Groups > comp.lang.forth > #43610 > unrolled thread

Forth News from Europe - Video - Olimex - ARM - VFX - Mecrisp and the question: Which other Forths - ANS or not - will run on this NXP LPC1114 platform?

Started by	JUERGEN <epldfpga@aol.com>
First post	2015-11-21 01:20 -0800
Last post	2016-03-02 18:38 +0000
Articles	20 on this page of 45 — 9 participants

Back to article view | Back to comp.lang.forth

Page 2 of 3 — ← Prev page 1 [2] 3 Next page →

#44003 — Re: Forth News from Europe ...

From	Albert van der Horst <albert@spenarnc.xs4all.nl>
Date	2015-12-01 10:53 +0000
Subject	Re: Forth News from Europe ...
Message-ID	<565d7c1e$0$23741$e4fe514c@news.xs4all.nl>
In reply to	#43999

Raimond Dragomir <raimond.dragomir@gmail.com> writes:

>> Baud rate 9600 - data 8 bit - parity none - stop 1 bit - flow control
>> none - transmit delay: 0 ms/char, 100 ms/line
>>
>That 100ms per line is a killer. That's why I said you need some PC
>utility to download the raw text... Something that takes care of the
>system "ok" response, stripping of comments waould be also nice...
>I think such utils even exists now, but I lost my interest on this
>long time ago.

It is shit-ugly but it isn't a killer. The authors of noforth have
for decades made interesting and working applications with this
kind of communication.

This being said, noforth accomodates a startup mode to force
a handshake (partly because of me).
I'm working on a portable communication program
(linux,apple,MS) that takes advantage of this.
I hesitate to publish it because its about equally unreliable
as the usual communication programs. Once I'm convinced
that it is the drivers I'm going to publish it.

>In my modern system as I said I will compile from internal files.
>The files will get there by either a tcp/ip protocol (I have tftp
>working right now but an NFS will be wonderfull) or an XMODEM protocol
>or ZMODEM for a mere serial line.

These protocols are overkill for a small turnkey IMO.

>Of course, nothing will prevent a user to download the raw text in the
>old style, but this will be just a "possibility".

Raw text is interactive development and test.
We don't want to give up on that.

Groetjes Albert
-- 
Albert van der Horst, UTRECHT,THE NETHERLANDS
Economic growth -- being exponential -- ultimately falters.
albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst

[toc] | [prev] | [next] | [standalone]

#44004 — Re: Forth News from Europe ...

From	Raimond Dragomir <raimond.dragomir@gmail.com>
Date	2015-12-01 03:33 -0800
Subject	Re: Forth News from Europe ...
Message-ID	<88528a4c-da02-461e-a0f8-d79afc54f09c@googlegroups.com>
In reply to	#44003

> >working right now but an NFS will be wonderfull) or an XMODEM protocol
> >or ZMODEM for a mere serial line.
> 
> These protocols are overkill for a small turnkey IMO.

XMODEM is simple. Overkill maybe the entire chain (file system + forth
include stack + comunication protocol).

And I didn't say it will be small :)

My first system (for a cortex-m4) compiled to 28.8K flash and about 5K ram.
But it has the flash file system, a multitasking os (still cooperative)
some drivers and of course the forth system which is in no way that simple
any more (quotations, local variables, automatic constant folding +
some other optimizations, include file stack, file system words including
"shell" like words like ls, rm, cp, cat, etc...)
In this size it has no tcp/ip. My simple tcp/ip stack compiles to about 13K.


> 
> >Of course, nothing will prevent a user to download the raw text in the
> >old style, but this will be just a "possibility".
> 
> Raw text is interactive development and test.
> We don't want to give up on that.
> 
> Groetjes Albert
> -- 
> Albert van der Horst, UTRECHT,THE NETHERLANDS
> Economic growth -- being exponential -- ultimately falters.
> albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst

[toc] | [prev] | [next] | [standalone]

#44008 — Re: Forth News from Europe ...

From	Mark Wills <markwills1970@gmail.com>
Date	2015-12-01 04:36 -0800
Subject	Re: Forth News from Europe ...
Message-ID	<3b4f8eef-6624-48d0-8706-85d5536a6c8e@googlegroups.com>
In reply to	#44004

On Tuesday, 1 December 2015 11:33:58 UTC, Raimond Dragomir  wrote:
> > >working right now but an NFS will be wonderfull) or an XMODEM protocol
> > >or ZMODEM for a mere serial line.
> > 
> > These protocols are overkill for a small turnkey IMO.
> 
> XMODEM is simple. Overkill maybe the entire chain (file system + forth
> include stack + comunication protocol).
> 
> And I didn't say it will be small :)
> 
> My first system (for a cortex-m4) compiled to 28.8K flash and about 5K ram.
> But it has the flash file system, a multitasking os (still cooperative)
> some drivers and of course the forth system which is in no way that simple
> any more (quotations, local variables, automatic constant folding +
> some other optimizations, include file stack, file system words including
> "shell" like words like ls, rm, cp, cat, etc...)
> In this size it has no tcp/ip. My simple tcp/ip stack compiles to about 13K.
> 
> 
> > 
> > >Of course, nothing will prevent a user to download the raw text in the
> > >old style, but this will be just a "possibility".
> > 
> > Raw text is interactive development and test.
> > We don't want to give up on that.
> > 
> > Groetjes Albert
> > -- 
> > Albert van der Horst, UTRECHT,THE NETHERLANDS
> > Economic growth -- being exponential -- ultimately falters.
> > albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst

Wow sounds great! Where can I get a copy? :-)

[toc] | [prev] | [next] | [standalone]

#44010 — Re: Forth News from Europe ...

