[Scummvm-devel] Contributing to ScummVM with an interpreter for Magnetic Scrolls?
Torbjörn Andersson
eriktorbjorn at telia.com
Wed May 1 17:13:55 CEST 2019
The Legend games split the screen into fixed areas, though. The Magnetic
Windows GUI was basically a window manager (in 640x480 pixels), with
movable, resizable, potentially overlapping windows. Here is an
admittedly horrifying screenshot of what it could look like if you
opened all the possible windows:
http://www.update.uu.se/~d91tan/magnetic-windows.png
I defy anyone to find a sensible layout for all of that! You may think
you can do it, if you close the "Compass", "Inventory", and "Items in
room" windows, and for a while you can... and then you realize that not
every image is the same size, and you have to move the "Graphpics"
window out of the way...
Torbjörn Andersson
On 01/05/2019 08.53, Paul Gilbert wrote:
> Thanks for information Torbjörn. It doesn't sound like the missing
> screen elements are things I'd miss all that much. Though maybe some of
> it could eventually be reintroduced as an option. Whilst Glk doesn't
> currently have any "listbox" window types, I do have an already
> implemented listbox control for my stalled Legend engine that could
> probably be adapted. And a main menu implementation could be useful for
> Frotz v6 as well as the Magnetic engine.
>
> It will be nice if they do release the original versions alongside the
> remastered versions. Being away from home, I'm not sure how many
> Magnetic Scrolls games I have, if any.. comes from having too large a
> collection.. I'm sure Strangerke among others can empathize :), So if I
> don't have all of them; it would be nice to eventually have a legitimate
> source to get any missing ones.
>
> Paul.
>
>
> On Wed, May 1, 2019 at 4:00 PM Torbjörn Andersson
> <eriktorbjorn at telia.com <mailto:eriktorbjorn at telia.com>> wrote:
>
> From what I remember, the main feature not supported by the Magnetic
> interpreter is some of the "Magnetic Windows" interface (on-screen map,
> compass rose, separate windows for inventory and room objects, horribly
> cluttered windowing system, ... oh, sorry, that last thing isn't really
> a feature, is it?) used by Wonderland and the Magnetic Scrolls
> Collection.
>
> At least as long as the front end implements the necessary hooks for
> displaying static and animated pictures, etc. I'm not sure if it loads
> the original data files, or if they have to be specially prepared. (The
> files on my Magnetic Scrolls Collection CD certainly don't look like
> anything the emulator would recognize.)
>
> Also, it doesn't try to emulate the original look of the games with its
> animated menus and pull-down graphics. Which is a bit of a shame
> or,, in
> the case of the Magnetic Windows interface, a huge relief.
>
> When I asked [1] on the Strand Games forum (the company in the process
> of remastering the Magnetic Scrolls games, after the source code was
> recovered from a backup tape [2]) if they were planning on release the
> original versions as well, just for nostalgia, I got a reply from Hugh
> Steers referring me to Stefan Meier's "Magnetic Scrolls Memorial" for
> data files for the emulator.
>
> Since "Strand Games was started by Hugh Steers - a founding member and
> core developer of Magnetic Scrolls and Stefan Meier, curator of
> if-legends.org <http://if-legends.org>" and they claim to have the
> support of "several members
> of the original Magnetic Scrolls team" [3], I guess those files are
> kiiind-of officially endorsed...? It's probably worth asking them again
> later.
>
> [1]
> https://strandgames.com/community/discussion/65/the-original-versions
> [2]
> https://strandgames.com/blog/magnetic-scrolls-games-source-code-recovered
> [3] https://strandgames.com/about
>
> Torbjörn Andesson
>
> On 01/05/2019 05.10, Paul Gilbert wrote:
> > Hi Thomas,
> >
> > I'm not sure of the status of the Gargoyle Magnetic Scrolls
> sub-engine,
> > and I don't currently have any Magnetic Scrolls games handy to
> test it
> > on. But unless it's significantly lacking in support for the
> later games
> > that your own work adds support for, it'll likely prove easier to
> work
> > on converting it, I'm afraid. Since, after all, someone else has
> already
> > done the work of hooking it up to Glk as well. So it'll be easier
> to do
> > minimal work (relatively speaking) to hook it up to the ScummVM
> version
> > of the Glk layer. And then I can keep testing it as I properly
> objectify
> > everything to make sure I don't break anything along the way.
