News for hackers. Stuff that matters.
[dead]
Re:Building Uber’s Go Monorepo with Bazel by CydeWeys on Thursday May 14, 2020 @12:12PM (#23181274)
It easily could help them on their way to profitability if this works better than their previous system and allows faster, less hassle builds, i.e. you get more productivity out of the same number of engineers.

This tech negativity seems out of place on HN.

Re:Building Uber’s Go Monorepo with Bazel by nindalf on Thursday May 14, 2020 @03:01PM (#23183859)
> This tech negativity seems out of place on HN.

Why? Cynicism tends to make the source look smart. And the HN comments section is mostly filled with people (me included) trying to look smart.

While it's possible to stick to commenting only on topics we know a lot about, that's not the most efficient way to farm karma. The best way is to take a big cynical dump on whatever new thing people are trying to do. And under no circumstances should we show optimism - we won't look smart if we do.

Here are two comments from this thread. Ask yourself, which person seems smarter.

1. "I wish my company was at the scale of Uber to work on all these kinds of tech. Makes me want to leave my current job just to go to Uber."

2. "This will surely bring them to profitability. Exactly the innovation the food delivery space was looking for."

Re:Building Uber’s Go Monorepo with Bazel by quest88 on Thursday May 14, 2020 @12:44PM (#23181820)
Do you really think so? Do you think the rate at which uber is able to ship new features is the reason why they're not profitable? If yes, then shouldn't Lyft or Flywheel or some new company that can ship much faster be better off than Uber?
Re:Building Uber’s Go Monorepo with Bazel by superworrier on Thursday May 14, 2020 @03:10PM (#23183991)
From what I've heard of Lyft, they definitely do not ship faster. One (probable) reason is that their infrastructure and tooling is a mess.
Re:Building Uber’s Go Monorepo with Bazel by CydeWeys on Thursday May 14, 2020 @02:09PM (#23183161)
I don't see this as a new feature, I see it as cost control. Fixing technical debt brings your long-term carrying costs lower. They're increasing the efficiency of their developers so they can do more with fewer of them, which helps reduce total compensation costs.
Re:Building Uber’s Go Monorepo with Bazel by bbeekley on Thursday May 14, 2020 @01:47PM (#23182851)
> then shouldn't Lyft or Flywheel or some new company that can ship much faster be better off than Uber?

Isn't that the general startup success story: small scrappy team adapts faster to changes and the market, giving them some advantage over slow-moving megacorp?

Re:Building Uber’s Go Monorepo with Bazel by pm90 on Thursday May 14, 2020 @12:05PM (#23181177)
I understand the snark but this is hacker news after all. Any kind of technical innovation is awesome to me as an HN reader even if it may not make business sense.

Also, I’m totally fine with them using all that investment money for R&D rather than lobbying or returning to shareholders among other things.

Re:Building Uber’s Go Monorepo with Bazel by dmoy on Thursday May 14, 2020 @12:13PM (#23181294)
I agree with you. When there was a SpaceX article about bazel, it'd be a bit unfortunate to see a top comment be something about an exploding rocket.

(In reality, there were zero comments, so w/e)

Re:Building Uber’s Go Monorepo with Bazel by quest88 on Thursday May 14, 2020 @12:48PM (#23181880)
What about profitability? R&D is fine if they're profitable, but they're not.
Re:Building Uber’s Go Monorepo with Bazel by bhupy on Thursday May 14, 2020 @01:12PM (#23182242)
Uber doesn't need to be profitable yet, they have $10B in cash.

While this doesn't mean that they can continue to be unprofitable forever, it means that they have latitude to engage in R&D (like this post) while also moving towards profitability, at the same time.

Re:Building Uber’s Go Monorepo with Bazel by lhorie on Thursday May 14, 2020 @01:30PM (#23182531)
Adopting an open source build system doesn't exactly strike me as R&D
Re:Building Uber’s Go Monorepo with Bazel by bhupy on Thursday May 14, 2020 @01:34PM (#23182618)
Sure, I mean that Uber can engage in whatever cool tech projects that they want and publish whatever blog posts that they want.

This documented project's usefulness to the tech community is totally independent from its balance sheet. It's not like NOT publishing this eng blog post makes some material difference in Uber's long run ability to turn a profit...

Re:Building Uber’s Go Monorepo with Bazel by lhorie on Thursday May 14, 2020 @02:09PM (#23183164)
By that logic, companies should never open source anything. Yet, they do. Typically, the reason is to attract talent. Not everything has to be focused on short-term ROI.
Re:Building Uber’s Go Monorepo with Bazel by bhupy on Thursday May 14, 2020 @02:51PM (#23183728)
Yes, we are in agreement. I'm attacking the notion that we ought to bring Uber's profitability into this, it seems totally irrelevant to the merits of using Bazel in a Go monorepo.
Re:Building Uber’s Go Monorepo with Bazel by eternalban on Thursday May 14, 2020 @02:14PM (#23183234)
Are you passionate about food delivery? Are you laser focused on food on wheels? Come join us at Uber to create the future of food delivery! We have an excellent team of engineers and are looking for more!

vs

At Uber we face technical challenges that afford our engineers the opportunity to design and build world class software, make a significant impact in OSS, and tell their peers about it in major conferences. We have an excellent team of engineers and are looking for more!

Re:Building Uber’s Go Monorepo with Bazel by tequila_shot on Thursday May 14, 2020 @12:05PM (#23181174)
Not sure if this is sarcasm. If not, how?