From	Raimond Dragomir <raimond.dragomir@gmail.com>
Date	2015-12-01 05:26 -0800
Subject	Re: Forth News from Europe ...
Message-ID	<f18cce59-59da-492e-bb48-7ec50b147475@googlegroups.com>
In reply to	#44008

marți, 1 decembrie 2015, 14:36:57 UTC+2, Mark Wills a scris:
> On Tuesday, 1 December 2015 11:33:58 UTC, Raimond Dragomir  wrote:
> > > >working right now but an NFS will be wonderfull) or an XMODEM protocol
> > > >or ZMODEM for a mere serial line.
> > > 
> > > These protocols are overkill for a small turnkey IMO.
> > 
> > XMODEM is simple. Overkill maybe the entire chain (file system + forth
> > include stack + comunication protocol).
> > 
> > And I didn't say it will be small :)
> > 
> > My first system (for a cortex-m4) compiled to 28.8K flash and about 5K ram.
> > But it has the flash file system, a multitasking os (still cooperative)
> > some drivers and of course the forth system which is in no way that simple
> > any more (quotations, local variables, automatic constant folding +
> > some other optimizations, include file stack, file system words including
> > "shell" like words like ls, rm, cp, cat, etc...)
> > In this size it has no tcp/ip. My simple tcp/ip stack compiles to about 13K.
> > 
> > 
> > > 
> > > >Of course, nothing will prevent a user to download the raw text in the
> > > >old style, but this will be just a "possibility".
> > > 
> > > Raw text is interactive development and test.
> > > We don't want to give up on that.
> > > 
> > > Groetjes Albert
> > > -- 
> > > Albert van der Horst, UTRECHT,THE NETHERLANDS
> > > Economic growth -- being exponential -- ultimately falters.
> > > albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst
> 
> Wow sounds great! Where can I get a copy? :-)

Thank you! Well, no where. :(
This is due to my lack of time doing:
- documentation
- presentation
- demo applications (like the simple arduino like ones, at least)
- whatever else.

For example, I have no clue, but no time either to study, of how or where to
post it on the web. Suggestions welcome!

Another thing is that it won't work on every "gadget" board which are 
hundreds now in the world. My system needs besides the MCU, a small serial
flash memory. Fortunately, there are some boards with that, and the
number of such boards seems to increase... 
Unfortunately, I don't own such an "of the shelf" board... yet. My system
works on a LPC4088 MCU with a serial flash chip with a quad spi connection.
If you have that, it will work for you. Else, there must be some porting...
The serial flash chip type is not that important but here also may be some
porting... Currently I tested two different manufacturers, just because I'm
using that chips...

My idea is to identify a few broadly available boards that fit my system and then
make the reference implementation for them.
One of my ideas is the ESP8266... but I need one first

[toc] | [prev] | [next] | [standalone]

#44014 — Re: Forth News from Europe ...

From	Albert van der Horst <albert@spenarnc.xs4all.nl>
Date	2015-12-01 14:42 +0000
Subject	Re: Forth News from Europe ...
Message-ID	<565db1b9$0$23771$e4fe514c@news.xs4all.nl>
In reply to	#44010

Raimond Dragomir <raimond.dragomir@gmail.com> writes:

>mar=C8=9Bi, 1 decembrie 2015, 14:36:57 UTC+2, Mark Wills a scris:
>> On Tuesday, 1 December 2015 11:33:58 UTC, Raimond Dragomir  wrote:
>> > > >working right now but an NFS will be wonderfull) or an XMODEM protoc=
>ol
>> > > >or ZMODEM for a mere serial line.
>> > >=20
>> > > These protocols are overkill for a small turnkey IMO.
>> >=20
>> > XMODEM is simple. Overkill maybe the entire chain (file system + forth
>> > include stack + comunication protocol).
>> >=20
>> > And I didn't say it will be small :)
>> >=20
>> > My first system (for a cortex-m4) compiled to 28.8K flash and about 5K =
>ram.
>> > But it has the flash file system, a multitasking os (still cooperative)
>> > some drivers and of course the forth system which is in no way that sim=
>ple
>> > any more (quotations, local variables, automatic constant folding +
>> > some other optimizations, include file stack, file system words includi=
>ng
>> > "shell" like words like ls, rm, cp, cat, etc...)
>> > In this size it has no tcp/ip. My simple tcp/ip stack compiles to about=
> 13K.
>> >=20
>> >=20
>> > >=20
>> > > >Of course, nothing will prevent a user to download the raw text in t=
>he
>> > > >old style, but this will be just a "possibility".
>> > >=20
>> > > Raw text is interactive development and test.
>> > > We don't want to give up on that.
>> > >=20
>> > > Groetjes Albert
>> > > --=20
>> > > Albert van der Horst, UTRECHT,THE NETHERLANDS
>> > > Economic growth -- being exponential -- ultimately falters.
>> > > albert@spe&ar&c.xs4all.nl &=3Dn http://home.hccnet.nl/a.w.m.van.der.h=
>orst
>>=20
>> Wow sounds great! Where can I get a copy? :-)

