Hi there!

It's a looong time since I posted something on AmigaWorld, as these days I am
not a big poster as I used to be in the BBS days [1980-1990's]... but anyways...

Also I do not know if it was already talked about recently, or maybe its the
1001th time this subject is brought up, but I was wondering about the
AmigaDE/AmigaAnywhere system and technology...

Once it was very popular and then Pfuitt! It just seemed to disappear on this
side of the planet... I have seen here that there are people that have some
software developer stuff to work on this, even AmigaDE version 2 [whatever that
is?]

If the "technology" was so popular, just why did all things fell apart, so to
speak? Who had and still have the legal rights on this technology? Seems to me
that it was something apparented to what Apple got on their Iphone/Ipad/Iwhatever
but as I saw it AmigaAnywhere was around 10 years ago AND seemed to be superior
to what is seen nowadays...

I for one, would VERY like to see something be done with AmigaAnywhere

Another thing: being a Canadian, I saw that RIM (Research In Motion) a local
company, had some troubles these past months... I think that if it would be
feasable, maybe adapting AmigaAnywhere for their stuff [Blackberry and all]
would be something to "start them up" again...?

Anyways, I may be preaching in the desert, but don't you think there are some
things that could be done?

Lastly, people on these forums that still dabble on the AmigaDE/
AmigaAnywhere, can something be done so we could discuss what could be done
with it all...? I would also be very interested in developing software
(games, utilities, whatever) AmigaDE compatible...

Anyways, lets talk about this, because it would be sad to not do something
with this kind of "platform" that seemed so promising...

Okay, before I bother you more with that, I await for your comments...

See ya all!

M.

@Megamania

"Very popular" compared to what, Appstore?

I guess it was "popular" on Amiga sites because the name contains "Amiga".

Basically it died.....and that was a good thing.

@Megamania

It kindof felt unsupported when it was relevant. Something about ncurses, Linux changed and Amiga Inc. called it a bug, refusing to follow the rest of the world. To me, it kindof sunk at that point as no one wanted to go find such an old Linux to run the dev kit on. I thought it was a decent idea, but should have been pushed a bit further rather than make claims that programming languages would fill in blanks such as memory protection. I think the original company that made the underlying engine folded a few years ago as well, so there was nothing left for DE to be based on.

Amiga Inc. is maybe still doing something, some game kit for BlackBerry devices, I doubt that's related to DE, but I'm not sure what it might actually be. But BlackBerry sounds like a quickly sinking ship, so I don't expect that to be a long-term Amiga Inc. product line.

_________________
All glory to the Hypnotoad!

@Megamania

Tao Group (the makers of intent, which is essentially all AmigaDE/AmigaAnywhere is) no longer exist. That doesn't necessarily stop Amiga Inc from using and building on what they have already, but AFAIK they no longer have any form of development team.

Whilst we're here, I started a thread recently asking if anybody has the intent 2 SDK (or at least something newer than the PartyPack version), as I'm looking for a tool that can extract and create Tao Native Archives (for a curiosity, non-Amiga-related project). If anybody does please can you PM me.

_________________
"Miracles we do at once, the impossible takes a little longer" - AJS on Hyperion
Avatar is Tabitha by Eric W Schwartz

@Megamania

Hi. It's nice to see someone else interested in Taos.

It is actually my primary programming focus.
I'm hoping to get back into it next year as last year was extremely busy.

It appears true that Tao ceased to be in its original operating form.
Last I knew, a while back, it was being run by administrators but I don't know for what purpose.
The people from Tao seem to have created instead http://antixlabs.com/
I don't know how that relates to the original idea if it does at all?

But, I like Taos. I'm using 1.5 which was generously provided to me by Francis Charig quite a while ago now. I'm not aware of an intent 2.0 being available?

I've created my very own object model for it even though it comes with an object model already. Originally I was provided with some specifications for the Amiga Objects model by Fleecy and some non-working example code. Apparently a Jesse was developing it for them.
It seemed to turn up later on the web as a separate project. I can't even remember what Amiga was calling their object model. Does anyone remember?

What I have developed looks nothing like what they were developing. It has gone through many versions and iterations and still has more to go to add in other features I wish to add to it. In many respects it doesn't look anything like other object models.

