“The Orwell Prize is delighted to announce that, to mark the 70th anniversary of the diaries, each diary entry will be published on this blog exactly seventy years after it was written, allowing you to follow Orwell’s recuperation in Morocco, his return to the UK, and his opinions on the descent of Europe into war in real time. The diaries end in 1942, three years into the conflict.” The first entry begins: “Caught a large snake in the herbaceous border beside the drive. About 2’ 6” long, grey colour, black markings on belly but none on back except, on the neck, a mark resembling an arrow head (ñ) all down the back. [N.b.: They seem to have screwed up the formatting, since an ñ does not resemble an arrowhead. –LH] Did not care to handle it too recklessly, so only picked it up by extreme tip of tail. Held thus it could nearly turn far enough to bite my hand, but not quite.” Today’s reads, in its entirety, “Drizzly. Dense mist in evening. Yellow moon.” Should be good reading. Thanks, Paul!


  1. Huh, that’s peculiar, when I look at that page I see an arrow, but when I copy and paste it into the comment field here (or indeed a text editing program) it appears as a ñ.

  2. Max Pinton says

    The Orwell page is encoded UTF-8, while for some reason LH is in ISO 8859-1. Which is strange because I know I’ve seen non-Latin glyphs here. Did something change?

  3. Max Pinton, with HTML it’s possible to enter non-Latin-1 characters on a Latin-1 website without trouble. Details here. This isn’t true of plain-text protocols like Usenet.

  4. They’re trying to use WingDings via a span with font-family (span isn’t allowed in comments here), so it’ll work in IE, but not other browsers / OSes. 0xF1 (ñ) in that font is more or less U+21E7 ⇧.

  5. Wasn’t something similar done with Pepys a coupla years ago?

  6. Looking at the html source:
    style=”font-size: small; font-family: Wingdings”
    It’s intended to be displayed in the Wingdings font. According to this page http://www.alanwood.net/demos/wingdings.html that’s a “LATIN SMALL LETTER N WITH TILDE” in other fonts.
    This is the appropriate line from that page:
    ñ 241 0xF1 ⇧ 8679 U+21E7 Upwards white arrow Arrows

  7. It’s supposed to be “LATIN SMALL LETTER N WITH TILDE” regardless of the font. Apparently Microsoft takes the position that Wingdings, rather than being a font that contains a lot of non-Latin-1 characters, is a font of Latin 1 characters that uses extremely idiosyncratic renderings — for example, the n-tilde looks like an up arrow.

  8. Thanks for the explanation, everyone!
    Wasn’t something similar done with Pepys a coupla years ago?
    At the start of 2003, to be exact; I wrote about it here, and have been following it (and commenting there) regularly. It’s a great way to read the diary, and I’ll be sad when it ends in a few more years.

  9. Krown, A.J.P. says

    I was surprised to find, when he was mentioned at LLog, that some (young?) people are very anti George Orwell. They think he’s treated with too much reverence, is all I can remember from what was written. Oh, and there was something about his having an old-fashioned, colonial attitude towards something or other: that really seemed a bit much to me. He probably was quite awful in many ways — great people usually seem to be — but writing was not one of them, thank God. So I will add that diary blog to my menu.

  10. @KCinDC: Well, that’s how Wingdings is implemented (you can’t really blame it: it predates Unicode, and back in the day that’s pretty much how alternative charsets worked).

  11. John J Emerson says

    It may just be because “Politics and the English Language” is required reading in too many classes. Like E.B. White, it’s become a pest.

  12. John J Emerson says

    It may just be because “Politics and the English Language” is required reading in too many classes. Like E.B. White, it’s become a pest.

  13. 2003 … Have I really been reading you for that long? I can’t imagine where else I might have seen it.

  14. Ran, the fact that Wingdings uses certain codepoints for certain characters doesn’t have anything to do with how Unicode codepoints that happen to be the same should be displayed. By that logic, all the characters in the upper half of the various “extended ASCII” sets would be hopelessly confused, depending on the font.
    I remember having a conversation years ago with someone who was coding Turkish characters as ð and ý and the like (eth and y-acute not being Turkish characters, of course) because Internet Explorer displayed the right Turkish characters when he applied a particular font.

  15. That blog is great–I read an excellent biography last year and this will augment what I know about him.

Speak Your Mind