>Thank you! Well, no where. :(
>This is due to my lack of time doing:
>- documentation
>- presentation
>- demo applications (like the simple arduino like ones, at least)
>- whatever else.

>For example, I have no clue, but no time either to study, of how or where t=
>o
>post it on the web. Suggestions welcome!

>Another thing is that it won't work on every "gadget" board which are=20
>hundreds now in the world. My system needs besides the MCU, a small serial
>flash memory. Fortunately, there are some boards with that, and the
>number of such boards seems to increase...=20
>Unfortunately, I don't own such an "of the shelf" board... yet. My system
>works on a LPC4088 MCU with a serial flash chip with a quad spi connection.
>If you have that, it will work for you. Else, there must be some porting...
>The serial flash chip type is not that important but here also may be some
>porting... Currently I tested two different manufacturers, just because I'm
>using that chips...

>My idea is to identify a few broadly available boards that fit my system an=
>d then
>make the reference implementation for them.
>One of my ideas is the ESP8266... but I need one first

On the other hand noforth is mature, over 5 years old, accomodates about
all 430 chips and is fiercely documented. It has been used for a number of
practical applications. My demonstration of big ben tune on the tingle tangle
was years ago. On noforth's website you can down load an other music
application that you can try after attaching a small loudspeaker to the
Launchpad (demonstrated at a German Forth convention.).

keywords : noforth nijhof ouwerkerk

Groetjes Albert
-- 
Albert van der Horst, UTRECHT,THE NETHERLANDS
Economic growth -- being exponential -- ultimately falters.
albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst

[toc] | [prev] | [next] | [standalone]

#44017 — Re: Forth News from Europe ...

From	Raimond Dragomir <raimond.dragomir@gmail.com>
Date	2015-12-01 08:16 -0800
Subject	Re: Forth News from Europe ...
Message-ID	<9053b38b-dbfa-421f-90f4-c51f31d58c36@googlegroups.com>
In reply to	#44014

> On the other hand noforth is mature, over 5 years old, accomodates about
> all 430 chips and is fiercely documented. It has been used for a number of
> practical applications. My demonstration of big ben tune on the tingle tangle
> was years ago. On noforth's website you can down load an other music
> application that you can try after attaching a small loudspeaker to the
> Launchpad (demonstrated at a German Forth convention.).
> 
> keywords : noforth nijhof ouwerkerk
> 
> Groetjes Albert
> -- 
> Albert van der Horst, UTRECHT,THE NETHERLANDS
> Economic growth -- being exponential -- ultimately falters.
> albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst

Sounds great, and looks great. 
But I always wanted to ask: why "noforth" name? "no" looks so negative...

[toc] | [prev] | [next] | [standalone]

#44020 — Re: Forth News from Europe ...

From	Albert van der Horst <albert@spenarnc.xs4all.nl>
Date	2015-12-01 19:27 +0000
Subject	Re: Forth News from Europe ...
Message-ID	<565df49e$0$23772$e4fe514c@news.xs4all.nl>
In reply to	#44017

Raimond Dragomir <raimond.dragomir@gmail.com> writes:

>> On the other hand noforth is mature, over 5 years old, accomodates about
>> all 430 chips and is fiercely documented. It has been used for a number of
>> practical applications. My demonstration of big ben tune on the tingle tangle
>> was years ago. On noforth's website you can down load an other music
>> application that you can try after attaching a small loudspeaker to the
>> Launchpad (demonstrated at a German Forth convention.).
>>
>> keywords : noforth nijhof ouwerkerk
>>
>> Groetjes Albert
>> --
>> Albert van der Horst, UTRECHT,THE NETHERLANDS
>> Economic growth -- being exponential -- ultimately falters.
>> albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst

>Sounds great, and looks great.
>But I always wanted to ask: why "noforth" name? "no" looks so negative...

There are so many Forths that people have to add their initials to
their Forth's. What can you do?

Groetjes Albert
-- 
Albert van der Horst, UTRECHT,THE NETHERLANDS
Economic growth -- being exponential -- ultimately falters.
albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst

[toc] | [prev] | [next] | [standalone]

#44005 — Re: Forth News from Europe ...

