Hacker News new | comments | show | ask | jobs | submit login

Preservation and reduction of bit rot is a great reason to open source Flash or ensure there is always a definitive version available, even if all the browsers should ban it because it is a messy, unperformant security liability.



Dop not worry. The SWF files will be preserved and long playable after 2020, using Flash Projector, Swiff Player or iSwiff. Just no longer in web browsers.


And I must add that open sourcing it would clear up a lot of situation where the distribution of flash player had an unclear status.


> Preservation and reduction of bit rot is a great reason to open source Flash

It's a reason, and a good point, but it's not a good reason. Improving the already-open-source players would be a better alternative. Adobe could even help facilitate that, by documenting the interfaces and enough of the internals so that other developers can achieve better compatibility. A canonical sandboxed version would also achieve the goal of preservation, without the security woes. Open-sourcing Flash is would just ensure its continued use as it is today, crowding out either of those efforts.


> Improving the already-open-source players would be a better alternative.

The open source players could be improved by comparing implementation with Adobe's once it is open sourced.


Yes, that would also work in the sense of improving compatibility, but it would still leave the other players competing with Adobe's own. Eventually they might prevail, due to being less security-bug-ridden, but in the interim we'd still be worse off.

Also, this whole discussion is moot because open-sourcing Flash ain't gonna happen. I can practically guarantee that it's too entangled with other companies' IP for that to happen. Maybe the people who are arguing that we should treat open source as an alternative to the trash can (which is pretty damn insulting to those of us who actually produce open-source software) could spend just a little bit of their obviously-copious free time lobbying to change that IP regime.


> but it would still leave the other players competing with Adobe's own.

You have that backwards.

The effort going into open source projects emulating a closed source product are wasted if that closed source project can be open sourced. That's a win for everybody, after that the effort can be directed to something more important because perfection is now attainable.

The only reason these open source projects exist in the first place is because the reference implementation isn't open sourced. I'm pretty sure the maintainers of those projects will welcome such a move by Adobe, there is no innate right to some kind of 'protected space' for open source projects that purposefully aim to provide an open implementation of a closed source product. Obviously one of the risks of embarking on such a project is that the original may one day be open sourced.


> The effort going into open source projects emulating a closed source product are wasted if that closed source project can be open sourced.

Yeah, that Linux thing just died when Solaris was open-sourced. Even more relevantly, IE and Chrome.

> perfection is now attainable.

That's cute. As somebody who has actually produced software for thirty years (but without ever hitching my wagon to Flash like some here) I'm a bit more skeptical about the likelihood of taking a gigantic long-lived ball of mud and making it perfect. In particular, security is hard to bolt on after the fact. I certainly wouldn't assume that the maintainers of other implementations would be overjoyed by the prospect. Just look at how many reimplementations there are of things that were already open source. I'd say it's unlikely that a significant number of people will work on Flash without being paid to do so.


> Yeah, that Linux thing just died when Solaris was open-sourced.

I fail to see the relevance. Linux never tried to be an 'open source Solaris clone' and of course it didn't die off, in fact, it flourished, which sort of undermines your argument.

> Even more relevantly, IE and Chrome.

What of them? I could see Mozilla vs IE.

> As somebody who has actually produced software for thirty years

Appeal to authority ;)

> but without ever hitching my wagon to Flash like some here

Join the club. I absolutely hated Flash and for many years resisted using it to deliver video.

> I'm a bit more skeptical about the likelihood of taking a gigantic long-lived ball of mud and making it perfect.

Perfect as in 'perfect emulation', not perfect as in perfect all across the board. Having the Flash source would allow the open source implementations to look at how things were done so they can emulate all behavior up to and including the bugs. If Adobe would open source it then they would not have to avoid looking at even disassemblies of the Flash code.

> In particular, security is hard to bolt on after the fact.

This is not about future developments in Flash, it is about digital archeology. If you start seeing it through that lens then maybe you will be a bit more relaxed about it.

Nobody is suggesting that Flash be given eternal life or that we will have another round of Flash content if Flash is open sourced. The sooner it is gone from the web the better. But that old stuff still remains and maybe one day someone wants to review it. That is what this is about. CF people using C64 Roms or Nintendo stuff 3 decades old.

> I certainly wouldn't assume that the maintainers of other implementations would be overjoyed by the prospect.

Did you ask them?

> Just look at how many reimplementations there are of things that were already open source.

