superyesh

I just meant any organization can/has an informal structure and has to mature into a remote work structure which is relatively new.

The points you highlight are very valid, but you cannot just say it had nothing to do with remote work.

Long after Wall Street ordered its bankers back to the office, the California-based lender’s chief executive, Greg Becker, at times worked from Hawaii, president Mike Descheneaux decamped to Florida, chief risk officer Laura Izurieta was based in a suburb of Washington and general counsel Mike Zuckert worked mostly from New York, according to several people close to the bank.

I have noticed quite often that a lot of things to fall through the cracks irrespective of how many meetings you have on zoom. A lot of catch up happens offline in informal conversations. This might be mitigated on an IC level mostly given that a good team might have a well groomed backlog being fed to them. But at a leadership/design level things are still more informal and might take a few years to get to an efficient level. So dismissing that remote work could have led to a gap is probably being very narrow. The way I read this was the leadership did not handle remote work well and feedback did not flow bidirectionally well across multiple levels. When in person sometimes luck/chance just happens and might be enough to mitigate huge issues.

edit: from the downvotes looks like remote work is an untouchable topic and has nothing wrong with it. /s

+1 Sometimes just because you can does not mean you should.

Heck you could even say the Amazon Whole Foods Acquisition was a loser -- they haven't leveraged the store network like Walmart has.

As someone who was a regular at Whole Foods even before the Amazon acquisition, from my viewpoint, it has been a win-win.

1. The online shopping experience has been amazing from the Amazon site/app. Target comes close. 1.a. The free delivery for Prime members was an awesome perk while it lasted and definitely made me buy from WF more than the alternatives I have.

2. I get 5% from the Prime card, I actually am incentivized to shop more at WF.

3. Amazon wise I can safely pick up my packages from the nearest store.

I understand it is subjective. But I use a forked version of https://github.com/puckel/docker-airflow on our managed K8s cluster and it points to a cloud managed Postgres. It has worked pretty well for over 3 years with no-one actually managing it from an infra POV. YMMV. This is driving a product whose ARR is well in the 100s of Millions.

If you have simple needs that are more or less set, I agree Airflow is overkill and a simple Jenkins instance is all you need.

Definitely interesting to look at the Pay Gap charts. But it might be interesting to see what the pay gap is after looking at hourly wage. I often hear this argument that men work longer hours. Given how much data they have presented, will definitely be worth looking at that dimension too.

Edit: In S/W that might not matter at all, but some business functions hours might matter. More data the merrier.

Absolutely spot on! Glad to see I am not the only one pondering about it. It has gotten so bad that I try to log out or use incognito to find new stuff. Tiktok on the other hand is amazing at surfacing stuff that I am actually find interesting.

Personally I think YouTube is only showing me what it knows for sure I am interested in vs TikTok actually trying to explore and exploit what I might be interested in.

Agreed, spent way too much time on the Tesla forums and did not consider the target audience :+1

I understand how the perusing the media gives this biased feeling, just feels like Tesla mentions make for a more engaged article. I ordered a MYLR without ever sitting in one due to how neat my friend's Model 3 was. After driving it for the first time, I doubt if I will buy another car. Tesla just makes amazing cars period. Much better than the 3 series I had before as a comparison.

No not a dumb question at all. Hard to give an exhaustive list. One example is say you are a Facebook/Insta influencer who is promoting a link to a product. You might have an incentive to inflate the engagement rate with that link. Just engagement will be likes/clicks on the post/link. By tracking conversions you can try to mitigate this type of behavior to an extent.

Click fraud is very real, but the big players ploughing money into digital ad spend are not idiots. There is a huge shift towards conversion/action driven advertising. Advertisers can see how much of an effect a particular ad channel had on the final conversion e.g purchase of a shoe/add to cart. Just talking about click/ad fraud without talking about counter measures/best practices is just click-baity usually.

