dmlittle

Since the element you're trying to apply the full-bleed to lives in a container that has a max-width, wouldn't you need to play with negative margins or absolute positioning to accomplish this with only margins?

This was one of Apple's first iOS games back in 2008[1]

[1] https://www.macworld.com/article/1134588/texasholdem.html

I'm sure there are more valid reasons but some that come to mind are the following:

From a security point of view scoping your cookies to a sub-domain means that other services in other subdomains might not have access to them (depending on how you set them). If you store the cookies at abc.com then any subdomain will have access to the cookies stored in abc.com and you might not want that. As an aside, cookie security is one the reason why companies usually have a separate domain for user-submitted assets/media.

You have separate services for your e-commerce platform and your main marketing website and you host them in separate domains. This is why sometimes you see abc.com and dashboard.abc.com or app.abc.com. This is not a technical limitation as you could do have path-based routing to each individual service but if you don't already have that functionality hosting each service in each domain is easier/simpler.

(A combination of the 2 above) Your marketing website might be hosted by a third-party website builder and since you don't fully control what they serve you want to make sure that cookies by your platform aren't accessible to the third-party vendor. If you're proxying the requests you can strip the cookies but that's a pain that might not be worth introducing if you're OK with having separate domains.

You like having separate subdomains because it makes it seem more "professional" ¯\_(ツ)_/¯

You don't want the SEO of one service to bundled with the other.

Not necessarily RSS feeds I'm subscribed to for personal reading but I created a Slack channel at work that is subscribed to different status pages RSS/Atom feeds. Whenever there's an outage Slack will pick it up and post it in the channel. It's a pretty convenient way of knowing when the services you use (GitHub, NPM, CircleCI, etc.) are having issues.

Instead of saying 100ms, you can also say "the total request time increased by 733% with the introduction of TLS (from 15ms to 110ms), something's probably wrong and we should take a further look at it".

I don't like playing these kind of games but sometimes you have to do it.

From my experience if you have a bucket with billions of objects it will still take a few days after the objects are deleted by the lifecycle rules to be able to delete the bucket. The objects are non-recoverable but I believe due to the eventual consistency nature of S3 you can't delete the bucket until the objects are all fully deleted.

Not OP but we also had to move a relatively large S3 bucket (significantly larger than 25TB) and unfortunately AWS doesn't have a way to change the ownership of the S3 bucket. My guess is it has to do with how the underlying system stores the objects.

We ended up writing a Go program to copy the objects from one bucket to another to help with the parallelization of migration. This was also before AWS had announced S3 Batch operations so I'm not sure how much better it would be to use that today. The deletion of the bucket also took us over a week. Even though we had deleted all objects in the bucket due to the eventual consistency nature of S3 we weren't allowed to delete the bucket until all objects were fully removed from S3. All we could get from AWS support was to wait a few more days and reach back if we couldn't delete it then.

Edit: depending on your object naming scheme you might also run into the S3 prefix rate limits.

S3 buckets and RDS instances probably also outlive your cluster if you're treating them as cattle. I want to make sure I don't delete my RDS instance while I'm doing a blue/green cluster upgrade. I can see defining SQS next to your application being handy (they're somewhat stateful while there are messages in the queue but I'm sure you can do something to ensure they're drained before destroying them).

A one-time[1] $99 fee for all of their icons is not greedy. If you had to do that work yourself you would spend a lot more just in your time alone.

[1] The Standard Pro License is perpetual, meaning that you'll still have a license if you don't renew your paid plan. You'll retain access to the last version available when your plan was active.

We've worked with Cure53 and NCC Group and both of them were great to work with.

They might be keying off SSL certs created in the past few days/hours. For example, this cert[1] for www.krsnalite.com was created a few minutes ago.

[1] https://crt.sh/?id=3120549998

Nice! I hadn't heard about R-Trees before. I guess you can potentially make do with a quadtree[1] but it seems R-Trees are better suited for geodetic data.

