Hacker News Reader

Comment by userbinator | Hacker News Reader

Fortunately the source code is available:

https://android.googlesource.com/platform/frameworks/base/+/...

If I had the time, I'd try to compile a binary of it that will run on Win95 just to give my fuckings to the planned obsolescence crowd.

View on HN

There is no point for Google to push planned obsolescence on the PC or server space. They don't have a market there.

It does benefit them to make it harder for competitors.

When you mention "competitors," what industries or markets are you referring to?

No one would write Android apps on a Chromebook, and making it harder to do so would only reduce the incentive for companies to develop Android apps.

How could Google benefit from pushing a newer instruction set standard on Windows and macOS?

The one moderately popular competitor is the project in the OP that is suffering directly from this upstream change.

I doubt Google even cares about F-Droid. The Play Store competes with the iOS App Store, Huawei's App Galery, and probably the Samsung Store long before F-Droid becomes relevant.

If they required a Google-specific Linux distro to build this thing or if they went the Apple route and added closed-source components to the build system, this could be seen as a move to mess with the competition, but this is simply a developer assuming that most people compiling apps have a CPU that was produced less than 15 years ago (and that the rest can just recompile the toolchain themselves if they like running old hardware).

With Red Hat and Oracle moving to SSE4.1 by default, the F-Droid people will run into more and more issues if they don't upgrade their old hardware.

F-Droid is so insignificantly small their entire userbase is smaller than the amount of users of each app in the top 500 of every single large store (Play, Galaxy, Huawei, etc.)

This happened because nobody gives a shit about F-Droid, not because it's somehow a "threat" with unmaintained apps.

While your perspective makes some sense, it's highly improbable. It's unlikely that Google was aware of F-Droid's infrastructure specs, or its inability to fix the issue in advance.

It seems you're suggesting a very specific, targeted attack.

> It seems you're suggesting a very specific, targeted attack.

Yes, just like it happened with Firefox: https://news.ycombinator.com/item?id=38926156

Former Chrome team member here. Nightingale's suspicions were plausible but incorrect. The primary cause of every one of these we looked into over the years (and there were indeed many) was teams not bothering to test against Firefox because its market share was low compared to the cost of testing for it. In many cases teams tried to reduce support burden by simply marking "unsupported" any browser they didn't explicitly test, which was sometimes just Chrome and Safari. We were distressed at this and wrote internal guidance around not doing things like the above, and tried to distribute it and point back to it frequently. Unfortunately Firefox' share continued to go down, engineering teams continued to be resource-constrained, and the problem continued to occur.

Several years ago I glumly opined internally that Firefox had two grim choices: abandon Gecko for Chromium, or give up any hope of being a meaningful player in the market. I am well aware that many folks (especially here) would consider the first of those choices worse than the second. It's moot now, because they chose the second, and Firefox has indeed ceased to be meaningful in the market. They may cease to exist entirely in the next five years.

I am genuinely unhappy about this. I was hired at Google specifically to work on Firefox. I was always, and still remain, a fan of Firefox. But all things pass. Chrome too will cease to exist some day.

Thank you for the interesting insight. This is more or less what I expected.

> suspicions were plausible but incorrect

The suspicions were not about the evil will of the engineers. It's the will of Google itself (or managers, if you want), which plays the main role here. This is exactly what causes the following:

> engineering teams continued to be resource-constrained

It reminds me a bit of Boeing: https://news.ycombinator.com/item?id=19914838

The resource constraints had nothing to do with intentionally not funding "support a competing browser properly", though, and everything to do with just not funding engineering and test work at all except to build Shiny Idea That Got A Promo.

Despite its size, Google does shoestring engineering of most things, which is why so much is deprecated over time -- there's never budget for maintenance.

So I mean in some sense yes, there's valid criticism of Google's "will" here, but that will was largely unaware of Firefox, and the consequences burned Google products and customers just as much or more in the long run. Nightingale looked past individual instances to see a pattern, but didn't continue to scale the pattern up to first-party products as well.

"If I had the time, I'd try to compile a binary of it that will run on Win95 just to give my fuckings to the planned obsolescence crowd"

The idea that not supporting a 20+ year old system is "planned obsolescence" is a bit shallow

But you don't, so you won't, scoring one for the planned obsolescence crowd.

And so won't anyone else who has time to complain about planned obsolescence, and that includes myself.

The Win95 API is pretty incomplete. That was actually a terrible OS. The oldest I'd go playing this game with anything serious is probably XP.

It can read files, write files, and allocate memory. Is there anything else you need to compile software?

Can it? Files on Windows 95 and files on most Unix-like OSes are very different things.

They're the same from the perspective of a stream of persistent bytes.

If you want "very different" then look at the record-based filesystems used in mainframes.

Do you have any recommended reading about record-based filesystems?

Like it is a one-off thing to support some system. You must maintain it and account it for all the features you bring in going forward.

Reply on news.ycombinator.com