offtopic - Critical of the Computer Industry

[Jason Damisch]

As a Forth programmer, I don't know how you feel about the computer industry in general, but for myself, I feel that there is much wrong with it. I won't go into all of the hundreds of complaints that I have about it. As a Forth programmer, personally, I feel disconnected from the mainstream of computer programming culture and I.T. culture. As such I feel that I can step back from it and take a critical view of it, while also having some knowledge about the technical aspects of it at the same time. This I feel gives me a unique perspective.

Yet, there is not a comp.complaints category for my gripes. I can laugh about that because why would the computer establishment create a venue for people to gripe about their work or lack thereof.

One time when I had an office for my little website business I rented an office in a building with other office spaces. I talked to a lady who worked for another business there. She told me that she wished that there was an alternative to Windows, Mac, Linux etc. I wish that I had a good answer for her. Of course I still don't.

I see that there is a disconnect between computer users and computer programmers. There still is. I also see that customer satisfaction varies widely.

There we are.

Jason

[Rod Pemberton]

"Jason Damisch" <jasond...@yahoo.com> wrote in message
news:acaaee13-82e0-4ef7...@googlegroups.com...

> As a Forth programmer, I don't know how you feel about the
> computer industry in general, but for myself, I feel that there
> is much wrong with it.

If one does not qualify with the stated constraint of "as a Forth
programmer," should one respond to this or was that a
pre-emptively exclusionary statement? ...

> I won't go into all of the hundreds of complaints that I have
> about it.

Okay, obviously this flame bait... But, I'll bite anyway.

Why did you bother to post mentioning your "hundreds of
complaints," about the "the computer industry in general," "as a
Forth programmer," if you weren't willing to discuss them with us?

> As a Forth programmer, personally, I feel disconnected from the
> mainstream of computer programming culture and I.T. culture.

...

> As such I feel that I can step back from it and take a critical
> view of it, while also having some knowledge about the
> technical aspects of it at the same time.

...

> This I feel gives me a unique perspective.

How so?

You clearly, just stated that you're ignorant of "mainstream ...
computer programming culture" while being knowledgeable about
Forth. That description seems to fit _ALL_ the Forth
programmers here. The opposite seems to fit _ALL_ the non-Forth
programmers not here, i.e., they're skilled in their domains and
ignorant of Forth. So, wouldn't you need to be knowledgeable
about _BOTH_ cultures to actually have a "unique perspective"?
I.e., your perspective, from my perspective, is that you fit into
the Forth crowd. Tell me, how is that unique?


Rod Pemberton

[Jason Damisch]

On Friday, May 3, 2013 1:54:30 PM UTC-7, Rod Pemberton wrote:
> "Jason Damisch" <jasond...@yahoo.com> wrote in message

> > As a Forth programmer, I don't know how you feel about the
> > computer industry in general, but for myself, I feel that there
> > is much wrong with it.
>
>
> If one does not qualify with the stated constraint of "as a Forth
> programmer," should one respond to this or was that a
> pre-emptively exclusionary statement? ...

You can respond to it or not as you see fit. It is an exclusionary statement so far as I am not a C programmer.

> > I won't go into all of the hundreds of complaints that I have
> > about it.

> Okay, obviously this flame bait... But, I'll bite anyway.

No, it is not flame bait. It is an expression of genuine anger, frustration and discontent. If you can't see it as being anything more than flame bait then perhaps it reflects on the values of mainstream computer culture.

> Why did you bother to post mentioning your "hundreds of
> complaints," about the "the computer industry in general," "as a
> Forth programmer," if you weren't willing to discuss them with us?

Because if I did, it would flood this outpost with hundreds of off-topic threads. I thought that one was enough. If I do this, then I will create a blog of my own for this topic.

> > As a Forth programmer, personally, I feel disconnected from the
> > mainstream of computer programming culture and I.T. culture.
> ...
> > As such I feel that I can step back from it and take a critical
> > view of it, while also having some knowledge about the
> > technical aspects of it at the same time.
> ...
> > This I feel gives me a unique perspective.

> How so?

Because I am not emotionally invested in the way things are.

> You clearly, just stated that you're ignorant of "mainstream ...
> computer programming culture" while being knowledgeable about
> Forth. That description seems to fit _ALL_ the Forth
> programmers here.

No, probably most of the people in here have an affinity with mainstream computer culture, and happen to also be Forth programmers.

> The opposite seems to fit _ALL_ the non-Forth
> programmers not here, i.e., they're skilled in their domains and
> ignorant of Forth.

I think that would be true of most C programmers, but not all.

> So, wouldn't you need to be knowledgeable
> about _BOTH_ cultures to actually have a "unique perspective"?
> I.e., your perspective, from my perspective, is that you fit into
> the Forth crowd. Tell me, how is that unique?
>
> Rod Pemberton

I might fit into the Forth crowd, but I am not economically nor emotionally invested in the current state of affairs. If I was then I would be happy and not posting this thread.

Jason

[visua...@rocketmail.com]

On Friday, May 3, 2013 2:17:27 PM UTC-4, Jason Damisch wrote:
> As a Forth programmer, I don't know how you feel about
> the computer industry in general, but for myself, I feel that there is
> much wrong with it. I won't go into all of the hundreds of complaints
> that I have about it. As a Forth programmer, personally, I feel
> disconnected from the mainstream of computer programming culture
> and I.T. culture.

Jason,
I understand your feelings. Everytime I sit on my Windows laptop, I have to deal with all these behaviours which I do not like, and no chance to fix it. In case my Windows doesn't work, I switch over to Linux, to get my Windows running again - Windows is easier to use for me. And then all these upgrades, which make Windows slower and worse.

Thanks for your offtopic - Critical of the Computer Industry - now I am feeling that I am not alone any more.

I cancelled IEEE, because there is nothing about Forth. Now I get IEEE letters once a week inviting me to come back, and emails nearly every day. I told them IEEE doesn't make sense for me, because they don't report anything about Forth.

Forth is the right environment for people like us, for scientists, and other people who like to go further and who like to have their own special computing environment.

I had several customers in DNA and Cancer research. One of them wrote his computer programs by himself. I showed him how Forth works and he worked with Forth ever since, working on the human Genome project. With Forth he could do things easy and fast which he couldn't have done that way with any other programming language.

The normal way for scientists to get their computer job done is to use software written by programmers which do not have the scientist's knowledge. With Forth the scientist is able to write his software himself. My customer proved it.

There is a way out.

Times are changing. With all these new Operating Systems around like Android etc. the computing landscape is changing. There is a niche for Forth computers, for people like you and my who like to enjoy their own instead of mainstream.

There is ForthOS, which could be a base, because it runs on a normal PC without any other OS, but it doesn't have any graphics. This could be a starting point, adding OpenGL to it, if possible. Today I got the OpenGL Programming Guide by mail. It will need a while to read and understand, yet it's a start.

PCs and laptops without OS are really cheap nowadays, so ForthOS with OpenGL and a suitable human interface would be great, I am sure of this.

If this strikes a cord for enough people, a crowdfunding could be started to built a special PC (or better laptop) with Forth machines inside instead of Intel/AMD. Chuck Moore once said that PCs are messed up from their beginning, so a new design based on one or several Forth RISC machines would be a better way. Chuck said:

" If you don't plan ahead you can get into a real mess trying to work things up afterwards. Incidentally that is what happened on the PC. You put an 8086 on a board and then add interface chips around to provide some kind of video and some serial and parallel busses and you end up with a pretty gerrigged mess. At some point you want to go back and redesign the whole thing and throw out all the interface chips and consolidate things and they never did that. They are trying to do it now with ASICs that handle all the horrible I/O processing that a PC requires and it is just a nightmare." We can do better.

The production doesn't have to go in the millions, a few or even a few hundred would be great. Production knowledge is not a problem. Problems are USB and Web Browser. USB could be solved easily, but Web Browser remains a problem. On the other hand - when this Forth Computer is available, people will show up to introduce a Web Browser to this machine, I am sure of this. Interesting thing is that Chuck's iTV did have a Web Browser, but this software seems to be proprietary.

How does this sound?

Jeff Fox, the visionary, one wrote: "I would like to see $5 computers that are better than PCs. I would like to see Forth enable a new generation of very smart and cheap computers and enable a new generation of very clever children."

Cheers,
Dirk.
http://www.visualforth.blogspot.com/

[Andrew Haley]

As you'd expect. The main problem, as I see it, is that we haven't
got to grips with the complexity of the systems we create. The
easiest way to solve a problem with a system is to add a layer of
indirection; and, indeed, this is often the best thing to do. But
over time there is an accretion of these layers, and underlying
mistakes don't get fixed but get worked around.

Many of us who programmed 8- and 16-bit microprocessors know how fast
they could be at responding to real-world events and we wonder how
fast computers would be today if they were programmed in the same way.
But computers are today required to do more and more things, some of
them useful but many of them not. I suspect that vast amounts of what
computers do is useless, but no-one knows which is the useless part.

I have more than once been guilty of failing to fix design mistakes,
either because of laziness or time pressure; it's a very human
failing. Those of us with experience need to do better, to set an
example, and to encourage the great mass of inexperienced developers
to do the right thing. If not us, then who will?

Andrew.

[Bernd Paysan]

Andrew Haley wrote:
> Many of us who programmed 8- and 16-bit microprocessors know how fast
> they could be at responding to real-world events and we wonder how
> fast computers would be today if they were programmed in the same way.
> But computers are today required to do more and more things, some of
> them useful but many of them not. I suspect that vast amounts of what
> computers do is useless, but no-one knows which is the useless part.

Oh, come on, if I look through the Android source code, I can quickly
identify the useless parts. It's just that the Android programmers don't
see it when they are creating useless abstractions. They have learned 42
design pattern at school, and now are trying to create the model-view-
container-database-interpreter-proxy-adapater-interface-proxy for everything
they see.

You like an example? To play hardware-accelerated video, you need two
shared memory buffers: One is for an encoded frame, the other is for the
decoded pictures (YUV in typically 4:2:0 format), of which you have at least
three when you have B frames. And a way to tell the DSP which codec you
have used. That's it. And that's essentially how all hardware decoders
work, and if you don't have an hardware decoder, it's also how all software
decoders work. And the same applies for audio decoding, too, all this is
essentially shared memory stuff, and it's all chopped into chunks called
"frames". Every frame has a timestamp: that's when you play it.

So, expose that interface in a generic way (with well-defined constants for
selecting the codecs, and "I updated the input buffer" + "wait for you
updating the picture buffer" calls+events), and you are done. It's just one
layer of abstraction.

When I read that Google is using OpenMAX AL, which is a Khronos group
standard, I thought, fine, that's probably going to look like that, because
OpenGL, another Khronos group standard, in the meantime really is a very
thin layer on top of the GPU hardware.

But it wasn't. OpenMAX AL is onion programming at its best, it takes an
MPEG transport stream. Nobody sane uses MPEG transport streams anymore for
anything but legacy (including the MPEG association, they switched to
Quicktime with MP4). All more recent formats are frame-oriented, because
that makes sense. What doesn't make any sense is extracting something and
then wrapping it up again, only to be extracted in the next step again.

To keep it simple is something you don't learn at school, for whatever
reason.

Companies like Google have a youth cult, and pretty strong age
discrimination, as the young programmers say, the old greybeards don't want
to program in Python and Java, and probably especially not programm onions,
but that's the way to go today. Also, the old greybeards are arrogant and
say "we had more responsive programs on the C64." And of course, they did
it all 10 miles through knee-deep snow, uphill in both directions ;-).

--
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://bernd-paysan.de/

[Paul Rubin]

Bernd Paysan <bernd....@gmx.de> writes:
> Companies like Google have a youth cult, and pretty strong age
> discrimination, as the young programmers say, the old greybeards don't
> want to program in Python and Java,

Greybeards do program in Lisp, and Python isn't that much different ;-).

I may not have reached grey-beardedness yet, but lately I've been using
Erlang, which has become hip recently but apparently has some
association with greybeards. There is a movie about more recent
effots to modernize its image:

https://www.youtube.com/watch?v=rRbY3TMUcgQ

;-)

[Andrew Haley]

Bernd Paysan <bernd....@gmx.de> wrote:
> Andrew Haley wrote:
>> Many of us who programmed 8- and 16-bit microprocessors know how fast
>> they could be at responding to real-world events and we wonder how
>> fast computers would be today if they were programmed in the same way.
>> But computers are today required to do more and more things, some of
>> them useful but many of them not. I suspect that vast amounts of what
>> computers do is useless, but no-one knows which is the useless part.
>
> Oh, come on, if I look through the Android source code, I can
> quickly identify the useless parts.

Oh, sure, there is low-hanging fruit; there always is. I still
suspect, though, that there are a lot of useless parts that no-one
knows about.

> Companies like Google have a youth cult, and pretty strong age
> discrimination, as the young programmers say, the old greybeards
> don't want to program in Python and Java, and probably especially
> not programm onions, but that's the way to go today.

I don't think so: I was at dinner with a bunch of Googlers a month or
two back, and there were a significant number of greybeards. From
what I've seen, Google values experience.

Andrew.

[rickman]

On 5/3/2013 5:30 PM, Jason Damisch wrote:
>
> I might fit into the Forth crowd, ...snip...

"The Forth crowd"... that might just be an oxymoron... How many people
do you need to have a crowd? lol

--

Rick

[rickman]

On 5/4/2013 2:24 AM, visua...@rocketmail.com wrote:
>
> I cancelled IEEE, because there is nothing about Forth. Now I get IEEE letters once a week inviting me to come back, and emails nearly every day. I told them IEEE doesn't make sense for me, because they don't report anything about Forth.

I quit IEEE because I learned how it is run. Or I should say, how the
local chapters are run. I'm sure the main body is run like a company...
well, somewhat anyhow. They do provide standards and other useful
information. But the local chapters are just a good ol' boys (and
girls) network. They don't even follow their own rules.

As to Forth content, that has to come from individuals. People need to
do research and produce papers. IEEE doesn't foster that, they just
publish the research that is sent to them and is worthy of
publication... by their standards of course.

> There is a way out.
>
> Times are changing. With all these new Operating Systems around like Android etc. the computing landscape is changing. There is a niche for Forth computers, for people like you and my who like to enjoy their own instead of mainstream.

It doesn't have to be a Niche. It should meet the needs of many people
*and* commercial interests. I won't pretend to understand the needs of
Google in pushing Andriod out the door, but if you don't meet the needs
of the big boys, an OS will never get much traction.

> There is ForthOS, which could be a base, because it runs on a normal PC without any other OS, but it doesn't have any graphics. This could be a starting point, adding OpenGL to it, if possible. Today I got the OpenGL Programming Guide by mail. It will need a while to read and understand, yet it's a start.

I suppose an alternative to Windows and Linux could gain traction by a
grassroots effort, but it is a long row to hoe. Linux happened because
it is a variation on Unix which had a lot of passionate developers.
ForthOS has... Gavino and he doesn't know how to program.