I'm also looking to develop my own compiler for the system. That is my primary goal for next year. There are things I want to add such as being able to start a new tool (callable code for the unknowing here) while another tool is still in the process of being created. At the moment this can't be done. I want to do this as part of improving my object model.

I saw that a fellow developed Reggae over at http://krashan.ppa.pl/mph/what-is-reggae-framework which was born out of the Amiga datatypes idea. It is a good solution but I believe that I will be able to produce a better solution with the Taos system, my object model and my new compiler when its done.

I think Taos is a great system that just needs some changes to be made to it.
That is what I am seeking to do.

Last edited by gonegahgah on 04-Dec-2012 at 04:41 AM.

@gonegahgah

" I can't even remember what Amiga was calling their object model. Does anyone remember?"

Fleecy wrote something about "Object Sea" once... But perhaps it's not the "name" you were looking.


Other than that.
Amiga Inc failed to deliver AmigaAnywhere.

But AmigaAnywhere exist via AmigaForever (AOS3 anywhere) and via HollyWood (compile once run everywhere).
Community will succeed where AmigaInc failed? Perhaps they should have listened the community in y 2000...2001....


+It seems there are some efforts to enable CPU agnostic binaries, at least for AROS devices.

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

@gonegahgah

I'm pretty sure the Amiga objects were just called AmigaObjects

Can you have a look to see if Intent 1.5 has the following:
/dev/tool/archive/util/jar2archive

If so, please can you list what else is in /dev/tool/archive[/util]?

Last edited by Chris_Y on 04-Dec-2012 at 01:59 PM.

_________________
"Miracles we do at once, the impossible takes a little longer" - AJS on Hyperion
Avatar is Tabitha by Eric W Schwartz

@KimmoK

I just had a look back through my old emails from back then.
They were calling it the Amiga Component Model (ACM).

Here's a web site where Fleecy was saying that they were close to releasing it in a couple of weeks back in 2000!
http://www.amiga.org/forums/archive/index.php/t-41877.html

I'm fairly certain that it ended up as the Advanced Foundation Classes at http://afc.sourceforge.net/ which appeared not long after in 2001.

Here's another blog message from Fleecy about 3 months after the AFC appeared:
http://tech.groups.yahoo.com/group/amigaone/message/1109

I recall that they had seemed to lose touch with Jesse who was working on the ACM for a time. I don't know why that happened.

As they didn't provide me with a working component model I wrote my own. However, through a series of evolutions I scrapped most of the approach that they were describing anyway and created something that I believe is more Tao like. I think my approach is ultimately also more programmer friendly while being much simpler overall.

@Chris_Y

I've had a look. It does contain that.

The other things in there are:
_new.00, api.html, class.00, class.fixup.00

The other things in the util directory are:
build.00, make.00, output.00, options.00, xjar2archive.00

@gonegahgah

Hmm, I was hoping for an unarchive of some description. Maybe that is elsewhere in /dev/tool?

Is it possible to send me a copy of the SDK? I understand if you don't want to do this, but it's difficult to obtain software from dead companies (and I can't find it on the web either)

_________________
"Miracles we do at once, the impossible takes a little longer" - AJS on Hyperion
Avatar is Tabitha by Eric W Schwartz

@Chris_Y

Most archivers usually do both don't they? Is this an exception?

If not I would have thought that maybe the xjar2archive would perhaps have been the opposite of jar2archive? Is it something different?

Probably the best guy to ask would the Francis Charig himself. He was quite a friendly fellow when I was emailing with him way back. You should maybe be able to get onto him at the Antix site: http://antixlabs.com/

@gonegahgah

Quote:

gonegahgah wrote: @Chris_Y Most archivers usually do both don't they? Is this an exception? If not I would have thought that maybe the xjar2archive would perhaps have been the opposite of jar2archive? Is it something different?

I don't know, I don't have the two tools so am unable to play around with them. jar2archive converts a .jar file to a native archive, so I wouldn't expect it to convert back, but to use archive2jar for that purpose.

I can't find an email address for Francis Charig.

Last edited by Chris_Y on 04-Dec-2012 at 03:03 PM.

_________________
"Miracles we do at once, the impossible takes a little longer" - AJS on Hyperion
Avatar is Tabitha by Eric W Schwartz

@Chris_Y

If you try http://antixlabs.com/contact/contact.php then maybe Francis will answer your query about those archive tools or maybe he will put you onto Chris Hinsley the principle programmer. You can only but try... I only have Francis' old Tao-Group email.