From	Mark Wills <markwills1970@gmail.com>
Date	2015-12-01 03:40 -0800
Subject	Re: Forth News from Europe ...
Message-ID	<29fd8862-c7ea-4ab9-a71b-41bf97979420@googlegroups.com>
In reply to	#44003

On Tuesday, 1 December 2015 10:53:19 UTC, Albert van der Horst  wrote:
> Raimond Dragomir <raimond.dragomir@gmail.com> writes:
> 
> >> Baud rate 9600 - data 8 bit - parity none - stop 1 bit - flow control
> >> none - transmit delay: 0 ms/char, 100 ms/line
> >>
> >That 100ms per line is a killer. That's why I said you need some PC
> >utility to download the raw text... Something that takes care of the
> >system "ok" response, stripping of comments waould be also nice...
> >I think such utils even exists now, but I lost my interest on this
> >long time ago.
> 
> It is shit-ugly but it isn't a killer. The authors of noforth have
> for decades made interesting and working applications with this
> kind of communication.
> 
> This being said, noforth accomodates a startup mode to force
> a handshake (partly because of me).
> I'm working on a portable communication program
> (linux,apple,MS) that takes advantage of this.
> I hesitate to publish it because its about equally unreliable
> as the usual communication programs. Once I'm convinced
> that it is the drivers I'm going to publish it.
> 
> >In my modern system as I said I will compile from internal files.
> >The files will get there by either a tcp/ip protocol (I have tftp
> >working right now but an NFS will be wonderfull) or an XMODEM protocol
> >or ZMODEM for a mere serial line.
> 
> These protocols are overkill for a small turnkey IMO.
> 
> >Of course, nothing will prevent a user to download the raw text in the
> >old style, but this will be just a "possibility".
> 
> Raw text is interactive development and test.
> We don't want to give up on that.
> 
> Groetjes Albert
> -- 
> Albert van der Horst, UTRECHT,THE NETHERLANDS
> Economic growth -- being exponential -- ultimately falters.
> albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst

Just implement Y modem uploads (from the PC to the controller) and
you're done. No need to re-invent the wheel. Those old serial protocols
were pretty good. Z Modem might be overkill, but Y-modem would be
just great.

[toc] | [prev] | [next] | [standalone]

#44308 — Re: Forth News from Europe ...

From	Paul Rubin <no.email@nospam.invalid>
Date	2015-12-21 23:06 -0800
Subject	Re: Forth News from Europe ...
Message-ID	<87y4cnhshk.fsf@nightsong.com>
In reply to	#43970

Raimond Dragomir <raimond.dragomir@gmail.com> writes:
>> >   https://www.adafruit.com/products/2772
>> A very beautiful board !  It may be a bit complicated to port
>> Mecrisp-Stellaris to this one, as the Atmel chip has a buffered flash
>> write page size of 64 bytes. ...
> No way. You cannot write to a page if it is not completely erased. 
> And there's more: if you want to erase a page you can't. You can erase
> only 4 pages at a time (a block) which translates to a 256 bytes
> working ram buffer for ANY flash writes.

The chip has enough ram (32k) that the 256 byte buffer might be
tolerable, especially if there's a way to multi-purpose it with the
terminal line buffer or other areas that Forth needs to keep available
some of the time.  Plus there is 256k of flash, so maybe it's possible
to use a 256 byte block as temp storage so you can rewrite a 64 byte
page by copying the surrounding pages to the temp block and back,
i.e. you need only 64 bytes of ram buffer instead of 256.

FWIW, there's now an alternate version of the board with a "hard disk"
(microSD slot): https://www.adafruit.com/products/2796 giving
essentially infinite storage for something like this.  I'm liking these
boards more and more.  They apparently include a 32 khz crystal so you
can have an RTC if you don't reset or unpower the board.  And there's
a cute OLED display that plugs on top: https://www.adafruit.com/products/2900

This is tempting for many things.

[toc] | [prev] | [next] | [standalone]

#44309 — Re: Forth News from Europe ...

From	Raimond Dragomir <raimond.dragomir@gmail.com>
Date	2015-12-22 00:07 -0800
Subject	Re: Forth News from Europe ...
Message-ID	<41c72150-e792-4008-aa8f-8fccfd02fe19@googlegroups.com>
In reply to	#44308

> The chip has enough ram (32k) that the 256 byte buffer might be
> tolerable, especially if there's a way to multi-purpose it with the
> terminal line buffer or other areas that Forth needs to keep available
> some of the time.  Plus there is 256k of flash, so maybe it's possible
> to use a 256 byte block as temp storage so you can rewrite a 64 byte
> page by copying the surrounding pages to the temp block and back,
> i.e. you need only 64 bytes of ram buffer instead of 256.

I didn't say 256 bytes is too much. It is just a technical info.

