Software that makes placemats

[jdaw1]

+CircletextsSameFontSizeIfRadiiShrunkToBeSame. The need for a feature like this had been on the placemat software’s to-do list since 2007.

[jdaw1]

Is The Bung Hole a confirmed venue or simply the default parameter now in the placemat software?!

It is the default of the placemat software’s main user but not of the software itself.

Though it’s not a bad idea. Why not write ‟The Bung Hole” rather than ‟London”?

Appropriate values also added to the default of ExternalLinks.

[jdaw1]

Very rarely one might want the centre heading non-centred (e.g.). This can be controlled with HeaderFooterCenterRelativeX, which is the centre of the HeaderCenterText, 0 being at the left margin, 1 being at the right, and 0.5 in the centre.

Also see the organisation thread for the tasting of a smattering of Graham.

[jdaw1]

The following quotations come from the parameters for the Smattering of Graham and Malvedos tasting.

Code: Select all

/HeaderFooterCenterRelativeX {TypeOfPagesBeingRendered /Glasses eq {GlassPositions SheetNum get 6 get 0 get MgnL sub PageWidth MgnL MgnR add sub div} {0.5} ifelse} def

Also the position of centre header needed to be changed, for which the code had been altered to add the parameter HeaderFooterCenterRelativeX. This was then assigned to some 132 characters of code, the complication of which might suggest that the form of this parameter was not optimally chosen.

The default value for HeaderFooterCenterRelativeX is plain ‟0.5”, as simple as can be. But the first time this parameter was used the value had to be changed to something computed from something deep in an array. Is that acceptable? Is this a question for the Taylor 1977 committee?

[DRT]

Would it be possible to print the left have portion of GM76 and G91 on the left hand sheet so that when the sheets are overlapped there is continuity in the spacing of the circles?

[jdaw1]

Would it be possible to print the left have portion of GM76 and G91 on the left hand sheet so that when the sheets are overlapped there is continuity in the spacing of the circles?

I thought about that, and it seems to me that would require a major reworking of the software.

What might perhaps be possible would be to print it to a custom paper size (about twice A3), and have a post-process to cut the print job into pieces.

But even if that could be done, in practice would the paper be positioned sufficiently accurately? And stay so positioned? Methinks not. And a similar question could be asked again with the answer presumed to be negative about the accuracy of the trimming of the margin area on which a printer won’t print.

[PhilW]

Would it be possible to print the left have portion of GM76 and G91 on the left hand sheet so that when the sheets are overlapped there is continuity in the spacing of the circles?

I thought about that, and it seems to me that would require a major reworking of the software.

What might perhaps be possible would be to print it to a custom paper size (about twice A3), and have a post-process to cut the print job into pieces.

But even if that could be done, in practice would the paper be positioned sufficiently accurately? And stay so positioned? Methinks not. And a similar question could be asked again with the answer presumed to be negative about the accuracy of the trimming of the margin area on which a printer won’t print.

Alternatively, just using A3, you could print the left-hand sheet as is, and print a right-hand sheet which included the right-hand 5 columns (i.e. from 75/90 inclusive). Then, any vertical slice on the right-hand sheet through 63/83 would create a page which would overlap directly onto the left page to provide the continuous image, without onerous precision on the cutting.

[jdaw1]

Alternatively, just using A3, you could print the left-hand sheet as is, and print a right-hand sheet which included the right-hand 5 columns (i.e. from 75/90 inclusive). Then, any vertical slice on the right-hand sheet through 63/83 would create a page which would overlap directly onto the left page to provide the continuous image, without onerous precision on the cutting.

That was completely obvious, but not quite so completely obvious that I thought of it. An excellent solution to the cutting problem. (Even if not quite to the precision-positioning problem.)

And the only change in the code needed, though not even needed in this case, would be in the internal variable TitleAboveBelowOverRandomSeeds, which would need to change such that the same item of Circlearrays/Titles/Belowtitles/etc in multiple locations would have the same ‘random’ positions of stars of flowers. Easily done. (Edit: done.)

[jdaw1]

There is the side-effect that it causes there to be two pre-pour sheets for the overlapping circles. The obvious and easy solution would be to have an extra boolean parameter, PrePourRemoveDuplicatesAndSortByWithinTitles (having the obvious interaction with PrePourReverseOrder), and StickyLabelsRemoveDuplicatesAndSortByWithinTitles.

