MathGroup Archive 2010

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Suggestions

  • To: mathgroup at
  • Subject: [mg111656] Re: Suggestions
  • From: "David Park" <djmpark at>
  • Date: Mon, 9 Aug 2010 05:13:40 -0400 (EDT)

One might have hoped that the algorithm would catch a simple case like this:

Graphics[{Table[Line[{{0, y}, {1, y}}], {y, 0, 1, 0.1}]},
 AspectRatio -> .4] 

Style[Graphics[{Table[Line[{{0, y}, {1, y}}], {y, 0, 1, 0.1}]},
  AspectRatio -> .4], Antialiasing -> False] 

David Park
djmpark at  

From: John Fultz [mailto:jfultz at] 

This is why we have an Automatic value as an option for Antialiasing. 
In some cases, antialiasing clearly contributes to image quality, and in 
some cases, it clearly degrades it.  Automatic ties into an algorithm 
which attempts to efficiently make determinations about what would look 
best.  Because it tries to do so efficiently, there are cases where it 
doesn't do an absolutely perfect job.

Also, something I'd like to clarify from your message....the 
Antialiasing option does not affect 3D graphics, and that 3D graphics 
cannot have antialiasing turned on and off at the level of individual 
primitives.  This is because 3D acceleration hardware doesn't support 
control of antialiasing at that fine of granularity.  Because of the 
fundamental differences in the way things work for 3D, 3D antialiasing 
is controlled by a different option 
(RenderingOptions->"HardwareAntialiasingQuality").  It's more easily 
controlled in the Appearance->Graphics pane of the Preferences dialog.


John Fultz
jfultz at
User Interface Group
Wolfram Research, Inc.

David Park wrote:
> One can control whether fractions change their script size and other
> features with the FractionBoxOptions.
> fraction = Style[(a + b)/c, 
>    FractionBoxOptions -> {AllowScriptLevelChange -> False}]; 
> Then to write a Text cell with the fraction:
> 1) Write the text and leave a place for the Inline cell.
> 2) Put fraction into the Inline cell and evaluate in place.
> You can also control text format with the LineIndent and LineSpacing
> options.
> Presentations used to have an Antialiasing command, but now that Version 6
> introduced it as the default standard the shoe is on the other foot and it
> has an Aliasing command! Many users might not be aware that Antialiasing
> be turned on and off, not just for an entire graphic but also for
> primitives. If you are drawing horizontal or vertical lines then
> Antialiasing will often cause them to be slightly fuzzy and thick
> on their exact placement on the screen. By turning Antialiasing off for
> these particular lines you can sharpen them up. You will notice that WRI
> does not use Antialiasing on Frames and Ticks.
> David Park
> djmpark at
> From: Murray Eisenberg [mailto:murray at] 
> First, one can never tell ahead of time what might emerge in a 
> subsequent version!
> Second, as to size of typeset fractions: the often unpleasant small size 
> is akin to the default behavior the "gold standard" of mathematical 
> typesetting, namely, LaTeX. There the size of fractions in in-line math 
> automatically shrinks, but if you want to resize it larger, you can 
> explicitly do so by inserting a markup command to say to use display 
> style, which sets numerator and denominator in normal size -- with the 
> result, of course, that the line with such an in-line fraction becomes 
> higher than normal and extra inter-line leading has to be introduced 
> (automatically), which can lead to unpleasant results.
> (So what mechanism would you want to control fraction size, and how 
> would you want the inter-line spacing to be handled. (AFAIK, Mathematica 
> does not a distinction between typesetting in-line math and display math.)
> Third, as to indicating where an error is: perhaps part of the 
> difficulty here the very syntax of Mathematica. Syntactical scanning of 
> an expression is not such a simple thing. Here I contrast with J or APL, 
> where there is a fairly strict left-to-right or right-to-left 
> syntactical scanning, so that it is essentially trivial to determine 
> where an error occurs.
> On 8/6/2010 6:56 AM, S. B. Gray wrote:
>> This was posted in 2003:
>> "Selwyn Hollis"<selwynh at>  wrote in message
>> news:bgcupq$9qi$1 at
>>   >  >  Print preview should certainly be near the top of any wish list.
> (But
>>   >  >  if you're fortunate enough to have a Mac w/OS X, printing to PDF
> is
>>   >  >  almost as good.)
>>   >  >
>>   >  >  To any such wish list I'd also add, in no particular order:
>>   >  >
>>   >  >  * antialiasing of graphics!!
>>   >  >  * flexible top/bottom cell margins
>>   >  >  * serious undo functionality
>>   >  >  * real unicode support (on the Mac anyway)
>>   >  >  * better find/replace (e.g., replace all in selection)
>>   >  >  * robust import of eps and pdf graphics
>>   >  >  * less need for tweaking to make expressions look right; namely:
>>   >  >        ** better typesetting of fractions (size should not be
> connected
>>   >  >  to sub/superscript multiplier)
>>   >  >** more consistent and reliable placement of sub/superscripts
>>   >  >** more consistent and reliable spacing
>>   >  >** more consistent and reliable sizing of summation and integral
> signs
>> To which I would add:
>> 1. Make word wrap work in a rational way. At present, it sucks.
>> 2. Allow noncontiguous selection of text like MS Word (finally) does
>> 3. Get a text menu displayed, again like MS Word
>> 4. Allow making actual margins around graphics variable
>> 5. Make the cell properties more visible (the stripes on the right edge)
>> 7. MAKE THE ERROR MESSAGES USEFUL(Mathematica knows where the error is;
> why
>> doesn't it tell you?)
>> 8. Give some indication that Wolfram listens to its customers.
>> 9. I expect a certain number of these suggestions and Hollis' to be
>> acted on, namely zero.
>> Steve Gray

  • Prev by Date: Re: A new graphic user interface
  • Next by Date: Re: ploting piecewise function in contourplot3d section
  • Previous by thread: Re: Suggestions
  • Next by thread: Re: Suggestions