From Wikipedia, the free encyclopedia
Jump to: navigation, search

TeX or TeX[edit]

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 "TeX". —Tamfang 08:11, 26 February 2007 (UTC)

This was discussed on Talk:TeX. Knuth's prefered form is not TeX; he wrote an entire article to explain how to typeset the logo properly (see TeX#_note-29), that is, where each letter subtly touches its neighbours. In addition, Knuth himself uses the simple form "TeX" in HTML, see for example [1] 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 TeX 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[edit]

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' - 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. —Tamfang 19:13, 3 May 2007 (UTC)

Note 1[edit]

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

Citation; genre[edit]

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)


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 TeX 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 TeX 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. (talk) 00:21, 28 June 2008 (UTC)


User:Thumperward added {{primarysources}}. 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 {{primarysources}} in three days. I'm not sure if the TUGboat journal meets the stringent criteria of say,, 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 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 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[edit]

User: changed

Curves in Metafont are defined not as conic sections but as cubic splines, for greater versatility and simpler arithmetic.


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)