> PCs and laptops without OS are really cheap nowadays, so ForthOS with OpenGL and a suitable human interface would be great, I am sure of this.

I've never found a source for laptops without an OS or even a Linux
laptop that was as cheap as a laptop which included the MS tax. Do OS
free laptops exist?

> If this strikes a cord for enough people, a crowdfunding could be started to built a special PC (or better laptop) with Forth machines inside instead of Intel/AMD. Chuck Moore once said that PCs are messed up from their beginning, so a new design based on one or several Forth RISC machines would be a better way. Chuck said:

Is the PC hardware really so badly messed up? I think it is a fairly
stable platform with reasonable tradeoffs between retaining
compatibility with older standards and embracing new. The old PC
motherboards were built out of TTL and had a few glitches. As they
pulled this logic onto chips they had to include the glitches to retain
compatibility. Ok, that sounds bad, but it works ok today since
everyone know how to program around the glitches. This logic resides in
a tiny corner of the ASICs that make up a PC motherboard today and is
basically ignored other than during booting.

Just how large an effort would it be to start from scratch? Sounds
enormous. If you just work with existing hardware for now and develop a
fully functional OS, it can always be ported to better hardware if that
becomes important.

> How does this sound?
>
> Jeff Fox, the visionary, one wrote: "I would like to see $5 computers that are better than PCs. I would like to see Forth enable a new generation of very smart and cheap computers and enable a new generation of very clever children."

Coming from Jeff, it sounds rational. I can't say that when it comes
from Gavio. I'm still not sure about Steve. Your comments sound fairly
well based in reality. Gavino is running from left field into the
twilight zone. Steve is somewhere in between.

--

Rick

[rickman]

On 5/4/2013 2:46 PM, Andrew Haley wrote:

> Bernd Paysan<bernd....@gmx.de> wrote:
>
>> Companies like Google have a youth cult, and pretty strong age
>> discrimination, as the young programmers say, the old greybeards
>> don't want to program in Python and Java, and probably especially
>> not programm onions, but that's the way to go today.
>
> I don't think so: I was at dinner with a bunch of Googlers a month or
> two back, and there were a significant number of greybeards. From
> what I've seen, Google values experience.

I've heard the opposite. A friend who is an excellent programmer
interviewed with them over the phone and they were very enthusiastic
about his experience and background. But when he came in for a face to
face interview and they could see how old he was, they lost interest in
a hurry.

There is a difference between celebrating your long time employees and
"valuing experience". I have heard repeatedly how Google wants *young*
programmers because of a feeling that only the young can innovate.

I won't argue that they are totally wrong. I remember hearing some
statistic that when it comes to Nobels and other awards, if you haven't
done something important by the time you are thirty, the chances are you
never will. I'm about to have my second shot at the thirty threshold so
I guess I'm really doomed to mediocrity. lol

--

Rick

[Andrew Haley]

rickman <gnu...@gmail.com> wrote:
> On 5/4/2013 2:46 PM, Andrew Haley wrote:
>> Bernd Paysan<bernd....@gmx.de> wrote:
>>
>>> Companies like Google have a youth cult, and pretty strong age
>>> discrimination, as the young programmers say, the old greybeards
>>> don't want to program in Python and Java, and probably especially
>>> not programm onions, but that's the way to go today.
>>
>> I don't think so: I was at dinner with a bunch of Googlers a month or
>> two back, and there were a significant number of greybeards. From
>> what I've seen, Google values experience.
>
> I've heard the opposite. A friend who is an excellent programmer
> interviewed with them over the phone and they were very enthusiastic
> about his experience and background. But when he came in for a face to
> face interview and they could see how old he was, they lost interest in
> a hurry.

Maybe it depends on the team. I find that a very difficult-to-
believe story because they would have had his CV.

> There is a difference between celebrating your long time employees
> and "valuing experience".

Few have been at Google for very long, though. The greybeards of whom
I speak hadn't been there all that long.

Andrew.

[rickman]

On 5/4/2013 4:58 PM, Andrew Haley wrote:
> rickman<gnu...@gmail.com> wrote:
>> On 5/4/2013 2:46 PM, Andrew Haley wrote:
>>> Bernd Paysan<bernd....@gmx.de> wrote:
>>>
>>>> Companies like Google have a youth cult, and pretty strong age
>>>> discrimination, as the young programmers say, the old greybeards
>>>> don't want to program in Python and Java, and probably especially
>>>> not programm onions, but that's the way to go today.
>>>
>>> I don't think so: I was at dinner with a bunch of Googlers a month or
>>> two back, and there were a significant number of greybeards. From
>>> what I've seen, Google values experience.
>>
>> I've heard the opposite. A friend who is an excellent programmer
>> interviewed with them over the phone and they were very enthusiastic
>> about his experience and background. But when he came in for a face to
>> face interview and they could see how old he was, they lost interest in
>> a hurry.
>
> Maybe it depends on the team. I find that a very difficult-to-
> believe story because they would have had his CV.

CV doesn't give an age. I guess you could estimate, but he may not have
included all of his background. He has a pretty good CV too. Didn't
matter. He is the sort that can always find a good job because of his
obvious skills and talent.

What part do you not believe?

>> There is a difference between celebrating your long time employees
>> and "valuing experience".
>
> Few have been at Google for very long, though. The greybeards of whom
> I speak hadn't been there all that long.

How did they get in? How do you explain all the reports of age
discrimination? This isn't the first time I have heard of this.

--

Rick

[babr...@gmail.com]

On Saturday, May 4, 2013 4:24:11 PM UTC-4, rickman wrote:
> I've never found a source for laptops without an OS or even a Linux
> laptop that was as cheap as a laptop which included the MS tax.
> Do OS free laptops exist?

I am sure these exist. I forgot to mention that these are refurbished laptops at eBay.

The OLPC Initiative tried to create a laptop for 100$ with OpenBoot PROM inside.
The final price was 50% higher or more.

At eBay you get your refurbished OS-free personal laptop for less than 100$ including shipping. These are the cheapest.

Here an example:
http://www.ebay.com/itm/Dell-Latitude-D600-1-4GHz-Pentium-Mobile-512MB-RAM-30GB-DVD-RW-WiFi-/151039442819?pt=Laptops_Nov05&hash=item232aa6ff83

Or this one - ten Dell Latitude D600 (with RS232 DSub) for 600$, that means 60$ each:
http://www.ebay.com/itm/Dell-Latitude-D600-Lot-of-10-complete-Just-need-OS-Good-Batteries-/251270172481?pt=Laptops_Nov05&hash=item3a80de8f41

[Andrew Haley]

rickman <gnu...@gmail.com> wrote:
> On 5/4/2013 4:58 PM, Andrew Haley wrote:
>> rickman<gnu...@gmail.com> wrote:
>>> On 5/4/2013 2:46 PM, Andrew Haley wrote:
>>>> Bernd Paysan<bernd....@gmx.de> wrote:
>>>>
>>>>> Companies like Google have a youth cult, and pretty strong age
>>>>> discrimination, as the young programmers say, the old greybeards
>>>>> don't want to program in Python and Java, and probably especially
>>>>> not programm onions, but that's the way to go today.
>>>>
>>>> I don't think so: I was at dinner with a bunch of Googlers a month or
>>>> two back, and there were a significant number of greybeards. From
>>>> what I've seen, Google values experience.
>>>
>>> I've heard the opposite. A friend who is an excellent programmer
>>> interviewed with them over the phone and they were very enthusiastic
>>> about his experience and background. But when he came in for a face to
>>> face interview and they could see how old he was, they lost interest in
>>> a hurry.
>>
>> Maybe it depends on the team. I find that a very difficult-to-
>> believe story because they would have had his CV.
>
> CV doesn't give an age. I guess you could estimate, but he may not
> have included all of his background. He has a pretty good CV too.
> Didn't matter. He is the sort that can always find a good job
> because of his obvious skills and talent.
>
> What part do you not believe?

That they didn't guess out how old he was. I can tell when I read a
CV.

>>> There is a difference between celebrating your long time employees
>>> and "valuing experience".
>>
>> Few have been at Google for very long, though. The greybeards of whom
>> I speak hadn't been there all that long.
>
> How did they get in?

I don't know, but people who want to work at Google have to do the
usual things, such as application and interviews. The people I know
certainly are experts: in computational linguistics, in compiler
technology, and so on. But of course that's the kind of people that
work there.

> How do you explain all the reports of age discrimination?

Dunno; I haven't heard any. Have you considered that there might be
other reasons that he didn't get the job?

> This isn't the first time I have heard of this.

Andrew.

[Paul Rubin]

rickman <gnu...@gmail.com> writes:
> I've never found a source for laptops without an OS or even a Linux
> laptop that was as cheap as a laptop which included the MS tax.

You can get Chromebooks (Linux based) for a bit less than entry level
Windows laptops. There is also a pretty expensive Chromebook
(Chromebook Pixel) that is frankly nicer hardware than you can get for
Windows. I'm sort of tempted to buy one, but it has some shortcomings
that will probably keep using conventional laptops.

> Do OS free laptops exist?

Just bare metal, sold to computers? I haven't seen that. I think you
can still get some Dell laptops set up with FreeDOS which is an MSDOS
clone. Dell has had Linux lap offerings on and off, but they tend to
cost as much as, or more than, the same laptop with Windows.

The Lenovo Thinkpad that I'm using now came with Suse Linux and did have
a nice discount compared to the same hardware with Windows, but they
don't sell it that way any more.

[Bernd Paysan]

rickman wrote:
>> Maybe it depends on the team. I find that a very difficult-to-
>> believe story because they would have had his CV.
>
> CV doesn't give an age.

Well, if your CV contains something like "wrote the first Unix kernel in
1970", it is a strong hint... BTW: German CVs even contain a picture,
married/single and other things you wouldn't include in an US CV.

> I guess you could estimate, but he may not have
> included all of his background. He has a pretty good CV too. Didn't
> matter. He is the sort that can always find a good job because of his
> obvious skills and talent.
>
> What part do you not believe?

I suppose Andrew was together with a greybeard-only department ;-). People
gathering around Ken Thompson or so.

>>> There is a difference between celebrating your long time employees
>>> and "valuing experience".
>>
>> Few have been at Google for very long, though. The greybeards of whom
>> I speak hadn't been there all that long.
>
> How did they get in? How do you explain all the reports of age
> discrimination? This isn't the first time I have heard of this.

I think Andrew had a meeting with the Go team. That's an exception. The Go
team is very atypical for Google. Go is the language where you replace 20
overloaded servers with Java+Php+Python+other modern crap with 2 servers,
one of them lightly loaded, and the other is the hot spare. I suppose they
didn't go through the normal interview procedure.

IMHO the "you need to be under 30 to be innovative" is a fallacy. You need
to have done something impressive under 30 to qualify, but most of the
people who had done something impressive at that age did more impressive
things later.

Many young people are mediocre, just like many old people are mediocre.
Experience isn't counted in years, either. If you have done many impressive
things, you are experienced. It's not related to the length of the facial
hair.

Some companies work down their employees by driving them into burnout.
These companies need fresh blood, and they usually assume that the wear of
their employees is "due to age", not due to the overload. It only takes a
few years to wear down an engineer to a Wally in such an environment, and as
a result, the bosses of such companies think that people are only productive
below 30 - because they don't have anyone productive above 30. Given how
many FOSS projects died when Google hired the developers (of which you never
hear again, even though they would have one day per week to work on their
project), I suspect that Google is such a company. Many American companies
are. They are men-eaters.

[Elizabeth D. Rather]

A completely bare-metal laptop would be useful if and only if you had
enough docs to write drivers for it. In my experience, that's even rarer!

However, Chuck & co. were putting native colorforths on various
computers, including some laptops, so it should be possible. Presumably
there are still documented BIOS calls or something. If you have a native
colorforth, you could maybe copy its drivers.

Cheers,
Elizabeth

--
==================================================
Elizabeth D. Rather (US & Canada) 800-55-FORTH
FORTH Inc. +1 310.999.6784
5959 West Century Blvd. Suite 700
Los Angeles, CA 90045
http://www.forth.com

"Forth-based products and Services for real-time
applications since 1973."
==================================================

[babr...@gmail.com]

On Saturday, May 4, 2013 6:50:17 PM UTC-4, Paul Rubin wrote:

> rickman <gnuarm> writes:
>
> I think you can still get some Dell laptops set up with FreeDOS
> which is an MSDOS clone.

Instead of FreeDOS I would prefer ForthOS.
But I need a Graphic Desktop Environment, I am used to it, and it makes working easier.

KDE has a nice story, see http://en.wikipedia.org/wiki/KDE - but I don't think that KDE is compatible with ForthOS, and I am sure that KDE is not programmed the Forth way.

Thanks for appreciating the idea!

http://en.wikipedia.org/wiki/Cross-platform is another interesting report.

[babr...@gmail.com]

On Saturday, May 4, 2013 7:45:39 PM UTC-4, Elizabeth D. Rather wrote:
> However, Chuck & co. were putting native colorforths on various
> computers, including some laptops, so it should be possible.
> Presumably there are still documented BIOS calls or something.
> If you have a native colorforth, you could maybe copy its drivers.

What about Chuck's iTV web browser? Is this available somewhere?

Cheers,
Dirk.

[Elizabeth D. Rather]

Not that I know of. It was mostly not Chuck's, and not very successful.

[Paul Rubin]

"Elizabeth D. Rather" <era...@forth.com> writes:
> A completely bare-metal laptop would be useful if and only if you had
> enough docs to write drivers for it. In my experience, that's even
> rarer!

I just meant a standard x86 laptop, shipped with no OS so you can
install your own. I don't think you can buy them that way (no OS) at
the consumer level. You instead often end up paying for an OS that you
don't want to use, even in the cases where it's a Linux version, since
the laptop manufacturer is usually paying for some supported version of
Linux (SuSe in the case of my Thinkpad), so they don't have to handle
support calls themselves. An exception to that might be FreeDOS. I
don't know if Dell is paying anything for that. They do tell people not
to expect much support.

Here, you can save a little money by choosing FreeDOS instead of Windows:

http://www.dell.com/us/fedgov/p/precision-m6600/fs

[Steve]

Jason, you could start a groups comp.complaints is Usenet by going
through the registration processes. Definitely think it is a great
idea, and anybody could apply for sub groups. It is up to anybody to
try to do it.

If you want an alternative to Windows you can do it like Linus did
with Linux. There have been a few Forth OS's and processors and fpga
processors, to base your desire on and rework. Go over to my Misc
expansion.... thread and start of project. You are part of the
change you want, even if a little bit, you can do something. It is
pretty much for everybody, there are gripers and there are doers, and
people that can see what to do, only the last two groups can really
get things done together, listening to the first group.


