Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.forth > #23656

Re: THANK YOU

Show all headers | View raw

On 6/12/2013 9:09 PM, Hugh Aguilar wrote:
> On Jun 12, 1:14 am, Mark Wills<markrobertwi...@yahoo.co.uk>  wrote:
>> On Wednesday, June 12, 2013 2:08:42 AM UTC+1, Hugh Aguilar wrote:
>>> It is very difficult to make money programming in Forth. Even if your
>>> Forth system is better than all of the other Forth systems, and better
>>> than the C systems, it is still very unlikely to be used
>>> professionally.
>>
>> No no, I fully understand where you're coming from and I fully agree, including the last point about the difficulty in making money from Forth.
>>
>> I think even the Forth shops such as Forth, Inc. and MPE aren't winning business on the basis that they are a Forth shop. They are winning business as a competitive software consultancy, and Forth just happens to be the particular "weapon of choice for them". It could be, of course, that Forth helps them to be competitive in the sense that they can turn projects around faster, and their small size as companies means lower overheads. It's actually not a bad position to be in, as long as you can keep winning business.
>
> If your only advantage is how fast you can write code, then you don't
> have much going for you. Selling speed of development only works if
> the application has already been done many times and there is no
> innovation involved --- you are just cranking out code like sausage,
> and your selling point is that your team of code monkeys can turn the
> crank faster than the competitor's team of code monkeys. This is
> pretty much the world of GUI software, which is why automatic code-
> generators are used --- the code monkeys program by pointing and
> clicking, with no thinking required.

Hugh, I know you think I am biased against you, but in reality, I just 
read what you write and respond to that just like I would respond to 
anyone else.

What you are saying is nonsense.  No matter what the project, if it is 
run properly, a lot of thinking is done up front with many decisions on 
how to implement things made before one line of application code was 
written.  There may be some test code developed to iron out issues and 
there can be scaffold code written that helps to decompose the 
architecture.  Coding always takes time and if the approach using a 
different language can actually help speed coding and testing time, that 
is a plus on nearly every project.


> If you are faced with an application that requires innovation, and you
> try to sell speed of development, your customer is going to think that
> you are full of baloney. This is essentially why Forth Inc. didn't get
> the job writing the compiler for the MiniForth. Elizabeth Rather gave
> her usual sales pitch about "turning months into weeks" --- this was
> obvious baloney, considering that nobody knew at that time what the
> MiniForth instruction set was going to look like, so there was no way
> of knowing what would be involved in getting the job done. The sales
> pitch was baloney as they were pretending to have experience in
> something that had never been done before, plus their price was about
> an order of magnitude more than I was asking, so they didn't get the
> job --- that is not really my fault.

Let's consider your price vs. the price quoted by Forth, Inc.  You were 
living in your van working at a rate barely above minimum wage, working 
when they wanted you and not charging for thinking, but only for banging 
on the keyboard.  Forth, Inc was a professional organization, paying 
good wages to their employees, offering benefits, paying people to not 
just type, but to think and do a good job, working regular 40 hour weeks 
and having a real life.  Yes, I can see why your employer was able to 
save so much money by letting you work like a dog rather than using the 
services of Forth Inc to get the job done professionally.


> If Forth is going to succeed, it has to be used for work that other
> languages aren't any good at. For the most part, I think this involves
> custom processors. I think that it is easier to build a custom
> processor that Forth can compile into, than custom processor that C
> can compile into. Once you have a custom processor, then you can have
> hardware support for what the application is doing --- this is where
> you get a speed boost of 3 or 4 orders of magnitude. There are not
> very many applications that lend themselves to being done on a custom
> processor with application-specific hardware support. Usually, this
> would be an application that does some fairly simple task
> repetitively, but has to do it extremely quickly.

I can see where Forth is good for supporting custom processors, but I 
don't see where that is a unique characteristic.

Check out the ZPU.  It is a MISC like soft processor for FPGAs designed 
specifically to be programmed in C.  It seems to do well enough that it 
has been used commercially.