> 
> FWIW, there's now an alternate version of the board with a "hard disk"
> (microSD slot): https://www.adafruit.com/products/2796 giving
> essentially infinite storage for something like this.  I'm liking these
> boards more and more.  They apparently include a 32 khz crystal so you
> can have an RTC if you don't reset or unpower the board.  And there's
> a cute OLED display that plugs on top: https://www.adafruit.com/products/2900
> 
> This is tempting for many things.

I prefer a serial flash chip for harsh industrial applications, but sd
cards are usually ok too. At least if you want such "unlimited" storage.
The problem is that you will want FAT32, and good luck with that.

This is a reason for why I'm developing and using my own forth: because
I write the kernel in C and I'm able to include in it any C library, in
this case Chan's fatfs file system (it compiles under 20K on a cortex-m)

And this is true for anything around. For example I needed some time ago
to encrypt some messages with a public 2048bit public key... on a 
cortex-m0 board @48MHz... With my best tryings I got something around
70-80 seconds... Then I found a stripped down version from openssl and 
voila! 170 milliseconds...

[toc] | [prev] | [next] | [standalone]

#44337 — Re: Forth News from Europe ...

From	Paul Rubin <no.email@nospam.invalid>
Date	2015-12-23 20:14 -0800
Subject	Re: Forth News from Europe ...
Message-ID	<87twn8h492.fsf@nightsong.com>
In reply to	#44309

Raimond Dragomir <raimond.dragomir@gmail.com> writes:
> I didn't say 256 bytes is too much. It is just a technical info.

Ah, ok.

> I prefer a serial flash chip for harsh industrial applications, but sd
> cards are usually ok too. At least if you want such "unlimited" storage.
> The problem is that you will want FAT32, and good luck with that.

I wonder what the erase page size is on microsd cards: it could end up
needing a lot of extra block copies or bigger ram buffers.  FAT32 is
pretty simple though, I thought.  If not, it's possible to implement an
old-fashioned Unix-style file system with a few hundred lines of code.

> This is a reason for why I'm developing and using my own forth: because
> I write the kernel in C and I'm able to include in it any C library, in
> this case Chan's fatfs file system (it compiles under 20K on a cortex-m)

20k seems way too much for a basic fat fs.  I wonder what's going on.  I
can see usefulness in being able to call C code from Forth, though that
wouldn't seem to require writing the Forth itself in C.  But, there's
an interesting eForth port to C:

   http://chiselapp.com/user/tehologist/repository/compc/home

The C part is only about 500 LOC.

> And this is true for anything around. For example I needed some time ago
> to encrypt some messages with a public 2048bit public key... on a 
> cortex-m0 board @48MHz... With my best tryings I got something around
> 70-80 seconds... Then I found a stripped down version from openssl and 
> voila! 170 milliseconds...

Again something really seems wrong there, if you mean low exponent RSA
encryption.  Maybe a badly chosen algorithm or something.  Even 170 msec
sounds kind of slow.  If you post the code, I can take a look.

[toc] | [prev] | [next] | [standalone]

#44310 — Re: Forth News from Europe ...

From	Albert van der Horst <albert@spenarnc.xs4all.nl>
Date	2015-12-22 11:17 +0000
Subject	Re: Forth News from Europe ...
Message-ID	<56793141$0$23814$e4fe514c@news.xs4all.nl>
In reply to	#44308

Paul Rubin <no.email@nospam.invalid> writes:

>Raimond Dragomir <raimond.dragomir@gmail.com> writes:
>>> >   https://www.adafruit.com/products/2772
>>> A very beautiful board !  It may be a bit complicated to port
>>> Mecrisp-Stellaris to this one, as the Atmel chip has a buffered flash
>>> write page size of 64 bytes. ...
>> No way. You cannot write to a page if it is not completely erased.
>> And there's more: if you want to erase a page you can't. You can erase
>> only 4 pages at a time (a block) which translates to a 256 bytes
>> working ram buffer for ANY flash writes.

>The chip has enough ram (32k) that the 256 byte buffer might be
>tolerable, especially if there's a way to multi-purpose it with the
>terminal line buffer or other areas that Forth needs to keep available
>some of the time.  Plus there is 256k of flash, so maybe it's possible
>to use a 256 byte block as temp storage so you can rewrite a 64 byte
>page by copying the surrounding pages to the temp block and back,
>i.e. you need only 64 bytes of ram buffer instead of 256.

>FWIW, there's now an alternate version of the board with a "hard disk"
>(microSD slot): https://www.adafruit.com/products/2796 giving
>essentially infinite storage for something like this.  I'm liking these
>boards more and more.  They apparently include a 32 khz crystal so you
>can have an RTC if you don't reset or unpower the board.  And there's
>a cute OLED display that plugs on top: https://www.adafruit.com/products/2900

>This is tempting for many things.

Clever programming may help to completely hide the block structure
of underlying flash rom. This is e.g. the case in noforth. The whole
meta source is available, and it is wel  documented.

(Sorry to plug it again, but it is splendid.)