Thanks


Steve

[Steve]

OK, lucky I read your post Dirk. Yes, you are right, Mr Jeff could
see, and eventually things may get cheap.

I can advise a little, because I spent many years digging for
solutions.


I advocated a simple management services OS that has access to open
API's to give common programming strategy and ease of incorporating
different hardware. The hardware could be interfaced and run by
Android system, and one day you translate Open API's and drivers to
Forth or Misc and just use all the software programs you have created
already on the Android, PC or whatever used to host the API's, you
then are very far along the life cycle faster because you started
easier without waiting for custom software and hardware.

Years ago there were open source universal driver initatives, which
would be suitable for you. You could at least code a compatible
system and auto translate the c code drivers to forth via a
translation program to get the first version of a driver running
before rewriting it in forth or misc. Now you have open API and
drivers on native misc or other hardware. If you do not wish to
translate and still interface and host it on a coprocessing system,
you could still have a wide choice of hardware to host Open API's and
drivers for forth to interface to. Ultimately, a misc C compiler
would allow these things to be compiled without translation (which is
needed, even to put Linux Java or Android on misc). So many ways to
skin a cat and solve a issue.

Dirk, GA is designing a self contained development system (hopefully
they expsnd it to include every valid interface as a reference design
over the years) that is a computer. You could negotiate to expand it
into a computer and take that to kickstarter?

Interfaces, unfortunately wirelesses is needed. I tried avoiding for
years, but now it is a must for the greater market, otherwise choices
would be simple. These days, you could almost get by doing a basic
comouter with one or two wired external interfaces. Thunderbolt is an
extension of PCI express, and could replace all other non wireless
interfaces if it had the support. You can run display port monitors ,
get HDMI conversion cables, do anything that USB could do, but I don't
know if there would be thunderbolt audio anytime soon. They wish to
put USB and firewire etc across it, but if that is still happening, or
cheap, I don't know. Going backwards, you can build in pci e slot
etc support. But in reality, you would still need USB and sata for
now. Wirelessly, in the short term (see my Misc Excpansion..... for a
bigger explanation) Bluetooth offers a hardware base and many
advantages, including another very low powered wireless standard now
for periphials, and cohosting 24Mb/s with wifi, and plans towards high
end speed. Unfortunately wireless USB, I have not seen any on
shelves, but it would be better. The other side of the coin is wifi
asnd wifi direct. With wifi direct it can host periphials. Wifi is
going to 7gb/s and maybe it will go higher like wihd one day (just my
idea). So, with bt and wifi you have a broad wireless basis, with one
of these or USB wireless as the only needed pan/devices wireless
interface in the future. So, you start to need only 2 or more
software and hardware interfaces each for hardware and wireless to
make a very functional system.


I think two big/biggest mistakes for Misc, is no real 18 bit+ address
bus, and no global access to that bus, or other busses, so nodes can
quickly report back or pull data distributedly, instead nodes are
specislised for streaming. With global access, a pool of nodes can
work independently buried in a chip, interfacing and sending results
as they wish. Much better. In this case even an 18x processor
becomes a competitor to all 16 bit-4 bit processors. Looking at it, I
would not mind designing my own processor version, I am more like an
architect or engineer than a builder laborer. I consider many ways to
do things. I think many ways that would make great processors. I
think Google sees something that can fill a need, and buy or fund it,
that is how we got Android. They used to have pages where yu can
submit proposals, they have the money. Maybe any of us could make
something they would need.

If every body did something towards something bigger than what we all
ready have, then we could have change. I'm looking at doing my own,
but everybody else could do what they could together. If you made a
successful programmabe product you could have a hundred new
permanent people here within months.


Thanks Dirk


Steve.

[Steve]

Rick, you have to see/feel suchbabfeikd to know where you are in it.

Gavino either has to find somebody that wants to do it, or do it
himself, the tools and the hardware are lwrgerly there and done. If
he used c to forth translation chain, he might be able to get firefox
OS running on misc. He could make a c compiler for misc, he could
make an interpreter engine to take the c code to an freindly
internediate and run that instead of eforth from the software bus.
Sure it might be very slow, but a step to his dreams instead of
waiting. Eventually misc may get faster wider better, and the
solution too.

I suspect that, just maybe, he is already working on such a solution,
and might be baiting to see our response.

Steve.
>
> --
>
> Rick

[Elizabeth D. Rather]

On 5/4/13 9:08 PM, Steve wrote:
...

> Gavino either has to find somebody that wants to do it, or do it
> himself, the tools and the hardware are lwrgerly there and done. If
> he used c to forth translation chain, he might be able to get firefox
> OS running on misc. He could make a c compiler for misc, he could
> make an interpreter engine to take the c code to an freindly
> internediate and run that instead of eforth from the software bus.
> Sure it might be very slow, but a step to his dreams instead of
> waiting. Eventually misc may get faster wider better, and the
> solution too.
>
> I suspect that, just maybe, he is already working on such a solution,
> and might be baiting to see our response.

You can't have been here very long. In between the repeated wishes for
somebody else to do it in Forth are complaints that Forth is too hard
for him to learn. This has been going on for years.

[Andrew Haley]

Bernd Paysan <bernd....@gmx.de> wrote:
>
> I think Andrew had a meeting with the Go team.

No, actually.

Andrew.

[Albert van der Horst]

In article <buSdnd-tBZE-ABjM...@supernews.com>,

Elizabeth D. Rather <era...@forth.com> wrote:
>On 5/4/13 12:50 PM, Paul Rubin wrote:
>> rickman <gnu...@gmail.com> writes:
>>> I've never found a source for laptops without an OS or even a Linux
>>> laptop that was as cheap as a laptop which included the MS tax.
>>
>> You can get Chromebooks (Linux based) for a bit less than entry level
>> Windows laptops. There is also a pretty expensive Chromebook
>> (Chromebook Pixel) that is frankly nicer hardware than you can get for
>> Windows. I'm sort of tempted to buy one, but it has some shortcomings
>> that will probably keep using conventional laptops.
>>
>>> Do OS free laptops exist?
>>
>> Just bare metal, sold to computers? I haven't seen that. I think you
>> can still get some Dell laptops set up with FreeDOS which is an MSDOS
>> clone. Dell has had Linux lap offerings on and off, but they tend to
>> cost as much as, or more than, the same laptop with Windows.
>>
>> The Lenovo Thinkpad that I'm using now came with Suse Linux and did have
>> a nice discount compared to the same hardware with Windows, but they
>> don't sell it that way any more.
>>
>
>A completely bare-metal laptop would be useful if and only if you had
>enough docs to write drivers for it. In my experience, that's even rarer!
>
>However, Chuck & co. were putting native colorforths on various
>computers, including some laptops, so it should be possible. Presumably
>there are still documented BIOS calls or something. If you have a native
>colorforth, you could maybe copy its drivers.

I'm afraid it will be more like copying the path Chuck followed, by
finding out how to call the BIOS for a few selected calls.
The exact example is only useful on the exact same hardware.

>
>Cheers,
>Elizabeth
--
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

[Alex McDonald]

On May 4, 7:46 pm, Andrew Haley <andre...@littlepinkcloud.invalid>
wrote:

Agreed. As an employee of a Silicon Valley company, I can vouch that
folks like me (I'm in my late 50s and I have a grey beard) are well
represented. Pick any of the household Valley names, and the same is
true. I've never met anyone that doesn't want to program in a specific
language because of age.

[Doug Hoffman]

On 5/3/13 2:17 PM, Jason Damisch wrote:

> I talked to a lady
> who worked for another business there. She told me that she wished
> that there was an alternative to Windows, Mac, Linux etc.

What were her specific complaints with OS X?

No experience with Linux so I can't comment. Plenty of experience with
Windows, but then we have a large IT department at work to support the
many problems we rncounter with our Windows machines and applications.
My lone home Windows XP machine sort of works, can't get the wi-fi to
connect, and it crashes a lot. I don't know anything about anti virus
software so I never use my Windows machine for internet access and
definitely never put sensitive info on it.

Lots of positive experience with OS X at home.

> I see that there is a disconnect between computer users and computer
> programmers. There still is. I also see that customer satisfaction
> varies widely.

At least for *personal computer* use I am hard pressed to fault OS X.
Is it perfect? Of course not. But it is darned good, IMO. One great
thing is my wife can use her MacBook Pro with almost zero support from
our IT department(me). She is a very smart gal but knows nothing about
how computers work. Still, she has no problem with email, web browsing,
buying lots of stuff via the internet, writing and printing letters with
a word processor, editing and printing photographs from the digital
camera or her iPhone, buying music and movies from iTunes, using
facetime with friends and family. She has even upgraded her computer's
software without my help. She can also drive and operate her car quite
well but has no idea about what makes a car work.

We have of course a variety of the typical computer related items which
simply plug and play and operate seamlessly/painlessly with our wireless
HP all-in-one printer/copier. We stream movies, music, and so forth via
our AppleTV for home theater use. She loves her iPhone (which I don't
use and don't need to learn how).

Things aren't perfect. But I remember the days of DOS back in the early
1980s and I think it's fair to say we have come a *very* long way since
then, at least with OS X. Windows too actually, but OS X gives me far
fewer problems. I use a lot more software on my Mac than my does
including a spreadsheet, CAD, and more. Frankly, it all amazes me.

-Doug

[Steve]

Elizabeth D. Rather wrote:
> On 5/4/13 9:08 PM, Steve wrote:
> ...
> > Gavino either has to find somebody that wants to do it, or do it
> > himself, the tools and the hardware are lwrgerly there and done. If
> > he used c to forth translation chain, he might be able to get firefox
> > OS running on misc. He could make a c compiler for misc, he could
> > make an interpreter engine to take the c code to an freindly
> > internediate and run that instead of eforth from the software bus.
> > Sure it might be very slow, but a step to his dreams instead of
> > waiting. Eventually misc may get faster wider better, and the
> > solution too.
> >
> > I suspect that, just maybe, he is already working on such a solution,
> > and might be baiting to see our response.
>
> You can't have been here very long. In between the repeated wishes for
> somebody else to do it in Forth are complaints that Forth is too hard
> for him to learn. This has been going on for years.
>
> Cheers,
> Elizabeth

So, you mean he's not some sort of hidden programming genius, like in
Kung Fu Hustle. Too hard to learn, is he a c or a lisp programmer? I
think if he puts in the time he could do it.

I've make a list that could help him:
http://www.forth.com/starting-forth/ This shouldn't be too difficult
unless he is a logo programmer, or basic.
http://thinking-forth.sourceforge.net/

Some of yours:
http://www.forth.com/forth/fph.html
http://www.forth.com/forth/fat.html
http://www.forth.com/swiftforth/dl.html

Down the end is some experimental c to forth translation code, kindly
put out by Stephen
http://www.mpeforth.com/arena.htm

I'm needing to brush up myself on the latest versions/thinking in
Forth. Would there be anything else you would recommend Elizabeth?


Thanks again,


Steve.

[AKE]

On Saturday, May 4, 2013 9:31:35 PM UTC+1, rickman wrote:
> I have heard repeatedly how Google wants *young*
> programmers because of a feeling that only the young can innovate.
>

> I remember hearing some
> statistic that when it comes to Nobels and other awards, if you haven't
> done something important by the time you are thirty, the chances are you
> never will. I'm about to have my second shot at the thirty threshold so
> I guess I'm really doomed to mediocrity. lol
>

There are numerous counter-examples: Knuth -- still productive into his eighties, Erdos -- another octagenerian mathematician who was tremendously productive. Going back a century, Gauss, Euler. ... Chuck Moore is not so young himself.

Age and important work I think are not correlated -- in fact, I think better work comes with more experience.

For a radically new approach, yes I agree that is more likely to happen with fresh genius, but IMO it is the freshness of perspective, not age per se, that matters.

[Mark Wills]

On May 4, 9:58 pm, Andrew Haley <andre...@littlepinkcloud.invalid>
wrote:

> rickman <gnu...@gmail.com> wrote:
> > On 5/4/2013 2:46 PM, Andrew Haley wrote:

> >> Bernd Paysan<bernd.pay...@gmx.de>  wrote:

CV's are always anonymised for legal reasons and should not contain
any information that gives away age, gender, race, sexuality etc.

[Mark Wills]

On May 5, 12:50 am, "visualfo...@rocketmail.com" <babru...@gmail.com>
wrote:

> On Saturday, May 4, 2013 6:50:17 PM UTC-4, Paul Rubin wrote:
> > rickman <gnuarm> writes:
>
> >  I think you can still get some Dell laptops set up with FreeDOS
> > which is an MSDOS clone.
>
> Instead of FreeDOS I would prefer ForthOS.
> But I need a Graphic Desktop Environment, I am used to it, and it makes working easier.
>

> KDE has a nice story, seehttp://en.wikipedia.org/wiki/KDE- but I don't think that KDE is compatible with ForthOS, and I am sure that KDE is not programmed the Forth way.

>
> Thanks for appreciating the idea!
>
> http://en.wikipedia.org/wiki/Cross-platformis another interesting report.

You don't need a graphic IDE, you just think you do! Did you ever try
Microsoft's QuickBasic or QuickC IDE's in the 80's? These ran under
DOS and were excellent. They were mouse or keyboard based and had
windows, albeit character based ones. Really great environment.
Borland's were equally as good.

I remember in the Microsoft environments there was a command line
switch you could use to run them in 80x50. It was really good in that
mode!

Ah, memories!

[rickman]

Of course we aren't mind readers. But like I said, he was very
qualified and they were very enthusiastic about him until they met him
in person. He is not the type that has issues when meeting people, just
the opposite, he tends to leave a very positive impression.
But like many of us, his age is not hidden in his appearance.

The interview is seldom about discovering new positives about a person.
The interview confirms that the resume is reasonably accurate, the
person has no personality issues, but just as important, that they fit
in with the "culture". In this case the word on the street is that the
culture includes youth.

>> This isn't the first time I have heard of this.

If it were an isolated instance, I would not think a lot of it. But
Google has a reputation. Also, this was some time ago. Perhaps your
greybeards being new hires indicate a change in thinking. Or perhaps
for the true experts they will ignore age, but for the other 90%, age
ends up being a factor. I can't say and since everyone knows the law,
no one will talk honestly about their hiring practices. I do firmly
believe that age discrimination exists. I have seen it any number of
times with myself and others.