Then there are any number of commercial soft cores that are supported by 
C, Microblaze, NIOS, the Lattice CPU32 (I forget the name).

Your analysis of custom extensions in processors is a bit off.  I can't 
think of any operation that hardware would speed up by four orders of 
magnitude.  That's 10,000 fold!  Unless it is a function where the added 
hardware would be 100 times larger than the processor itself, I can't 
see how you would get that large of a speed up.  If the processor is 
executing code at 1 clock cycle per instruction to get a 10,000 fold 
speed up would require an operation that takes 10,000 instructions to be 
done in 1 clock cycle.  That's a pretty big piece of hardware!

Forth has its advantages and its disadvantages.  Unfortunately the 
disadvantages stand out like a sore thumb and it can be hard to show a 
newbie what the advantages are.  It is even harder to document the 
advantages in a way that will be believed by the managers controlling 
the purse strings.

-- 

Rick

Back to comp.lang.forth | Previous | Next — Previous in thread | Next in thread | Find similar

Thread

THANK YOU Richard Owlett <rowlett@pcnetinc.com> - 2013-06-05 22:01 -0500
  Re: THANK YOU "Elizabeth D. Rather" <erather@forth.com> - 2013-06-05 19:08 -1000
    Re: THANK YOU Mark Wills <markrobertwills@yahoo.co.uk> - 2013-06-06 00:10 -0700
    Re: THANK YOU Brad Eckert <hwfwguy@gmail.com> - 2013-06-06 16:01 -0700
  Re: THANK YOU stephenXXX@mpeforth.com (Stephen Pelc) - 2013-06-06 11:19 +0000
  Re: THANK YOU rickman <gnuarm@gmail.com> - 2013-06-06 16:46 -0400
    Re: THANK YOU "Rod Pemberton" <do_not_have@notemailnotq.cpm> - 2013-06-06 23:28 -0400
      Re: THANK YOU rickman <gnuarm@gmail.com> - 2013-06-07 18:07 -0400
    Re: THANK YOU Alex McDonald <blog@rivadpm.com> - 2013-06-07 03:27 -0700
      Re: THANK YOU anton@mips.complang.tuwien.ac.at (Anton Ertl) - 2013-06-07 11:44 +0000
        Re: THANK YOU Alex McDonald <blog@rivadpm.com> - 2013-06-07 14:16 -0700
      Re: THANK YOU rickman <gnuarm@gmail.com> - 2013-06-07 18:11 -0400
        Re: THANK YOU visualforth@rocketmail.com - 2013-06-08 21:37 -0700
          Re: THANK YOU rickman <gnuarm@gmail.com> - 2013-06-09 14:42 -0400
          Re: THANK YOU Mark Wills <markrobertwills@yahoo.co.uk> - 2013-06-11 00:51 -0700
            Re: THANK YOU rickman <gnuarm@gmail.com> - 2013-06-11 20:57 -0400
              Re: THANK YOU Hugh Aguilar <hughaguilar96@yahoo.com> - 2013-06-11 18:12 -0700
                Re: THANK YOU rickman <gnuarm@gmail.com> - 2013-06-11 23:27 -0400
              Re: THANK YOU Hugh Aguilar <hughaguilar96@yahoo.com> - 2013-06-17 00:04 -0700
            Re: THANK YOU Paul Rubin <no.email@nospam.invalid> - 2013-06-12 00:22 -0700
        Re: THANK YOU Alex McDonald <blog@rivadpm.com> - 2013-06-09 03:59 -0700
    Re: THANK YOU Hugh Aguilar <hughaguilar96@yahoo.com> - 2013-06-08 19:02 -0700
      Re: THANK YOU rickman <gnuarm@gmail.com> - 2013-06-09 14:45 -0400
