dates & notation (WAS Re: coding a data range) 
Author Message
 dates & notation (WAS Re: coding a data range)

Even the order of year month day is a mess.  Dates are one of the very
few places (others?) where we do not write digits in significance order.

12/10/02    10/12/02  02/10/12  02/12/10  ...

Ted



Fri, 03 Jun 2005 01:02:30 GMT  
 dates & notation (WAS Re: coding a data range)

Quote:






>>>> P.S. Btw., what does 'sgrifennodd' mean? Is it in Welsh?

>>>   'wrote'
>>> yes

>> The start of the word bears similarity to Dutch "schrijf" (presesnt
>> tense) and the more simple Afrikaans (skryf). Probably has roots in
>> Old Frisian :-)

> You jest!

Well spotted John...

Quote:
> It has roots in good old Latin: scribo, scribere, scripsi, scriptum, and
> has gone through the usual processes with words that we see in other
> languages: an initial schwa because the "scr" was unpronounceable by the
> natives without it, the c and the b were softened to g and v
> respectively, etc., together with the usual Welsh tense and personal
> endings.

Yes, the OED bears out the Latin connection (I checked the words "scibe" and "script" and
a few derivations) although it suggests that this word entered late Middle English from
Old French (escript). Of course, Welsh may have got it directly from Latin instead, I
don't know. But I wouldn't be surprised if the word was brought in after 1066 - how many
of the locals could write when the Romans were here I wonder?

Damn, I must stop this. Far too much fun...

--
-----BEGIN GEEK CODE BLOCK-----
Version: 3.1
GO/! d- s++:+ a+ C++(++++) US++++$ UB++ U*++ P+++ L+++ E--- W+++ N++ w--- O-
M+ V- PS+ PE+ Y+ PGP t+ 5++ X R* tv+ b+ DI++ D G e(*) h++/-- r+++ y?
------END GEEK CODE BLOCK------

-----------------------------------------------------
Bob Hoekstra:   APL & Unix Consultant
Tele:           +44 (0)1483 771028
                 http://www.HoekstraSystems.com

-----------------------------------------------------



Fri, 03 Jun 2005 04:32:00 GMT  
 dates & notation (WAS Re: coding a data range)

 >> ...I use workstation in the proper sense - not a PC...
 >
 > My laptop has a 1.6 Ghz CPU and 384Mb of RAM, and very fast graphics. It IS a
 > workstation...

It is definitely true that the modern PC has closed the gap between it and a
"workstation". From FOLDOC:

workstation:
        <computer> A general-purpose computer designed to be used by one
        person at a time and which offers higher performance than normally
        found in a personal computer, especially with respect to graphics,
        processing power and the ability to carry out several tasks at the
        same time.

However, I still think that a "real" one runs AIX, HP-UX, Irix, SunOS, etc. I admit I'm
bigotted in this respect.

 > ... In any case the Linux version is no different than the Solaris version.

Not entirely true. As I understood it, there were a number of refinements made in the port
to Linux, specially w.r.t. standardization, that may or may not have found themselves in
the mainstream Unix versions. This was apparently the case when I reviewed the products
(http://www.vector.org.uk/v163/bob163.htm and http://www.vector.org.uk/v164/bob164.htm).

 > I would certainly pay a reasonable price for IBM APL2. I would have 10 years ago.

So would I. The question is, what is "a reasonable price". Your reaction indicated that
IBM's current price is unreasonable in your opinion. I don't necessarily find it
unreasonable, merely unjustifiable (i.e. can I do something with it to recover the cost
and hopefully turn a profit).

 >> Thirdly, returning to Danny's original posting, I have qualms the statement "...I'm
 >> gonna try to build a website with Soliton APL as the scripting language. How, I have
 >> no idea...". The first question is "why?". There are several perfectly adequate
 >> scripting languages for this purpose that do not suffer from the high start-up cost
 >> of an APL interpreter.
 >
 > I don't like any of the environments. I was going to see if I'd like APL in that role.
 >
 >> Please note that I have nothing against someone who wants to run APL as a backend
 >> calculation engine, e.g. to provide actuarial or financial calculations on line, but
 >> to use it as a general purpose scripting language seems crazy.
 >
 > Why? This would be Apache mod-APL. I find the operator syntax very natural and I don't
 > see why it can't be adapted to the web environment.

