Re: Draft v2: London and Reiser's UNIX/32V paper, reconstructed

[Oliver Corff]

Hi Branden,

On 12/06/2024 18:22, G. Branden Robinson wrote:
> Hi Oliver & Larry,
>
> At 2024-06-12T17:02:22+0200, Oliver Corff via wrote:
> > thank you very much for this effort.
> My pleasure--if I can call it a pleasure to discover defects in groff's
> mm package scrambling out like roaches when the light's turned on.  :-O
>
> > One question: the appearance of
> > your reconstruction is slightly different from the original.
> Oh, definitely.  Pixel-perfection is not a goal, and even if it were, I
> think it would need to wait for at least a few more revisions.
Absolutely reasonable.
> > Your lines contain approx. 10% more characters, in result a 10-line
> > paragraph in the original text is about 9 lines long in the
> > reconstruction.
> I noticed.  Also I am not certain that:
>
> 1.  groff mm and DWB 3.3 mm use the same page margins;
> 2.  DWB 3.3 mm and PWB (or whatever London & Reiser used) mm use the
>      same page margins;
>
> Each of these present challenges for fidelity, particularly the second.
>
> In principle, difficulties can be worked around by setting groff mm
> registers that affect these parameters.


Do you know whether your scan is scaled 1:1? In this case only, direct
measures could be taken from the image, assuming that the paper size is
letter.

> > May I assume that reconstruction of appearance was not your primary
> > objective?
> It was an objective, up to isomorphism^Wfont and margin issues.
>
> > Do we know about the original font metrics?
> I personally do not.  Width metrics can be obtained for the fonts used
> by DWB 3.3--they're right there in the troff font description files--but
> we don't even know what version of the formatting software London and
> Reiser used, nor which device they typeset the document that was scanned
> and made its way into Ritchie's hands.
>
> These uncertainties pose challenges for a perfect reconstruction.
>
> > What would be the parameters to .S in order to create an exact match?
> If the fonts aren't absolutely metrically compatible, this question may
> not have an answer.
>
> I won't rule out the fun/misery of fine tuning type size and margins to
> reproduce 32vscan.pdf's line and page breaks exactly; I got far closer
> to this goal with Kernighan & Cherry than I had even dreamed would be
> possible.
>
> But:
>
> Before tackling that objective I think it's important that we iron out
> any "macro" issues, in multiple senses.
>
> 1.  The macro package needs to behave very nearly identically.  There's
>      _some_ wiggle room here; things that don't affect breaking or
>      adjustment decisions aren't a big deal.
>
> 2.  The document _content_ needs to be correct; no missing or extraneous
>      words.  Any solecisms in the original, like crowded ellipsis dots or
>      "incorrect" hyphen/minus substitutions, must be preserved.  We can't
>      expect A and B to typeset the same if A and B are not equivalent in
>      terms of glyph sequences sent to the output driver.


Which, in return, makes visual identity an ideal tool to check for
undiscovered glitches which have the potential to cause inconsistent
line breaks.

When working in a negotiation team quite a few years ago, we would take
pages of claimed-to-be identical copies or transcripts of text,
superimpose them and check them against the light of a strong lamp for
gray areas --- mismatches in print. This helped us discover a good few
issues like altered digits etc., and the method was *much* faster than
reading side by side.

Best, Oliver.

--

Dr. Oliver Corff
Mail:oliver.corff@email.de