Comment by sonofhans | Hacker News Reader

Comment by sonofhans | Hacker News Reader

Flash was one of the best new technologies of the digital age, allowing nearly effortless creativity with easy tools and low-bandwidth, reliable delivery. It was glorious.

It was overused, terribly, and often used for questionable purposes (e.g., rendering a static website, or the dreaded SIFR). But it was like MySpace in how well and quickly it democratized creativity on the web, except it also afforded motion and games and video. Flash was great.

Adobe is awful, though. I’m not sure if Flash is the most useful thing they’ve killed outright, but it might be. I almost wish I could believe it was intentional, that they bought it to kill it. I really do think it was pure greed and stupidity, though. Most of what Adobe has done in the last 20 years is rent-seeking.

Imagine if they’d been a good steward of the tech, made it stable and performant on low-power devices. Yeah, Steve Jobs put the final nail in the coffin. Adobe administered the poison and tailored the funeral suit. Screw Adobe.

>rendering a static website

2advanced was the epitome of flash design. I don't think I was ever really paying much attention to the content, but it was fun to interact with the UI.

https://www.webdesignmuseum.org/gallery/2advanced-studios-v1...

EDIT: They're still going!

All on canvas now, but same kinda feel. Makes me think of command and conquer.

https://www.2advanced.com/

Imo Adobe didn't kill it, WC3 or the plugin design did, take your pick.

Flash was also the go-to technology for animation given the bandwidth constraints of that era. Those six episodes were created using Flash and broadcast on TV: http://swain.webframe.org/zeek.html

It wouldn't be too difficult to make something even better than flash that emitted wasm files, that could run everywhere. It would give lots of folks a creative outlet that could be infinitely shared.

If it's so easy, build it.

A portable runtime isn't that hard. An effective and approachable authoring tool is the real challenge.

Yes. The authoring tool was the brilliance of flash. It allowed non-coders to make interactive things really easily.

Anyone whose made an Instagram filter or TikTik effect recently - it was like that, though even easier, but allowing you to do actual coding right down to very low level stuff should you desire.

and you could get push your creation to any screen in the known universe with essentially zero testing.

Flash is the only thing that needlessly ate my battery worse than javascript though.

Pretty sure half the internet surfers had plugins that made flash run only when user activated by the time it started to wane.

Personally I only activated it for games i wanted to play, so most of that creativity was lost on me.

That was just shonky code, nothing to do with flash itself.

All of it ?!?

I remember the time when pc fans hearable meant 95% that i opened a web page with a flash component by mistake...

Heh, I’ve used that test to guess if someone was running npm.

The flash runtime was pretty efficient but the authoring tools were so easy to use that anyone could produce "working" abominations filled with the most good awful code that got the result they wanted.

I guess there is nothing wrong with javascript either?

Humanity has proven it can't handle anything else than static pages. Then the developers themselves have to pay for their bloat rather than the users.

I’m taking about a bit earlier - ie before mobile phones.

Obviously flash was not designed for small screens or small batteries.

But it could have been adapted! If Adobe had been in any way competent or caring the flash runtime could have been converted to html5

But Adobe suck

> before mobile phones.

It was the same even before mobile phones. Laptops had batteries too. CPU fans sounded even worse than today when pegged at 100%.

That program shaped my brain in ways I can't articulate. Took things I learned about computers and animation as an 11yo for granted as basic knowledge.

Having never used the authoring tool, can you point me towards something to model it after?

I think a Bret Victor, "Inventing on Principal" style editor would be perfect, combined with some sort of scratch like Python IDE where each element is defined in terms of its reactive behaviors with other elements on a timeline.

I didn't say easy, I said not too difficult. Order of magnitude difference. :)

https://youtu.be/PUv66718DII?t=634

With some http://lighttable.com/

You can watch some tutorials on Macromedia Flash on YouTube. Here's an example https://www.youtube.com/watch?v=sfHNexSSqfg&ab_channel=JGtut....

Building the tech isn’t the problem; getting 90% of Internet users to adopt it is.

Building a good tool is the problem. For instance Git was born in a very niche community (kernel programming), yet it spread like wildfire. Despite the shortcomings of its CLI (according to some), and despite the fact there was other well'established tools.

As someone that would rather use Mercurial and was perfectly fine with Subversion, that wild fire was "Linux kernel uses it, was made by Linus, and I want to be cool like Linus".

Rather than a weird "cult of personality", the fact that Git demonstrated with Linux that it could handle a relatively large amount of source code and allow dozens (hundred?) of programmer to work on it "asynchronously" were key factors for adoption, I believe. If SVN was that good, as it was already well-established online (Github was not the first online VCS frontend, far from it), it should have remained the king of the hill. But it was decapitated instead.

I am been using FOSS and commercial source management tooling since 1998, and the very fact that Linus only bothered with git due to BitKeeper changing their licence, proves that there was nothing special about it, other than personality and adoption wave.

SVN is still king of the hill in game development unless I'm mistaken. Because git is quite bad at handling large/binary files.

Game development often uses commercial solutions like perforce that have central servers so you can lock out your binary files and everyone will know not to touch them at the same time as you.

I think I’ve seen SVN in a single studio in the past decade. It’s almost all Perforce, because of better performance and tooling and the Unreal integration.

