This Financial Times opinion piece is by turns satirical and mocking of grammar Nazis, and punctilious and a bit mournful about linguistic changes imposed at speed by the demands of our...
This Financial Times opinion piece is by turns satirical and mocking of grammar Nazis, and punctilious and a bit mournful about linguistic changes imposed at speed by the demands of our technologies.
A person of my acquaintance regularly has to advise, "Don't bother with the apostrophe" when giving his surname, because it invariably causes database snafus.
Longer digression on limitations in use of punctuation for e-mail
As a former mail admin, I can't count the number of times I had to tell people, "E-mail addresses can't have your preferred punctuation." [RFC 5322, the original ASCII specification for SMTP headers, allows apostrophes as printable characters, but doesn't require support of all printable characters in local names. As an example, Windows Live Hotmail still only permits dots, underscores, and hyphens.]
This doesn't even begin to account for non-English punctuation. Internationalized Unicode in e-mail addresses still isn't universal as of 2019, so forget apostrophes, dingbats, guillemets, jù hào, non-English alphabets, etc. if you want your e-mail to be delivered with certainty. Gmail, Microsoft Exchange, and the other big players can support UTF-8 encoding for local names and domains, but there's no guarantee it's enabled for smaller providers.
[And, you kids, I do end sentences in text messages with full stops, so FYVM if you think that's passive-aggressive.]
When Catholique-Valpy tried to register her daughter's birth, the [Northwest Territories]'s vital statistics division told her it couldn't use the glottal stop. She went more than a year without legally registering her baby as her complaint was processed, paying Sahaiʔa's medical expenses out of pocket because of her inability to file for a territorial health card. She eventually settled for a birth certificate with an amended spelling when the need to register Sahaiʔa's birth became pressing.
Aren't most of these issues solved with autocorrect, when it comes to typing messages? As most mobile keyboards are developed in the US, I've found that English auto completion is really good, and...
Aren't most of these issues solved with autocorrect, when it comes to typing messages? As most mobile keyboards are developed in the US, I've found that English auto completion is really good, and typing a Youre automatically leads to a You're.
That being said, the bane of ASCII is probably going to haunt us for quite some time, as backends are slow to update. I can imagine that this might change in the future, as countries using different alphabets entirely or simply additional letters like the German ö, ü, ä become more relevant in the internet.
Not necessarily. Autocorrect doesn't choose the right word, it just chooses the best match in a list of existing words. Its ability to match words depends on the typer's ability to at least get...
Aren't most of these issues solved with autocorrect, when it comes to typing messages?
Not necessarily. Autocorrect doesn't choose the right word, it just chooses the best match in a list of existing words. Its ability to match words depends on the typer's ability to at least get close to the correct word.
"Youre" gets corrected because it's not an existing word. However, if I type "your a nice person", autocorrect won't do anything to correct the wrong "your" in that phrase, because "your" is an existing word in the word list.
If I type "your definatly a nice person", autocorrect doesn't know whether the correct word is "definitely" or "defiantly", and is just as likely to choose the latter as the former (which is why so many people are "defiantly sure" about things on the internet!).
Just have to be a bit of a pedant here, but I promise it supports your point! RFC 5322 is not the "original ASCII specification for SMTP headers"; it is merely the latest incarnation in the LONG...
[RFC 5322, the original ASCII specification for SMTP headers, allows apostrophes as printable characters, but doesn't require support of all printable characters in local names. As an example, Windows Live Hotmail still only permits dots, underscores, and hyphens.]
Just have to be a bit of a pedant here, but I promise it supports your point! RFC 5322 is not the "original ASCII specification for SMTP headers"; it is merely the latest incarnation in the LONG history of RFCs that goes back to 1982. The original RFC, good ol'^1 RFC 822, set out everything and spelled the end from the start for email. Every RFC since then (2822 and thence 5322^2) have been backwards compatible.
^1 in the sense that TECO or DOS are "good ol'"
^2 ... which annoyingly breaks the numbering pattern, grmbl
This Financial Times opinion piece is by turns satirical and mocking of grammar Nazis, and punctilious and a bit mournful about linguistic changes imposed at speed by the demands of our technologies.
A person of my acquaintance regularly has to advise, "Don't bother with the apostrophe" when giving his surname, because it invariably causes database snafus.
Longer digression on limitations in use of punctuation for e-mail
As a former mail admin, I can't count the number of times I had to tell people, "E-mail addresses can't have your preferred punctuation." [RFC 5322, the original ASCII specification for SMTP headers, allows apostrophes as printable characters, but doesn't require support of all printable characters in local names. As an example, Windows Live Hotmail still only permits dots, underscores, and hyphens.]
This doesn't even begin to account for non-English punctuation. Internationalized Unicode in e-mail addresses still isn't universal as of 2019, so forget apostrophes, dingbats, guillemets, jù hào, non-English alphabets, etc. if you want your e-mail to be delivered with certainty. Gmail, Microsoft Exchange, and the other big players can support UTF-8 encoding for local names and domains, but there's no guarantee it's enabled for smaller providers.
[And, you kids, I do end sentences in text messages with full stops, so FYVM if you think that's passive-aggressive.]
Makes me think of this:
Aren't most of these issues solved with autocorrect, when it comes to typing messages? As most mobile keyboards are developed in the US, I've found that English auto completion is really good, and typing a Youre automatically leads to a You're.
That being said, the bane of ASCII is probably going to haunt us for quite some time, as backends are slow to update. I can imagine that this might change in the future, as countries using different alphabets entirely or simply additional letters like the German ö, ü, ä become more relevant in the internet.
I think you mean You’re with an actual apostrophe :).
Not necessarily. Autocorrect doesn't choose the right word, it just chooses the best match in a list of existing words. Its ability to match words depends on the typer's ability to at least get close to the correct word.
"Youre" gets corrected because it's not an existing word. However, if I type "your a nice person", autocorrect won't do anything to correct the wrong "your" in that phrase, because "your" is an existing word in the word list.
If I type "your definatly a nice person", autocorrect doesn't know whether the correct word is "definitely" or "defiantly", and is just as likely to choose the latter as the former (which is why so many people are "defiantly sure" about things on the internet!).
Just have to be a bit of a pedant here, but I promise it supports your point! RFC 5322 is not the "original ASCII specification for SMTP headers"; it is merely the latest incarnation in the LONG history of RFCs that goes back to 1982. The original RFC, good ol'^1 RFC 822, set out everything and spelled the end from the start for email. Every RFC since then (2822 and thence 5322^2) have been backwards compatible.
^1 in the sense that TECO or DOS are "good ol'"
^2 ... which annoyingly breaks the numbering pattern, grmbl