@Thread

What I would like to see is the Sheep language released in a form or another, at least if it can be used on OS4.

_________________
Philippe 'Elwood' Ferrucci
Sam460 1.10 Ghz
AmigaOS 4 betatester
Amiga Translator Organisation

@elwood

Sheep looks like a basic variant.

http://strlen.com/sheep/sheep.png

If you like nice scripting languages try looking at Ruby, it handles arrays, hash’s and strings extremely well.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@scripting

AREXX, Python, Ruby, Sheep, Lua/HW,....
Too many options..

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

Although it's not a scripting language I always think it is sad that fewer people took the time to get to know the Virtual Processor (VP) language supplied with intent. It is quite clever in the way it operates. For example, it allows more than a single value to be returned by a coded function in such a straight forward manner without having to 'hack' work-a-rounds. For example you can code:

qcall /mystuff/myapp/dosomething,(iArg1,iArg2:iAnswer1,iAnswer2)

Just to show an example of why you would want this you might do:

qcall /mystuff/myapp/getmouse,(-:iMouseX,iMouseY)

In this example you get both the X and Y values returned.
That makes a great deal more sense then having to pass structures C fashion to get back more than one value at once.

Also VP allowed for the creation of new 'commands' easily with the use of the .macro command. You can craft your own shortcuts which is wonderful.

Despite all the things I find wonderful about intent VP, it is my view that some of the implementation, although well thought out, is still a tad adhoc on some things. As part of my writing a replacement compiler I have been generalising those aspects.

For example VP uses the % symbol to indicate that a compile time value is to be expressed. You could do things like:
.eval 3 + __counter
printf '%e\n'
where %e was replaced by the value evaluated previously by the 3 + __counter.
Nothing wrong with this but as I said I wanted to generalise the idea further.

I have done a whole design around replacing these single letter replacers with multi-letter replacers. So for example you could have printf '%sqr(6)\n' which would expand to printf '36\n' and then compile as that.

This particular improvement will require some fixes to be made to the supplied include files because at the moment the compiler assumes that only a single letter follows the % symbol and the include files were written with that assumption.

So if you have %Tabc and %T evaluates to xyz then this will be treated in the current compiler as xyzabc. Under my compiler the entire %Tabc will be treated as a label to be replaced. If it has no set value then it will expand to Tabc. If the value of %Tabc is la-di-da then it will expand to that. This will happen whether %T has a value or not.

I personally believe that this byproduct requirement to upgrade the include files is way worth the effort because of the usefulness of the increased utility and generalisation this allows. It was fun working out all the ins and outs of this.

I'm working on making the VP compiler even more outstanding than it already is in my opinion. There are many more improvements that will be added than just those explained so far.

Last edited by gonegahgah on 05-Dec-2012 at 02:10 PM.
Last edited by gonegahgah on 05-Dec-2012 at 02:01 PM.
Last edited by gonegahgah on 05-Dec-2012 at 01:57 PM.
Last edited by gonegahgah on 05-Dec-2012 at 01:56 PM.
Last edited by gonegahgah on 05-Dec-2012 at 01:55 PM.
Last edited by gonegahgah on 05-Dec-2012 at 01:51 PM.
Last edited by gonegahgah on 05-Dec-2012 at 01:49 PM.

Another nice thing about the VP language - although nobody seems to deeply care - is that it had this idea of two streams of commands interweaving with each other. Something like a hospital which has staff corridors and public corridors that weave around each other.

In VP this took the form of commands that logically emit compiled code and commands that didn't emit compiled code. For example cpy iFrom, iTo would emit code but .define MYDEF s-o-m-e-t-h-i-n-g didn't actually generate anything at the time.

There are exceptions to this such as the record command for starting a structure that don't emit code but I guess this made the language look more conventional.

You can see the difference more clearly between the two different commands .if and if.
The .if would tell the compiler to only generate the following code if the conditions were met whereas if emits actual branching code.

When you create your own commands using .macro you can use both forms ie. with a leading dot or without ie. .macro mymacro or .macro .mymacro. Again it is best to follow the convention of using the dotted form when no code is actually emitted. The two forms are distinct from each other.

You could see this also in some of their in-compiler functions though this is an area that is a little adhoc to my mind and which I want to generalise. At the moment they have .defined(mydef) which returns true if mydef was defined with .define and ?def(mydef) which returns true if it was set via mydef = 99, mydef := 99 or mydef:.

