[Hopperxx] converting .dir and/or .dcr to HTML5 and/or other languages

[Paul Catanese]

Dear Jim-
Just wanted to pipe up that I'm fascinated and very interested in this
direction; it would be great if it could bear fruit.  I don't have time to
focus on this deeplyy, but I am cheering you on, and wanted you to know
it, and wanted to share some thoughts. The idea that an old director
project could be translated to html5 seems like a great dream. Would love
to see it.
I was not aware of the projectorrays project - this certainly could be of
interest.

I was also fascinated about Danny's comment about .dir's being.zip's - I
didn't know that, but I had effectively shifted away from Director by
around 2006.
Googling a bit, I did come across this post that speaks to the internal
structure as an RIFX container.
https://nosamu.medium.com/a-tour-of-the-adobe-director-file-format-e375d1e063c0
This direction is a bit lower-level than I'd want to get mixed up in, but
could be useful in thinking about what needs to be extracted.
There's a few references at the bottom of the document that may also be of
interest, including the projectorrays project.

Please forgive some preaching to the choir, but wanted to jot down some
general thoughts:
* Recalling that some projects used 1-frame loops w/parent-child scripts,
some used "timeline only", and majority used a hybrid approach e.g.
approaches where instances carried behavioral scripts - I would suspect
that these different authored-project architectures might require differing
translation approaches for HTML5.
* Large multi .dir projects would likely derive back down to one of those
cases.
* Functionalities provided by 3rd party xtras seem like another area of
particular wrinkles to be unwound, as would some internal xtras like
quicktime.
* webgl should certainly be robust enough to handle shockwave 3d scenegraph
& related functionalities

*emulation?*
In the net.art context, I had spoken with the Rhizome folks years ago, and
their approach to preserving (flash) works was the emulation as a service
model.  They also did some interesting preservation techniques for director
CD-Rom's in which old hardware would run projects natively, and then
results streamed to clients - but the costs were personally way out of
reach. I'm not even clear if they are still involved in anything like that.
The approach you're talking about is a refreshing new angle for director
works. That said, I did start to wonder what it would mean to have a
software emulator, perhaps even a browser-based emulator that could run
director projects, rather than converting them.

*Bigger picture still*
Jon Ippollito & Richard Rinehart's book: Re-Collection offers another
thought - that new media art projects might be better considered as
notation based (e.g. like music or theatrical productions) - so that
recreating media artworks with new tools is part of their nature. I was at
the WRO art center in Poland a few weeks ago, and they had recreated pieces
like Camille Utterback's "text rain",  as well as an early Paul Sharits
piece, and a rutt-etra style video rasterization demo, among others, that
fit into this framework - check them out here <https://wrocenter.pl/pl/mi/>.
The guy who did the heavy lifting: Pawel Janicki might be interested in
your project.

-paul


On Sat, Jan 24, 2026 at 7:46 PM Jim Andrews <jim at vispo.com> wrote:

> Hi Danny,
>
> I expect you already know about ProjectorRays:
> https://github.com/ProjectorRays/ProjectorRays . It isn't fantastic but
> it's something.
>
> This friend of mine I mentioned says "I never used copilot. I am in the
> IntelliJ ecosystem - WebStorm. They have in there "junie" coding agent,
> which as far as i know built on top of Claude model."
>
> People are raving about Claude Code. Anyway, my friend seems to think that
> this sort of AI could well convert something like .dir files to HTML5. I'm
> not sure he appreciates the difficulty of the problem--though he is a pro
> programmer. Maybe he does.
>
> Anyway, I suspect that if we could get some specs and whatnot on Director,
> that would help in the process.
>
> All of us spent eons creating work with Director. Anyone else interested
> or have info to share on this question??
>
> ja
> On 2026-01-24 3:27 p.m., Danny Kodicek wrote:
>
> Adobe / Macromedia always insisted there was no way to decompile .dcrs
> (although that always sounded unlikely to me). But IIRC .dirs were like
> .docs - just zip files with a different name. Theoretically I think you
> should be able to rename them to .zip and open them up. That might only be
> true for later editions of Director, though - and it might not be true at
> all! Maybe I'm just making it up...
>
> On Sat, 24 Jan 2026 at 21:50, Jim Andrews <jim at vispo.com> wrote:
>
>> Hello dear Hopper folks,
>>
>> I have a question and issue to raise with you about conversion of
>> Director stuff to HTML5 (or other formats).
>>
>> I am a newbie at using AI in my programming practice--just started the
>> day before yesterday with Copilot in VSCode. So far so good.
>>
>> But a friend of mine is intent upon using more powerful AI tools to
>> convert the Arteroids (https://arteroids.vispo.com) source code .dir
>> into html5.
>>
>> Not surprisingly, the thing he's having problems with is the dir format.
>> There is a published format for SWF, and there's a decompiler for
>> Director, but I guess AIs are going to have big problems trying to
>> 'translate' dir or dcr files.
>>
>> Has anyone approached Adobe about making some (at least) documentation
>> about Director public so we can try to salvage and convert the IMMENSE
>> work many of us did in creating Director work?
>>
>> And what is out there now, if anything at all?
>>
>> ja
>> https://vispo.com
>>
>>
>> _______________________________________________
>> Hopperxx mailing list
>> Hopperxx at vispo.com
>> http://vispo.com/mailman/listinfo/hopperxx_vispo.com
>>
>
> _______________________________________________
> Hopperxx mailing listHopperxx at vispo.comhttp://vispo.com/mailman/listinfo/hopperxx_vispo.com
>
> _______________________________________________
> Hopperxx mailing list
> Hopperxx at vispo.com
> http://vispo.com/mailman/listinfo/hopperxx_vispo.com
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://vispo.com/pipermail/hopperxx_vispo.com/attachments/20260125/8648a0d4/attachment-0001.html>