[RAYC]

What proportion of people own magnum decanters?

Not I, and therefore i vote for a fancy bit of code that produces 2 decanter labels whenever the word "magnum" appears in titles, subtitles or overtitles etc.!

But i might look out for a magnum decanter also (now that i own 7 double magnums of port and quite a few more of wine, perhaps also a double magnum decanter...)

[jdaw1]

What proportion of people own magnum decanters?

Everybody who is anybody, surely?

Not I

Oh.

But i might look out for a magnum decanter also (now that i own 7 double magnums of port and quite a few more of wine, !)

Splendid.

perhaps also a double magnum decanter...

No: too heavy to use. A lot of wrist strength is needed to have the required careful control. Indeed, some have said the same of even a magnum decanter.

therefore i vote for a fancy bit of code that produces 2 decanter labels whenever the word "magnum" appears in titles, subtitles or overtitles etc.!

‟/DecanterLabelsNumCopies 0 def” admits the possibility of ‟2”. It doesn’t auto-detect the word ‟Magnum”, but can be set easily enough by hand.

[RAYC]

therefore i vote for a fancy bit of code that produces 2 decanter labels whenever the word "magnum" appears in titles, subtitles or overtitles etc.!

‟/DecanterLabelsNumCopies 0 def” admits the possibility of ‟2”. It doesn’t auto-detect the word ‟Magnum”, but can be set easily enough by hand.

thank you - the auto-detect feature was slightly tongue-in-cheek!

[jdaw1]

Alternatively, just using A3, you could print the left-hand sheet as is, and print a right-hand sheet which included the right-hand 5 columns (i.e. from 75/90 inclusive). Then, any vertical slice on the right-hand sheet through 63/83 would create a page which would overlap directly onto the left page to provide the continuous image, without onerous precision on the cutting.

That was completely obvious, but not quite so completely obvious that I thought of it. An excellent solution to the cutting problem. (Even if not quite to the precision-positioning problem.)

It worked well:

[jdaw1]

There can, rarely, be an additional complication. If a placemat is made from several A3 placemats, guillotined and overlapped !, then the same glass will appear on two sheets. But it should not have twice as many pre-pour sheets. This technicality is fixed with PrePourRemoveDuplicatesByWithinTitles, which, if true, removes duplicates, equivalence being determined by GlassesOnSheets having the same value of WithinTitles. The guillotine-and-overlap trick also messes with the order of the pre-pour sheets, sometimes fixable with PrePourSortByWithinTitles.

[DRT]

There can, rarely, be an additional complication. If a placemat is made from several A3 placemats, guillotined and overlapped !, then the same glass will appear on two sheets. But it should not have twice as many pre-pour sheets. This technicality is fixed with PrePourRemoveDuplicatesByWithinTitles, which, if true, removes duplicates, equivalence being determined by GlassesOnSheets having the same value of WithinTitles. The guillotine-and-overlap trick also messes with the order of the pre-pour sheets, sometimes fixable with PrePourSortByWithinTitles.

In summary: ask JDAW to produce the placemats when things get complicated.

[jdaw1]

In summary: ask JDAW to produce the placemats when things get complicated.

The new better behaviour is on by default, so you can do it.

[jdaw1]

At the Smattering of Graham and Malvedos, the consensus was that for the glasses that had to be lifted out, so those not at the edge, the radius should not be any smaller.

Radii = 120.2, all of them

Radius = 120.2 pt   ⇒   diameter ≈ 3.34″ ≈ 84.8 mm.

For comparison, five or six glasses on A4 (with 30pt margins all round) have a radius of 133.978, so just over 11% more.

[PhilW]

There can, rarely, be an additional complication. If a placemat is made from several A3 placemats, guillotined and overlapped !, then the same glass will appear on two sheets. But it should not have twice as many pre-pour sheets. This technicality is fixed with PrePourRemoveDuplicatesByWithinTitles, which, if true, removes duplicates, equivalence being determined by GlassesOnSheets having the same value of WithinTitles. The guillotine-and-overlap trick also messes with the order of the pre-pour sheets, sometimes fixable with PrePourSortByWithinTitles.

In summary: ask JDAW to produce the placemats when things get complicated.

Perhaps a simplification (in documentation) might be helpful; e.g. "For some layouts, it may be desirable for a specific bottle to appear on more than one glasses sheet (e.g. when using overlapping sheets). To avoid additional consequent duplication on subsequent sheets, define
PrePourRemoveDuplicatesByWithinTitles as true." Provided the duplication does not also occur on the tasting notes sheets, I don't think anyone would worry about the different pre-pour sheet order.

