5 ms·

We can be compassionate about the folks getting laid off while still realizing that the company was grossly inefficient and needed to restructure -- these are n

by type-r 4y ago

We can be compassionate about the folks getting laid off while still realizing that the company was grossly inefficient and needed to restructure -- these are not contradictory feelings. For me personally, I have no idea what "technology and development pace" these layoffs even could effect, as I have seen no real change in the service in the past ~5 years I've used it. Actually, I have noticed that in the last year or two they refuse to show me more than a tweet or two until trying to force me to log in.

411111111111111 4y ago
> Actually, I have noticed that in the last year or two they refuse to show me more than a tweet or two until trying to force me to log in. Things like that are usually UX driven, developers often don't have any recourse beyond implementing what's demanded in such a setting.
x0x0 4y ago
Ironically, I think Twitter started actually building product for the first time in a long time recently. One frame through which you could view Twitter is it was the place where famous / popular / important (athletes, movie stars, government officials) talk to the public. Spaces, enabling audio conversations for those people, makes tons of sense. So also Twitter Blue letting you read certain news w/o a paywall. If you're a serious consumer of news on Twitter, wsj/nyt/wapo/latimes/ft probably cost $1500/year. If the economics worked for the news orgs, there was huge value there. Of course, they really sat on the product for 5+ years, so oof...
yolovoe 4y ago
> as I have seen no real change in the service in the past ~5 years I've used it. You underestimate the effort it takes to keep a service as large as Twitter running. Constant software upgrades to maintain high standard of security, inevitable technological break-age, deprecated dependencies, operations, root-causing failures, migrations (you can't avoid hardware migrations; hardware will die out), planning fallacy, etc, etc.
- killdozer 4y ago
  Yes it's incredible to see so many people on _hacker news_ who don't understand how much work it is to keep even moderately large (low millions of users) services running.
  - wink 4y ago
    Yes, and repeatedly people point out that high dozens or low hundreds should be enough but they have thousands (allegedly). I think the point you are missing that especially tech people might have been using twitter for 10 years like a public IRC client. Open App/Browser, read messages, write messages, complain about lack of linear timeline, close app. Yes, it has grown, but they added a million features that probably none of the people here asked for. We don't necessarily follow Justin Bieber where the system needs to be able to handle 2m users connected, we follow a kernel developer with 2000 followers. I guess it's a marvel of engineering that's mostly needed for whatever method they found to make money off, I'm still asking myself this after these years.
- shaklee3 4y ago
  I thought Twitter was mostly on AWS?
  - boardwaalk 4y ago
    I don't think most of the issues mentioned are mitigated by AWS, sadly.
  - everfrustrated 4y ago
    According to whistleblower Peiter Zatko they had two datacentres. He reports that they were worried that if a DC went offline they might not be able to bring it back up again! Their servers in their DCs were running unpatched operating systems with no disk encryption! I highly recommend reading his full report. The level of disfunction he reports is insane.
- anthropodie 4y ago
  Here's story for you A team of 30 works at zerodha for serving 6.2 million users. They are largest trading broker in India. There story is worth a read https://zerodha.tech/blog/hello-world/ https://zerodha.tech/blog/hello-world/ Edit: Apparently they have 1200 employees for operations and everything else. Thanks to child comment for pointing that out.

by type-r 4y ago

411111111111111 4y ago

> Actually, I have noticed that in the last year or two they refuse to show me more than a tweet or two until trying to force me to log in. Things like that are usually UX driven, developers often don't have any recourse beyond implementing what's demanded in such a setting.

x0x0 4y ago

Ironically, I think Twitter started actually building product for the first time in a long time recently. One frame through which you could view Twitter is it was the place where famous / popular / important (athletes, movie stars, government officials) talk to the public. Spaces, enabling audio conversations for those people, makes tons of sense. So also Twitter Blue letting you read certain news w/o a paywall. If you're a serious consumer of news on Twitter, wsj/nyt/wapo/latimes/ft probably cost $1500/year. If the economics worked for the news orgs, there was huge value there. Of course, they really sat on the product for 5+ years, so oof...

yolovoe 4y ago

> as I have seen no real change in the service in the past ~5 years I've used it. You underestimate the effort it takes to keep a service as large as Twitter running. Constant software upgrades to maintain high standard of security, inevitable technological break-age, deprecated dependencies, operations, root-causing failures, migrations (you can't avoid hardware migrations; hardware will die out), planning fallacy, etc, etc.

killdozer 4y ago
Yes it's incredible to see so many people on _hacker news_ who don't understand how much work it is to keep even moderately large (low millions of users) services running.
- wink 4y ago
  Yes, and repeatedly people point out that high dozens or low hundreds should be enough but they have thousands (allegedly). I think the point you are missing that especially tech people might have been using twitter for 10 years like a public IRC client. Open App/Browser, read messages, write messages, complain about lack of linear timeline, close app. Yes, it has grown, but they added a million features that probably none of the people here asked for. We don't necessarily follow Justin Bieber where the system needs to be able to handle 2m users connected, we follow a kernel developer with 2000 followers. I guess it's a marvel of engineering that's mostly needed for whatever method they found to make money off, I'm still asking myself this after these years.
shaklee3 4y ago
I thought Twitter was mostly on AWS?
- boardwaalk 4y ago
  I don't think most of the issues mentioned are mitigated by AWS, sadly.
- everfrustrated 4y ago
  According to whistleblower Peiter Zatko they had two datacentres. He reports that they were worried that if a DC went offline they might not be able to bring it back up again! Their servers in their DCs were running unpatched operating systems with no disk encryption! I highly recommend reading his full report. The level of disfunction he reports is insane.
anthropodie 4y ago
Here's story for you A team of 30 works at zerodha for serving 6.2 million users. They are largest trading broker in India. There story is worth a read https://zerodha.tech/blog/hello-world/ https://zerodha.tech/blog/hello-world/ Edit: Apparently they have 1200 employees for operations and everything else. Thanks to child comment for pointing that out.