And yes, locking unmergeable files is essential either way.

Ah. I forgot about Perforce. I stand corrected

Mercurial was slower than a decaffeinated sloth. It lost because it sucked. Sorry.

A side effect of the great idea to write application software in Python that perdures to this day, still it was fast enough for purpose, it isn't like doing SCM operations is akin to 120 FPS for a game engine.

It isn't great that are now so many Git GUI clients written in Electron?

Can you configure them as much as you can configure Git TUI clients? :P

The problem with Python CLI tools is startup time. Just firing up the interpreter and importing modules can take a frustratingly long time if you're a twitchy CLI user. That's much less of a problem for a GUI tool that tends to sit open for a while.

As someone who's worked at a company using Mercurial, the wild fire was that everyone was begging to get onto something better and less opinionated.

Made by and used by one of the largest open source projects in the world. And built in 5 days.

The new Flash is a whole other pair pants. Also if people want to make easy things for the web in a closed platform, Unity already exists.

I wonder, how close does Godot get with its web export support? The "authoring tool" seems pretty good, and it exports to WASM.

It doesn't in high-school visual basic seemed to have had a 50% pickup rate based on final project quality (closest I can think of to Godot that I actually saw) , Flash nearly everyone/group had a viable thing to show off for the final project.

Godot is too complicated compared to Flash. Have you ever used Flash? In Flash, you could literally take the brush from the toolbox, draw a circle, and that was your player character. That was it. No importing assets, no creating sprites, no nothing. You literally draw the circle right there with vector art tools, and it can be whatever you want.

And those were THE best vector art tools, because when you drew a shape and the shape overlapped with another, it automatically erased the other shape like you would expect it to in a raster graphics editor. In pretty much every other software I tried, e.g. Inkscape, Affinity, Corel Draw, Illustrator, you just get two separate shape objects one on top of the other. They seem to be designed for drawing the outlines, not to actually paint with brushes. Flash understood what was intuitive for artists.

Honestly, the older I get the stranger I feel about the fact that there was a brilliance in creating interface for people to be productive with back then that seems to be completely gone. I think this may be in part because desktop applications are unusual nowadays, but it's just really strange that the things I remember seeing have been done exactly once and never copied by anybody despite how well they worked.

IMO, it’s not the technology that matters but the authoring tools. Flash was easy to pirate, sometimes free with school, and easy to use. Thinking of applications as “animations with scripting” was a great idea because it let people only learn what they wanted to.

If you just want to make animations, skip actionscript entirely. If you want some extra Easter eggs, just turn any symbol into a button and you can trigger any movie clip you want. Want to make this a whole game, heres the whole action script editor.

I don't think so, Flash Player was extremely stable and portable, swfs ran accurately (although sometimes slowly) on all targets, Safari, IE7, Nintendo Wii, didn't matter.

Maybe wasm+Canvas is getting close to being stable. But if you code up a webgl game and test it on desktop Chrome, odds are it'll just be a black screen on anything but the most popular Android devices, and you'll need to do tons of work and testing to get it to run on an iPhone.

And WebGPU is probably at least 5 years away from being stable enough to be considered to "run everywhere".

I strongly considered trying to do this a few times, but ended up writing an swf parser in C++ and used Skia to render. It's probably the best we can do right now, but we still regularly get players who complain about just seeing a black screen.

Adobe tried really hard to make Flash work well on mobile.

To the point they gave a free flagship Android smartphone with Flash installed on it to every employee, to try forcing some internal dog-fooding.

No they didn't. I was on the CAB (community advisory board) at the time (from the community side, received an invitation due to being very active on several ActionScript forums etc.)

What was happening is this: Adobe inherited Flex Framework from Macromedia. It was atrocious. Bloated, slow, unusable for games or simple interactive applets which made Flash popular in the first place. They tried to market that for creation of admin tools like what VSPhere used to have. The community was crying and complaining to Adobe that they aren't making any effort to support what Flash was good at.

Then smartphones appeared. And Adobe's engineers came up with Flash Catalyst. Conceptually, it was supposed to work similar to Flash or Fireworks, but it was supposed to render its output into generated Flex Framework code. Tons of unjustifiable bloat. Everything was too slow and too big. And whenever the people able to get their hands on the preview complained about it, the answer was always "it's not going to get faster / smaller".

Catalyst came with templates for "typical" phone apps and they were a disaster to deal with. So slow users would probably understand them to be simply broken / frozen. Oh, but Adobe assigned a bunch of career managers to these projects. People came made a lot of promises and got promoted into another department. It was painful to watch.

At the same time, I wrote a project that I called "Frameworkless MXML", it was more of a demonstration than an actual useful library, but the goal was to say that the original inherited framework might be stripped off a lot of bad components and made to be a lot faster and smaller. But it was just that, a PoC that never went anywhere.

I had no interest in working with Flash on smartphones because I see smartphones as proprietary jails and am not interested in contributing to that market. Few people in Flash OSS community probably also felt distaste for the smartphones... and the vast majority were the "non-contributing" users. They waited until the tower completely crumbled and jumped ship.

The urinal choice game was formative for me.

Reply on news.ycombinator.com