[jdaw1]

Documentation will be simplified thank you.

Which glasses are on which TN sheets is controleld by GlassesOnTastingNotePages. If the glasses are on A3, so sixteen-ish per page, then GlassesOnTastingNotePages shouldn’t be at its default value of equalling GlassesOnSheets. So the need to change the TNs is unaffected by the new functionality.

[jdaw1]

By request of DRT, yesterday evening, TastingNotesCirclesBehindTopX and TastingNotesCirclesBehindBottomX added. Documentation not yet updated, but see the eight TN sheets in the draft of the placemats for the March 2013 Taylor vertical. Comments welcomed.

[jdaw1]

Documentation updated with PhilW’s suggestion, and with new feature suggested by DRT.

see the eight TN sheets in the draft of the placemats for the March 2013 Taylor vertical. Comments welcomed.

Comments still welcomed.

[DRT]

Documentation updated with PhilW’s suggestion, and with new feature suggested by DRT.

see the eight TN sheets in the draft of the placemats for the March 2013 Taylor vertical. Comments welcomed.

Comments still welcomed.

My preference is for the layout of circles shown on the page starting at 1908, albeit that the positioning of the circles within the horizontal separator a is not quite perfect.

[RAYC]

I like the page starting 1942 but don't quite understand why it can't be spaced horizontally across the entire page (or at least centred), notwithstanding that the circles would not "touch".

[jdaw1]

I ! don't quite understand why it can't be spaced horizontally across the entire page (or at least centred), notwithstanding that the circles would not "touch".

Fitting circles tightly is part of the mojo of the whole software. I could add a boolean TastingNotesCirclesBehindFitAndCentreInRow, but don’t like it.

[jdaw1]

Fitting circles tightly is part of the mojo of the whole software. I could add a boolean TastingNotesCirclesBehindFitAndCentreInRow, but don’t like it.

Added. On the usual five-glass A4 example, this being true reduces the radius of the circles on the TN sheets from 87.7071 to 77.7381. Equivalently, setting TastingNotesCirclesBehindFitAndCentreInRow to false would increase the area of these circles by +27.3%.

The following come from the draft of the placemats for the March 2013 Taylor vertical (thread). The first image is old-style, with /TastingNotesCirclesBehindFitAndCentreInRow false def.

The next eight images are new-style, with /TastingNotesCirclesBehindFitAndCentreInRow true def.

(The exact selection of vintages has not yet been finalised. Some others might be added; perhaps a few of these will prove to be unavailable.)

[jdaw1]

The effect of the boolean TastingNotesCirclesBehindFitAndCentreInRow can be seen in the following two-frame GIF.

I prefer the bigger circles (radius 87.7 rather than 77.7); others don’t.

[jdaw1]

At the Bell AHB and RAYC did not like the layout of the circles on the TN sheets.

(Circles are usually fainter, at 0.2× the ink density of the glasses sheets, but on a small bitmap that was too faint, so here have been darkened to 0.5×.)

IIRC, the original complainants liked the first two sheets, diagonals \ and /, but not next two verticals |, and I don’t recall a comment on the fifth. Perhaps they could say more, and others could concur or otherwise.

[Alex Bridgeman]

My comment was that it caused my OCD to react badly when I was met with lines of circles which did not continue from the same point when moving to the next sheet in the tasting notes.

[jdaw1]

continue from the same point when moving to the next sheet in the tasting notes.

Does that mean that you would have been happy with an ordering like the following? (With obvious changing of which glasses on which sheets.)






Note to self: 0→1→0→½→½→1→1→½→0 gives the eight page types seen in the draft of the Taylor placemats for 22 March 2013, but in an order satisfying the ‘Bridgeman condition’.

[Alex Bridgeman]

Does that mean that you would have been happy with an ordering like the following?
Note to self: 0→1→0→½→½→1→1→½→0 gives the eight page types seen in the draft of the Taylor placemats for 22 March 2013, but in an order satisfying the ‘Bridgeman condition’.

Yes! This would have been much more comfortable for my aesthetics!

[jdaw1]

0→1→0→½→½→1→1→½→0

Or 0→1→1→0→½→½→1→½→0.
Or 0→1→1→0→½→1→½→½→0.
Do express a preference or I’ll have to choose. (Obviously, 0=Left, 1=Right, and ½=Centre.)