--

Rick

[rickman]

On 5/5/2013 5:38 AM, Albert van der Horst wrote:
> In article<buSdnd-tBZE-ABjM...@supernews.com>,
> Elizabeth D. Rather<era...@forth.com> wrote:
>> On 5/4/13 12:50 PM, Paul Rubin wrote:
>>> rickman<gnu...@gmail.com> writes:
>>>> I've never found a source for laptops without an OS or even a Linux
>>>> laptop that was as cheap as a laptop which included the MS tax.
>>>
>>> You can get Chromebooks (Linux based) for a bit less than entry level
>>> Windows laptops. There is also a pretty expensive Chromebook
>>> (Chromebook Pixel) that is frankly nicer hardware than you can get for
>>> Windows. I'm sort of tempted to buy one, but it has some shortcomings
>>> that will probably keep using conventional laptops.
>>>
>>>> Do OS free laptops exist?
>>>
>>> Just bare metal, sold to computers? I haven't seen that. I think you
>>> can still get some Dell laptops set up with FreeDOS which is an MSDOS
>>> clone. Dell has had Linux lap offerings on and off, but they tend to
>>> cost as much as, or more than, the same laptop with Windows.
>>>
>>> The Lenovo Thinkpad that I'm using now came with Suse Linux and did have
>>> a nice discount compared to the same hardware with Windows, but they
>>> don't sell it that way any more.
>>>
>>
>> A completely bare-metal laptop would be useful if and only if you had
>> enough docs to write drivers for it. In my experience, that's even rarer!
>>

>> However, Chuck& co. were putting native colorforths on various

>> computers, including some laptops, so it should be possible. Presumably
>> there are still documented BIOS calls or something. If you have a native
>> colorforth, you could maybe copy its drivers.
>
> I'm afraid it will be more like copying the path Chuck followed, by
> finding out how to call the BIOS for a few selected calls.
> The exact example is only useful on the exact same hardware.

I don't think that is exactly right. The purpose of the BIOS is to
provide hardware independent calls for standard I/O functions. The BIOS
is customized to match the hardware so that the software can be hardware
independent.

The limitation is that this only gets you a minimum subset of features
for the common denominator of hardware. For more advanced features you
need specific drivers which won't exist for a "rogue" OS, if I can call
it that. The other major limitation for such a machine would be the
total lack of driver support for nearly all peripherals such as
printers, etc. In the old days you just sent text, but we are *far*
beyond that now and every brand of printer, if not every model of
printer, needs a different driver that knows the specifics of how it
words. Some time back they had processors in printers to generate the
graphics. I think that is all done in the PC now and something more
like an image is sent to the printer. Or maybe the pendulum has swung
back to the other side and an ARM9 is more than enough processing to do
it all in the printer again.

But you get the point... drivers are critical and they are a moving
target. Even using Windows you may have to buy a new printer if you buy
a new machine with a new version of Windows. It has happened more than
once. It would be nearly impossible to support just one brand of
printers, not to mention all the other devices you would want to use.

--

Rick

[rickman]

On 5/5/2013 3:08 AM, Steve wrote:
>
> rickman wrote:
>> On 5/4/2013 2:24 AM, visua...@rocketmail.com wrote:
>>>
>
>>
>> Coming from Jeff, it sounds rational. I can't say that when it comes
>> from Gavio. I'm still not sure about Steve. Your comments sound fairly
>> well based in reality. Gavino is running from left field into the
>> twilight zone. Steve is somewhere in between.
>
> Rick, you have to see/feel suchbabfeikd to know where you are in it.
>
> Gavino either has to find somebody that wants to do it, or do it
> himself, the tools and the hardware are lwrgerly there and done. If
> he used c to forth translation chain, he might be able to get firefox
> OS running on misc. He could make a c compiler for misc, he could
> make an interpreter engine to take the c code to an freindly
> internediate and run that instead of eforth from the software bus.
> Sure it might be very slow, but a step to his dreams instead of
> waiting. Eventually misc may get faster wider better, and the
> solution too.

Uh, C to Forth running on MISC? Really, that's all there is to it?

To be useful, you can't just run an emulator. That would be absurdly
slow. I think they said the emulator runs 1 VAX MIPS. I don't know
what that is in modern terms, but I know it is *slow*. I think the
processor in my 12 year old GPS runs faster than that from batteries.

You need to tap the power of the many processors in the GA144. TI makes
a DSP that uses VLIW. It has eight CPUs that all need a 32 bit
instruction at the same time. But each CPU is not really a full blown
CPU. They are micro CPUs with full blown ALU capability and not much
else. To access memory, you use a CPU to generate the address and fetch
the data into the register file for another CPU, etc.

MISC is similar. You have to decompose your problem into tiny pieces
that can run in the extremely limited memory on each CPU and they all
have to coordinate. An X86 or ARM CPU has similar micro CPUs that are
hard coded in terms of where they are in the processing chain, what
functions they can execute and who they talk to, so you don't need to
think about that part, it is done for you. In MISC to have to start
from scratch and figure it all out yourself.

A C to Forth translator won't begin to scratch the surface writing a
useful browser on a GA144.

> I suspect that, just maybe, he is already working on such a solution,
> and might be baiting to see our response.

lol...

--

Rick

[Coos Haak]

Op Sun, 5 May 2013 13:02:04 -0700 (PDT) schreef Mark Wills:

Vista supports 16 bit programs in a 80x50 window. And nicer: black
characters on white background. Alas, no full screen mode as in the older
Windows XP.

--
Coos

CHForth, 16 bit DOS applications
http://home.hccnet.nl/j.j.haak/forth.html

[Rod Pemberton]

"rickman" <gnu...@gmail.com> wrote in message
news:km6ho1$9c1$1...@dont-email.me...

> On 5/5/2013 3:08 AM, Steve wrote:

> > Gavino either has to find somebody that wants to do it, or do
> > it himself, the tools and the hardware are lwrgerly there and
> > done. If he used c to forth translation chain, he might be
> > able to get firefox OS running on misc.
>

> Uh, C to Forth running on MISC? Really, that's all there is to
> it?
>

I've converted a ancient, very simple, incomplete C compiler to
emit Forth. I'm not sure which C compiler Firefox needs to
compile, but I can easily guess Firefox needs a powerful, modern C
compiler to compile. A project to modify a powerful ISO C
compiler to emit Forth is going to be as large as coding a Firefox
clone in Forth. The conversion can be done, but that's generally
not a good solution. Although, it could be, if there is enough
of a market to make money off of Forth-to-C conversions, after the
fact, since one spent so much time and energy to create it. But,
is that what a Forth programmer wants? I.e., C code which is
badly translated to Forth. It'd be nearly impossible for Forth
programmers to maintain.


Rod Pemberton

[Rod Pemberton]

"Elizabeth D. Rather" <era...@forth.com> wrote in message
news:p9idnR5YCa-8jBvM...@supernews.com...

> In between the repeated wishes for somebody else to do it in

> Forth are complaints that Forth is too hard for [Gav] to learn.

Yes, that does beg the question why he wants a PC based on Forth
in the first place...


RP

[Paul Rubin]

"Rod Pemberton" <do_no...@notemailnotq.cpm> writes:
> I can easily guess Firefox needs a powerful, modern C
> compiler to compile.

Firefox is written in C++. There is an experimental effort aimed at
eventually rewriting it in Rust (rust-lang.org).

[Rod Pemberton]

"Paul Rubin" <no.e...@nospam.invalid> wrote in message
news:7xsj222...@ruckus.brouhaha.com...

> "Elizabeth D. Rather" <era...@forth.com> writes:

> > A completely bare-metal laptop would be useful if and only if
> > you had enough docs to write drivers for it. In my experience,
> > that's even rarer!
>
> I just meant a standard x86 laptop, shipped with no OS so you
> can install your own.

I want that, but I build my computers from parts. You can build a
faster machine or a cheaper machine than offered by the major
companies that way. However, the majority of the market doesn't
want the ability to build computers from parts. It's comprised of
ordinary people. They buy computers like they buy cars or a
microwave: complete. The market is not comprised of programmers
and advanced computer users. There is no way that my parents
could ever correctly install Linux, noting that I haven't tried
out Ubuntu yet... It takes hours of me fine tuning new machines
with factory Windows installs for them, be it from Dell or
Samsung, etc. Those are already highly customized for ease of
use.

> I don't think you can buy them that way (no OS) at
> the consumer level. You instead often end up paying for an OS

> that you don't want to use, [...]

Many people want Windows since they can play games with it with
full video card support. Many people want a machine that "just
works". As of 2013, that's not possible with Linux, IMO. There
is so much hardware that just won't work with it or won't use
Windows drivers. Windows is the primary OS that's installed on
machines and will remain so for just that reason.

So, how do you come to the conclusion of people "often" buying an
OS they don't want? Was that just *your* perspective projected as
everyone's? If it was your opinion, why did you say "you" twice
instead of "I" ... ?


Rod Pemberton

[rickman]

On 5/5/2013 6:51 PM, Rod Pemberton wrote:
> "Paul Rubin"<no.e...@nospam.invalid> wrote in message
> news:7xsj222...@ruckus.brouhaha.com...
>> "Elizabeth D. Rather"<era...@forth.com> writes:
>
>>> A completely bare-metal laptop would be useful if and only if
>>> you had enough docs to write drivers for it. In my experience,
>>> that's even rarer!
>>
>> I just meant a standard x86 laptop, shipped with no OS so you
>> can install your own.
>
> I want that, but I build my computers from parts. You can build a
> faster machine or a cheaper machine than offered by the major
> companies that way.

Yes, you can build a desktop machine from parts, even *I* have done
that. But you can't build a laptop.

--

Rick

[Elizabeth D. Rather]

On 5/5/13 10:02 AM, Mark Wills wrote:
> On May 5, 12:50 am, "visualfo...@rocketmail.com" <babru...@gmail.com>
> wrote:
>> On Saturday, May 4, 2013 6:50:17 PM UTC-4, Paul Rubin wrote:
>>> rickman <gnuarm> writes:
>>
>>> I think you can still get some Dell laptops set up with FreeDOS
>>> which is an MSDOS clone.
>>
>> Instead of FreeDOS I would prefer ForthOS.
>> But I need a Graphic Desktop Environment, I am used to it, and it makes working easier.
>>
>> KDE has a nice story, seehttp://en.wikipedia.org/wiki/KDE- but I don't think that KDE is compatible with ForthOS, and I am sure that KDE is not programmed the Forth way.
>>
>> Thanks for appreciating the idea!
>>
>> http://en.wikipedia.org/wiki/Cross-platformis another interesting report.
>
> You don't need a graphic IDE, you just think you do! Did you ever try
> Microsoft's QuickBasic or QuickC IDE's in the 80's? These ran under
> DOS and were excellent. They were mouse or keyboard based and had
> windows, albeit character based ones. Really great environment.
> Borland's were equally as good.

It depends what this machine is being used for. Most of the serious
discussion of misc processors has been focused on embedded applications.
For this purpose, an interactive cross-compiler running on a standard PC
or Mac is the most practical, most powerful, and most convenient.

If you're trying to make an alternative misc-based PC with a Forth-based
OS, you're taking on a significant task beyond the GUI: configuring
peripherals, writing drivers, etc. What is the market for this device? I
really don't see it as a viable project.

[Rod Pemberton]

"rickman" <gnu...@gmail.com> wrote in message

news:km6nhs$74p$1...@dont-email.me...

Yes, you can, or could... OCZ was selling DIY laptop shells. I
don't see them on their site today. AIR, the "shell" was far more
complete than a normal PC, just needing a few parts. I also
recall the "shell" not being quite top-of-the-line performance
wais as compared to other complete laptops, e.g., looked like
surplus new-old-stock to me...

E.g., articles from a few years back:
http://www.brighthub.com/computing/hardware/articles/36914.aspx
http://www.engadget.com/2008/06/06/oczs-diy-gaming-laptop-now-available/


Rod Pemberton

[Paul Rubin]

AKE <assadebr...@gmail.com> writes:
> There are numerous counter-examples: Knuth -- still productive into
> his eighties,

Knuth (born January 10, 1938) turned 75 this year.

[Paul Rubin]

Alex McDonald <bl...@rivadpm.com> writes:
> I've never met anyone that doesn't want to program in a specific
> language because of age.

How many teenage Forth programmers do you know?

[Steve]

rickman wrote:
> On 5/5/2013 3:08 AM, Steve wrote:
> >
> > rickman wrote:
> >> On 5/4/2013 2:24 AM, visua...@rocketmail.com wrote:
> >>>
> >
> >>
> >> Coming from Jeff, it sounds rational. I can't say that when it comes
> >> from Gavio. I'm still not sure about Steve. Your comments sound fairly
> >> well based in reality. Gavino is running from left field into the
> >> twilight zone. Steve is somewhere in between.
> >
> > Rick, you have to see/feel suchbabfeikd to know where you are in it.

Hmm, that one got past me, slightly off screen when I quickly proof
read, sorry. "see/feel such a field to know..." that should have been.

> >
> > Gavino...

>
> Uh, C to Forth running on MISC? Really, that's all there is to it?

No, we are not psychic but we can fill in the detail.

.>

> To be useful, you can't just run an emulator. That would be absurdly
> slow. I think they said the emulator runs 1 VAX MIPS. I don't know
> what that is in modern terms, but I know it is *slow*. I think the
> processor in my 12 year old GPS runs faster than that from batteries.

Which emulator Mr Rick?

I think as much as you think Gavino whines about about not having
things, you whine about how something is not possible rather than
finding or even seeing solutions. You don't do engineering by just
looking up books of others complete designs to use, you have to work
out a list of things and make the best solution you can. You are on
the opposite side of feild to Gavino, and I constantly wait for you to
come out of early morning twilight. Now I have answered adequately
many of your objections before. You know by now, that I see such
emulation, and you skipped most of the other several alternatives to
emulation I gave, more for a future enhanced misc cores more than
current. That on the current you take more of a hit, and you only
need to get a product and foot in the door to get started. That even
something equivalent to a 66mhz performance is useful, but expect at
least a couple of hundred. Plus, you are blind to the obvious
solutions to get that working a lot faster than you think. You find
fault not solutions. It is as if both of you are getting pleasure
baiting the conversation. You maybe subconsciously, Gavino maybe
consciously, like ying and yang twins, though Gavino's posting looks
like it could be written by somebody who just had a good night out on
the town. I'd rather not spend an hour or two answering obvious
questions per post, I've already put 60-100 hours into this that are
needed on other work.

> You need to tap the power of the many processors in the GA144. TI makes
> a DSP that uses VLIW. It has eight CPUs that all need a 32 bit
> instruction at the same time. But each CPU is not really a full blown
> CPU. They are micro CPUs with full blown ALU capability and not much
> else. To access memory, you use a CPU to generate the address and fetch
> the data into the register file for another CPU, etc.
>
> MISC is similar. You have to decompose your problem into tiny pieces
> that can run in the extremely limited memory on each CPU and they all
> have to coordinate. An X86 or ARM CPU has similar micro CPUs that are
> hard coded in terms of where they are in the processing chain, what
> functions they can execute and who they talk to, so you don't need to
> think about that part, it is done for you. In MISC to have to start
> from scratch and figure it all out yourself.