Groetjes Albert
-- 
Albert van der Horst, UTRECHT,THE NETHERLANDS
Economic growth -- being exponential -- ultimately falters.
albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst

[toc] | [prev] | [next] | [standalone]

#44311 — Re: Forth News from Europe ...

From	Matthias Koch <matthias.koch@hot.uni-hannover.de>
Date	2015-12-22 12:31 +0100
Subject	Re: Forth News from Europe ...
Message-ID	<n5bc5f$buo$1@newsserver.rrzn.uni-hannover.de>
In reply to	#44310

> Clever programming may help to completely hide the block structure
> of underlying flash rom. This is e.g. the case in noforth. The whole
> meta source is available, and it is wel  documented.
> 
> (Sorry to plug it again, but it is splendid.)
> 
> Groetjes Albert

Hi Albert,

of course :-) It is just two things: I hope to get it without erase cycles,
and I am busy with other ports and features. The EFM32 Giant Gecko should be
the next addition my target list.

Perhaps we meet at 32C3 in Hamburg next week ?

Matthias

[toc] | [prev] | [next] | [standalone]

#45715 — Re: Forth News from Europe ...

From	Paul Rubin <no.email@nospam.invalid>
Date	2016-02-26 19:19 -0800
Subject	Re: Forth News from Europe ...
Message-ID	<87d1rivo06.fsf@jester.gateway.pace.com>
In reply to	#43968

Matthias Koch <matthias.koch@hot.uni-hannover.de> writes:
>>   https://www.adafruit.com/products/2772
> A very beautiful board !

This might also be of interest:

  http://oshchip.org/
  https://www.tindie.com/products/OSHChip/oshchip-v10/

It's a Cortex M0 / BLE radio board in 16 pin DIP format, breadboard
friendly, should fit in the Tictac box nicely, 16 Mhz, 256K flash, 32K
SRAM, RTC and other nice features.  A bit expensive though, 25 USD.
Uses Nordic nRF51822-CFAC-A0 SOC.

[toc] | [prev] | [next] | [standalone]

#45756 — Re: Forth News from Europe ...

From	JUERGEN <epldfpga@aol.com>
Date	2016-03-01 04:57 -0800
Subject	Re: Forth News from Europe ...
Message-ID	<df069b4d-9643-4626-93f3-ad64472ae155@googlegroups.com>
In reply to	#45715

On Saturday, February 27, 2016 at 3:19:25 AM UTC, Paul Rubin wrote:
> Matthias Koch <matthias.koch@hot.uni-hannover.de> writes:
> >>   https://www.adafruit.com/products/2772
> > A very beautiful board !
> 
> This might also be of interest:
> 
>   http://oshchip.org/
>   https://www.tindie.com/products/OSHChip/oshchip-v10/
> 
> It's a Cortex M0 / BLE radio board in 16 pin DIP format, breadboard
> friendly, should fit in the Tictac box nicely, 16 Mhz, 256K flash, 32K
> SRAM, RTC and other nice features.  A bit expensive though, 25 USD.
> Uses Nordic nRF51822-CFAC-A0 SOC.

Thanks Paul, very much on the radar already, and I assume there will be a Forth for it soon, then a subboard will be built to have a Bluetooth enabled MicroBox. it will have the same small HW interface as the 430 / ARM system: 3 switches, 5 LEDS and the Minimum Forth to interface to them. From then on a wireless interface via BT should be the next step.

Matthias, are you working on a Mecrisp for this chip already?

[toc] | [prev] | [next] | [standalone]

#45758 — Re: Forth News from Europe ...

From	rickman <gnuarm@gmail.com>
Date	2016-03-01 09:43 -0500
Subject	Re: Forth News from Europe ...
Message-ID	<nb49lm$49u$1@dont-email.me>
In reply to	#45756

On 3/1/2016 7:57 AM, JUERGEN wrote:
> On Saturday, February 27, 2016 at 3:19:25 AM UTC, Paul Rubin wrote:
>> Matthias Koch <matthias.koch@hot.uni-hannover.de> writes:
>>>>    https://www.adafruit.com/products/2772
>>> A very beautiful board !
>>
>> This might also be of interest:
>>
>>    http://oshchip.org/
>>    https://www.tindie.com/products/OSHChip/oshchip-v10/
>>
>> It's a Cortex M0 / BLE radio board in 16 pin DIP format, breadboard
>> friendly, should fit in the Tictac box nicely, 16 Mhz, 256K flash, 32K
>> SRAM, RTC and other nice features.  A bit expensive though, 25 USD.
>> Uses Nordic nRF51822-CFAC-A0 SOC.
>
> Thanks Paul, very much on the radar already, and I assume there will be a Forth for it soon, then a subboard will be built to have a Bluetooth enabled MicroBox. it will have the same small HW interface as the 430 / ARM system: 3 switches, 5 LEDS and the Minimum Forth to interface to them. From then on a wireless interface via BT should be the next step.
>
> Matthias, are you working on a Mecrisp for this chip already?