[Alex Bridgeman]

Depends on how many pages of tasting notes there are.

But say I prefer 0->.5->1->0.5->0

[PhilW]

Peronally I much prefer the diagonals over the verticals (and am not one of those previously expressing such view). I think probably because much of the page already contains grid-based elements, and the diagonal layout with circles adds good counterpoint. I also prefer the slightly larger circles which overlap the horizontal lines to the smaller ones which fit within; these are both purely aesthetic judgements.

[jdaw1]

Peronally I much prefer the diagonals over the verticals (and am not one of those previously expressing such view). I think probably because much of the page already contains grid-based elements, and the diagonal layout with circles adds good counterpoint. I also prefer the slightly larger circles which overlap the horizontal lines to the smaller ones which fit within; these are both purely aesthetic judgements.

I agree with your first point (diagonals), and for the reason you give. Which suggests 0→1→0→½→1→1→½→½→0.

I agree with your second point (larger rather than smaller), but the consensus was otherwise.

[jdaw1]

As illustrated in the following example, the default arrangement is now 0→1→0→½→1→1→½→½→0→0↻ (which includes all nine elements of {0, ½, 1} × {0, ½, 1}).









If there are more than nine TN sheets, the default pattern starts again at the beginning. Henceforth his should be considered a desideratum.

Parameters used to make this example:

Code: Select all

/Circlearrays [
	1  1  6 9 mul {[ exch 2 string cvs ]} for
] def
/Titles [ Circlearrays {0 get} forall ] def
/Belowtitles [ Circlearrays length {()} repeat ] def

/Names [
	[(JDA) {-0.08 Kern} (W)]
] def

/TastingNotesCirclesBehind true def
/TastingNotesCirclesBehindFadingFactor 0.35 def

/HeaderLeftText (TastingNotesCirclesBehind: a test) def
/HeaderCenterText [] def
/HeaderRightText [
	(www) {-0.092 Kern} (.jdawiseman.com)
] def

/VoteRecorders false def
/NonDecanterLabelGlassesNumCopies 0 def

[jdaw1]

I’m considering adding two more parameters:
• CopyrightOfPlacematsText
• CopyrightOfPlacematsLink
with default values
• (This work by Julian D. A. Wiseman is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported Licence.)
• (http://creativecommons.org/licenses/by- ... deed.en_GB) (here clickable)

Are there any objections, particularly to the choice of default values? (And if anybody knows how to use pdfmark to add a machine-readable CC licence to a PDF, please reveal. [Edit: question asked])

[Alex Bridgeman]

No objection from me.

[jdaw1]

• Recent changes

December 2012

• +CopyrightStatementPlacemats, +LicensingAgreementTextPlacemats and +LicensingAgreementLinkPlacemats.

Addition to log page:

That link, clickable: creativecommons.org/licenses/by-nc-sa/3.0/deed.en_GB.

[Alex Bridgeman]

As illustrated in the following example, the default arrangement is now 0→1→0→½→1→1→½→½→0→0↻ (which includes all nine elements of {0, ½, 1} × {0, ½, 1}).









If there are more than nine TN sheets, the default pattern starts again at the beginning. Henceforth his should be considered a desideratum.

Parameters used to make this example:

Code: Select all

/Circlearrays [
	1  1  6 9 mul {[ exch 2 string cvs ]} for
] def
/Titles [ Circlearrays {0 get} forall ] def
/Belowtitles [ Circlearrays length {()} repeat ] def

/Names [
	[(JDA) {-0.08 Kern} (W)]
] def

/TastingNotesCirclesBehind true def
/TastingNotesCirclesBehindFadingFactor 0.35 def

/HeaderLeftText (TastingNotesCirclesBehind: a test) def
/HeaderCenterText [] def
/HeaderRightText [
	(www) {-0.092 Kern} (.jdawiseman.com)
] def

/VoteRecorders false def
/NonDecanterLabelGlassesNumCopies 0 def

I do like this new default

[DRT]

I do like this new default

I like it so much I don't know how I managed to live without it.

[jdaw1]

Roy Hersh has given the software a plug, on page 5 of issue 73 (December 2012) of his newsletter.

[DRT]

Roy Hersh has given the software a plug, on page 5 of issue 73 (December 2012) of his newsletter.