My answer to anyone who asks for kafka: Show me that you can't do what you need with a beefy Postgres.

Sorry thats just a clickbait-y statement. I love Postgres, try handling 100-500k rps of data coming in from various sources reading and writing to it. You are going to get bottlenecked on how many connections you can handle, you will end up throwing pgBouncers on top of it.

Eventually you will run out of disk, start throwing more in.

Then end up in VACCUUM hell all while having a single point of failure.

While I agree Kafka has its own issues, it is an amazing tool to a real scale problem.

Hmmm not sure if this is an actual problem or bad configuration. Might be based on Org settings, my org locks this down since it uses ldap as source of truth for display names.

Have a Herman Miller in the office, got myself a https://www.autonomous.ai/?lang=en-US&quickview=ergonomic-co... for home use. At the 3 month mark I can say I love the Ergo Chair Core more and will definitely recommend since its not as expensive as HM, but comfortable enough.

Author here, any feedback is appreciated :+1

got burned out by XML and design pattern heavy frameworks

Most succinct summary I have heard lately :) I think a good bunch of blame is also on the developers who blindly adhere to said patterns. https://www.quora.com/What-are-the-most-ridiculous-Java-clas... to rest the case.

Agreed. Perspective is important :+1 In hindsight, should not have commented and brushed it off :)

Probably an overreaction on my part, but the comment made it seem like Adobe was a boring existence and not much happens.

Hmmm, I kept reading your comment multiple times and could not put a finger on what bothered me.

Want an example? Adobe Sign. Here is a product that they can never stop supporting. The software doesn't need more features, but they still need security updates and maintenance.

I do work at Adobe but not on Sign, and I think it just ruffled my feathers to see an offhand comment putting down some great work.

Please do check out https://acrobat.adobe.com/us/en/sign/features.html

You will be utterly surprised how many FAANG level salary companies use Java EE :)

Disclaimer: I do hate Java with a passion

Neat! One production ready python library I love to use in this space is https://radimrehurek.com/gensim/ It is quite mature and can handle a good amount of data well! It offers topic modeling too and can b helpful to find similar documents also.

Arc is an opinionated framework for defining predictable, repeatable and manageable data transformation pipelines;

I am confused by the title `Arc, an open-source Databricks alternative `. One of the main benefits of Databricks is the managed Spark. This isn't replacing Databricks as such probably giving an alternative to one of the features in Databricks.

Being a child aged 5 to 17 is 99.9 percent protective against the risk of death and 98 percent protective against hospitalization. For children 0 to 4, these numbers are 99.9 percent (death) and 96 percent (hospitalization).

I am genuinely curious what the raw numbers are behind this. Mainly to eliminate a data bias for rare events. I would assume parents of newborns and toddlers to be over cautious and hence might indicate a low exposure to people in general causing lower numbers showing up here.

Massive Scheduler performance improvements

THIS! Mainly looking forward to the scheduler HA improvements! That has been the biggest pain in a self managed Airflow environment.

Great article! One of my favorite data-structures lately is the https://github.com/tdunning/t-digest. Would be great to see how it compares and what additional accuracy we get but storing some of the data.

How about joins? Thats usually where things get tricky. Don't see it in the README, so probably missing something?

Too bad there's not really a ML/AI ecosystem around (besides Spark, which seems to be losing favor partially because of its embarrassing performance), because using Python is such a drag.

Even if Spark might not be the best tool for ML, it is still the tool to beat for data processing and custom ETL.

Nice! When I wanted to drill a bit deeper past docker, this was a great additional resource. https://jvns.ca/blog/2016/10/10/what-even-is-a-container/ definitely demystifies some of the magic.

Great effort in visualizing the Pivot breakdown. Excel is amazing at what it does, though where it fails is when you have > 1 Million rows. R comes in handy in those situations. Recently came across http://pivottabler.org.uk/index.html which is fantastic and makes me not miss Excel!