Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Dear Google Cloud: Your Deprecation Policy Is Killing You

medium.com

Post reply on HN
Print view View original

81–90 of 417 posts

Re: Dear Google Cloud: Your Deprecation Policy Is Killing You

#81

Post by Animats »

physicles wrote: post #56

The mishandled Python 3 transition sure created a lot of pain (I experienced some of it, even though I only started with Python in 2016), but I never realized how much it hurt the language’s momentum. This must be one of the reasons why JavaScript is so popular today. JS benefits here from being run in browsers, which the author points out are fantastic at back compat, even as the dev culture surrounding the language…

I never realized how much it hurt the language’s momentum.

Yes. The combination of a botched transition with a high arrogance level was hell. Here's something I wrote in 2015 after porting a medium-sized system from Python 2 to Python 3.[1] I listed a number of package problems encountered during the conversion. Ones that indicated those libraries weren't being used much in production. I discovered that the SQL connector broke if you loaded a database of tens of thousands of records, for example. And oh, did I get hate mail. You can see the comments below mine.

That's what led me to Go. Go is a mediocre language. That's a strength when you just need to get something done. It comes with a set of libraries which are heavily used internally within Google. So, if you're doing something that is server-side for some web-related task, the libraries for that are probably both present and well-debugged. And they don't seem to be deprecated rapidly. I'm not using them for any Google-specific services, so I can't speak to that.

[1] https://lists.archive.carbon60.com/python/python/1187081/?pa...

Re: Dear Google Cloud: Your Deprecation Policy Is Killing You

#82

Post by mamurphy »

fh973 wrote: post #75

I think Google couldn't do better, even if they wanted to. The reason is the engineering culture and the promotion system. People expect to switch to new projects, and work towards to a promotion within 2-3 years. This shows in all products, and presumalby also the reason for the deprecation. New people come on the project, want and need to do a major overhaul, and do not have resource or incentives to support the ol…

My armchair thought: what if the CEO said "wow, Yegge is right, we've gotten really off trail here. you need to show leadership in maintaining a 3+ year old system at Google for promotions?" Could that happen and would it work?

Re: Dear Google Cloud: Your Deprecation Policy Is Killing You

#83

Post by logicslave12 »

tassl wrote: post #23
lowiqengineer wrote: post #2

The answer to why this is, I’m beginning to realize after talking to dozens of Googlers working at Google as acquaintances, friends, former coworkers, HN comments and twitter personalities is - Google has an active disrespect if you’re not part of their in group . If you haven’t passed their tests, you’re basically an object of derision - a mark, an object just used for incrementing a CPM metric, or an incompetent, l…

I have worked at a few orgs and I've heard so many time the "because we are Google" to explain why a product will be used. I've never understood that mentality; I've never understood people being proud basically about a product somebody else built 20 years ago. I do believe leadership is failing Google, and their lack of vision is starting to affect the company. There was quite a bit of inertia, but I fear search mig…

NLP is rapidly improving, a time may come when someone else can create good search just because nlp tech has gotten so good

Re: Dear Google Cloud: Your Deprecation Policy Is Killing You

#84

Post by sidkshatriya »

KingOfCoders wrote: post #41

"Fuck yooooouuuuuuuu. Fuck you, fuck you, Fuck You. Drop whatever you are doing because it’s not important. What is important is OUR time." Most NPM packages.

> “[...] Drop whatever you are doing because it’s not important. What is important is OUR time”

> Most NPM packages

NPM packages tend to be free and open source so morally they don’t owe us anything. So yes what is important is the time of the authors of the packages. I am grateful someone took the trouble to release them and many popular ones are actually well maintained.

But cloud services that you are paying for are a totally different ball game. The company does need to maintain backwards compatibility for a good amount time and they need to value your time.

Re: Dear Google Cloud: Your Deprecation Policy Is Killing You

#85

Post by glogla »

joshuamorton wrote: post #70
CJefferson wrote: post #58

Earlier quoted context omitted.