Instead I intend to change these to ?defined() and ?set(). I like the convention of using the ? leader to indicate that the function returns true or false.

I will also be making it possible to create your own in-compiler functions. At the moment you can do this using .define as for the c language but to my mind it is too limiting. I will be adding something much more empowering. For macro commands I will still only support dotted and undotted commands but for functions I will support question mark, maybe dotted and still non-symbol lead names.

Seeing .if ?dbg is much more explanatory than seeing .if .dbg.

I'm also making it possible to create your own compile flow controllers which is near impossible in VP at the moment (or probably any language). For example the above will be able to be made into the macro .ifdbg with my new compiler if you chose to.

Intent was a nice small footprint plaform.
That was called amigaanywhere (basically all the stuff from TAO group, and some more peaces of software from Ainc/tao/thirst party contactors).

That was a project venture, that was just a bit ahead of it's time (like Amiga OS when it was released in 80's), meanwhile there was a lack of many things. The idea was good (be on every handheld devices), but Ainc could not complete amigaanywhere with what was missing to be the killer thing (money, money, money).

At that time I had some Mb of self written code on this platform, and to be honnest, that was a pleasure to write code dor tao's system.
Maybe I still have some SDK package, this is something I have to check.

iOS came and killed a lot of things, then android, now the whole thing is completely outdated.
Then TAO went bankrupt, and that was clear AmigAnywhere could not evolve more without being able to adapt and modify tao technology.

Amiga Inc done a great job to try redoing the thing from scratch, and they done a good attempt with amigaanywhere v2 (apparently, that was mainly work from fleecy), that thing seemed to me to be a "game layer" for handheld devices (although I haven't used it a lot), certain developers managed to adapt their amigaanywhere games to amigaanywhere v2 apparently without big headhaches.

And there again, money, money money, that was frankly an excellent try (amigaanywhere v2), but then main investor in amiga inc died, and then a complete silence came. With a lot of things still to be added (network support for example, relationnal database, and many other things), Amiga was certainly out of money to finish the product.

I think they just continued improve their amigaanywhere v2 game layer, and this is probably now the basis of the game they release on blackberry, but I may be completely wrong there, this is a complete speculation of my own.

Oen thing I can say for sure

At that time (Amigaanywhere v1), I completed a tool for tracking daily pro activity expenses, the whole thing was less than 500Kb, with gui, and some advanced functions (bdd storage, export as csv, an xls export was even in alfa stage). That app was used (by me alone in this world) since 2 years ago on an old ipaq device, but was rock solid;)
the ipaq died (and people were moking me with that old device).
Then I rewrote the thing from scratch with modern tools (pcsoft windev mobile)
I just needed 2 days to rewrite the whole thing for android device while discovering the pcsoft software ( the tool costs 1600€ and spammed me a package that size is 7 Mb!)
Well it's ok, eveything is builtin, xls export, html export, pdf export, sync with my pro accounting software etc, all those functions were up and running in just two days.

Sad to see that now, a developer just calculates what tool can bring the app it need in the fastest way possible (money, money money), and no matter the size of the exe, neither the CPU it eat while working. Don't care, there is a 16Gb sdcard in the phone, and the 1Ghz proc can deal with it without any problem.

And thinking about it, this is really efficiant in a 'how many it costed me to to the thging in two days', and how many hundreds of hours i've spent years ago writing the same thing with C lines of code, while debugging amidb (amigaanywhere database), finding glitches, fighting with C code bugs of my own and so on.

All in one, I have the feeling that Carl Sassenrath was attempting the same thing as amigaanywhere (vision on how to do was different, but the goal to achieve was the same), and it seems both products although good and visionary were epic fails! (with all the respect I have for Carl).

I even wonder if there was not a discution between Carl and AInc before Ainc started to code their AmigaAnywhere v2 engine, I guess, but if I give you the name of the guiy who told me about this, you'll tell me it's bullshit (but I'am almost sure something was in discution at that time, don't confuse discussing of something, and passing an agreement).

That was a good old time, where I met some nice coders and nice people.

Even talking with fleecy daily was fun, I just start NOW to see what was his idea and vision, and see how he was ahead of it's time....but to be in business, you must be there "at the right moment", not before, not after....