Why these original FORTRAN quirks?
in [Fortran]
|
From: glen herrmannsfeldt on 27 Oct 2006 13:21 In comp.lang.fortran jmfbahciv(a)aol.com wrote: (snip on 6 character symbol length) > Once you expand the limit, you'll always end up with bit leak > which rapidly becomes bit rot. > Six was sufficient. We used the characters dot and percent > and their position within the six-char field to have meaning. DEC did like 6 character symbols, with sixbit on 36 bit machines, and RAD50 on 16 bit machines. One thing, though, was that DEC assemblers had local symbols. Between any two normal alphanumeric symbols, you could put local symbols, which I believe where a digit followed by a $. That saves a lot of symbol space for real symbols. IBM S/360 assemblers use eight character symbols, with a way to generate a unique temporary symbol when needed. The temporary symbol is some letters followed by a number that is incremented on each request. -- glen > You could at a glance whether the variable name was an error > code, a word index, a field, an address, a bit mask or a bit > definition.
From: Carl Lowenstein on 27 Oct 2006 15:10 In article <rh03k25v6mjv03tdmgnfhr0i5tvgaus9if(a)4ax.com>, Brian Inglis <Brian.Inglis(a)SystematicSW.ab.ca> wrote: > >Long term trend analysis charts? >15 minute samples for a year is over 32K. >You'd have to put each year on a separate chart, or another line on >the same chart with a good possibility of interference even on a large >format plotter. >Sometimes just seeing being able to see the overall trend(s) visually >allows you to see factors you wouldn't otherwise spot. >That's what data warehousing is all about, and major database enabled >visualization packages. >AutoCAD recognized that high end graphics market a couple of decades >ago. Yeahbut. If your plotter has a resolution of 300 points/inch, which is pretty good, your plot comes out at 0.32 inches/day or 9.73 feet/year. Slightly more on leap years. I suppose that the modern 1200bpi inkjets could cut this down to manageable size. Time was when one had to write special software to produce CalComp plots more than 32K points long. carl -- carl lowenstein marine physical lab u.c. san diego clowenst(a)ucsd.edu
From: Peter Flass on 27 Oct 2006 18:00 Brian Inglis wrote: > fOn Thu, 26 Oct 2006 01:40:05 -0700 in alt.folklore.computers, glen > herrmannsfeldt <gah(a)ugcs.caltech.edu> wrote: > > >>Brian Inglis wrote: >> >>(snip on limits in VAX compilers) >> >> >>>>Excel through 2003 still has a limit of 256 columns and 65536 rows, >>>>no matter how much memory, real or virtual, you have. >> >>>>I understand this will be increases in 2007. >> >>>They went up from 32K to 64K rows did they? >> >>Not official, but it looks like 1M rows and 16K columns. >> >> >>>Gave up on Excel years ago for high volume data. >> >>I used it for a project that required it, but mostly I don't use it. >>I would say that most problems that require that much data >>(more than 256 columns or 64K rows) are better done by processing data >>while reading it, not first reading it all in. No matter how big >>main memory gets, data sources will always be larger. >> >>(When I was working on DNA sequence data I once calculate that >>the exponential growth in DNA data was faster than Moore's law. >>It has been a consistent 1%/week for some years now.) >> >>There is no visual advantage to 1M rows and 16K columns. There is no >>possible way to view it. > > > Long term trend analysis charts? > 15 minute samples for a year is over 32K. > You'd have to put each year on a separate chart, or another line on > the same chart with a good possibility of interference even on a large > format plotter. I hate to say this, but using a spreadsheet for this is like using a trowel to dig Lake Mead. I know that "if all you have is a screwdriver...", but there are lots of graphics programs that could handle this easily. Think SAS, for example.
From: jmfbahciv on 28 Oct 2006 09:53 In article <ehtlih$fkk$1(a)news1.ucsd.edu>, cdl(a)deeptow.ucsd.edu (Carl Lowenstein) wrote: >In article <rh03k25v6mjv03tdmgnfhr0i5tvgaus9if(a)4ax.com>, >Brian Inglis <Brian.Inglis(a)SystematicSW.ab.ca> wrote: >> >>Long term trend analysis charts? >>15 minute samples for a year is over 32K. >>You'd have to put each year on a separate chart, or another line on >>the same chart with a good possibility of interference even on a large >>format plotter. >>Sometimes just seeing being able to see the overall trend(s) visually >>allows you to see factors you wouldn't otherwise spot. >>That's what data warehousing is all about, and major database enabled >>visualization packages. >>AutoCAD recognized that high end graphics market a couple of decades >>ago. > >Yeahbut. If your plotter has a resolution of 300 points/inch, which is >pretty good, your plot comes out at 0.32 inches/day or 9.73 feet/year. >Slightly more on leap years. I suppose that the modern 1200bpi inkjets >could cut this down to manageable size. > >Time was when one had to write special software to produce CalComp plots >more than 32K points long. And then the pen would leak. /BAH
From: Brian Inglis on 28 Oct 2006 11:15
On Fri, 27 Oct 2006 19:10:09 +0000 (UTC) in alt.folklore.computers, cdl(a)deeptow.ucsd.edu (Carl Lowenstein) wrote: >In article <rh03k25v6mjv03tdmgnfhr0i5tvgaus9if(a)4ax.com>, >Brian Inglis <Brian.Inglis(a)SystematicSW.ab.ca> wrote: >> >>Long term trend analysis charts? >>15 minute samples for a year is over 32K. >>You'd have to put each year on a separate chart, or another line on >>the same chart with a good possibility of interference even on a large >>format plotter. >>Sometimes just seeing being able to see the overall trend(s) visually >>allows you to see factors you wouldn't otherwise spot. >>That's what data warehousing is all about, and major database enabled >>visualization packages. >>AutoCAD recognized that high end graphics market a couple of decades >>ago. > >Yeahbut. If your plotter has a resolution of 300 points/inch, which is >pretty good, your plot comes out at 0.32 inches/day or 9.73 feet/year. >Slightly more on leap years. I suppose that the modern 1200bpi inkjets >could cut this down to manageable size. Standard wide format resolution seems to be 720dpi, 48in/yr, but varies between 600 and 2400 maximum, 14.6in/yr: could do 5 years across a 72in bed, or a couple of strips of 36in. >Time was when one had to write special software to produce CalComp plots >more than 32K points long. ....and ISTR their standard drivers went haywire if you didn't clip or scale the vectors to the paper width: drew right down the edge of the paper, then started drawing from wherever they ended up, a few feet away from where they should be plotting. I'm a big fan of previewing, to avoid including insignificant outlying data, and autoscaling, to be able to show all of the data, regardless of output media size. Nice to be able to view project plans a few feet high and umpty feet long on the wall. -- Thanks. Take care, Brian Inglis Calgary, Alberta, Canada Brian.Inglis(a)CSi.com (Brian[dot]Inglis{at}SystematicSW[dot]ab[dot]ca) fake address use address above to reply |