If someone is working on a Forth for the OSHchip I would be very 
interested in helping.  I need exactly this device for a project.  I was 
looking at using a much larger FPGA with MCU and Bluetooth board based 
on the SmartFusion 2 device, but the OSHchip would be small enough for 
the final design rather than having to spin a smaller board.

BTW, does the OSHchip look like it uses an unusually thick PCB?  Or does 
it just look that way because it is so small?  I was wondering if the 
holes for the pins being so close to the edge of the board required the 
board to be extra thick?

-- 

Rick

[toc] | [prev] | [next] | [standalone]

#45773 — Re: Forth News from Europe ...

From	Raimond Dragomir <raimond.dragomir@gmail.com>
Date	2016-03-01 23:10 -0800
Subject	Re: Forth News from Europe ...
Message-ID	<08c665e1-6266-4835-958f-e2d53e0ba24a@googlegroups.com>
In reply to	#45758

marți, 1 martie 2016, 16:43:58 UTC+2, rickman a scris:
> On 3/1/2016 7:57 AM, JUERGEN wrote:
> > On Saturday, February 27, 2016 at 3:19:25 AM UTC, Paul Rubin wrote:
> >> Matthias Koch <matthias.koch@hot.uni-hannover.de> writes:
> >>>>    https://www.adafruit.com/products/2772
> >>> A very beautiful board !
> >>
> >> This might also be of interest:
> >>
> >>    http://oshchip.org/
> >>    https://www.tindie.com/products/OSHChip/oshchip-v10/
> >>
> >> It's a Cortex M0 / BLE radio board in 16 pin DIP format, breadboard
> >> friendly, should fit in the Tictac box nicely, 16 Mhz, 256K flash, 32K
> >> SRAM, RTC and other nice features.  A bit expensive though, 25 USD.
> >> Uses Nordic nRF51822-CFAC-A0 SOC.
> >
> > Thanks Paul, very much on the radar already, and I assume there will be a Forth for it soon, then a subboard will be built to have a Bluetooth enabled MicroBox. it will have the same small HW interface as the 430 / ARM system: 3 switches, 5 LEDS and the Minimum Forth to interface to them. From then on a wireless interface via BT should be the next step.
> >
> > Matthias, are you working on a Mecrisp for this chip already?
> 
> If someone is working on a Forth for the OSHchip I would be very 
> interested in helping.  I need exactly this device for a project.  I was 
> looking at using a much larger FPGA with MCU and Bluetooth board based 
> on the SmartFusion 2 device, but the OSHchip would be small enough for 
> the final design rather than having to spin a smaller board.
> 
> BTW, does the OSHchip look like it uses an unusually thick PCB?  Or does 
> it just look that way because it is so small?  I was wondering if the 
> holes for the pins being so close to the edge of the board required the 
> board to be extra thick?
> 
> -- 
> 
> Rick

It's just the usual 1.6mm pcb (compare with the 2.54mm pitch of the pins).
I bet they made it normal thikness (1.6mm) because of the pins mounting stability. At this size a 0.8mm or 1mm pcb thickness is more usual but there are the pins that are soldered on the pcb and can have mechanical problems (and thermal also - try to solder these pins on your pcb and you can melt the solder on the OSH pcb also if you are not carefull enough...)

[toc] | [prev] | [next] | [standalone]

#45781 — Re: Forth News from Europe ...

From	JUERGEN <epldfpga@aol.com>
Date	2016-03-02 06:18 -0800
Subject	Re: Forth News from Europe ...
Message-ID	<090897a3-8caa-43bc-9108-1874e6881c0d@googlegroups.com>
In reply to	#45773

On Wednesday, March 2, 2016 at 7:10:22 AM UTC, Raimond Dragomir wrote:
> marți, 1 martie 2016, 16:43:58 UTC+2, rickman a scris:
> > On 3/1/2016 7:57 AM, JUERGEN wrote:
> > > On Saturday, February 27, 2016 at 3:19:25 AM UTC, Paul Rubin wrote:
> > >> Matthias Koch <matthias.koch@hot.uni-hannover.de> writes:
> > >>>>    https://www.adafruit.com/products/2772
> > >>> A very beautiful board !
> > >>
> > >> This might also be of interest:
> > >>
> > >>    http://oshchip.org/
> > >>    https://www.tindie.com/products/OSHChip/oshchip-v10/
> > >>
> > >> It's a Cortex M0 / BLE radio board in 16 pin DIP format, breadboard
> > >> friendly, should fit in the Tictac box nicely, 16 Mhz, 256K flash, 32K
> > >> SRAM, RTC and other nice features.  A bit expensive though, 25 USD.
> > >> Uses Nordic nRF51822-CFAC-A0 SOC.
> > >
> > > Thanks Paul, very much on the radar already, and I assume there will be a Forth for it soon, then a subboard will be built to have a Bluetooth enabled MicroBox. it will have the same small HW interface as the 430 / ARM system: 3 switches, 5 LEDS and the Minimum Forth to interface to them. From then on a wireless interface via BT should be the next step.
> > >
> > > Matthias, are you working on a Mecrisp for this chip already?
> > 
> > If someone is working on a Forth for the OSHchip I would be very 
> > interested in helping.  I need exactly this device for a project.  I was 
> > looking at using a much larger FPGA with MCU and Bluetooth board based 
> > on the SmartFusion 2 device, but the OSHchip would be small enough for 
> > the final design rather than having to spin a smaller board.
> > 
> > BTW, does the OSHchip look like it uses an unusually thick PCB?  Or does 
> > it just look that way because it is so small?  I was wondering if the 
> > holes for the pins being so close to the edge of the board required the 
> > board to be extra thick?
> > 
> > -- 
> > 
> > Rick
> 
> It's just the usual 1.6mm pcb (compare with the 2.54mm pitch of the pins).
> I bet they made it normal thikness (1.6mm) because of the pins mounting stability. At this size a 0.8mm or 1mm pcb thickness is more usual but there are the pins that are soldered on the pcb and can have mechanical problems (and thermal also - try to solder these pins on your pcb and you can melt the solder on the OSH pcb also if you are not carefull enough...)

