pmelendez

> Incredibly US-centric comment and responses.

Just because you might have a different experience doesn't mean this is US centric. The same happens across Canada and in several big cities in South America.

>Or do the whole redeployment all the time and you'll see the "risk" of doing so was psychological or a few better tests away.

YMMV but I do have personal experience where the risks weren't psychological. Teams stepping into each other shoes and broken each other features are a real problem, solved by communication but that overhead has real costs.

>Modularize Your Monolith

It is true that probably any monolith can be break down into components, that won't prevent the full redeployment (and all the risks that it brings) though.

I think in reality no one needs Microservices, or Monolith for that matter. You would pick the poison that adjust the best to your needs.

> I don't like to stick to one project for a long time, which is visible on my resume, and recruiters don't like that.

Have you considered the consulting industry? Project rotation is high (2-3 projects/year) but you don't look jumpy in your resume (although to be honest, I have done the same before and only once a potential employer brought it up as a red flag)

It sounds like the problem is with its dependency with headless-gl, but still you have a point

I would opt for balance, there is a reason why some languages compile in debug and release mode, because of the tradeoffs.

Having code that is optimal in performance often implies a tradeoff in debuggability. If debugging helpfulness is the major design decision of a programming language, that designer is trading off performance.

I came too late to the party and just got:

"We're very sorry, but we hit the rate limit of the model... please try again shortly."

> Y’know what headphones really did “just work” though? Wired headphones, circa 2017.

And you know what's a major annoyance right? Untangling wired headphones.

Perl comes to mind, given the fact that was trained as a linguistic: https://en.wikipedia.org/wiki/Larry_Wall

Usually when clients need that level of vigilance, they provide their own laptops. I wouldn't say is a "new" normal, I think it has been common practice for a while at least in the consulting industry

It makes me sad that ~75% of the discussion is around YouTube thumbnails and not the post.

A $5k RPi NAS is indeed ridiculous, and was hoping to see some good insights in here.

I see people responding to this thread with examples of high-volume workflows that use a small number of machines for the database layer.

I think those examples are missing the point. One of the examples uses a Redis cluster, which is a sign that the DB alone can't support the workflow.

Another example is using a database for certificate issuance. Which I suspect is using one table or multiple that can be shard and it doesn't suffer from lock contention or joints of multiple tables.

I wouldn't be brave enough to count the number of cases where horizontal scaling is needed but I would say it is definitely not zero, especially for read replicas when high availability is needed

Several things in this article resonated with me negatively. But I think these two are the biggest ones:

>values stored in the files are typed. >most importantly, these formats trade human readability and writability for precision

Those two properties are advantageous only when CSV is used in cases meant to be just parsed and as a data transfer format. But in reality, CSV files are being used in many different contexts. For instance, data scientists love to leverage and chain Unix tools to create a subset of data to test models. Also, it is often used as an export format to validate outputs quickly.

I think the problem is that CSV is often the subject of abuse. In the same way that a spreadsheet is a subpar database, I don't believe we are nearly close to the time when we will retire Excel.

>And it's bi-directional. But it's not "cool" technology or "web-based" so developers won't consider it.

I might be missing a point or two here, but I don't see how SMTP can work for this case at all. You would require every API consumer to setup a SMTP server (which is another piece of infrastructure to maintain), and then somehow have a layer of authentication so the recipient can control who post messages on that server (overhead for the publisher per new customer). Then we still haven't resolved the issues on the customer side (a bad code that could pop all messages and now we might require the publisher to replay them again).

I haven't even started to think about security and network hardening challenges yet. Again, I might be missing the point but this is not a case of cool tech overuse to me.

>No, the implication is that we believe that these algorithms are non-faulty

I think that falls under OC's argument and it seems eventually solvable. Still, what you are pointing is a human bias. Humans have been taking similar unfair decisions without possibility for recourse or review since forever, so it would be interesting to have a review mechanism in general.

The incident appears to be acknowledged now: https://status.slack.com/2021-01/9ecc1bc75347b6d1

I came here to state the same.

I am not sure when we changed the terms, but back in the day, this would happily fall into machine learning. As he mentioned, if you want a good driver you would execute thousands of experiments to pick a good set of parameters

> the setup over with correctly you never have an issue, unless you can't get set it up correctly.

Eh... I don't know about that. I still have problems with GPU drivers when updating Ubuntu (and I don't think that's significantly different for other distros)

Is the web version feature-complete? What's included in the desktop version that is not available in the online demo?

As a non-native English speaker living in a English speaking country, I appreciate the phonetic guidance of a fairly uncommon word

There was a time when I used to think the same, then I found my backups were corrupted (or stopped because run out of space, etc) just in the moment when I needed them.

To my knowledge, adversarial networks are actually two different networks, one “correcting” the output of the other one. On the other hand, CNN consist of just one architectural model that internally uses convolution.

Also, the original authors are not necessarily the ones that posted it in here.

Some people would disagree with you.

https://twitter.com/id_aa_carmack/status/569688211158511616?...

I don't think there are many definitions of machine learning that claim the models to be intelligent. Most of them limit the term to models that can be built from data.

Learning is a skill that not necessarily comes with an "intelligent" label attached to it.

There is a popular trend where people will read the comments first and depending on the discussion they move to read the article

I guess it depends on the definition of done of each team.

We normally can't complete tasks until they are code reviewed, so checking out locally would be the analogous of working offline in a CVCS

That's a bit bold statement... In Spanish is Piña and in Portuguese is Abacaxi, but a good amount of laguanges does call it ananas

"My wife's Windows machine kept slowing down. I wanted to put an SSD in it but we didn't have the ability to do a reinstall on a new drive. So we put an SSD, more RAM, and Xubuntu on it. She's been smooth sailing for 6 months."

That probably has more to do with the hardware upgrade rather than the OS

That's a very unusual definition of OS. Last time I checked the operating system was any piece of software that worked as an abstraction for the underlying hardware. Your list looks more like a wish list instead of OS requirements.