I've never been hired or approached by Google. However, as a University lecturer I know a lot of people Google have employed. In my opinion, they have rejected some of the greatest students I have ever taught, and accepted some idiots who know how to speak well. While they have employed some good people, I believe they purposefully taret the type of people who think working at Google makes you a fundamentally better…

> In my opinion, they have rejected some of the greatest students I have ever taught There's no question that Google is happy to have false-negatives in interviews. > I believe they purposefully taret the type of people who think working at Google makes you a fundamentally better person than anyone else why would they do this? And, if they do this, how are they generally speaking so successful? (this is also an amusi…

> And, if they do this, how are they generally speaking so successful?

They have almost complete monopoly on online search, web ads, online video, on phone OS, on browsers. And they are not afraid to abuse those to get more, and are getting away with this.

You don't need to do anything right when you rent-seek most of the online world.

Re: Dear Google Cloud: Your Deprecation Policy Is Killing You

#86

Post by londons_explore »

mamurphy wrote: post #82
fh973 wrote: post #75

I think Google couldn't do better, even if they wanted to. The reason is the engineering culture and the promotion system. People expect to switch to new projects, and work towards to a promotion within 2-3 years. This shows in all products, and presumalby also the reason for the deprecation. New people come on the project, want and need to do a major overhaul, and do not have resource or incentives to support the ol…

My armchair thought: what if the CEO said "wow, Yegge is right, we've gotten really off trail here. you need to show leadership in maintaining a 3+ year old system at Google for promotions?" Could that happen and would it work?

Yes, but there might be other downsides... Suddenly every developer turns into 'maintenance man' just refactoring existing code to make it neater without adding much.

Re: Dear Google Cloud: Your Deprecation Policy Is Killing You

#87

Post by bobbiechen »

bickeringyokel wrote: post #69

> One fun bit of trivia about Bigtable is that they had these internal control-plane entities (as part of the implementation) called tablet servers, which had large indexes, and at some point they became a scaling bottleneck. Anyone understand the technical jargon here? I'm a bit lost.

The BigTable paper is available here: https://research.google/pubs/pub27898/

> Bigtable maintains data in lexicographic order by row key. The row range for a table is dynamically partitioned. Each row range is called a tablet, which is the unit of dis- tribution and load balancing.

And later...

>The master is responsible for assigning tablets to tablet servers, detecting the addition and expiration of tablet servers, balancing tablet-server load, and garbage col- lection of files in GFS. In addition, it handles schema changes such as table and column family creations. Each tablet server manages a set of tablets (typically we have somewhere between ten to a thousand tablets per tablet server). The tablet server handles read and write requests to the tablets that it has loaded, and also splits tablets that have grown too large.

Re: Dear Google Cloud: Your Deprecation Policy Is Killing You

#88

Post by waheoo »

I recently migrated off of digital ocean into aws.

I was originally planning to go to go for the native k8s experience.

Then I read somewhere online that was doesn't sunset services where as gcp does.

I was half way through the migration process to gcp and bailed on a dime.

Re: Dear Google Cloud: Your Deprecation Policy Is Killing You

#89

Post by dilyevsky »

bickeringyokel wrote: post #69

> One fun bit of trivia about Bigtable is that they had these internal control-plane entities (as part of the implementation) called tablet servers, which had large indexes, and at some point they became a scaling bottleneck. Anyone understand the technical jargon here? I'm a bit lost.

Check out the original bigtable paper and it’ll make more sense. I think the author maybe misremembering some technical details here though.

Re: Dear Google Cloud: Your Deprecation Policy Is Killing You

#90

Post by jiggawatts »

ryanobjc wrote: post #14

I used to work with Steve. He’s a good story teller. But a story teller tells good stories by having a fairly casual relationship with facts and the truth. For example, emacs lisp suffers from bitrot. Some core things haven’t changed in forever but also you can’t pick up elisp config from 15 years ago and it’s fine today. All live systems evolve and change. In this case emacs is just so slow that Steve thinks that it…

Nobody pays EMACS a few million a year and runs their core business through it.
Post reply on HN