So what? Fragmentation is the very power of open source at work.

> I'd say it's unlikely that a significant number of people will work on Flash without being paid to do so.

Exactly. So that is - again - why open sourcing the old Flash player is a good thing.


> Improving the already-open-source players would be a better alternative.

Between the options of open-sourcing an already-complete and canonically "correct" implementation, and having Adobe document their code and interfaces and improving the partial (but open) implementations that we have, it seems like the latter involves a lot more work. I guess I don't see the benefit.


Open sourcing it won't help preserve content - that's far more effectively achieved via VMs and using the archived players which Adobe provide (and will hopefully continue to provide or allow preservation of on abandoware or other sites).

Maintaining the build systems, compatibility layers and so forth won't be easy (Adobe often couldn't manage it), and bug fixing in the code is a risky prospect for preservation as it could render older content unplayable or differently playable.

The swf file format is almost open - it would be nice if Adobe updated the published specification to the latest format. Again that would go some way towards ensuring the content remained available in some way.

Edit: genuinely interested in the downvotes here - am I missing something fundamental about content preservation? It seems logical to me that the file spec being available + original players is far more valuable than the source, as the source will then bitrot without continuous maintenance. In comparison the players will always run in the VMs.


Not sure if you know but Adobe open source the Flex framework, is now part of Apache and it continued to be used in projects. If Flash and AIR could be open sourced, we could ahve an alternative for GUI application development(nothing wrong with choice) and AIR could compete with Electron


I've had the misfortune to work with Flex - I really didn't enjoy it. It was also incredibly difficult to make Flex apps accessible in a meaningful way.

AIR did hold promise, but ultimately suffered from being terribly cumbersome to extend when you wanted to access native features or platform standard libraries (I've created several AIR native extensions over the years).

However, one of the big problems with using Flex or AIR after Adobe have end-of-lifed Flash is that you're going to lose access to the content creation tools. There are some alternatives (FlashDevelop and FDT for coding, I forget the name of the attempt at an open source Flash timeline editor), but nothing is close to being as good as Adobes own tools when you want to.

If you want to leverage knowledge of ActionScript, people may be better off looking at something like Haxe for cross platform gui app dev.


I would not start a new project in Flex/AIR but I am stil bug fixing existing projects. I use Intelli IDEA, I do not need the drag and drop GUI editor I learned how to make good layouts editing the XML and it is much more simpler to layout components using Flex(or other GUI toolkits , I used Qt and WPF ) then using divs and css(thought he new flexbox in css improved the situation), So in Flex(Qt,and others) you have components like listboxes, dropdowns, DataGrids that are very flexible , you want an item to look different you create an item renderer and reimplement the paint method in AS3 or if is simple enough in plain XML(or using the GUI Designer) In HTML5 land you do not have such flexible components as standard so you need to find an open source or buy this compoenents or implement one. You can say it is easy to implement a dropdown(built in one can't show icons and can't be customized), you hack some divs,lis and some css and it is done. But later you find that you did not consider corner cases, your implementation opens outside of the screen if dropdown is on the bottom of the page, the dropdown has no property where you can ask "show max 6 element, probably you did not add all the standard hover effects, and not implemented all the events, My point is that in Flex and other toolkits you can do a good GUI with good layout using the built in stuff and no hacks(I am thinking at css where some obvious thing did not worked for me until I found on SO that I need to set min-width to 0 so the css engine finding this useless rule would take a different path and display the expected thing). Can you share what GUI toolkits you used and what features you loved from each one? Thanks


I tend to use the native UI for each platform I work on. My general opinion of cross platform UI frameworks these days is that they for developer convenience - they don't help the users. However I'm also aware that persuading the budget holder that users are worth spending money on is hard :)

For the drop downs - when I used Flex I had the same issues over window bounds, in that nothing could extend beyond the limits of the flash rendered window. Did they fix that in later versions?


You can extend over window bounds, what I meant is the dropdown should open up or give you a property to allow you this option, also Flex dropdowns had the property you can set how many items are visible at a time. I learned Adobe AIR because my employer liked it better then Qt at that time and he also trusted Adobe products, moving from c++/Qt (and previous C# Winforms and WPF) to AIR was a downgrade in some aspects but after you learn it you can do your job and implement most of the things fast. What do you use for creating native apps for Windows, I do not know what is the prefered thing this days?




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: