On Sunday, November 2, 2003 at 2:49:51 AM UTC+5:30, Marno wrote:

I have been experimenting with various authoring tools, and have come
down to three: TADS, Hugo, and ADRIFT. I don't want to go into why I
have narrowed it down to these particular three, but one of the
criteria I am looking at is the system's popularity and use in the IF community.

Hugo looks like a very good system. But I found no Hugo entries for
the past several years' comps, and no recent (2003) games at the IF
archive. I have to wonder... why not?

Any ideas?

Mike

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

[email protected] wrote:

On Sunday, November 2, 2003 at 2:49:51 AM UTC+5:30, Marno wrote:

I have been experimenting with various authoring tools, and have come
down to three: TADS, Hugo, and ADRIFT. I don't want to go into why I
have narrowed it down to these particular three, but one of the
criteria I am looking at is the system's popularity and use in the IF community.

Hugo looks like a very good system. But I found no Hugo entries for
the past several years' comps, and no recent (2003) games at the IF archive. I have to wonder... why not?

Any ideas?

Mike

Since you have quoted a message without adding anything more, I guess that
you might just have the same question, almost sixteen years later.

As far as I know, TADS, Hugo, and ADRIFT are not "Free VMs" (although I see some conflicting information; also, Hugo seems to have other problems such
as assuming 16-bit integers on 32-bit systems, and if the license says the
file cannot be modified, that is problematic). (Free VMs include Glulx, Z-machine, OASYS, and possibly some others. Note that this is unrelated to whether the authoring system is Free software; ZILF is, but I think Inform7 isn't Free software.)

Of course, just because many other people do not use it, does not mean
that you do not have to use it yourself.

(I wonder if you can convert a Hugo story file to other VMs or implement
it in other VMs? For example, I implemented Z-machine in Glulx; maybe it
is possible to implement Hugo in Glulx?)

--
Note: I am not always able to read/post messages during Monday-Friday.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 2019-08-12, [email protected]d <[email protected]d> wrote:

[email protected] wrote:

On Sunday, November 2, 2003 at 2:49:51 AM UTC+5:30, Marno wrote:

I have been experimenting with various authoring tools, and have come
down to three: TADS, Hugo, and ADRIFT. I don't want to go into why I
have narrowed it down to these particular three, but one of the
criteria I am looking at is the system's popularity and use in the IF
community.

Hugo looks like a very good system. But I found no Hugo entries for
the past several years' comps, and no recent (2003) games at the IF
archive. I have to wonder... why not?

Any ideas?

Mike

Since you have quoted a message without adding anything more, I guess that you might just have the same question, almost sixteen years later.

As far as I know, TADS, Hugo, and ADRIFT are not "Free VMs" (although I see some conflicting information; also, Hugo seems to have other problems such
as assuming 16-bit integers on 32-bit systems, and if the license says the file cannot be modified, that is problematic). (Free VMs include Glulx, Z-machine, OASYS, and possibly some others. Note that this is unrelated to whether the authoring system is Free software; ZILF is, but I think Inform7 isn't Free software.)

Of course, just because many other people do not use it, does not mean
that you do not have to use it yourself.

(I wonder if you can convert a Hugo story file to other VMs or implement
it in other VMs? For example, I implemented Z-machine in Glulx; maybe it
is possible to implement Hugo in Glulx?)

Inform 6 and the 6/12 are under an Artistic license since a few years.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

As far as I know, TADS, Hugo, and ADRIFT are not "Free VMs" (although I see some conflicting information; also, Hugo seems to have other problems such
as assuming 16-bit integers on 32-bit systems, and if the license says the file cannot be modified, that is problematic).

The above statement is a bit confusing.

TADS is the only one of those three systems that actually uses a VM (the T3-VM), which is documented here:

https://www.tads.org/t3doc/doc/techman/t3spec.htm

As to those systems being not free, I'm not quite sure what you mean, for all three are open source (although TADS doesn't allow derivative works, it allows porting the T3-VM to other systems).

ADRIFT has become fully open source since quite a while, and the full source code is available on GitHub under BSD-3-Clause:

https://github.com/jcwild/ADRIFT-5

Hugo has also been released under BSD-2-Clause since quite some years, the full code being available in various places, including the IF Archive. Here's my own repository for Hugo sources:

https://github.com/tajmone/hugo

Possibly the confusion with Hugo being open sources is that the copyright notice was never changed in the source files, mainly for historical preservation reasons. But the license prevails on these copyright notices, and you are free to change Hugo as
you like.

Hugo seems to have other problems such
as assuming 16-bit integers on 32-bit systems

Could you please expand on that? I'm interest in learning more about these limits, since I'm working on/with Hugo right now.

Since under Windows I'm not experiencing any problems with these, I'm assuming that you are referring to limits applying to other architectures (which I'm interested in learning about).

Best regards,

Tristano

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

Tristano <[email protected]> wrote:

As far as I know, TADS, Hugo, and ADRIFT are not "Free VMs"

The above statement is a bit confusing.

TADS is the only one of those three systems that actually uses a VM (the T3-VM), which is documented here:

https://www.tads.org/t3doc/doc/techman/t3spec.htm

I was using a broader meaning by the term "VM", since I was talking about a different consideration than normally is considered by "VM".

As to those systems being not free, I'm not quite sure what you mean, for all three are open source (although TADS doesn't allow derivative works, it allows porting the T3-VM to other systems).

