pouët.net

Does anyone know of any method(s) of extracting scroll-text from demos (all systems, all types) or a repository (website, newsgroup whatever) of texts that already exists? OK, so there can be a lot of guff that we can all do without (the same greets list every time? no thanks) but there's a lot of history locked away in them I'm sure (I've read a fair few, mainly Atari ST ones, in my time and have learnt things here and there) if only they can be read more conveniently. Short of starting the demo, pausing and copying down three words infinitum I personally don't know how extracting all those miss-spelt words can be done; I imagine it involves hex-editors and shit right? I'd be most grateful for any information and I may make it into a little project if it can be a simple enough task to do (I'm not that clever see). Cheers, John.

I once imagined a demosite with a big database of all scrolltexts in txt or scrollers java applets to read, from all demos ever, or slowly slowly extracting them (but it's painful) and,. ok I was dreaming of ultra cool demoscene sites I'd like to see. Because in the scrollers there can be a lot of interesting demo history and having the scrollers extracted on simple text on a website would be great resource for someone to find more funny issues on demo history.

And yes you have to find the address of the scrolltext, it could be plain text you can see with hex editor and extract easilly, it could be compressed, it could be encrypted, it could have other symbols for special fonts or not ASCII format but others the coders just came into when drunken (In one of my old demos ABCDE.. was not from 65 but started from 1 and I had 0 for space, for example), so you really have to use your knowledge and perception to rip the text. Well, except from some demos where the text is straight forward ASCII unpacked and you can easilly see it with a hex editor.

MagikGimp: You usually find the texts in plain text with a hex-editor in the demo-binary, yes. Sometimes you need to apply some magic to get special characters working right, sometimes you don't.

omg.. who needs so much text!?!?

I've been having this idea for a long time, and I even dared to write some code that automatizes some of the ripping.

By the way, here's the best scrolltext ever:

http://low.fi/~viznut/newlimits.txt

It was an easy rip. Just run the demo in the emulator, save the memory state in a file and run the UNIX command "strings" for this file. The text happened to be stored in petscii (A is 65), so this tool was able to recognize and extract it. In many other cases the text is in screen codes (A is 1) so it needs a dedicated tool to recognize automatically.

i could swear rich k's world record scroll text was available on the net somewhere..

excellent idea! perhaps demozoo.org could employ that?

If someone does a system like that, I have a crapload of ripped scrolltexts from old Amiga demos offline somewhere.. I'd be happy to upload them.

How did you rip them?

i've done a bunch of conversions for http://c64mags.untergrund.net , would be nice with a site that makes text from all sort of productions and platforms readable and searchable.

I'll consider storing the ASCII characteres instead of the bitmap offsets for my next demos.

OMG!

Enough with the databases!

I like this idea, but since someone might not want their scrolltexts to be ripped it would introduce a new challenge for coders - ripping protection for scrolltexts. Encryption isn't enough of course, since the ripper could just use OCR or type it in manually, so the font also had to be totally unreadable, both for OCR and humans.

Btw, all diskmags should be ripped too.

This has to stop!

no, it has not to stop! i'd like to see such a database, too..

Yep, diskmag rips are also nice, and perhaps we could even combine the demo lyrics collection with this project and collect some scenepoetry as well. However, what is particularly important in old scrolltexts is that they provide a lot of interesting information about the scene especially in the pre-diskmag era. Announcements, partyreports, opinions, hype about new routines, etc. etc.

And by the way, I don't see why Pouet couldn't be expanded for this. We already have the possibility to add infofiles and file_id.diz files for prods, we should just have an additional "texts" link behind which we could put the ripped texts.

MagikGimp: I do it with the magic of manual labor. Execute demo in WinUAE -> save memory dump into a file -> use your favourite text/hex editor to search for a string you just saw in the scroller -> copy/paste into a txt file. Doesn't get much easier than that.. As for fancy automation, I haven't got a clue.

Cruzor: That's hardly a new challenge, "protections" for scrollers have been around for ages.. Usually by just encrypting them in a simple fashion (XOR or something) so they can't be read directly from the memory dump. :)

Cruzor: I just understood what you said about encryption.. Sorry, early in the morning and I'm still half asleep ;)

If you're curious about scrolltexts from past releases, why don't you just watch the releases? That way you get them in the intended context as well. I know many of you are the "diskmag crowd" and as such don't care enough about demos to actually want to watch them, I just wish you'd take that one step further and ignore the scene completely (the way the scene tries to ignore you).

I know trolls are supposed to be ignored, but just this once..

For myself, the point of extracting scrollers is the ability to find strings related to subject X that interests me by doing a simple search (grep for example), instead of having to watch 75 intros/demos just for that bit of data. Outside of that, I actually do prefer reading scrollers as they're shown in the releases.

Now then, if you're religiously against ripping scrolltexts off memory (for other reasons than trolling that is), why not just ignore threads like this one and continue playing with yourself?

Now I'm curious, please tell me which ones who are in that crowd.

Influence Device: are you Timbaland?

@Influence Device: and so you are the last hardcore scener on planet earth coming to save the scene or what??

No, I'm a scener who is tired of everyone trying to document and catalogue everything I do. When I put scrolltext in a demo, that's a message to the people watching the demo. If I want something indexed or just readable outside the context of the demo, that's what readme files are for. Sure, ripping scrolltext is no different from ripping music and graphics, but THAT'S LAME TOO!

This relentless fixation that some people seem to have with building databases is why I keep changing my handle. See this means the database people always end up listing me under several names, because despite their information fetish they don't care enough about who I am or what I do/say/release to keep up. The people who do care will know. To the rest, fuckings.

I think we should have a database of who has sent fuckings to whom in the past 20 years. Then we could cross-reference that with some of all the other scene databases and see if any interesting patterns emerge.

Or maybe the demoscene is more fun if you try to be part of it.