[1] https://en.wikipedia.org/wiki/Quadtree

I figured some universities would start doing things like this. You can even make the class special in that it's pass/fail, doesn't affect your GPA, doesn't appear in your transcript, it's a special class that is free and doesn't increase your tuition, etc. Glad to know it's no longer going to be required to pull moves like this.

The best tech stack is highly dependent on who's asking the question. Use the one that you're most familiar/comfortable with. Don't feel compelled to use the "latest flashy trend" just because people are talking about it. For example, using HTML with vanilla JS is still a perfectly valid approach over React/Vue/Angular/etc..

I would probably say MySQL and Postgres is apples to apples just different types of apples (granny vs delicious). Their differences are not as big as a regular relation DB vs. a columnar DB.

Not sure about all of them but different databases have different use cases. It's not necessarily Postgres vs. MySQL where they are competing databases for the same use case. Snowflake, for example, is a data warehouse that (I believe) stores data in a columnar format. This makes it much better at performing aggregate queries (how many shows were watched on netflix today?) but a lot slower for specific queries (where in this episode is gavinray currently at?).

Not sure if you've heard of Sheet2Site[1] which I believe is a direct competitor.

[1] https://sheet2site.com/

Some websites will restrict search or other behavior if your VPN is in a major public cloud, such as AWS, as it's an indicator that you might be trying to scrape their site. For example, you can't watch Netflix or purchase from Nike from an AWS IP. I believe in Nike's case they are trying to prevent people from automating the buying of limited edition sneakers.

My understanding (and personal opinion too) is that hybrid teams (half remote, half onsite) don't tend to work well. Most likely, the remote portion of the team will be left out of a lot small impromptu discussion that happen within the office. A solution is to force all conversation and discussions to happen via Slack/email but that makes a lot of small discussion not happen as the activation energy might be too high. Ultimately if you have a small portion of your team geographically removed from the rest of the team they might end up being left out a bit.

For example, let's say you have a team of 5 people, 4 people live in city A and 1 person lives in city B. It might be the case that the 4 people in city A meet up every week in person. They'll eventually form a closer working relationship with each other potentially leaving the person in city B out. Hashicorp's decision might be to try to prevent these kind of situations (for what it's worth I've heard about this Hashicorp rule by talking to people but I can't find anything online about it).

Zapier, GitLab and Hashicorp are a few that come to mind. I believe Hashicorp even has a rule were for a given team you can't have 3 or more people in the same city.

Looking at Vitess' roadmap[1] it's in their medium term roadmap but I'm not aware of any other project that currently support Postgres.

[1] https://vitess.io/docs/resources/roadmap/

They probably use MySQL for Basecamp and it's worked for them so they're also using it for Hey (same team developing both). As for Vitess it makes sense since it's made to work nice with Kubernetes (which they are using). I don't know of a Postgres equivalent.

Not only is it mature, but it's his framework. Sure he's not the only maintainer but he's the original author. It makes sense to use your own tools; specially if they're successful.

The benefit is being able to keep your existing maven/npm/pip workflows as well as use the same workflow for both internal and public dependencies.

Does anyone know what happens to the 2nd stage engine? Does it stay in orbit or does it eventually fall back to earth/go into space?

What's the purpose of the methane purge?

It's not a gremlin and at the end of the day it's probably a misconfiguration or bug somewhere but it doesn't help anyone to point fingers and blame. A system probably went through a design phase, why didn't the reviewers of the design didn't catch this? The implementation was probably reviewed by peers, why didn't they catch this? It might have gone through QA, why didn't they catch this? And the list can go on and on.

At the end of the day a mistake was made and it should be fixed but being enraged doesn't help anyone. Figure out what went wrong, how to make sure it doesn't happen again and do it.

Mistakes happen, highly available systems fail, being enraged and demanding heads isn't going to make anyone happy but just scared. A full investigation is probably going to happen and steps will be taken to make sure this doesn't happen again but it's hard to blame it on a particular individual. Ideally you'd have steps such that one person alone can't screw up and it also requires other people to let that happen.

I know the AWS status page is known for delayed updates but AWS itself is up. Amazon.com (retail) is what is down.