> >
> > Though I'll be unlikely to be spending much time beyond basic
> polishing
> > up of the sub-engine afterwards. So based on your experience with
> the
> > games & engine involved, there might be opportunities afterwards
> for you
> > to contribute if you're interested, and to do pull requests to
> improve
> > the functionality of the engine based on your work.
> >
> > Regards,
> >
> > Paul.
> >
> >
> > On Wed, May 1, 2019 at 2:55 AM dettus <dettus at dettus.net
> <mailto:dettus at dettus.net>
> > <mailto:dettus at dettus.net <mailto:dettus at dettus.net>>> wrote:
> >
> > Hello.
> >
> >
> > I must admit that I have not heard about Gargoyle or GLK before.
> >
> > Judging from the github repository, the Gargoyle engine for
> Magnetic
> > Scrolls is using the original magnetic interpreter from Niclas
> > Karlson. What he did was, as far as I know, reverse-engineer
> the C64
> > ports of the Magnetic Scrolls games in a painstaking process
> to make
> > them run on DOS. What I did was starting from scratch and
> using his
> > code as a reference. His code is probably more settled than mine.
> > Plus, I'd hate to step on somebody's toes who has already
> invested
> > copious amounts of time and energy into a project, so if you
> prefer
> > to use his code: Please, please do so!
> >
> > What I can offer you is an engine using a more modern coding
> style.
> > Even though it is C, I am sure that it can be integrated
> easily with
> > C++ projects. I have worked with projects that use both languages
> > and I am experienced in designing APIs for that.
> >
> >
> > Thus far, I have the following function pointers for callbacks:
> >
> >
> > typedef int (*cbLineAOutputChar)(void* context,char
> c,unsigned char
> > controlD2,unsigned char flag_headline);
> > typedef int (*cbLineAOutputString)(void* context,char*
> > string,unsigned char controlD2,unsigned char flag_headline);
> > typedef int (*cbLineAInputString)(void* context,int*
> len,char* string);
> > typedef int (*cbLineADrawPicture)(void* context,tPicture*
> > picture,int mode);
> > typedef int (*cbLineASaveGame)(void* context,char* filename,void*
> > ptr,int len);
> > typedef int (*cbLineALoadGame)(void* context,char* filename,void*
> > ptr,int len);
> >
> > As well as those API functions:
> >
> > // configuration functions
> > int lineA_getsize(int* size);
> > int lineA_init(void* hLineA,void* pSharedMem,int
> > *sharedmemsize,void* pMag,int magsize,void* pGfx,int gfxsize);
> > int lineA_getVersion(void* hLineA,int* version);
> > int lineA_setCBoutputChar(void* hLineA,cbLineAOutputChar pCB,void
> > *context);
> > int lineA_setCBoutputString(void* hLineA,cbLineAOutputString
> > pCB,void* context);
> > int lineA_setCBinputString(void* hLineA,cbLineAInputString
> pCB,void*
> > context);
> > int lineA_setCBDrawPicture(void* hLineA,cbLineADrawPicture
> pCB,void*
> > context);
> > int lineA_setCBSaveGame(void* hLineA,cbLineASaveGame
> pCB,void* context);
> > int lineA_setCBLoadGame(void* hLineA,cbLineALoadGame
> pCB,void* context);
> >
> >
> > You typically initialize the engine with pointers to the game
> > binaries, set the function pointers, and let it run. When it
> reaches
> > a point where an input is required, the engine generates a
> callback
> > and pauses until the callback returns.
> >
> >
> > Regarding pictues: They are also being rendered internally, and
> > another callback is triggered. An upper layer must draw it in
> > whatever way it deems appropriate. So, it does not really
> care where
> > it sits, or if it will even be displayed. The data format
> itself is
> > very basic, almost .xpm.
> >
> >
> > typedef struct _tPicture
> > {
> > unsigned int palette[16];
> > int height;
> > int width;
> > char pixels[65536];
> > } tPicture;
> >
> >
> > I am sure I can design an extra function for animations, such
> as this:
> >
> > typedef int (*cbLineADrawAnimation)(void* context,tPicture*
> > frames[],int mode,int framenum,int delay);
> >
> >
> >
> > Please let me know if this is of interested to you.
> >
> > Thomas
> >
> >
> >
> > On 4/30/19 2:20 AM, Paul Gilbert wrote:
> >> Hi guys,
> >>
> >> Thanks for the excellent summary of the current status,
> Filippos.