I'm sure it can, but at what cost? And why?

 >> Perhaps SAX is more suitable for this than most, in that you could have a continuous
 >> n-task running with a TCP/IP port open to the web server, and which does a split ws
 >> to service requests. This would avoid much of the start-up cost, but would never give
 >> the efficiency of e.g. apache+mod-perl. The alternative of writing an entire
 >> webserver in APL has also been done.
 >>
 >> Personally, while apache is free I think this is a pointless excercise.
 >
 > I agree. I'm talking about script-glue for websites. And I said, it's an experiment.

I'd be interested to hear how you progress.

 > The vast majority of coding I've done is in C and C++. I actually like APL and think I
 > could be productive in it in any environment.

This is certainly true. However, a web server, especially a busy (commercial) one, can go
through some tough times where every clock cycle is important. Here is the crux - can your
APL code provide the efficiency required? Can it go into a large production environment?

 >
 > -drl
 >

--
-----BEGIN GEEK CODE BLOCK-----
Version: 3.1
GO/! d- s++:+ a+ C++(++++) US++++$ UB++ U*++ P+++ L+++ E--- W+++ N++ w--- O-
M+ V- PS+ PE+ Y+ PGP t+ 5++ X R* tv+ b+ DI++ D G e(*) h++/-- r+++ y?
------END GEEK CODE BLOCK------

-----------------------------------------------------
Bob Hoekstra:   APL & Unix Consultant
Tele:           +44 (0)1483 771028
                 http://www.HoekstraSystems.com

-----------------------------------------------------



Fri, 03 Jun 2005 05:03:15 GMT  
 dates & notation (WAS Re: coding a data range)

Quote:
>  > ... In any case the Linux version is no different than the Solaris
version.

> Not entirely true. As I understood it, there were a number of refinements
made in the port
> to Linux, specially w.r.t. standardization, that may or may not have found
themselves in
> the mainstream Unix versions. This was apparently the case when I reviewed
the products
> (http://www.vector.org.uk/v163/bob163.htm and

http://www.vector.org.uk/v164/bob164.htm).

Just to be precise, my point was that the Linux version was in no way
crippled. The same is true of IBM's version AFAIK. In fact on Solaris one
must first prepare the UNIX sesssion that will host Sax.

Quote:
>  > I would certainly pay a reasonable price for IBM APL2. I would have 10
years ago.

> So would I. The question is, what is "a reasonable price". Your reaction
indicated that
> IBM's current price is unreasonable in your opinion. I don't necessarily
find it
> unreasonable, merely unjustifiable (i.e. can I do something with it to
recover the cost
> and hopefully turn a profit).

I think a reasonable price would be, say, $200 with other options. About
like any other professional development environment.

Quote:
>  > The vast majority of coding I've done is in C and C++. I actually like
APL and think I
>  > could be productive in it in any environment.

> This is certainly true. However, a web server, especially a busy

(commercial) one, can go
Quote:
> through some tough times where every clock cycle is important. Here is the
crux - can your
> APL code provide the efficiency required? Can it go into a large

production environment?

The only way to know is to try it.



Fri, 03 Jun 2005 07:25:34 GMT  
 dates & notation (WAS Re: coding a data range)

Quote:

> I would certainly pay a reasonable price for IBM APL2. I would have 10 years
> ago.

Then why aren't you running APL2 for OS/2 under OS/2.  I think it's
getting hard to get now but I went that route several years ago and have
never regreted it.

Oh, I see you are using Outlook Infect - a virus propagator with a
little NG misreading facility.

Seriously, if you can find a copy, you'll enjoy it.

