Talk:Metafont

== TeX or Te X ==

I'm reverting user:Schutz's last change, until he or she defends the implicit claim that "TeX" is more correct than Knuth's own preferred form "Te X ". &mdash;Tamfang 08:11, 26 February 2007 (UTC)
 * This was discussed on Talk:TeX. Knuth's prefered form is not Te X ; he wrote an entire article to explain how to typeset the logo properly (see TeX), that is, where each letter subtly touches its neighbours. In addition, Knuth himself uses the simple form "TeX" in HTML, see for example on his web site, which seems to indicate that this is his preferred form, at least in HTML. Schutz 12:47, 26 February 2007 (UTC)

As far as I know, TeX should be used when you can't use a proper "full" typographic approach (e.g. using a typewriter or in simple text file); but it seems to me that Te X is very close to how the logo looks like when I write \TeX\ and print the document... (but I haven't magnified the on-screen page and measured the characters and their relative positions!) Ittakezou0 (talk) 12:12, 3 November 2008 (UTC)

Douglas Hofstadter
I recall an attack by Hofstadter on Knuth in relation to the nature of universality specifically on the basis of 'metafonts'. I believe the article in question is in 'Metamagical Themas' - 203.206.251.225 12:27, 29 April 2007 (UTC)


 * Yes, I remember it too. Hofstadter's point was that the diversity of fonts cannot be captured by a manageably small number of variables.  Doesn't invalidate the metafont concept for subsets of that variety, though.  &mdash;Tamfang 19:13, 3 May 2007 (UTC)

Note 1
The note is wrong, at least in the online file. That cite doesn't appear in the document.

Citation; genre
The citation about getting an artist to design a font with 60 parameters does not occur in the cited document.


 * Indeed. Someone needs to provide a proper citation. VasileGaburici (talk) 13:48, 2 September 2008 (UTC)


 * Fixed now. VasileGaburici (talk) 14:02, 2 September 2008 (UTC)

The genre of MetaFont is definitely not "computer font". I changed it to "computer language", but if there is a more accurate genre, please correct me. --Slashme 12:43, 11 October 2007 (UTC)

Definition
According to me what is written in the preamble is wrong: METAFONT interpreter runs METAFONT code (as interpreters do), and the output is (well, can be since the interpreter can output strings to the console too, but this is not the real purpose of METAFONT) a bitmapped collection of glyphs into a single file in a specific format (gf that can be compressed in pk). The METAFONT interpreter interprets, and the result of this interpretation is not a convertion, but a generated bitmap font (and a metric file, TFM, that is the only thing Te X needs to do its job).

Then, this generated bitmap font in the specific format (gf or crunched pk), can be used by any software able to use it, i.e. able to read the bitmap and embed it someway, somewhere. E.g. a DVI reader must be able to load gf/pk bitmap fonts (which are not embedded into DVI), and so must do drivers like dvips, dvipdf (these ones do convert from DVI format into other format), or engines that do not output DVI but a format that can embed bitmaps (as pdftex... and as far as I know is the only one).

So the statement that can be included in PostScript is unclear and not fully correct. It should changed to something like that can be used by softwares to show/print glyphs or something like that.

The METAFONT system and METAFONT outputs (both tfm and gf/pk) are used maily by the Te X system (I mean by this that TFMs are used by (vir)tex program and bitmaps by DVI converters/readers), but this does not mean that the only place where they can be embedded is PostScript files... There's e.g. PDF too at least.

Or at least a e.g. should be added to stress the fact that the use in PostScript files is just an example among several possibility.

151.32.17.100 (talk) 00:21, 28 June 2008 (UTC)

User:Thumperward added. This is rather ludicrous for a piece of software. 90%+ of software pages on Wikipedia do not have any peer-reviewed references. Unless User:Thumperward can come up with a good reason why METAFONT needs to meet this standard, I will remove in three days. I'm not sure if the TUGboat journal meets the stringent criteria of say, lwn.net, cough, cough, but it surely isn't edited by DEK. There's one in-depth comparison paper for the PostScript conversion methods by K. Píška. I moved it to references. VasileGaburici (talk) 18:18, 14 August 2008 (UTC)


 * All articles need to meet our standards of sourcing. That most software articles fail is because most software articles are low-quality. This article should not be low-quality. We need more than one decent secondary source (preferably a lot more) before this article should be considered to be adequately referenced. Chris Cunningham (not at work) - talk 13:48, 3 November 2008 (UTC)
 * I have news for you: the only books on Metafont are those written by Knuth, and they are the most comprehensive reference. These are not self-published books. So, I'm removing the tag; it just scare readers away and its presence is not going to result in any improvement to the article. If you think this article is not sufficiently sourced for Wikipedia, start an AfD. VG &#x260E; 16:15, 3 November 2008 (UTC)


 * I don't see that brinkmanship is a particularly constructive response. This article is unlikely to improve to GA status if it is deemed that the reference manual and two papers is sufficient sourcing for it. The article currently contains practically nothing which establishes the notability of the language. Daring me to AfD is would seem to be the way least likely to result in improvement, closely followed by ensuring that the article is not added to any relevant cleanup categories. Chris Cunningham (not at work) - talk 16:26, 3 November 2008 (UTC)


 * I've added a third party reference: "Fonts and Encodings" by Haralambous. VG &#x260E; 16:33, 3 November 2008 (UTC)


 * Thanks. I'll see whether I can add any too. Chris Cunningham (not at work) - talk 16:39, 3 November 2008 (UTC)

quadratic vs conic
User:67.119.194.133 changed
 * Curves in Metafont are defined not as conic sections but as cubic splines, for greater versatility and simpler arithmetic.

to
 * Curves in Metafont are defined as cubic splines rather than quadratic, for greater versatility with similarly simple arithmetic.

with the edit note
 * might be splitting hairs, but quadratic spline arithmetic isn't *that* tricky

I'm not going to revert, but a point was missed: a quadratic spline (parabola) is easier than a cubic spline, but a general conic (ellipse or hyperbola) is not. —Tamfang (talk) 20:00, 15 August 2010 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on Metafont. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20110927042453/http://www.tex.ac.uk/ctan/systems/knuth/dist/mf/mf.web to http://www.tex.ac.uk/ctan/systems/knuth/dist/mf/mf.web

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 05:55, 3 December 2017 (UTC)

Language details
A short note, to explain my thinking in adding details on Metafont as a language. Metafont does a number of things quite differently than the contemporary mainstream of programming languages. Some of that is just old and quirky, whereas other things could be features that are quite attractive, especially for certain kinds of domain-specific languages (and exactly which is which will furthermore vary depending on the reader's point of view). The point of reporting on these details here in Wikipedia is to simplify finding out about these details and making comparisons between different systems/languages—learning that someone made a system where things worked one way could be the pivotal step towards making a new system that provides some desirable feature of the old system, even if other features are quite different. The point is not to teach Metafont (except perhaps to point out some detail that might confuse the mainstream reader, such as the use of  in certain Metafont variable names). 130.243.68.158 (talk) 14:44, 24 September 2019 (UTC)