      Re: THANK YOU "Rod Pemberton" <do_not_have@notemailnotq.cpm> - 2013-06-10 16:54 -0400
        Re: THANK YOU "Elizabeth D. Rather" <erather@forth.com> - 2013-06-10 11:33 -1000
          Re: THANK YOU "Rod Pemberton" <do_not_have@notemailnotq.cpm> - 2013-06-11 19:50 -0400
            Re: THANK YOU Mark Wills <markrobertwills@yahoo.co.uk> - 2013-06-12 00:58 -0700
              Re: THANK YOU Paul Rubin <no.email@nospam.invalid> - 2013-06-12 01:08 -0700
                Re: THANK YOU Mark Wills <markrobertwills@yahoo.co.uk> - 2013-06-12 01:16 -0700
              Re: THANK YOU "Rod Pemberton" <do_not_have@notemailnotq.cpm> - 2013-06-17 20:27 -0400
                Re: THANK YOU Elizabeth D Rather <erather@forth.com> - 2013-06-17 14:29 -1000
        Re: THANK YOU Hugh Aguilar <hughaguilar96@yahoo.com> - 2013-06-11 17:25 -0700
          Re: THANK YOU "Rod Pemberton" <do_not_have@notemailnotq.cpm> - 2013-06-17 21:19 -0400
            Re: THANK YOU "WJ" <w_a_x_man@yahoo.com> - 2013-06-18 02:05 +0000
              Re: THANK YOU Alex McDonald <blog@rivadpm.com> - 2013-06-17 21:27 -0700
                Re: OT was, [Re: THANK YOU] "Rod Pemberton" <do_not_have@notemailnotq.cpm> - 2013-06-18 21:34 -0400
            OT Re: THANK YOU albert@spenarnc.xs4all.nl (Albert van der Horst) - 2013-06-18 14:53 +0000
      Re: THANK YOU Brad Eckert <hwfwguy@gmail.com> - 2013-06-10 14:11 -0700
        Re: THANK YOU rickman <gnuarm@gmail.com> - 2013-06-10 19:08 -0400
          Re: THANK YOU <fred@example.com> - 2013-06-11 13:02 +0100
            Re: THANK YOU albert@spenarnc.xs4all.nl (Albert van der Horst) - 2013-06-11 12:49 +0000
      Re: THANK YOU Mark Wills <markrobertwills@yahoo.co.uk> - 2013-06-11 00:44 -0700
        Re: THANK YOU Hugh Aguilar <hughaguilar96@yahoo.com> - 2013-06-11 18:08 -0700
          Re: THANK YOU Mark Wills <markrobertwills@yahoo.co.uk> - 2013-06-12 01:14 -0700
            Re: THANK YOU "Elizabeth D. Rather" <erather@forth.com> - 2013-06-12 08:03 -1000
            Re: THANK YOU Hugh Aguilar <hughaguilar96@yahoo.com> - 2013-06-12 18:09 -0700
              Re: THANK YOU albert@spenarnc.xs4all.nl (Albert van der Horst) - 2013-06-13 10:45 +0000
                Re: THANK YOU Mark Wills <markrobertwills@yahoo.co.uk> - 2013-06-13 03:53 -0700
                Re: THANK YOU "Elizabeth D. Rather" <erather@forth.com> - 2013-06-13 07:37 -1000
                Re: THANK YOU Hugh Aguilar <hughaguilar96@yahoo.com> - 2013-06-13 19:01 -0700
                Re: THANK YOU "Rod Pemberton" <do_not_have@notemailnotq.cpm> - 2013-06-17 20:12 -0400
              Re: THANK YOU rickman <gnuarm@gmail.com> - 2013-06-15 11:45 -0400
              Re: THANK YOU "Rod Pemberton" <do_not_have@notemailnotq.cpm> - 2013-06-17 20:25 -0400
  Re: THANK YOU Joshua Litt <jalitt@gmail.com> - 2013-06-06 18:49 -0700
  Re: THANK YOU Alex McDonald <blog@rivadpm.com> - 2013-06-07 14:19 -0700
  Re: THANK YOU malcox0@gmail.com - 2013-07-03 17:12 -0700
    Re: THANK YOU visualforth@rocketmail.com - 2013-07-03 19:50 -0700

csiph-web