Ted



Fri, 03 Jun 2005 14:26:32 GMT  
 dates & notation (WAS Re: coding a data range)


Quote:
> Even the order of year month day is a mess.  Dates are one
of the very
> few places (others?) where we do not write digits in
significance order.

> 12/10/02    10/12/02  02/10/12  02/12/10  ...

> Ted

well, not everybody has a problem  --  in earlier versions
of Windows it was possible to specify Swedish are the
preferred convention for displaying dates

also, names (except the Koreans, except for Koreans resident
in the west, when it is difficult to know whether Mr Pong
Bae Park wouldn't prefer to be known as Park Bae Pong)

addresses (except in China, which appears to use major/minor
order consistently)

and have you ever tried to read out a 6-digit telephone
number in German?

cheers   . . .   /phil



Fri, 03 Jun 2005 17:03:38 GMT  
 dates & notation (WAS Re: coding a data range)



Quote:



sgrifennodd Bob



sgrifennodd Dragan

> Yes, the OED bears out the Latin connection (I checked the

words "scibe" and "script" and
Quote:
> a few derivations) although it suggests that this word

entered late Middle English from
Quote:
> Old French (escript). Of course, Welsh may have got it

directly from Latin instead, I
Quote:
> don't know. But I wouldn't be surprised if the word was

brought in after 1066 - how many
Quote:
> of the locals could write when the Romans were here I

wonder?

according to Latin sources, the Druids knew all about
writing but disavowed its use on the grounds that relying
too much on the written word was bad for the memory

Quote:
> Damn, I must stop this. Far too much fun...

aah  --  is this the Protestant Work Ethic making us all
feel guilty again?

stay happy, have fun   . . .   /phil



Fri, 03 Jun 2005 17:15:44 GMT  
 dates & notation (WAS Re: coding a data range)

Quote:

> So would I. The question is, what is "a reasonable price". Your reaction
indicated  that
> IBM's current price is unreasonable in your opinion. I don't necessarily
find it
> unreasonable, merely unjustifiable (i.e. can I do something with it to
recover the cost
> and hopefully turn a profit).

Bob, well said...a friend of mine in the US was complaining the other day
about the cost of APL+Win 4.0. I'm afraid it went in one ear and out the
other, his son has $3000 worth of computer games!

fred



Fri, 03 Jun 2005 18:34:11 GMT  
 dates & notation (WAS Re: coding a data range)


Quote:


> > Even the order of year month day is a mess.  Dates are one
> of the very
> > few places (others?) where we do not write digits in
> significance order.

> > 12/10/02    10/12/02  02/10/12  02/12/10  ...
> well, not everybody has a problem  --  in earlier versions
> of Windows it was possible to specify Swedish are the
> preferred convention for displaying dates

(since i am the only Scandinavian here):

Earlier?

I've heard it's also possible to choose US settings in addition to Finnish
in later versions of Linus' little program :-).



Fri, 03 Jun 2005 19:56:09 GMT  
 dates & notation (WAS Re: coding a data range)

Quote:

> > I would certainly pay a reasonable price for IBM APL2. I would have 10
years
> > ago.

> Then why aren't you running APL2 for OS/2 under OS/2.  I think it's
> getting hard to get now but I went that route several years ago and have
> never regreted it.

Be real. I've developed for OS/2 at IBM. I love OS/2, but it's dead.

-drl



Fri, 03 Jun 2005 20:08:09 GMT  
 dates & notation (WAS Re: coding a data range)


Quote:
> "phil chastney"


Quote:



> > > Even the order of year month day is a mess.  Dates are
one
> > of the very
> > > few places (others?) where we do not write digits in
> > significance order.

> > > 12/10/02    10/12/02  02/10/12  02/12/10  ...
> > well, not everybody has a problem  --  in earlier
versions
> > of Windows it was possible to specify Swedish as the
> > preferred convention for displaying dates

> (since i am the only Scandinavian here):

> Earlier?