Maybe you are psychic, and just started to get it and say something
progressive after what I just wrote because you somehow knew I was
going to write it ;-) . No, truly, thank you, this is exactly what I
am thinking, way beyond what you think, ways to leverage out as much
CPU potential in parallel. I even mentioned a little, like what you
said here, to spark people's interest. Now you are thinking like an
architect. You start to see the field to see where you are in it. My
ideas, as a solution without hand rewriting c to forth, is right smack
in a viable play zone of the field.

> A C to Forth translator won't begin to scratch the surface writing a
> useful browser on a GA144.

You know I have said a translator that translates smoothly with little
hands in, to get quickly through the room of OS code, not one that
requires constant intervention, because it is not practical to do it
that way, or rewrite everything.

>
>
> > I suspect that, just maybe, he is already working on such a solution,
> > and might be baiting to see our response.
>
> lol...
>

More light :-)
But honestly, what do really know of him, he could be taunting us but
really working for all we know.
Gavino, come forth and reveal some of your code to us?

> --
>
> Rick


Thanks Mr Rick


Steve.

[Steve]

Rod Pemberton wrote:
> "rickman" <gnu...@gmail.com> wrote in message
> news:km6ho1$9c1$1...@dont-email.me...
> > On 5/5/2013 3:08 AM, Steve wrote:
>

> > > Gavino...

> >
> > Uh, C to Forth running on MISC? Really, that's all there is to
> > it?
> >
>
> I've converted a ancient, very simple, incomplete C compiler to
> emit Forth. I'm not sure which C compiler Firefox needs to
> compile, but I can easily guess Firefox needs a powerful, modern C
> compiler to compile. A project to modify a powerful ISO C
> compiler to emit Forth is going to be as large as coding a Firefox
> clone in Forth. The conversion can be done, but that's generally
> not a good solution. Although, it could be, if there is enough
> of a market to make money off of Forth-to-C conversions, after the
> fact, since one spent so much time and energy to create it. But,
> is that what a Forth programmer wants? I.e., C code which is
> badly translated to Forth. It'd be nearly impossible for Forth
> programmers to maintain.

Hi Rod, thanks for this, could you post some linjsvto your project,
and any others you know, we have Stephen Pelc's translator over at the
Misc Expansion... thread.

When you say modern C compiler is too difficult to modify, are you
talking about the project code management side, or c++?

Hmm, I seem to be mixing up concepts from two theads here myself,
firefox +OS, firefox OS, and full android OS to forth or misc. I'm
more for the misc perspective, others for PC. If he really waned a
forth PC with firefox, he could do a forth OS/openfirmware expansion
with universal open driver solution from Linux, and open api's, and
Linux firefox in a runtime solution.

Thanks

Steve

[Jason Damisch]

https://i.chzbgr.com/maxW500/7143000576/h840DC5EF/

lol

[Steve]

Elizabeth D. Rather wrote:
> On 5/5/13 10:02 AM, Mark Wills wrote:
> > On May 5, 12:50 am, "visualfo...@rocketmail.com" <babru...@gmail.com>
> > wrote:
> >> On Saturday, May 4, 2013 6:50:17 PM UTC-4, Paul Rubin wrote:
> >>> rickman <gnuarm> writes:


> It depends what this machine is being used for. Most of the serious
> discussion of misc processors has been focused on embedded applications.
> For this purpose, an interactive cross-compiler running on a standard PC
> or Mac is the most practical, most powerful, and most convenient.
>
> If you're trying to make an alternative misc-based PC with a Forth-based
> OS, you're taking on a significant task beyond the GUI: configuring
> peripherals, writing drivers, etc. What is the market for this device? I
> really don't see it as a viable project.
>
> Cheers,
> Elizabeth
>

Very carefully. Any new system is on a knife edge of existence these
days, so it pays to get the bluntest lowest profile blade so if you
fall off you can climb back on. Blunt as in you have very
significant appeal for people to use you, low profile as in you need
very little investment to do and progress, so if you mis step you can
more easily climb back on board again. For appeal, you either have
it, and/or for an eco system you have money to produce it, or open
source community to produce it. Seriously, if you had ten thousand
node 32 bit processor, with good distributed parallel programming
model running at up to 10 Ghz each node, available at affordable cost
to developers today (as you should) people would be lining up to
develop open source for such a thing. Making it compatible with
existing ego system makes it a natural leap progression for those ego
systems.

As it is, you have to very carefully strategically plan you entry into
the market, for a must system to attract customers. Misc offers
natively low powered high execution advantages that are hard to sell
even to the technology savy these days, so the product these enable is
what you sell, the processor becoming part of the back story of its
success and effectiveness. Ring computers being a perfect example
that is hard for others to compete as effectively in. Having Android
on it gives a ready market to sell into, but also places the potential
to leverage native code and system software on these devices, to build
up a eco system ifnthey back ground where eventually you can go native
devices in years to come. The reason it is worth doing, is as far as
non optical non quantum silicon use goes, Misc is effectively an end
game array processing product, able to outstrip existing mainline
competitors when processing performance gains start to get hard to
very hard to achieve. It should outstrip a minimal Arm core. As far
as a Forth PC with normal x86, sold as a Forth PC as the main feature,
today hardly any market, maybe a few thousands. You might as well
advance a forth OS, as an open software distribution instead, and go
and buy a PC to use it on. I think the GA might sell many more
development systems to an enthusiast/electronics market than a Forth
PC. A Forth OS, or PC, is not about the practicality, but the love,
and sometimes, just sometimes, that love pays off, as in Linux's case,
but a forth OS Forth PC is going to need to be Misc alongside PC forth
OS version to make it special (universal driver architecture is worth
checking out).


Thanks

Steve.

[Elizabeth D. Rather]

I'm sorry, but I really can't make heads or tails out of the paragraph
above. That has nothing to do with misc or the discussion at hand.

> As it is, you have to very carefully strategically plan you entry into
> the market, for a must system to attract customers. Misc offers
> natively low powered high execution advantages that are hard to sell
> even to the technology savy these days, so the product these enable is
> what you sell, the processor becoming part of the back story of its
> success and effectiveness.

On the contrary, very low-powered devices have a considerable market.
But it's mainly in embedded systems. This is a vast, rapidly growing
market. Here is just one recent article on it:
http://www.cio.com/article/689563/IDC_Embedded_Systems_Market_to_Double_By_2015

What it's hard to realize for folks whose main exposure to computers is
PCs and smartphones, is that most of these units do not look or act like
consumer computers, and have very different needs. Although some
internet connectivity is relevant, browsers and GUIs generally aren't.
Yes, the article mentions Windows Embedded OS, but that's a joke to most
people making microprocessor controlled widgets.

> Ring computers being a perfect example
> that is hard for others to compete as effectively in. Having Android
> on it gives a ready market to sell into, but also places the potential
> to leverage native code and system software on these devices, to build
> up a eco system ifnthey back ground where eventually you can go native
> devices in years to come. The reason it is worth doing, is as far as
> non optical non quantum silicon use goes, Misc is effectively an end
> game array processing product, able to outstrip existing mainline
> competitors when processing performance gains start to get hard to
> very hard to achieve. It should outstrip a minimal Arm core.

Again, the GA devices are aimed at a very, very different market from Arms.

> As far
> as a Forth PC with normal x86, sold as a Forth PC as the main feature,
> today hardly any market, maybe a few thousands. You might as well
> advance a forth OS, as an open software distribution instead, and go
> and buy a PC to use it on. I think the GA might sell many more
> development systems to an enthusiast/electronics market than a Forth
> PC. A Forth OS, or PC, is not about the practicality, but the love,
> and sometimes, just sometimes, that love pays off, as in Linux's case,
> but a forth OS Forth PC is going to need to be Misc alongside PC forth
> OS version to make it special (universal driver architecture is worth
> checking out).

GA is trying to sell embedded system processors, not Forth PCs, Forth
OSs (in the conventional sense of an OS with a user interface, etc.), or
development systems. You are taking gavino's pleas much too seriously.
Spend some time in GA's web site and look at where their interests
really lie.

[Steve]

The reality is that it is possible to do that and desirable. That the
other market might be mire difficult. Elizabeth, remember what you
said about Misc design and markets.

>
> > As far
> > as a Forth PC with normal x86, sold as a Forth PC as the main feature,
> > today hardly any market, maybe a few thousands. You might as well
> > advance a forth OS, as an open software distribution instead, and go
> > and buy a PC to use it on. I think the GA might sell many more
> > development systems to an enthusiast/electronics market than a Forth
> > PC. A Forth OS, or PC, is not about the practicality, but the love,
> > and sometimes, just sometimes, that love pays off, as in Linux's case,
> > but a forth OS Forth PC is going to need to be Misc alongside PC forth
> > OS version to make it special (universal driver architecture is worth
> > checking out).
>
> GA is trying to sell embedded system processors, not Forth PCs, Forth
> OSs (in the conventional sense of an OS with a user interface, etc.), or
> development systems. You are taking gavino's pleas much too seriously.
> Spend some time in GA's web site and look at where their interests
> really lie.

But that is not where most of interests lay, and what we do, and what
we pay for are different from that. So, yes, the future Misc is said
to be a different beast, and as has happened before, great successes
in the industry where made in areas that parts were not originally
intended for. Integration is mapping up external logic parts on
boards, so in the end won't you do compete with arms? So, we are
trying to make a go at a separate market. If you have the money GA
designs custom versions, but it sounds like the 32 bit version, maybe
pretty good as is.

>
> Cheers,
> Elizabeth
>

Thanks agan.

[Steve]

Sorry, I missed this. Everything at hand, if you want to make a
desktop machine it GA to be reliable to justify it like has been said,
but that is very difficult. In order to do it, your require strategy
to take advantages of natural advantages and to build up eco systems
and product appeal to attract people. It is very simple. As it is, a
forth pc offers very little potential in agreement with what you said.

About the embedded market, unfortunately I have deleted the quote.
But as for the conventional market, as I said things are drying up,
but I was thinking of mentioning the new embedded market on the other
hand, like you have, but forgot. That yes there is a big market
there, but large sections of that will ve crushed by conventional
standardized asic's in time. Some sections will need cores for
decision making, but it is not like a desktop much of it is simple
standardized tasks, but there is bound to be a ssizable section that
still requires on array for more dynamic purposes, but an array with a
controller like function will be one of these options. So, the
embedded sector in the end might not be as big an opportunity as it
looks, but conventional run anything pressing has a sizable life left
in it.

Thanks


Steve.

[visua...@rocketmail.com]

On Sunday, May 5, 2013 2:06:29 PM UTC-4, AKE wrote:
> There are numerous counter-examples: Knuth -- still productive
> into his eighties, Erdos -- another octagenerian mathematician
> who was tremendously productive. Going back a century, Gauss, Euler. ...
> Chuck Moore is not so young himself.

> Age and important work I think are not correlated -- in fact,
> I think better work comes with more experience.

Long ago I read a book "My Success Story" or something link this, written by a German author, claiming that men have their highest mentally productive times in their eighties. His famous example was Konrad Adenauer, who was elected as the first German Chancellor when he was 73, and served fourteen years as German Chancellor.

So I am looking forward relaxed for my next thirty years.

[Steve]

Well, that explains it for me.

[Elizabeth D. Rather]

Yes, people can also be extremely productive for many years. However, I
think it's also true that the major innovations are done by
young-to-middle aged people. Chuck did his best work in his late 30's,
early 40's, IMO.

That's not to say, though, that older people can't continue to
contribute at an extremely high level. A company of all hot, young
innovators with few seasoned managers and implementers will get in
trouble sooner or later.

[visua...@rocketmail.com]

On Sunday, May 5, 2013 4:02:04 PM UTC-4, Mark Wills wrote:
> You don't need a graphic IDE, you just think you do!
> Did you ever try Microsoft's QuickBasic or QuickC IDE's in the 80's?
> These ran under DOS and were excellent. They were mouse or keyboard based
> and had windows, albeit character based ones. Really great environment.
> Borland's were equally as good.

I never used Microsoft products others than their OS.

You may be right.

In 1993 I programmed a graphical application displaying insole force measurements for the Yokohama Showa Hospital using UR/Forth.
We had only sixteen capacitive sensors back then. They are still doing research: http://www.rounenkango.m.u-tokyo.ac.jp/pdf/jrm_24_5_insole_type.pdf

Graphical programming was really easy back then with UR/Forth, I added windows and zooming, worked great, but the resolution was low - and it looked only good with white (or colors) on black. Black on white didn't look good.

I do not know if it is possible to show an oscilloscope screen looking real that way, lets say like LabView, but ForthView instead.

[Andrew Haley]

Mark Wills <markrob...@yahoo.co.uk> wrote:

> CV's are always anonymised for legal reasons and should not contain
> any information that gives away age, gender, race, sexuality etc.

Let's say you had the CV of someone called Roshana Ahmed, who did a
BSc in 1980. How much would you know of that person's age, race, and
gender?

Andrew.

[AKE]

On Monday, May 6, 2013 2:56:06 AM UTC+1, Paul Rubin wrote:

> AKE writes:
>
> > There are numerous counter-examples: Knuth -- still productive into
>
> > his eighties,
>
>
>
> Knuth (born January 10, 1938) turned 75 this year.

Good point ;) I should have been a bit vaguer with the "he's old and still good" sentiment :)

[Mark Wills]

On May 6, 11:19 am, Andrew Haley <andre...@littlepinkcloud.invalid>
wrote:

I totally agree. However, in the CV's that come from our HR dept, we
would see that "the person" did a BSc in 1980 (because that may be
relevant to the experience required - that's our call to make) but the
name: Roshana Ahmed, would have been removed, that's HR's call. So
would the date of birth, gender, etc.

I can also tell you that during our interviews, we have a "permitted"
list of questions that we are allowed to ask candidates. We are not
allowed to deviate from the list, and we are not allowed to follow up
on the replies to questions. HR's view is that if we ask "random"
questions we might ask some candidates "better" questions than others,
and that wouldn't be fair. So, to avoid getting sued, we have to ask
every candidate the same questions, and then sit there like nodding
dogs.

You really couldn't make it up; and I'm not. I swear the above is
absolutely true.

[Andrew Haley]

Mark Wills <forth...@gmail.com> wrote:
> On May 6, 11:19?am, Andrew Haley <andre...@littlepinkcloud.invalid>

> wrote:
>> Mark Wills <markrobertwi...@yahoo.co.uk> wrote:
>> > CV's are always anonymised for legal reasons and should not contain
>> > any information that gives away age, gender, race, sexuality etc.
>>
>> Let's say you had the CV of someone called Roshana Ahmed, who did a
>> BSc in 1980. How much would you know of that person's age, race, and
>> gender?
>

> I totally agree. However, in the CV's that come from our HR dept, we
> would see that "the person" did a BSc in 1980 (because that may be
> relevant to the experience required - that's our call to make) but the
> name: Roshana Ahmed, would have been removed, that's HR's call. So
> would the date of birth, gender, etc.

But once you have the name, you know thw rest. And I certainly would
know the person's name before the interview.

