world of difference between saying that targeted advertising as it is done now is unacceptable, and saying that targeted advertising is fundamentally unacceptable.

The first one I'm actually sympathetic to. The second I vehemently disagree with.

> Never allow dynamic ad content! Ads should be individually approved by publishers and not subject to change after approval.

Partially agree. Publishers should definitely be able to blacklist advertisers. However, networks like Google get thousands of ads that run for only a handful of impressions. Asking publishers to manually approve every single ad that might run is onerous.

I would like to see approval required for anything executable: JavaScript, Flash, etc. Waking up to find that your website has become an attack vector through malicious banner ads is not okay.

> Ban user tracking and targeting.

Mostly disagree.

I click on a meme site. There's not a lot of relevant ads for that. I could get served ads for lolcats-themed merchandise that I don't want, or low-value broad-targeted ads that probably still aren't relevant to me. Nobody wins.

Or, based on my search history, I could get ads for computer parts that I'm actually interested in. I get ads that are relevant, the publisher gets more revenue, and the advertiser has a real chance at a sale. Everyone wins.

> It's a red queen's race and no one benefits.

That couldn't be less true.

Now, systems that try to track individual people (rather than blocks with similar demographic/behavioral traits) are a serious problem. We do need stricter controls to prevent this.

> Pay for ad time slots, not user impressions.

Here we go. You want to throw away the very thing that democratized digital advertising. You want a system where either you pay five figures to reserve a time block, or you don't get to play.

Hard disagree.

> Use anti-trust to prevent link aggregators from being advertisers.

I'm not sure how anti-trust ties into this. Are you alleging that, say, Fark has a monopoly status, and is abusing it to gain an unfair advantage in an adjacent market?

Now, I think you're on to a very real problem here--advertising dollars are being diverted away from the organizations that produce content--but severing the only real revenue source for most aggregators seems like an inappropriate response.

Taste might not be a concern, but toxicity would be even more if a concern if the crops bioconcentrate the chemical.

That, and feedback that they genuinely enjoyed things that the rest of us hate.

Microsoft heard from hundreds of users who were heartbroken when Clippy was removed from Microsoft Word.

I wouldn't be surprised if that's intentionally done to force users to leave enough space around the console for airflow.

If so, one or two companies could handle fix this problem by dedicating an employee to working on Postgres.

Unfortunately, most of the likely candidates with money to throw around have their own pet database systems, and don't have in interest in Postgres.

Maybe one of the mid-sized cloud providers like Digital Ocean would be interested?

> but it comes from a no-reply address.

I consider that an antipattern in and of itself.

Unless you have a truly abysmal average CLV, replies should be directed to a system that first filters out (and acts on) bounce messages, simple unsubscribe requests, and obvious spam, then dumps everything else into a ticketing system.

Exposing an avenue of communication (which is what email fundamentally is) then trying to pretend that it doesn't exist is just obnoxious.

I personally strongly agree, but I want to point out that HN is not a remotely representative sample of email users.

Study marketing a bit, and you'll find loads of things that email power users loathe that still get an overall very positive response from the general public.

Know your userbase. Cater to them, not some random people on the internet with strong opinions.

It's primarily used in the context of reaching out to users who just signed up for a service.

After they're signed up/subscribed, you send an "onboarding" email with general information on how to get started using the service, and encouraging them to do things that affect whatever KPIs your boss cares about.

They can be genuinely helpful for tools that have a steep learning curve, either by serving as an entry point into the documentation or by providing contact information for support.

Otherwise, at best they serve as a reminder in case you get distracted before you finish setting up your account.

If settle for just being able to disable animated emoji in the subject line.

(I was shocked the first time I got spam that started playing animations before I opened it.)

Even without electricity, there's plenty of horror stories about sex toys--particularly dildos--made with unsafe materials. Some materials can cause allergic reactions, others are carcinogenic.

In this particular case, I think everything compute-intensive happens in the GPU anyway, so there's not much to lose by using a high level language on the CPU side.

Here's an idea: Make a pill that cures any viral disease! Don't bother me with questions about how to make it work. I'm an idea guy, not a chemist. Are you interested in the idea or not?

The reason people don't buy ideas is because ideas are worthless without execution. An idea with execution is already a business.

> Bipolar patients tend to struggle with unlimited free time, unstructured days, and no accountability to their work peers.

I was thinking more broadly in terms of people with disabilities when I made the remark about social safety nets.

In the specific case of bipolar disorder, I have no reason to disagree with you.

I saw this behavior under Ubuntu.

Poking around the Firefox wiki, it looks like it uses it's own native backends on basically every platform except Windows.

I have no idea whether any of this is relevant to the original Firefox support issue.

If you're curious:

I don't have the machine in front of me, but it's an eee PC with I believe an Intel n550 CPU and an Intel GMA 3150 graphics chipset. I saw this behavior under a recent version of Ubuntu, probably 19.10.

(The "fast" framerate was something like one frame very 1-2 seconds, which was acceptable for running an image filter. The slow framerate was one frame every 30-40 seconds, during which Chrome locked up completely.)

Whoops, you're right, I got my acronyms mixed up.

Funny enough, I also used to know a (different) person who had borderline personality disorder. I can confirm firsthand that they are very different!

Chrome seems to have a more full-featured shim layer for handling older graphics drivers.

I have a netbook that Firefox won't even try to launch WebGL on. Chrome seems to run WebGL just fine, though shaders with more than about twelve instructions fall off a performance cliff.

(Add one multiplication operation over the limit, and performance decreases twentyfold. Keep the new line, but delete any other single line of code in the shader, and performance goes back to normal.)

Meanwhile, I know of a practicing therapist in my community who stops just short of encouraging patients to experiment with hallucinogenics. (She's also fairly vocally anti-vaccine.)

The range of attitudes within the profession just astounds me.

I knew someone with BPD.

She might have been the hardest-working person I've known. She also couldn't hold down a job to save her life.

She's the reason I can't ever see eye to eye with anyone who wants to cut our social safety nets or who sees welfare recipients as somehow subhuman.

The article is indeed garbage.

It smells like there's actually some interesting research being done here using graphene to extract energy from the ambient environment (possibly harvesting the work done as the graphene expands during hearing?) but the press release was written by someone who does not understand the subject at all.

> the author seems to be out for blood

Sometimes a really over-the-top reaction is the only way to get the attention of a large organization.

Exactly what regulations are preventing new broadband providers from entering the market?

Over half. 52%, actually.

https://www.pewresearch.org/fact-tank/2020/09/04/a-majority-...

More to the point: Getting people killed could make the Russian government stop turning a blind eye to cybercrime.

The researcher apparently joined Snap after the paper was published.

A paper like this would look pretty good on your resume if you're applying to a social media company.

> Don't people rely on Nvidia for deep learning workflows? I thought that stuff ran on Linux?

It all comes down to there always being two ways to do things when interacting with a GPU under Linux: The FOSS/vendor-neutral way, and the Nvidia way.

The machine learning crowd has largely gone the Nvidia way. Good luck getting your CUDA codebase working on any other GPU.

The desktop Linux crowd has largely gone the FOSS route. They have software that works well with AMD, Intel, VIA, and other manufacturers. Nvidia is the only one that wants vendor-specific code.