5 ms·

If my experience is any indication of general industry practices, businesses are very risk averse when it comes to upgrading libraries or dependencies as it usu

If my experience is any indication of general industry practices, businesses are very risk averse when it comes to upgrading libraries or dependencies as it usually triggers costly regression testing. At my current gig, we still run on angular 1.3.something and there is no way the business will allow us to migrate to v2 as we struggle to justify exactly what added value will we get from essentially a top-down rewrite.

Touche 10y ago
> there is no way the business will allow us to migrate to v2 We as engineers should not accept this. Business should not be involved in technical decisions like library upgrades. Business and engineering should negotiation development scheduling, nothing more. This means what new features are worked on, what bugs are prioritized, etc. All remaining time should be up to engineering's discretion. And if business has you working on features full-time, then that's a (really) bad business and you should find a more acceptable company to work for.
- natoliniak 10y ago
  I dont't disagree with you, but after some years in the industry I noticed certain patterns in how businesses operate: as they get larger and stable they get more risk averse and more in love with legacy systems. But I never worked for Google or the likes, so maybe somewhere out there, the grass is greener :)
- coding123 10y ago
  I'd have to say I agree with you generally in "library upgrades" but I don't consider AngularJS and Angular "2+" as the same library.
- tomc1985 10y ago
  Except when said library upgrades send costs spiraling out of control, because the refactor took 3x longer than anticipated...
tiengtinh 10y ago
similar to your case, our product was also built on angular 1.x. However, we (and the product as well) have had a better fate in which we successfully convince the big guy to use React for any new feature that we want to build or code we need to fix. Now, we are having more and more react components, and lesser angular 1 code every day.

Touche 10y ago

> there is no way the business will allow us to migrate to v2 We as engineers should not accept this. Business should not be involved in technical decisions like library upgrades. Business and engineering should negotiation development scheduling, nothing more. This means what new features are worked on, what bugs are prioritized, etc. All remaining time should be up to engineering's discretion. And if business has you working on features full-time, then that's a (really) bad business and you should find a more acceptable company to work for.

natoliniak 10y ago
I dont't disagree with you, but after some years in the industry I noticed certain patterns in how businesses operate: as they get larger and stable they get more risk averse and more in love with legacy systems. But I never worked for Google or the likes, so maybe somewhere out there, the grass is greener :)
coding123 10y ago
I'd have to say I agree with you generally in "library upgrades" but I don't consider AngularJS and Angular "2+" as the same library.
tomc1985 10y ago
Except when said library upgrades send costs spiraling out of control, because the refactor took 3x longer than anticipated...

tiengtinh 10y ago

similar to your case, our product was also built on angular 1.x. However, we (and the product as well) have had a better fate in which we successfully convince the big guy to use React for any new feature that we want to build or code we need to fix. Now, we are having more and more react components, and lesser angular 1 code every day.