I talked to them and it seems it is all open source anyway, so you can do your own layout, board  and manufacturing.

I had asked as well for availability in Europe - as shipping is more than the board. But not for now.

The quickest way forward might be if you do a deal with Matthias and Mecrisp: You send him 2 kits and then it might speed up to have Mecrisp running on it.
This would take cost and overheads away and would help on motivation if he has time to slot it in. 
I am waiting for the Forth on it as it is planned for the educational side of the MicroBox - control it via PC and Mobile without wires. Just what we need.

[toc] | [prev] | [next] | [standalone]

#45784 — Re: Forth News from Europe ...

From	Raimond Dragomir <raimond.dragomir@gmail.com>
Date	2016-03-02 09:37 -0800
Subject	Re: Forth News from Europe ...
Message-ID	<cccaa2c6-3131-418c-977a-76ce0a6afd80@googlegroups.com>
In reply to	#45781

> I talked to them and it seems it is all open source anyway, so you can do your own layout, board  and manufacturing.
> 
Im not clear, what is the added value of this board (except putting together
a few components on a pcb, probably a very simple schematic recommended by
the chip manufacturer) ?
I always wonder this same question when I see a lot of this kind of boards,
just the mcu with the normal harness around on a simple pcb...

[toc] | [prev] | [next] | [standalone]

#45819 — Re: Forth News from Europe ...

From	rickman <gnuarm@gmail.com>
Date	2016-03-04 04:08 -0500
Subject	Re: Forth News from Europe ...
Message-ID	<nbbj48$qrs$1@dont-email.me>
In reply to	#45784

On 3/2/2016 12:37 PM, Raimond Dragomir wrote:
>> I talked to them and it seems it is all open source anyway, so you can do your own layout, board  and manufacturing.
>>
> Im not clear, what is the added value of this board (except putting together
> a few components on a pcb, probably a very simple schematic recommended by
> the chip manufacturer) ?
> I always wonder this same question when I see a lot of this kind of boards,
> just the mcu with the normal harness around on a simple pcb...

In general any wireless interface must have regulatory agency approvals 
to be sold.  I'm not sure of the details, it may be ok to see unapproved 
units for hobby or lab use.  But often modules such as this are used 
because they provide regulatory approvals as long as you use the module 
without modifications and design your system to not muck up the on board 
antenna.

If you use an external antenna or in any way modify the radio, you will 
need to get your own regulatory approvals which is a bear.

It is also nice just to have a board that has already been designed and 
tested when it contains wireless interfaces as they are not hard to 
screw up.

-- 

Rick

[toc] | [prev] | [next] | [standalone]

Page 2 of 3 — ← Prev page 1 [2] 3 Next page →

csiph-web

Forth News from Europe - Video - Olimex - ARM - VFX - Mecrisp and the question: Which other Forths - ANS or not - will run on this NXP LPC1114 platform?

Contents

#44003 — Re: Forth News from Europe ...

#44004 — Re: Forth News from Europe ...

#44008 — Re: Forth News from Europe ...

#44010 — Re: Forth News from Europe ...

#44014 — Re: Forth News from Europe ...

#44017 — Re: Forth News from Europe ...

#44020 — Re: Forth News from Europe ...

#44005 — Re: Forth News from Europe ...

#44308 — Re: Forth News from Europe ...

#44309 — Re: Forth News from Europe ...

#44337 — Re: Forth News from Europe ...

#44310 — Re: Forth News from Europe ...

#44311 — Re: Forth News from Europe ...

#45715 — Re: Forth News from Europe ...

#45756 — Re: Forth News from Europe ...

#45758 — Re: Forth News from Europe ...

#45773 — Re: Forth News from Europe ...

#45781 — Re: Forth News from Europe ...

#45784 — Re: Forth News from Europe ...

#45819 — Re: Forth News from Europe ...