> >> As he said, I'm currently in the middle of working my way
> through
> >> converting all the Gargoyle engines to ScummVM, cleaning them up
> >> and making them conform to our coding requirements, as well as
> >> making them proper C++ classes. If you do decide to work on the
> >> Gargoyle sub-engine, please let me know.. I'll hold off
> working on
> >> it until after I've converted the remaining sub-engines. That'll
> >> give you plenty of time.. the TADS & Hugo sub-engines are big
> >> enough that that'll likely take me a while to finish
> converting :).
> >>
> >> If you have any questions about the Glk engine specifically,
> feel
> >> free to drop me a line. There's also the technical forums at
> >> intfiction.org <http://intfiction.org> <http://intfiction.org>.
> Though my schedule is a
> >> bit packed for the next few weeks, so if you do email, I may
> take
> >> a bit to get back to you.
> >>
> >> As far as music and animation goes, music at least should be
> easy,
> >> so long as it's a standard format ScummVM understands. Currently
> >> the engine supports raw sound, wav files, aiff, and mp3. As for
> >> graphics, that's somewhat straightforward as well, as long
> as the
> >> graphics can be segmented into completely separate windows than
> >> the text, and it looks like the Magnetic Scroll games do.. when
> >> you create windows in Glk, you can specify what portion of the
> >> screen each will use (for text vs graphics). I've had a lot of
> >> trouble with the Frotz sub-engine, but only because graphics and
> >> text can intermingle in the same windows, which Glk doesn't
> allow.
> >>
> >> Regards,
> >>
> >> Paul Gilbert (aka DreamMaster).
> >>
> >>
> >>
> >>
> >>
> >> On Mon, Apr 29, 2019 at 7:18 PM Filippos Karapetis
> >> <bluegr at gmail.com <mailto:bluegr at gmail.com>
> <mailto:bluegr at gmail.com <mailto:bluegr at gmail.com>>> wrote:
> >>
> >> Sorry, forgot to say that the glk engine in ScummVM can be
> >> found here:
> >> https://github.com/scummvm/scummvm/tree/master/engines/glk
> >>
> >> There are currently subengines for the alan2, frotz, glulxe,
> >> magnetic, scott, and tads interpreters, in various stages of
> >> development. dreammaster is mainly working on these. You can
> >> also find information about the glk engine in our wiki:
> >> https://wiki.scummvm.org/index.php?title=ScummGlk
> >>
> >> Regards
> >> Filippos Karapetis
> >>
> >> On Mon, Apr 29, 2019 at 12:14 PM Filippos Karapetis
> >> <bluegr at gmail.com <mailto:bluegr at gmail.com>
> <mailto:bluegr at gmail.com <mailto:bluegr at gmail.com>>> wrote:
> >>
> >> Hello, and glad to hear about your interest in ScummVM,
> >> and your work on the Magnetic Scrolls games!
> >>
> >> It's great that you are working on this. Bear in mind,
> >> though, that there has been extensive work on the
> Magnetic
> >> Scrolls games, which has been integrated into the
> Gargoyle
> >> interpreter:
> >> http://ccxvii.net/gargoyle/
> >>
> >> Gargoyle is being integrated as a subengine into
> ScummVM,
> >> in the glk engine. The glk engine is a port of the
> >> Gargoyle engine into ScummVM. There's an extensive wiki
> >> for Gargoyle, here:
> >> http://ifwiki.org/index.php/Gargoyle
> >>
> >> So, my advice would be to make your engine part of
> the glk
> >> engine, and base your work on the already working
> Magnetic
> >> Scrolls interpreter from the Gargoyle engine. The
> >> interpreter is open source, and can be found here:
> >> https://github.com/garglk/garglk/tree/master/terps/magnetic
> >>
> >> The list of supported Magnetic Scrolls games can be
> found
> >> here:
> >>
> https://github.com/garglk/garglk/blob/master/terps/magnetic/Generic/games.txt
> >>
> >> And there's an archive about interactive fiction games,
> >> where a lot of resources about the Magnetic Scrolls
> games
> >> can be found, here:
> >> http://www.ifarchive.org/indexes/if-archive/magnetic-scrolls/
> >>
> >> Thanks for your interest in these games! It would be
> great
> >> to see your work on them :)
> >>
> >> Regards
> >> Filippos Karapetis
> >>
> >>
> >> On Mon, Apr 29, 2019 at 10:08 AM dettus
> <dettus at dettus.net <mailto:dettus at dettus.net>
> >> <mailto:dettus at dettus.net
> <mailto:dettus at dettus.net>>> wrote:
> >>
> >> Hello.
