From ... From: Erik Naggum Subject: Re: strings and characters Date: 2000/03/21 Message-ID: <3162629683063483@naggum.no>#1/1 X-Deja-AN: 600651358 References: <3162477100853468@naggum.no> <8666uioc8n.fsf@g.local> <3162506888111902@naggum.no> <_ssB4.33$SW5.633@burlma1-snr2> mail-copies-to: never Content-Type: text/plain; charset=us-ascii X-Complaints-To: newsmaster@eunet.no X-Trace: oslo-nntp.eunet.no 953686790 3314 195.0.192.66 (22 Mar 2000 00:59:50 GMT) Organization: Naggum Software; vox: +47 8800 8879; fax: +47 8800 8601; http://www.naggum.no User-Agent: Gnus/5.0803 (Gnus v5.8.3) Emacs/20.5 Mime-Version: 1.0 NNTP-Posting-Date: 22 Mar 2000 00:59:50 GMT Newsgroups: comp.lang.lisp * Barry Margolin | Wouldn't those characters be of type CHARACTER? Mustn't a vector | specialized to type CHARACTER be able to hold all objects of type | CHARACTER? Isn't such a vector a subtype of STRING? what was the _intent_ of removing string-char and making fonts and bits implementation-defined? has that intent been carried forward all the way? | Where does the standard ever give license for a value to change during | assignment? 16.1.2 Subtypes of STRING, and I qoute: However, the consequences are undefined if a character is inserted into a string for which the element type of the string does not include that character. | Well, I was there and you weren't, so I think I can comment on the intent, | to the best of my recollection. that's appreciated, but I must say I find "I was there and you weren't" to be amazingly childish as "arguments" go. | What we wanted to remove from the standard were the API and UI that dealt | with the nature of specific attributes. We didn't want to distinguish | these specific attributes (bits and fonts), which often didn't make sense | in many implementations or applications. But I don't think we intended | to destroy the notion that attributes are part of the objects, and are | thus included in assignments just like any attributes and slots of other | data types. They could get lost during I/O, due to the fact that the | language can't specify the nature of external file formats, but as long | as you stay within the Lisp environment they should stick. perhaps you, who were presumably there for the duration, could elaborate on the intended _consequences_ of the removal of the string-char type and the change to strings from being made up of a subtype of character that explicitly excluded fonts and bits to a character type that didn't need to include fonts and bits? #:Erik