> I can also tell you that during our interviews, we have a
> "permitted" list of questions that we are allowed to ask
> candidates. We are not allowed to deviate from the list, and we are
> not allowed to follow up on the replies to questions. HR's view is
> that if we ask "random" questions we might ask some candidates
> "better" questions than others, and that wouldn't be fair. So, to
> avoid getting sued, we have to ask every candidate the same
> questions, and then sit there like nodding dogs.
>
> You really couldn't make it up; and I'm not. I swear the above is
> absolutely true.

Thank God I don't work for your employer. There's a lot of people who
make totally bogus rules on the grounds of "health and safety" and "to
avoid getting sued", but when you ask actual lawyers it turns out that
there's no such requirement.

Andrew.

[rickman]

On 5/6/2013 10:48 AM, Andrew Haley wrote:
> Mark Wills<forth...@gmail.com> wrote:
>> On May 6, 11:19?am, Andrew Haley<andre...@littlepinkcloud.invalid>
>> wrote:
>>> Mark Wills<markrobertwi...@yahoo.co.uk> wrote:
>>>> CV's are always anonymised for legal reasons and should not contain
>>>> any information that gives away age, gender, race, sexuality etc.
>>>
>>> Let's say you had the CV of someone called Roshana Ahmed, who did a
>>> BSc in 1980. How much would you know of that person's age, race, and
>>> gender?
>>
>> I totally agree. However, in the CV's that come from our HR dept, we
>> would see that "the person" did a BSc in 1980 (because that may be
>> relevant to the experience required - that's our call to make) but the
>> name: Roshana Ahmed, would have been removed, that's HR's call. So
>> would the date of birth, gender, etc.
>
> But once you have the name, you know thw rest. And I certainly would
> know the person's name before the interview.

Really, you can tell gender and race from a name?

Pat Carver, what gender, what race?

Jay Marancik, what gender, what race?

Kim Papadopoulos, what gender, what race?

--

Rick

[Mark Wills]

On May 6, 3:48 pm, Andrew Haley <andre...@littlepinkcloud.invalid>
wrote:

> Thank God I don't work for your employer.  There's a lot of people who
> make totally bogus rules on the grounds of "health and safety" and "to
> avoid getting sued", but when you ask actual lawyers it turns out that
> there's no such requirement.
>
> Andrew.
>
>

I'm with you on that one, but it's like herding cats. I've given up!

[rickman]

Ok, if all these things you talk about are possible, much less
practical, please show me. I will be happy to acknowledge your
accomplishments. :)

--

Rick

[Andrew Haley]

rickman <gnu...@gmail.com> wrote:
> On 5/6/2013 10:48 AM, Andrew Haley wrote:
>> Mark Wills<forth...@gmail.com> wrote:
>>> On May 6, 11:19?am, Andrew Haley<andre...@littlepinkcloud.invalid>
>>> wrote:
>>>> Mark Wills<markrobertwi...@yahoo.co.uk> wrote:
>>>>> CV's are always anonymised for legal reasons and should not contain
>>>>> any information that gives away age, gender, race, sexuality etc.
>>>>
>>>> Let's say you had the CV of someone called Roshana Ahmed, who did a
>>>> BSc in 1980. How much would you know of that person's age, race, and
>>>> gender?
>>>
>>> I totally agree. However, in the CV's that come from our HR dept, we
>>> would see that "the person" did a BSc in 1980 (because that may be
>>> relevant to the experience required - that's our call to make) but the
>>> name: Roshana Ahmed, would have been removed, that's HR's call. So
>>> would the date of birth, gender, etc.
>>
>> But once you have the name, you know thw rest. And I certainly would
>> know the person's name before the interview.
>
> Really, you can tell gender and race from a name?

No, and I didn't claim that you always could. What I did claim was
that if you had the CV of someone called Roshana Ahmed, who did a BSc
in 1980, you would know. And you'd probably be close.

Not always, no, but you can get some idea. Let's see what was said:

>>>>> CV's are always anonymised for legal reasons and should not contain
>>>>> any information that gives away age, gender, race, sexuality etc.

Just for emphasis, in case you missed it:

>>>>> any information

And I still don't believe that a Google interview panel would have no
advance idea how old a candidate was.

> Pat Carver, what gender, what race?
>
> Jay Marancik, what gender, what race?
>
> Kim Papadopoulos, what gender, what race?

Zaphod Beeblebrox, what gender, what race?

Andrew.

[Mark Wills]

On May 6, 11:21 pm, Andrew Haley <andre...@littlepinkcloud.invalid>
wrote:
>

> Zaphod Beeblebrox, what gender, what race?
>
> Andrew.
>
>

Which head?

Male. Beetlegeusean.

Age? Holy zarquon man, how am I supposed to know?! Ask The Guide!

[Rod Pemberton]

"Steve" <nosp...@gmail.com> wrote in message
news:33d51745-defc-4d50...@h9g2000pbr.googlegroups.com...

> Hi Rod, thanks for this,

Thanks for what ... ? The post?

> [...] could you post some linjsvto your project, [...]

What is "linjsvto"?

It's just a small private project. It hasn't been released.

> When you say modern C compiler is too difficult to modify, are
> you talking about the project code management side, or c++?
>

I'm just saying it'll be alot of work to modify a major C (or C++)
compiler to emit Forth. It's an undertaking for a corporation or
team of people, but not a person. It'll also be alot of work to
code a browser from scratch in Forth. However, code from scratch
will be clean and understandable by humans. Programmatically
translated code won't be. It'll work but be a total mess.


Rod Pemberton

[rickman]

On 5/7/2013 6:53 PM, Rod Pemberton wrote:
> "Steve"<nosp...@gmail.com> wrote in message
> news:33d51745-defc-4d50...@h9g2000pbr.googlegroups.com...
>
>> Hi Rod, thanks for this,
>
> Thanks for what ... ? The post?
>
>> [...] could you post some linjsvto your project, [...]
>
> What is "linjsvto"?

What? You don't read "Steve"? "Links to"... There is something with
his computer that makes an auto spell checker into an auto mistake
maker. Or maybe it is just working in a different language.

> It's just a small private project. It hasn't been released.
>
>> When you say modern C compiler is too difficult to modify, are
>> you talking about the project code management side, or c++?
>>
>
> I'm just saying it'll be alot of work to modify a major C (or C++)
> compiler to emit Forth. It's an undertaking for a corporation or
> team of people, but not a person. It'll also be alot of work to
> code a browser from scratch in Forth. However, code from scratch
> will be clean and understandable by humans. Programmatically
> translated code won't be. It'll work but be a total mess.

--

Rick

[rickman]

On 5/6/2013 6:21 PM, Andrew Haley wrote:
> rickman<gnu...@gmail.com> wrote:
>> On 5/6/2013 10:48 AM, Andrew Haley wrote:
>>> Mark Wills<forth...@gmail.com> wrote:
>>>> On May 6, 11:19?am, Andrew Haley<andre...@littlepinkcloud.invalid>
>>>> wrote:
>>>>> Mark Wills<markrobertwi...@yahoo.co.uk> wrote:
>>>>>> CV's are always anonymised for legal reasons and should not contain
>>>>>> any information that gives away age, gender, race, sexuality etc.
>>>>>
>>>>> Let's say you had the CV of someone called Roshana Ahmed, who did a
>>>>> BSc in 1980. How much would you know of that person's age, race, and
>>>>> gender?
>>>>
>>>> I totally agree. However, in the CV's that come from our HR dept, we
>>>> would see that "the person" did a BSc in 1980 (because that may be
>>>> relevant to the experience required - that's our call to make) but the
>>>> name: Roshana Ahmed, would have been removed, that's HR's call. So
>>>> would the date of birth, gender, etc.
>>>
>>> But once you have the name, you know thw rest. And I certainly would
>>> know the person's name before the interview.
>>
>> Really, you can tell gender and race from a name?
>
> No, and I didn't claim that you always could. What I did claim was
> that if you had the CV of someone called Roshana Ahmed, who did a BSc
> in 1980, you would know. And you'd probably be close.
>
> Not always, no, but you can get some idea. Let's see what was said:

Some idea? How do you get "some idea" of gender? Either you know or
you don't know.... This is ridiculous.

>>>>>> CV's are always anonymised for legal reasons and should not contain
>>>>>> any information that gives away age, gender, race, sexuality etc.
>
> Just for emphasis, in case you missed it:
>
>>>>>> any information
>
> And I still don't believe that a Google interview panel would have no
> advance idea how old a candidate was.
>
>> Pat Carver, what gender, what race?
>>
>> Jay Marancik, what gender, what race?
>>
>> Kim Papadopoulos, what gender, what race?
>
> Zaphod Beeblebrox, what gender, what race?

I don't remember the race, but the gender was male and male, counting
each head separately. lol

So you acknowledge that your claim is ridiculous? Even with the name
Roshana Ahmed you don't know for sure. He/she may be US born with
ethnic parents. I honestly don't know if this is a female name or not.
As to age, I don't give the year of my degree anymore, why should I?
The year has no significance. But then I don't send out resumes either.

--

Rick

[Steve]

Rod Pemberton wrote:
> "Steve" <nosp...@gmail.com> wrote in message
> news:33d51745-defc-4d50...@h9g2000pbr.googlegroups.com...
>
> > Hi Rod, thanks for this,
>
> Thanks for what ... ? The post?

Posting about your C effort.

>
> > [...] could you post some linjsvto your project, [...]
>
> What is "linjsvto"?

From the famous language of the cat herders, referig to fur lint. No,
I'm joking. :-) Not to many cats here fortunately.

No, late nights, bad typing and a bad keyboard spell system is the
cause Rod, sorry.

It was "information" I meant.

>
> It's just a small private project. It hasn't been released.

Have you looked at Stephen Pelc's c compiler?

http://www.mpeforth.com/arena.htm.
Down the end of the page.

>
> > When you say modern C compiler is too difficult to modify, are
> > you talking about the project code management side, or c++?
> >
>
> I'm just saying it'll be alot of work to modify a major C (or C++)
> compiler to emit Forth. It's an undertaking for a corporation or
> team of people, but not a person. It'll also be alot of work to
> code a browser from scratch in Forth. However, code from scratch
> will be clean and understandable by humans.

Why wouldn't you program at least the underlying hardware/code
production portion from scratch in Forth?

> Programmatically
> translated code won't be. It'll work but be a total mess.

Why would you want to maintain outputted code in Forth instead of C?
The code is usually always going to be written, updated and maintained
in C by someone else. The only real use of such a product is to get
in a portion or system of completed code to a target that does not
normally have support for it, to use it instead of rewriting it from
scratch. It would not be meant to be a development environment
solution for outputed code. Why do people keep thinking I would mean
this, it is not very realistic? Maintaining the outputted code,
wouldn't that be like thinking we had to maintain threaded forth in
the outputed assembler, no, you just maintain it in the original
language.

I encourage you Rod, just to think of it as a potentially good tool to
conveniently pull in large amounts of C code not a Forth processor
environment short term, despite performance efficiency.

Steve.

>
>
> Rod Pemberton

[Paul Rubin]

"Rod Pemberton" <do_no...@notemailnotq.cpm> writes:
> I'm just saying it'll be alot of work to modify a major C (or C++)
> compiler to emit Forth. It's an undertaking for a corporation or
> team of people, but not a person.

From http://www.yosefk.com/blog/my-history-with-forth-stack-machines.html :

Speaking of which. I started to miss a C compiler. I downloaded
LLVM. (7000 files plus a huge precompiled gcc binary. 1 hour to build
from scratch. So?) I wrote a working back-end for the stack machine
within a week. Generating horrible code. Someone else wrote an
optimizing back-end in about two months.

[Andrew Haley]

rickman <gnu...@gmail.com> wrote:
> On 5/6/2013 6:21 PM, Andrew Haley wrote:
>> rickman<gnu...@gmail.com> wrote:
>>> On 5/6/2013 10:48 AM, Andrew Haley wrote:
>>>> Mark Wills<forth...@gmail.com> wrote:
>>>>> On May 6, 11:19?am, Andrew Haley<andre...@littlepinkcloud.invalid>
>>>>> wrote:
>>>>>> Mark Wills<markrobertwi...@yahoo.co.uk> wrote:
>>>>>>> CV's are always anonymised for legal reasons and should not contain
>>>>>>> any information that gives away age, gender, race, sexuality etc.
>>>>>>
>>>>>> Let's say you had the CV of someone called Roshana Ahmed, who did a
>>>>>> BSc in 1980. How much would you know of that person's age, race, and
>>>>>> gender?
>

> So you acknowledge that your claim is ridiculous?

Of course I don't!

> Even with the name Roshana Ahmed you don't know for sure.

You don't know anything for sure; they question is whether you have
any information.

> He/she may be US born with ethnic parents. I honestly don't know if
> this is a female name or not.

It is.

> As to age, I don't give the year of my degree anymore, why should I?
> The year has no significance. But then I don't send out resumes
> either.

Well, Q E flamin' D. Speaking as someone who reads many CVs, I
usually see a full history. In fact, I can't think of any that
didn't.

Andrew.

[Rod Pemberton]

"Andrew Haley" <andr...@littlepinkcloud.invalid> wrote in message
news:Fc-dnQWRMOpfHhrM...@supernews.com...

Unfortunately, most probably, you know everything, and correctly
too.

There is a high probability the "virtual" candidate is female, of
Arabic heritage, 54 to 57 years old, born around 1958. It's not
an absolute. The assumption is the candidate fits the norm for
the information provided. I.e., assuming "did a BSc" means
completed a BSc in 1980 and not started one in 1980:

-candidate is female, since Roshana is typically a female given
name

-candidate is Arabic, since Ahmed is the most common Arabic
surname

-candidate took the planned 4 years to complete a BSc, or maybe 5,
but not more ...

-candidate completed high school in the year they turned 18, with
18 being the norm for the US

-candidate entered a university without a delay after high school,
at age 17 or 18

There is much that can go wrong with that, i.e., graduated H.S.
young or late, took more than four years on a degree, e.g., night
school, mistake on birth name, inherited or assigned a surname
that doesn't represent nationality or insignificant part, etc.


Rod Pemberton

[Rod Pemberton]

"rickman" <gnu...@gmail.com> wrote in message
news:km8gqc$blg$1...@dont-email.me...

> Really, you can tell gender and race from a name?
>
> Pat Carver, what gender, what race?

I assume "Pat" is female, unless otherwise indicated. I assume
Carver is white. However, many white names were assumed by a
former slaves in the US.

> Jay Marancik, what gender, what race?

I assume "Jay" is male, unless otherwise indicated.

> Kim Papadopoulos, what gender, what race?

I assume "Kim" is female, unless otherwise indicated. I assume
Papdopoulos is Greek.

> [follow up post by Andrew]

I assume "Zaphod" is male, unless otherwise indicated.


