sciurus

Are they still on to of everyone else in a meaningful way, though?

There was discussion yesterday [0] that pointed out that there are 29MB Debian [1] and Ubuntu [2] images.

[0] https://news.ycombinator.com/item?id=30633670

[1] https://hub.docker.com/layers/debian/library/debian/stable-s...

[2] https://ubuntu.com/blog/minimal-ubuntu-released

This is still important, e.g. the Navy teaches celestial navigation in case GPS is disrupted.

https://www.thedrive.com/the-war-zone/3865/this-is-likely-wh...

You can pay for it today via https://hubs.mozilla.com/cloud

> Is it say possible to run a production quality server that can handle hundreds of thousands of concurrent users on digital ocean lets say?

Absolutely. I used to work as an SRE on Mozilla's WebPush infrastructure. Every running Firefox in the world establishes a connection to it; that means it peaks at tens of millions of concurrent connections every day. We could easily handle hundreds of thousands of connections on a couple CPU cores and gigs of RAM.

Although most connections were usually idle, we also regularly pushed messages to every connected client (e.g. when a collection in our remote settings service was updated).

It's written in Rust using Actix.

https://github.com/mozilla-services/autopush-rs

I think two things happened that are covered in that blog post

1) The performance per partition increased

2) The way AWS created partitions changed

When I was at Mozilla, one thing I worked on was Firefox's crash reporting system. It's S3 storage backend wrote raw crash data with the key in the format `{prefix}/v2/{name_of_thing}/{entropy}/{date}/{id}`. If I remember correctly, we considered this a limitation since the entropy was so far down in the key. However, when we talked to AWS Support they told us their was no longer a need to have the entropy early on; effectively S3 would "figure it out" and partition as needed.

EDIT: https://news.ycombinator.com/item?id=30373375 is a good related comment.

Their solution was to introduce entropy into the beginning of the object names, which used to be AWS's recommendation for how to ensure objects are placed in different partitions. AWS claims this is no longer necessary, although how their new design actually handles partitioning is opaque.

"This S3 request rate performance increase removes any previous guidance to randomize object prefixes to achieve faster performance. That means you can now use logical or sequential naming patterns in S3 object naming without any performance implications."

https://aws.amazon.com/about-aws/whats-new/2018/07/amazon-s3...

It's an interesting question if the percentage of users with telemetry disabled has increased a meaningful amount over time.

Mozilla has actually done some work to measure users with telemetry disabled; see the "telemetry coverage" section of https://blog.mozilla.org/data/2018/08/20/effectively-measuri...

I'm not sure how frequently they've done this or if any of the data has been made public.

Those are all valid points, but Mozilla's own telemetry shows their userbase shrinking.

https://data.firefox.com/dashboard/user-activity

Your timeline is off, FWIW. Brendan Eich resigned in 2104 and was replaced by Chris Beard as CEO from 2014 through 2019. Mitchell Baker became CEO in 2020.

As someone who used to work for a bioinformatics lab, I'm confused by this tweet.

The storage now being limited is for Gsuite- Classroom, Meet, Gmail, Calendar, Drive, Docs, Sheets, Slides, etc.

These aren't the products you'd store a lab's datasets in. They simply aren't designed for that and won't provide acceptable APIs, performance, etc. Thus although this change may stink for university IT, I don't see an alternate world where this lab had embraced Gsuite storage and was crippled by it.

For a purely mechanical change like this, someone could run black against the same revision of Django and verify the changes they see locally match the changes in this PR.

For anyone looking for more explanation of this feature:

https://michaelheap.com/git-ignore-rev/

I haven't used it, but on the surface AWS seems to be finally targeting Heroky's simplicity directly with App Runner.

https://aws.amazon.com/apprunner/

> My argument is that Firefox should have functionality equivalent to uBlock Origin by default

It largely does.

https://support.mozilla.org/en-US/kb/enhanced-tracking-prote...

They recently announced readable standby instances.

https://aws.amazon.com/blogs/database/readable-standby-insta...

Three people sharing a personal account, with one of them frequently traveling internationally, is such an unusual usage pattern that I'd be really disappointed with a service provider if they _didn't_ flag it for extra verification.

I think extensible open-source load testing tools like https://k6.io and https://locust.io are already replacing JMeter.

Approaches vary company-to-company, but https://response.pagerduty.com/ is a good resource for understanding how it often looks.

As someone else said, your math is off. Your point is still reasonable, though.

The uptime.is website is a handy resource for these calculations. For example, http://uptime.is/99.9 says

"SLA level of 99.9 % uptime/availability results in the following periods of allowed downtime/unavailability:

    Daily: 1m 26s
    Weekly: 10m 4s
    Monthly: 43m 49s
    Quarterly: 2h 11m 29s
    Yearly: 8h 45m 56s"

This is remarkably different from how a previous employer of mine treated HN. When we were on the front page, the most frequent internal response was to quote from https://digital.library.unt.edu/ark:/67531/metadc1279277/ to encourage people to stay away.

My wife and I have the Dropbox app installed on our phones, which automatically uploads all photos to that service (Remote Copy 1).

I have the Dropbox app installed on a desktop computer which is always running and syncing (Local Copy 1).

Once a week I hook an external drive up to the desktop and take a backup via windows-built in backup utility (Local Copy 2).

I have Backblaze installed on the desktop continually backing it up (Remote Copy 2).

Hugs to all the Firefox developers and SREs who investigated this.

Bring on the front lines for such a crucial and complex piece of software and associated infrastructure is tough.

My employer-provided computer runs multiple agents. They're only used for security auditing as far as I can tell.

One of them is reporting all the processes I'm running. Certain keywords will trigger IT to reach out to investigate.

Another of them is intercepting all my web traffic, even going as far as installing its own CA and decrypting SSL. It's fun when that hiccups and I start getting SEC_ERROR_REUSED_ISSUER_AND_SERIAL errors everywhere.

This provides great incentive for me to keep personal usage off my work computer.

FWIW on multiple browsers on several different computers I've never had Google docs take anywhere close to 15 seconds to load.

> There are some engineers that are so good at solving problems that they can be dropped into any crisis situation and help fix it... Those are your PEs.

I agree that "the solver" is one flavor of the role, but I think there are a few other valid ones too. Many of the staff+ engineers I've interacted with fall in the "tech lead" or "architect" category.

https://staffeng.com/guides/staff-archetypes

You might find https://staffeng.com/guides/what-do-staff-engineers-actually... interesting.

This is cool.

I also like https://www.solarham.net/

See for instance the client request rejection probability equation at https://sre.google/sre-book/handling-overload/