OK, so I was mistaken about Hugo and ADRIFT. However, since TADS does not
allow derivative works that does not qualify as open source or Free software (although perhaps a new implementation could be made by reading the documentation about the T3-VM and implementing that, maybe).

ADRIFT has become fully open source since quite a while, and the full source code is available on GitHub under BSD-3-Clause:

https://github.com/jcwild/ADRIFT-5

OK, so that is good.

However, it seems to be written in VB.NET and doesn't use Glk. (Fortunately, there is Mono to run .NET executables on non-Windows systems, and I read the documentation and it says there is a Mono Runner, so that will work, so that
is good. I personally would prefer in C or Glulx, but Mono works too.)

I also see no mention on the web site about the source code; it is found
only in GitHub as far as I can tell. (Maybe I missed it, but I cannot find it so easily, if it is there.)

I also cannot find documentation about the file format.

Hugo has also been released under BSD-2-Clause since quite some years [...]

https://github.com/tajmone/hugo

Possibly the confusion with Hugo being open sources is that the copyright notice was never changed in the source files, mainly for historical preservation. But the license prevails on these copyright notices, and you are free to change Hugo as you like.

Yes, it is confusing, but thank you for clarifying that.

Still I would want to see if Glk Hugo can be compiled into Glulx code. (I thought of some ideas about compiling C to Glulx, and there are some difficulties with some programs, such as dynamic local allocation within reentrant functions. Some other things would likely have to be altered too, such as the file handling into the Glk file handling. There is also the difficulty that the Glk "S" type does not correspond to C strings in Glulx; although most functions also have versions that use address/length which can
be used instead (this can be done automatically, so the program need not be altered to use this), this is not the case for file handling functions.)

Hugo seems to have other problems such
as assuming 16-bit integers on 32-bit systems

Could you please expand on that? I'm interest in learning more about these limits, since I'm working on/with Hugo right now.

Last time I checked I thought I noticed something like that, but I cannot
find it now. It looks like it is designed to work with 16-bit integers but allows the system to use larger integers, and would not malfunction on a
system with 32-bit or larger integers. I suppose I must have made a mistake, because what I wrote before seems to be wrong (either that, or it was the
case in some older version that is no longer in use, but has been fixed now).

So, I was mistaken. Sorry for my mistakes, and thank you to clarify!

--
Note: I am not always able to read/post messages during Monday-Friday.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

However, since TADS does not
allow derivative works that does not qualify as open source or Free software (although perhaps a new implementation could be made by reading the documentation about the T3-VM and implementing that, maybe).

Indeed, these restrictions are somewhat limiting, especially now that
many OSs are officially abandoning 64-bit support. I wish the author
might re-license TADS under a more permissive license, to ensure its
survival in time. His original intention was to prevent the proliferation of TADS variants incompatible with themselves, but this
goal can be achieved by any FOSS license that requests changing the
name of derivative software.

ADRIFT has become fully open source since quite a while, and the full source >> code is available on GitHub under BSD-3-Clause:

https://github.com/jcwild/ADRIFT-5

OK, so that is good.

However, it seems to be written in VB.NET and doesn't use Glk. (Fortunately, there is Mono to run .NET executables on non-Windows systems, and I read the documentation and it says there is a Mono Runner, so that will work, so that is good. I personally would prefer in C or Glulx, but Mono works too.)

The main problem with ADRIFT source code is that it relies on commercial third party components for the GUI, without which it won't
compile. These are rather expensive components, and to keep the
project FOSS they would have to be replaced by native VS controls
somehow.

I also see no mention on the web site about the source code; it is found
only in GitHub as far as I can tell. (Maybe I missed it, but I cannot find it so easily, if it is there.)

The new was broken out in a single brief post on the ADRIFT forum:

http://forum.adrift.co/viewtopic.php?f=14&t=11875

I also cannot find documentation about the file format.

There isn't one unfortunately, and the ADRIFT 5 Manual isn't complete
either. But the source code is well documented enough to allow re-implementing various modules in other languages if required. I had
a brief look at the file format to understand better how it compresses
XML files, etc., and my first impression is that it's quite doable
even without an official specs documentation — but of course, writing
one would make the whole endeavour easier.

Hugo seems to have other problems such
as assuming 16-bit integers on 32-bit systems

Could you please expand on that? I'm interest in learning more about these >> limits, since I'm working on/with Hugo right now.

Last time I checked I thought I noticed something like that, but I cannot find it now. It looks like it is designed to work with 16-bit integers but allows the system to use larger integers, and would not malfunction on a system with 32-bit or larger integers. I suppose I must have made a mistake, because what I wrote before seems to be wrong (either that, or it was the case in some older version that is no longer in use, but has been fixed now).

So, I was mistaken. Sorry for my mistakes, and thank you to clarify!

You're right about Hugo using 16 bit data types internally, either
signed or unsigned (depending on need), and in a recent email exchance
with Hugo author (Tessman) he mentioned that in case Hugo were to see
a new edition the first thing he'd be changing was dropping the 16-bit
integers in favour of a wider words (he even mentioned 64-bit data
types).

So chances are that these 16-bit integers might be a problem on some platforms, I'm just no aware which ones since I work mostly on x86 machines, and Hugo has been ported to so many platforms in the course
of the years that it's hard to keep track of them.

Anyhow, if you do stumble on the 16-bits limitation again, by chance,
please do buzz me for I'm interested in following it, since I'm
planning to work on a Hugo port in Rust sometime in Spring. Since
Rust can compile to multiple hardwares, I'll have to keep into account
the 16-bit issues.

Thanks again,

Tristano

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)