Rod Pemberton

[Rod Pemberton]

"Paul Rubin" <no.e...@nospam.invalid> wrote in message
news:7xobcmc...@ruckus.brouhaha.com...

> "Rod Pemberton" <do_no...@notemailnotq.cpm> writes:

> > I'm just saying it'll be alot of work to modify a major C (or
> > C++) compiler to emit Forth. It's an undertaking for a
> > corporation or team of people, but not a person.
>

> From [link]

>
> Speaking of which. I started to miss a C compiler. I
> downloaded LLVM. (7000 files plus a huge precompiled
> gcc binary. 1 hour to build from scratch. So?) I wrote a
> working back-end for the stack machine within a week.
> Generating horrible code. Someone else wrote
> an optimizing back-end in about two months.

Yes, but a stack-machine is not Forth. To convert C to Forth,
you're going to have to do lots of things: implement stack frames
in Forth for C procedures, convert C's function parameters to
0-operand, and convert *ALL* of C's syntax, from simple to
convoluted, to equivalent Forth, etc. I still think it's a large
undertaking. IMO, a C subset could be done in reasonable time.


Rod Pemberton

[Rod Pemberton]

"Steve" <nosp...@gmail.com> wrote in message

news:55140771-e5cd-44c2...@fz1g2000pbb.googlegroups.com...

> http://www.mpeforth.com/arena.htm.
> Down the end of the page.
>

I seem to recall something about it years ago...

However, I mostly remember Rob Chapman's Timbre when the topic of
C to Forth and Forth to C comes up. Although, I always have to
look up which direction the conversion takes place ...

My Forth interpreter has a small C backend today, but had a large
one when starting out. So, I've considered recycling the C code
to produce a Forth to C converter. I also have another simple,
rough Forth to C converter project which not based on a compiler,
interpreter, or parser. One of these might develop over time into
something useful.

> > > When you say modern C compiler is too difficult to modify,
> > > are you talking about the project code management side, or
> > > c++?
> > >
> >
> > I'm just saying it'll be alot of work to modify a major C (or
> > C++) compiler to emit Forth. It's an undertaking for a
> > corporation or team of people, but not a person. It'll also
> > be alot of work to code a browser from scratch in Forth.
> > However, code from scratch will be clean and understandable
> > by humans.
>
> Why wouldn't you program at least the underlying hardware/code
> production portion from scratch in Forth?
>

... of the C compiler or the browser?

Are you saying to code and maintain a separate backend in Forth
for one of them? If so, with a C compiler modified for Forth, I
don't see a point to that... The code is all C. That might be a
solution to handle bootstrapping a Forth port/conversion of the
browser. That depends on how good the C to Forth conversion is,
whether interpreted or compiled Forth, etc.

> > Programmatically
> > translated code won't be. It'll work but be a total mess.
>
> Why would you want to maintain outputted code in Forth instead
> of C?

You wouldn't. That was the point about coding it from scratch in
Forth...

Well, I was assuming that after the project was converted from C
to Forth that Forth programmers would take over maintaining Forth.
However, they could just convert C to Forth with every C update...
What's the point of that though?

> The only real use of such a product is to get
> in a portion or system of completed code to a target that does
> not normally have support for it, to use it instead of rewriting
> it from scratch.

Forth and C are the most ubiquitous languages. Support for both
is likely. I.e., the real need for conversion is only if Forth
programmers need existing C code in Forth, and don't want to start
from scratch with Forth code.


Rod Pemberton

[Steve]

Rod Pemberton wrote:
> "Steve" <nosp...@gmail.com> wrote in message
> news:55140771-e5cd-44c2...@fz1g2000pbb.googlegroups.com...
>
> > http://www.mpeforth.com/arena.htm.
> > Down the end of the page.
> >
>
> I seem to recall something about it years ago...

> > Why wouldn't you program at least the underlying hardware/code
> > production portion from scratch in Forth?
> >
>
> ... of the C compiler or the browser?

It was about the c compiler in this case.

> Are you saying to code and maintain a separate backend in Forth
> for one of them? If so, with a C compiler modified for Forth, I
> don't see a point to that... The code is all C. That might be a
> solution to handle bootstrapping a Forth port/conversion of the
> browser. That depends on how good the C to Forth conversion is,
> whether interpreted or compiled Forth, etc.

Separately from the browser again. If was referring to that you can
get a c compiler written mainly in c, so only a small back end has to
be written in forth/misc to get it to handle the structural
differences and produce forth from c, rather than the whole
compiler. Now you can try to get Firefox os to compile, or a number
of other OS's and browsers (hope I still have the right thread here).
I personally see limited advantage apart from a hobby statement, until
you have a competitive reason to implement these on misc (when misc
gets fast or functional enough, or has some need).

> > > Programmatically
> > > translated code won't be. It'll work but be a total mess.
> >
> > Why would you want to maintain outputted code in Forth instead
> > of C?
>
> You wouldn't. That was the point about coding it from scratch in
> Forth...

But we are talking about if you had to out c code into forth
(presumably for good reason, like if it was too much to redo it, like
hundreds of megabytes by yourself, and the machine had no c compiler
or other means of getting it on there, like a misc machine) surely you
would just maintain the c source (presuming it might be created and
maintained elsewhere) and recompile as needed.

> Well, I was assuming that after the project was converted from C
> to Forth that Forth programmers would take over maintaining Forth.
> However, they could just convert C to Forth with every C update...
> What's the point of that though?

Yes, that is what I was meaning above. If you are going to create
something new and your target is a misc device, you would just do it
in the best way for the device, and not c, but if your main target was
various Linux based devices on standard processors, you would do it in
c and recompile. In our case what we should look at is that a lot of
substantial work, and what could be used freely, are open source Linux
programs maintained elsewhere in c, so there is no advantage
maintaining them ourselves.

>
> > The only real use of such a product is to get
> > in a portion or system of completed code to a target that does
> > not normally have support for it, to use it instead of rewriting
> > it from scratch.
>
> Forth and C are the most ubiquitous languages. Support for both
> is likely. I.e., the real need for conversion is only if Forth
> programmers need existing C code in Forth, and don't want to start
> from scratch with Forth code.

Like on a theoretical misc computer, or a real misc device.

>
>
> Rod Pemberton


Steve.

[Alex McDonald]

On May 10, 1:08 am, "Rod Pemberton" <do_not_h...@notemailnotq.cpm>
wrote:

> "rickman" <gnu...@gmail.com> wrote in message
>
> news:km8gqc$blg$1...@dont-email.me...
>
> > Really, you can tell gender and race from a name?
>
> > Pat Carver, what gender, what race?
>
> I assume "Pat" is female, unless otherwise indicated.  I assume
> Carver is white.  However, many white names were assumed by a
> former slaves in the US.

English (specifically Cornish) surname. Pat is commonly both male &
female; in Ireland, almost exclusively male.

>
> > Jay Marancik, what gender, what race?
>
> I assume "Jay" is male, unless otherwise indicated.
>
> > Kim Papadopoulos, what gender, what race?
>
> I assume "Kim" is female, unless otherwise indicated.  I assume
> Papdopoulos is Greek.

Kim is a male name in Asia. Papadopoulos is almost certainly male;
Papadopoulou is the feminine form.

>
> > [follow up post by Andrew]
>
> I assume "Zaphod" is male, unless otherwise indicated.

And has two heads?

>
> Rod Pemberton

Guessing is never a good strategy.

[rickman]

On 5/9/2013 8:08 PM, Rod Pemberton wrote:
> "rickman"<gnu...@gmail.com> wrote in message
> news:km8gqc$blg$1...@dont-email.me...
>
>> Really, you can tell gender and race from a name?
>>
>> Pat Carver, what gender, what race?
>
> I assume "Pat" is female, unless otherwise indicated. I assume
> Carver is white. However, many white names were assumed by a
> former slaves in the US.

Wrong, this Pat is Male.

>> Jay Marancik, what gender, what race?
>
> I assume "Jay" is male, unless otherwise indicated.

Jay is quite the female.

>> Kim Papadopoulos, what gender, what race?
>
> I assume "Kim" is female, unless otherwise indicated. I assume
> Papdopoulos is Greek.

Papadopoulos is a Greek name, but Kim is not Greek, he was born in the
US about the same time as myself and dated a Korean gal who almost got
him disowned when they got engaged. Seems the parents are rather
orthodox.

>> [follow up post by Andrew]
>
> I assume "Zaphod" is male, unless otherwise indicated.

That's the point, any conclusions drawn are "assumptions". Andrew seems
to confuse "information" with the conclusions he draws from the
information.

--

Rick

[rickman]

On 5/9/2013 8:08 PM, Rod Pemberton wrote:

The assumptions here are wide ranging. In the US we have *many*
nationalities and ethnicities who have taken up residence since day one.
Names simply don't convey even remotely accurate information about
where a person is from, their gender and certainly nothing about how
good an employee they would make.

I simply don't provide information to a potential employer that is not
relevant to the job. I remember once being asked to provide financial
info and a release to have a full background investigation. I think
they may have asked for permission to access my medical history too. I
told them no and they said they wanted to do an interview and then we
could talk about the other info. Obviously it was negotiable. Since
then I just don't bother to continue the process if they ask for
excessive info. I've heard some employers ask for Facebook passwords
and such. That's just crazy!

--

Rick

[Paul Rubin]

"Rod Pemberton" <do_no...@notemailnotq.cpm> writes:
>> downloaded LLVM... wrote a

>> working back-end for the stack machine within a week.

> Yes, but a stack-machine is not Forth.

From what I could tell, it was a misc-like architecture programmed in

Forth.

> To convert C to Forth, you're going to have to do lots of things:
> implement stack frames in Forth for C procedures, convert C's function
> parameters to 0-operand, and convert *ALL* of C's syntax, from simple
> to convoluted, to equivalent Forth, etc. I still think it's a large
> undertaking. IMO, a C subset could be done in reasonable time.

Eh? C's syntax is already gone by the time one is dealing with the LLVM
retargeting process. The C front end produces a machine-independent
assembler-like intermediate representation (IR). The back end then
translates the IR to the actual instructions of the target machine. In
the case of a stack machine, these map very directly to Forth words. C
to Forth is comparatively trivial if you're willing to make free use of
locals.

[rickman]

That all sounds right except for the assumed equivalence between the
target machine instructions and Forth. Just because a CPU design uses a
stack does not mean it is equivalent to Forth code.

Maybe I misunderstand your point?

--

Rick

[Rod Pemberton]

"rickman" <gnu...@gmail.com> wrote in message

news:kmj32k$kmn$1...@dont-email.me...

> On 5/9/2013 8:08 PM, Rod Pemberton wrote:

> > I assume [...]

>
> That's the point, any conclusions drawn are "assumptions".
> Andrew seems to confuse "information" with the conclusions
> he draws from the information.
>

So, you're saying probability and statistics DOES NOT
provide any useful information ... ? Of course, it does. It's
just imperfect information.

I think you're trying to make the distinction between 100%
certainty ("information") and situations with a high probability
("assumptions"). Typically, they're the same, but not quaranteed
to be.


Rod Pemberton

[Rod Pemberton]

"rickman" <gnu...@gmail.com> wrote in message

news:kmj45h$kmn$2...@dont-email.me...

> Names simply don't convey even remotely accurate information

> about where a person is from, their gender [...]

Really ... ? Did you even look at my name before you wrote that?
What about your name?

You just said that you can't tell that I'm a male, whose surname's
heritage is "English, Anglo Saxon" (as Wikipedia describes it)...
No, my family hasn't lived in England, apparently for many
generations, but you can assume that I'm living somewhere someone
with English heritage would live. You know I write in English.

What you mean to say is that you can't tell with 100% certainty
anything from a person's name. That's true. But, you can tell
with a high degree of probability what is true for the vast
majority people.

How many guys do you know who are named "Peggy"? If you know of a
"Peggy", which is typically a female's name, that "Peggy" is
female, yes? Even though "Joe" is sometimes given to a girl, have
you ever met a girl name "Joe"? If you've met someone named "Joe"
wasn't it a guy? etc

> [...] certainly nothing about how good an employee they would
> make.

True. But, what does?

You don't know if they're a hard worker until they're hired and
been working that job. You don't know if they have experience
until they've worked a job for a while. You don't know if they're
smart until they've worked a job for a while longer. You don't
know if they're truly competent until they've worked a job for a
long while.

A degree is of no use in making this judgement either. 80% of
people with a certain degree are incompetent in that degree.
Listed experience is of no use too since 80% of people exaggerate
(or lie) to get the job. You have to hire a person, place them in
the job, and review their performance to confirm, unless it's
blatantly obvious they're not qualified from "good faith"
information provided by the person.


Rod Pemberton

[Paul Rubin]

rickman <gnu...@gmail.com> writes:
> That all sounds right except for the assumed equivalence between the
> target machine instructions and Forth. Just because a CPU design uses
> a stack does not mean it is equivalent to Forth code.

The first sentence in the article is:

This is a personal account of my experience implementing and using
the Forth programming language and the stack machine architecture.

Later it says:

I defined the instruction set in a couple of hours. It mapped to Forth
words as straightforwardly as possible, plus it had a few things Forth
doesn't have that C might need, as a kind of insurance (say, access to
16-bit values in memory).

I think the thing about 16-bit values just means it has something like
c! and c@ but for 16 bit fields. The data cell size is apparently 64
bits.

[rickman]

I don't see the connection. The fact that someone designed a stack
machine to run something very Forth like does not mean the stack machine
executes Forth. None of the above quotes contradict this statement.

To generate Forth from a C compiler means you need to generate Forth.
I'm not saying this would be hard, I think just the opposite. But any
given stack machine is not necessarily the same as the Forth VM so that
they are not equivalent problems. Equivalent in complexity perhaps, but
obviously there are some issues which the author of the link you refer
to seemed to think required some special instructions which are not part
of Forth.

--

Rick

[rickman]

I'm not interested in arguing this point. Your words above contradict
themselves. If you want to believe what you wrote, feel free.

--

Rick

[rickman]

On 5/11/2013 3:56 PM, Rod Pemberton wrote:
> "rickman"<gnu...@gmail.com> wrote in message
> news:kmj45h$kmn$2...@dont-email.me...
>
>> Names simply don't convey even remotely accurate information
>> about where a person is from, their gender [...]
>
> Really ... ? Did you even look at my name before you wrote that?
> What about your name?
>
> You just said that you can't tell that I'm a male, whose surname's
> heritage is "English, Anglo Saxon" (as Wikipedia describes it)...
> No, my family hasn't lived in England, apparently for many
> generations, but you can assume that I'm living somewhere someone
> with English heritage would live. You know I write in English.

You are making no sense whatsoever. A person with an English last name
can likely live *anywhere* on the planet that is habitable. The bloody
English ruled half the world and explored much of the other half. Are
you just writing at random?

> What you mean to say is that you can't tell with 100% certainty
> anything from a person's name. That's true. But, you can tell
> with a high degree of probability what is true for the vast
> majority people.