I'm working from memory here, but Windows 3.1 (and I think
3.11 was pretty much the same)allowed you to select your
date format from a menu, and the "Swedish" style had years,
months, days in the right order  (so did Japanese, I
believe)

IIRC, Win95 and Win98 required you to choose your language
(or was it your locale?), and following that, you were
offered a restricted set of format options, all of them
"wrong"

which didn't matter too much, because I only ever used those
OSs at clients' premises  --  NT (finally!) allowed you to
"roll your own" date format, and that's what I had on my own
machine

hence "earlier"  --  you're now going to tell me I've got
that all wrong, I expect  . . .   /phil



Fri, 03 Jun 2005 20:45:46 GMT  
 dates & notation (WAS Re: coding a data range)


Quote:


> > Earlier?

> I'm working from memory here, but Windows 3.1 (and I think
> 3.11 was pretty much the same)allowed you to select your
> date format from a menu, and the "Swedish" style had years,
> months, days in the right order  (so did Japanese, I
> believe)

> IIRC, Win95 and Win98 required you to choose your language
> (or was it your locale?), and following that, you were
> offered a restricted set of format options, all of them
> "wrong"

> which didn't matter too much, because I only ever used those
> OSs at clients' premises  --  NT (finally!) allowed you to
> "roll your own" date format, and that's what I had on my own
> machine

> hence "earlier"  --  you're now going to tell me I've got
> that all wrong, I expect  . . .   /phil

Absolutely not. Only one detail, which might be rather hard to see at the
first glance: The Combo boxes where you choose formats are editable combos,
i.e. they use an Edit field for the selected list alternative. The lists
contain pre-made suggestions which are suitable for the "global" language
selection, but you can modify them by typing your own combinations of YY MM
DD ?? VV JJ etc., depending on the chosen language. (Win98, don't know about
the others, but i assume they haven't taken steps backwards with these)

Reflecting those settings in application software, that's another story...
but they are available through winapi.

/ Tomas



Fri, 03 Jun 2005 23:13:18 GMT  
 dates & notation (WAS Re: coding a data range)

Quote:

> Be real. I've developed for OS/2 at IBM. I love OS/2, but it's dead.

Checkout some of comp.os.os2.* and you'll see it's far from dead.  The
rumours keep flying and I wouldn't be a bit surprised if that isn't
encouraged by M$ but wasn't it Mark Twain who said, "The rumours of my
demise have been greatly exagerated."

YMMV but I have only three apps that require me to boot Windoze and I
NEVER leave the phone jack connected when I do.

Ted



Sat, 04 Jun 2005 02:07:36 GMT  
 dates & notation (WAS Re: coding a data range)

"Bob Hoekstra" escribed <snip>

Quote:
> Tropical Year = 365.24220 days (interval between successive solar transits
of the vernal
> equinox)

> Sidereal Year = 365.25636 days (interval between successive passes of a
fixed star)

> Anomalistic Year = 365.25964 days (interval between successive perigee

transits)
<snip>

Back in the "good old days", when we had to stick to natural integers
(Silicone Block Computers
had no floating point co-centric crop circles in those days), we were able
to achieve an accuracy
of 1 day in 300 years (equipment, Stonehenge Mk IIb, 31688pHz).

Ray

P.S. For those of you not familiar with speed rating of SBC's before
the full introduction of  "SI" units
31688pHz  <=> 31.688nHz <=>  1 cycle per 365.2510122 days
give or take a day or two depending on the cloud cover, crop circles
and  height of the trees on the western horizon.

PPS Moors Law does not apply to SBC.
A speed increase from 31688pHz to 3GHz in 4412 years relates to a
factor of 2 every 822 months, not 18 months.

PPPS. The first major speed increase occurred with the over-clocking of the
mother-goddess cycle, by changed from a Solar base to a Lunar Cycle, with an
approximate 12x improvement in speed. (Please no jokes about "Moon-bikes")



Sun, 05 Jun 2005 08:45:33 GMT  
 
 [ 132 post ]  Go to page: [1] [2] [3] [4] [5] [6] [7] [8] [9]

 Relevant Pages 
 

 
Powered by phpBB® Forum Software