> >>
> >>
> >> My name is Thomas, I am a big fan of ScummVM.
> >>
> >> Currently, I am in the process of writing an
> >> interpreter for classic
> >> text adventures from Magnetic Scrolls, such as "The
> >> Pawn" or "The Guild
> >> Of Thieves".
> >>
> >> Even though I am not saying that it SHOULD be
> >> integrated into ScummVM, I
> >> would like to think that it COULD. It is WAY to
> early
> >> to really start
> >> integrating it, but I like to be prepared.
> >>
> >>
> >>
> >> So, the reason why I am writing this mail is
> because I
> >> would be
> >> interested as on how I would have to design my
> code to
> >> make it possible.
> >> Currently, my interpreter needs a couple of
> hooks into
> >> any form of GUI:
> >>
> >> - Initialization, where pointers to the original
> >> binaries are being provided
> >>
> >> - Callbacks for textual input (strings)
> >> - Callbacks for textual output (strings)
> >> - Callbacks for textual output (characters)
> >> - Callbacks for drawing a picture (XPM)
> >>
> >> My code is purely C, so callbacks I am using
> function
> >> pointers. Said
> >> functions are, at the moment, doing nothing more
> than
> >> just a simple
> >> printf() and fgets(). The games I can play so
> far are
> >> "The Pawn", "The
> >> Guild of Thieves" and "Jinxter". HOWEVER, later
> games,
> >> such as
> >> "Corruption", "Myth", "Fish!" and "Wonderland" added
> >> features such as
> >> animation and music.
> >>
> >> SO MY QUESTION is this: How would I best design the
> >> callbacks for
> >> animation and music in way so that it MIGHT be
> >> integratable into ScummVM?
> >>
> >>
> >>
> >> Dettus
> >>
> >>
> >>
> >> P.S.: My interpreter is available at
> >> http://www.dettus.net/dMagnetic.
> >>
> >> (Hopefully this link, and the fact that this is my
> >> first email will not
> >> make this spam. ;) )
> >>
> >>
> >>
> >>
> >>
> >> _______________________________________________
> >> Scummvm-devel mailing list
> >> Scummvm-devel at lists.scummvm.org
> <mailto:Scummvm-devel at lists.scummvm.org>
> >> <mailto:Scummvm-devel at lists.scummvm.org
> <mailto:Scummvm-devel at lists.scummvm.org>>
> >> https://lists.scummvm.org/listinfo/scummvm-devel
> >>
> >>
> >>
> >> --
> >> "Experience is the name every one gives to their
> mistakes"
> >> - Oscar Wilde
> >>
> >>
> >>
> >> --
> >> "Experience is the name every one gives to their mistakes" -
> >> Oscar Wilde
> >> _______________________________________________
> >> Scummvm-devel mailing list
> >> Scummvm-devel at lists.scummvm.org
> <mailto:Scummvm-devel at lists.scummvm.org>
> >> <mailto:Scummvm-devel at lists.scummvm.org
> <mailto:Scummvm-devel at lists.scummvm.org>>
> >> https://lists.scummvm.org/listinfo/scummvm-devel
> >>
> >>
> >> _______________________________________________
> >> Scummvm-devel mailing list
> >> Scummvm-devel at lists.scummvm.org
> <mailto:Scummvm-devel at lists.scummvm.org>
> <mailto:Scummvm-devel at lists.scummvm.org
> <mailto:Scummvm-devel at lists.scummvm.org>>
> >> https://lists.scummvm.org/listinfo/scummvm-devel
> > _______________________________________________
> > Scummvm-devel mailing list
> > Scummvm-devel at lists.scummvm.org
> <mailto:Scummvm-devel at lists.scummvm.org>
> <mailto:Scummvm-devel at lists.scummvm.org
> <mailto:Scummvm-devel at lists.scummvm.org>>
> > https://lists.scummvm.org/listinfo/scummvm-devel
> >
> >
> > _______________________________________________
> > Scummvm-devel mailing list
> > Scummvm-devel at lists.scummvm.org
> <mailto:Scummvm-devel at lists.scummvm.org>
> > https://lists.scummvm.org/listinfo/scummvm-devel
> >
>
>
> _______________________________________________
> Scummvm-devel mailing list
> Scummvm-devel at lists.scummvm.org
> https://lists.scummvm.org/listinfo/scummvm-devel
>
More information about the Scummvm-devel
mailing list