There is also the fact that a name is just a name. If you got your last
name from someone 20 generations ago and there have been not only any
number of emigrations, but 20 marriages, what does that tell you about
the person? Especially in the US, virtually NOTHING.

> How many guys do you know who are named "Peggy"? If you know of a
> "Peggy", which is typically a female's name, that "Peggy" is
> female, yes? Even though "Joe" is sometimes given to a girl, have
> you ever met a girl name "Joe"? If you've met someone named "Joe"
> wasn't it a guy? etc

Knowing that a person's name is Peggy doesn't tell me *anything* I find
important in a job interview.

>> [...] certainly nothing about how good an employee they would
>> make.
>
> True. But, what does?
>
> You don't know if they're a hard worker until they're hired and
> been working that job. You don't know if they have experience
> until they've worked a job for a while. You don't know if they're
> smart until they've worked a job for a while longer. You don't
> know if they're truly competent until they've worked a job for a
> long while.
>
> A degree is of no use in making this judgement either. 80% of
> people with a certain degree are incompetent in that degree.
> Listed experience is of no use too since 80% of people exaggerate
> (or lie) to get the job. You have to hire a person, place them in
> the job, and review their performance to confirm, unless it's
> blatantly obvious they're not qualified from "good faith"
> information provided by the person.

Now you are catching on. The entire interview process is very imperfect
and many people are hired into jobs to which they are not well suited
and end up leaving. Doing triage on resumes by reading their names does
nothing to help this.

I'm not interested in discussing this further. Feel free to reply and
I'll likely read it, but I don't think I'll be responding on this topic.

--

Rick

[Paul Rubin]

rickman <gnu...@gmail.com> writes:
> I don't see the connection. The fact that someone designed a stack
> machine to run something very Forth like does not mean the stack
> machine executes Forth.

Did you read the article? The machine was (initially) programmed in
Forth. He showed sample code that it ran, and it was Forth code.

> link you refer to seemed to think required some special instructions
> which are not part of Forth.

IIRC they ended up implementing or optimizing some memory access
primitives to make C code run better. Obviously they could have done
without those if they were willing to take the efficiency hit.

[Albert van der Horst]

In article <kmm7e0$otc$1...@speranza.aioe.org>,

--
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

[Albert van der Horst]

In article <kmmq7f$3ti$1...@dont-email.me>, rickman <gnu...@gmail.com> wrote:
>On 5/11/2013 3:56 PM, Rod Pemberton wrote:

<SNIP>

>
>Now you are catching on. The entire interview process is very imperfect
>and many people are hired into jobs to which they are not well suited
>and end up leaving. Doing triage on resumes by reading their names does
>nothing to help this.

Unfortunately it does. If one has 100 cv's to wade through, after throwing
hundreds in the bin, it helps to cut that in half by the simplest arguments.
Unless the remaining 50 cv's is not enough to find a suitable candidate.

>
>--
>
>Rick

[Elizabeth D. Rather]

On 5/12/13 2:29 AM, Albert van der Horst wrote:
> In article <kmmq7f$3ti$1...@dont-email.me>, rickman <gnu...@gmail.com> wrote:
>> On 5/11/2013 3:56 PM, Rod Pemberton wrote:
> <SNIP>
>
>>
>> Now you are catching on. The entire interview process is very imperfect
>> and many people are hired into jobs to which they are not well suited
>> and end up leaving. Doing triage on resumes by reading their names does
>> nothing to help this.
>
> Unfortunately it does. If one has 100 cv's to wade through, after throwing
> hundreds in the bin, it helps to cut that in half by the simplest arguments.
> Unless the remaining 50 cv's is not enough to find a suitable candidate.

A CV tells you a lot of important things having nothing to do with
gender, age, or ethnicity. The combination of the CV and cover letter
tell you whether the person can write coherently (important for
documentation), is careful with things like spelling (a programmer has
to be careful about spelling, in order to write accurate code), be
responsive to the requirements of the job being applied for (as he/she
needs to be responsive to the requirements of a project), and many other
things of this nature.

Sure, there are some professionally-written CVs, but they're usually
easy to spot, too.

Cheers,
Elizabeth

--
==================================================
Elizabeth D. Rather (US & Canada) 800-55-FORTH
FORTH Inc. +1 310.999.6784
5959 West Century Blvd. Suite 700
Los Angeles, CA 90045
http://www.forth.com

"Forth-based products and Services for real-time
applications since 1973."
==================================================

[Rod Pemberton]

"Elizabeth D. Rather" <era...@forth.com> wrote in message
news:GOmdnYj30boNfRLM...@supernews.com...

> A CV tells you a lot of important things having nothing to do
> with gender, age, or ethnicity. The combination of the CV and
> cover letter tell you whether the person can write coherently
> (important for documentation),

If a manager requires a programmer to document much, he/she
should hire a technical writer. It's a bit like asking an
electrical engineer to patch concrete. It should never be done.
I.e., documentation is way, way, way outside the skillset of the
employee and the standard job descriptions for programmers too,
e.g., programmer analyst, software developer, ... A little bit of
documentation is fine. Unfortunately, I'm seeing all this stuff
online where managers nowadays expect subordinates to do whatever
they ask, if it's not illegal. That's total BS, IMO. It strongly
indicates the manager is incompetent, i.e., 1) doesn't know how to
manage people respectfully 2) doesn't know what the skillset of
their employee is 3) doesn't know what their employee's job
actually encompasses 4) doesn't actually care about the employee
or their relationship with him/her 5) isn't concerned about any
negative outcomes resulting from their insensitivity.

> [...] is careful with things like spelling (a programmer has

> to be careful about spelling, in order to write accurate code),

I'd say that true for Fortran and Cobol, but definately not for C.
C's operators are all symbolic, e.g., like Forth's ! @ +! + - * /
and not wordy like Forth's AND OR XOR. C uses symbols & | ^ for
those and other operators. For C, all you need to do is be able
to correctly retype later on whatever set of characters, numbers,
and/or underscore that you happened to type before for a variable
or procedure name. So, memory is more important than spelling.
So, my decent spelling is for naught in C. If the spelling is
"incorrect", i.e., different from the "correct" version, the
compiler will warn that the "incorrectly" spelled version doesn't
have a correct declaration or use. Of course, a company may have
a coding guideline requiring more than that, e.g., awkward Pascal
naming for C... Personally, for my private code, I like to use 3
letter names whenever or wherever possible, e.g., idx for index,
sts for status, tmp for temporary, in and out for file streams,
etc. Names with a long length just clutter up the code. If you
understand what the code is doing, the names don't matter much,
although they can provide context to an English speaker if in
English, etc. E.g., you should be able to easily read most C code
even when someone coded it in Spanish, French, Chinese, etc.


Rod Pemberton

[Rod Pemberton]

"Paul Rubin" <no.e...@nospam.invalid> wrote in message

news:7xbo8hk...@ruckus.brouhaha.com...
> "Rod Pemberton" <do_no...@notemailnotq.cpm> writes:

(Sorry, I thought I sent this out a few days ago.)

> > To convert C to Forth, you're going to have to do lots of
> > things: implement stack frames in Forth for C procedures,
> > convert C's function parameters to 0-operand, and convert
> > *ALL* of C's syntax, from simple to convoluted, to equivalent
> > Forth, etc. I still think it's a large undertaking. IMO, a C
> > subset could be done in reasonable time.
>
> Eh? C's syntax is already gone by the time one is dealing with
> the LLVM retargeting process. The C front end produces a
> machine-independent assembler-like intermediate
> representation (IR). The back end then translates the IR to the
> actual instructions of the target machine. In the case of a
> stack machine, these map very directly to Forth words. C to
> Forth is comparatively trivial if you're willing to make free
> use of locals.

Well, they map very directly to stack operations, some of which
are also Forth words. But, that subset of Forth is not Forth
either... Stack operations are a Forth subset. I.e., if you were
converting C to equivalent Forth - not just stack operations which
may have Forth definitions - it'd be a larger task. In that case,
you couldn't just emit Forth stack operations. You'd need to emit
higher level Forth words too, like ALLOC or CONSTANT or
PARSE etc. In theory, it's possible to build these back up to
higher level Forth words from the stack operations. However, it's
generally more convenient to convert directly to high level Forth.


Rod Pemberton

[Paul Rubin]

"Rod Pemberton" <do_no...@notemailnotq.cpm> writes:
> I.e., if you were converting C to equivalent Forth - not just stack
> operations which may have Forth definitions - it'd be a larger task.

The idea is just to use Forth as a type of assembly language at the
compiler backend. The compiler output would not be expected to be
comprehensible by humans to any greater extent than normal compiler
output is.

[Elizabeth D. Rather]

It certainly wouldn't be Forth. The machine instructions in question are
a microscopic subset of Forth: maybe a dozen or so words out of the 300+
in the CORE wordset. And most "Forth processors" have several additional
instructions at a lower level, e.g. managing an A register.

The misc processors that have been discussed here have an instruction
set that would support a very efficient Forth. But a C compiler will
have a backend that generates those instructions, not anything you'd
recognize as Forth. Talking about a "C to Forth" translator for this
purpose is absurd and betrays a lack of understanding of the subject.

[Paul Rubin]

"Elizabeth D. Rather" <era...@forth.com> writes:
> It certainly wouldn't be Forth. The machine instructions in question
> are a microscopic subset of Forth: maybe a dozen or so words out of
> the 300+ in the CORE wordset.

Are you saying that

: square ( n -- n ) dup * ;
5 square . \ print square of 5

is not Forth, because it uses only a handful of those 300+ words?

> And most "Forth processors" have several additional instructions at a
> lower level, e.g. managing an A register.

And the Forth implementations for those processors have words that
use those instructions.

> The misc processors ... Talking about a "C to Forth" translator for this

> purpose is absurd and betrays a lack of understanding of the subject.

I think the subject was compiling C code for Misc processors. Yes, just
like any other compiler output, the result would not look like
human-written code.

These days there are compilers out there for all kinds of languages
(including C) that produce Javascript output, so you can code in your
favorite language and have the result run in web browsers. This Misc
thing doesn't seem any worse ;-).

[Elizabeth D. Rather]

On 5/14/13 11:23 PM, Rod Pemberton wrote:
> "Elizabeth D. Rather" <era...@forth.com> wrote in message
> news:GOmdnYj30boNfRLM...@supernews.com...
>
>> A CV tells you a lot of important things having nothing to do
>> with gender, age, or ethnicity. The combination of the CV and
>> cover letter tell you whether the person can write coherently
>> (important for documentation),
>
> If a manager requires a programmer to document much, he/she
> should hire a technical writer. It's a bit like asking an
> electrical engineer to patch concrete. It should never be done.

Perhaps we're talking about different activities. A programmer needs to
document the code at a level to support other programmers who will need
to use and maintain it. That's not the same as writing user manuals,
which would be a proper task for a technical writer.

I have hired and managed many programmers and tech writers. A programmer
who cannot document code coherently is useless, because that code will
be unmaintainable. And it's my experience that everyone, in technical
fields as well as less-technical ones, is more productive in a team if
he/she is able to communicate well.

> I.e., documentation is way, way, way outside the skillset of the
> employee and the standard job descriptions for programmers too,
> e.g., programmer analyst, software developer, ... A little bit of
> documentation is fine.

I think we're disagreeing about what's "a little bit".

> Unfortunately, I'm seeing all this stuff
> online where managers nowadays expect subordinates to do whatever
> they ask, if it's not illegal. That's total BS, IMO. It strongly
> indicates the manager is incompetent, i.e., 1) doesn't know how to
> manage people respectfully 2) doesn't know what the skillset of
> their employee is 3) doesn't know what their employee's job
> actually encompasses 4) doesn't actually care about the employee
> or their relationship with him/her 5) isn't concerned about any
> negative outcomes resulting from their insensitivity.

I'm not sure where this rant is coming from, but it's not what I was
talking about at all.

>> [...] is careful with things like spelling (a programmer has
>> to be careful about spelling, in order to write accurate code),
>
> I'd say that true for Fortran and Cobol, but definately not for C.
> C's operators are all symbolic, e.g., like Forth's ! @ +! + - * /
> and not wordy like Forth's AND OR XOR. C uses symbols & | ^ for
> those and other operators. For C, all you need to do is be able
> to correctly retype later on whatever set of characters, numbers,
> and/or underscore that you happened to type before for a variable
> or procedure name. So, memory is more important than spelling.

What's important is attention to detail, and spelling is a good
indicator of that.

> So, my decent spelling is for naught in C. If the spelling is
> "incorrect", i.e., different from the "correct" version, the
> compiler will warn that the "incorrectly" spelled version doesn't
> have a correct declaration or use. Of course, a company may have
> a coding guideline requiring more than that, e.g., awkward Pascal
> naming for C... Personally, for my private code, I like to use 3
> letter names whenever or wherever possible, e.g., idx for index,
> sts for status, tmp for temporary, in and out for file streams,
> etc. Names with a long length just clutter up the code. If you
> understand what the code is doing, the names don't matter much,
> although they can provide context to an English speaker if in
> English, etc. E.g., you should be able to easily read most C code
> even when someone coded it in Spanish, French, Chinese, etc.

Abbreviations make code harder to read for the programmer's colleagues
and code maintainers. I don't like excessively wordy names (particularly
long concatenated names), but simple English words (assuming an
English-speaking company) make code much more readable.

[Elizabeth D. Rather]

On 5/15/13 7:39 AM, Paul Rubin wrote:
> "Elizabeth D. Rather" <era...@forth.com> writes:
>> It certainly wouldn't be Forth. The machine instructions in question
>> are a microscopic subset of Forth: maybe a dozen or so words out of
>> the 300+ in the CORE wordset.
>
> Are you saying that
>
> : square ( n -- n ) dup * ;
> 5 square . \ print square of 5
>
> is not Forth, because it uses only a handful of those 300+ words?

"Forth" includes the : compiler, etc., which the processor instruction
set does not. The programming language is not the set of machine
instructions, and vice-versa.

>> And most "Forth processors" have several additional instructions at a
>> lower level, e.g. managing an A register.
>
> And the Forth implementations for those processors have words that
> use those instructions.
>
>> The misc processors ... Talking about a "C to Forth" translator for this
>> purpose is absurd and betrays a lack of understanding of the subject.
>
> I think the subject was compiling C code for Misc processors. Yes, just
> like any other compiler output, the result would not look like
> human-written code.

Correct. And it's inappropriate to call it Forth.

> These days there are compilers out there for all kinds of languages
> (including C) that produce Javascript output, so you can code in your
> favorite language and have the result run in web browsers. This Misc
> thing doesn't seem any worse ;-).

Sure. You could write a genuine C to Forth translator, but the result
wouldn't run on a misc chip without further processing at the Forth
level. If your objective isn't Forth as a language, what you want is a
C-to-misc compiler.