Excellent

[Andy Velebil]

Roy Hersh has given the software a plug, on page 5 of issue 73 (December 2012) of his newsletter.

Excellent

+1
This months newsletter with this plug is available for free on the FTLOP home page. Though you do have to register on the site, which is also free.

[jdaw1]

Indeed, the pictures sent to RAH for that page also recently appeared on his Facebook account (in case you were wondering about their source).

[jdaw1]

O.k., I'll throw this out there for Julian or an expert on this. I once made a sheet where there was one row across. I went to make a new sheet based on that, but I see the baseline has changed significantly, so I should probably use the new baseline ... How so I make an 11x17, landscape, with one row of six glasses across the top. I want there to be blank space below. Any pointers? (FYI, the 6 are 1955, 1970, 1977, 1985, 1992, and 2000 Taylor )

New version has extra possible values for use in PermittedPackingStyles: /TopRow /MiddleRow /BottomRow.

Please always start from the latest version. Things improve.

Understood. But I was hacking, and my attempts to hack the new one weren't working so well, so I started with something I had already modified before. But the new version is really cool with the error report at the end when you render it.

The used arrangement was /Gaia on US Ledger = 11″×17″ = /USL2.

My reason for using 11x17, and also for wanting to try to run the circles across the top was to accommodate a dinner plate. After we tasted the Port, I cooked up duck breasts served with a Port-cherry sauce. People could keep drinking what they had left with the dinner. With the Gaia arrangement it was necessary to move the glasses slightly, but you could still keep them arranged and know which was which.

[jdaw1]

Somebody asked me about the placemats. I sent an email of which a slightly edited version follows, perhaps for future reuse.

You asked about the placemats. It is all completely free (not even freemium the version on the website is the same thing I use). Start at the manual, at www.jdawiseman.com/papers/placemat/placemat.html.

A little scrolling down brings you to a green box entitled Making Your First Placemat: Advice for Beginners. That has all the instructions, and even a link to the program at www.jdawiseman.com/papers/placemat/placemat.ps.

Don’t decide to start this one day before the tasting. You’ll be too busy. Please do have a play and experiment and test much further in advance.

[jdaw1]

+TitleMaxHeightProportionInnerRadius, the default value of which might lessen how often it is worth changing FontSizesRatioTitlesMin.

[jdaw1]

Following discussion at the Doty Cup, default value of TastingNotesCirclesBehindFadingFactorIfAllBlack changed to 0.125 (from 0.1).

[Glenn E.]

Since the last time that I made placemats for a large tasting (18 glasses) on USL2 (11" x 17"), a parameter has changed so that the software now defaults to put only 6 glasses on each sheet. That is rather silly.

My first thought was that GlassesOnSheets would be the culprit, but I find the code for that indecipherable. To wit:

/GlassesOnSheets [
[ 0 Titles length 1 sub /e exch def /s exch def s 1 e {dup e lt e s sub 6 ge and {dup s sub e s sub 1 add dup 6 div ceiling cvi 3 -1 roll 3 copy mul exch idiv 4 1 roll 1 add mul exch idiv ne {][} if} if} for ]
] def % /GlassesOnSheets

How can I tell the current version of the software to please put all 18 glasses on a single sheet of USL2? We have done it before and we did not find the spacing to be too tight.

Worst case I can simply re-use an old version of the software, but I would prefer to keep up-to-date.

[jdaw1]

Code: Select all

/GlassesOnSheets [ 
	[ 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 ]
] def

or

Code: Select all

/GlassesOnSheets [ 
	[ 0 1 17 {} for ]
] def

I also don’t like the default of GlassesOnSheets. But, mostly, people want glasses divvied over however many sheets it takes, up to six glasses on each. The default does that. Please suggest an alternative parameterisation that would capture the two possible modes: auto ≤6; and truly custom.

Edit: Maybe something like

Code: Select all

/GlassesOnSheetsMode /SpreadOverSheets def    % /SpreadOverSheets /Custom
/GlassesOnSheetsMaxPerSheet 6 def    % For all glasses on one sheet set to a big integer
/GlassesOnSheetsCustom [  [ 0 1 2 3 ]  [ 4 5 6 7 ]  [ 8 9 10 11 ]  ] def

That seems to come much closer to the intuition of what’s wanted, and keeps some nasty code out of the parameters. Objections?

Actually, maybe ‟GlassesOnSheets” could have a better name. Suggestions?