I personally don't like this. An atom won't suddenly change value without your knowledge because to get its value, you must use @ or deref (which should be a big warning that, yes, this value might change between calls to deref).
Adding sigils, in my opinion, adds to the noise and makes it harder to read. I personally find sigils to be a noisy mistake in other languages (perl, php etc) and in my opinion you can get more benefit through a suitable naming convention such as a -state prefix, eg: foo-state I think, mainly my dislike for sigils is on variables and not so much on functions as I'm ok with foo? being a predicate and foo! being unsafe in STM. I think that's because they tell you useful meta information about what the function does, but to use a variable, I already have to know what data it's representing in order to call the correct functions on it and annotating it with sigils doesn't help much (unless perhaps you go full blown Hungarian notation, but even that isn't rich enough to adequately describe the nested data structures we use in Clojure - good descriptive variable names are much better at conveying content and purpose). I guess it may largely just be personal taste, although I would also take the wider community into account: there's a lot of code out there not using this convention - will that become confusing if you rely on symbols to tell you that a variable is an atom? On Mon 7 Dec 2015 at 00:27 Mars0i <marsh...@logical.net> wrote: > On Sunday, December 6, 2015 at 5:52:02 PM UTC-6, Glen Mailer wrote: >> >> I saw some sample code that prefixed the atom name with a ! recently, >> seemed to look sensible to me. >> >> (swap! !state conj :whatever) >> >> And so on. >> > > This idea is conceptually elegant, since the marker, !, is the same as the > related function suffix. > > I worry that having two bangs with a space between them, as in the swap! > example above, is a little bit visually confusing. Also, I wonder whether > placing @ next two another non-alphanumeric character is visually confusing > or messy. > > As an experiment, I just started using & as a suffix for variables whose > values are atoms.: > (def state& (atom 1)) > (swap! state& inc) > @state& > > Using only a suffix character means that you don't have a punctuation > character next to @, which I prefer. > > (I wouldn't want to use ! as the suffix for variables that are not > functions, though. (swap! state! inc) potentially very confusing. It's > not that someone reading it couldn't figure it out, but if I saw that, I'd > have to think for a second, and then keep remembering that state! is not a > function.) > > It occurred to me that I've seen both & and $ used in some languages to > suggest pointer dereferencing, so there's some vague harmony to using one > of them as indicators as markers of statefulness. "Ampersand" sounds > vaguely like "atom". If one wanted to have separate marker characters for > atoms, refs, and agents, maybe & is a good choice for atoms. Not sure it's > necessary to have different conventions for these three distinct uses, > though. > > -- > You received this message because you are subscribed to the Google > Groups "Clojure" group. > To post to this group, send email to clojure@googlegroups.com > Note that posts from new members are moderated - please be patient with > your first post. > To unsubscribe from this group, send email to > clojure+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/clojure?hl=en > --- > You received this message because you are subscribed to the Google Groups > "Clojure" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to clojure+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. > -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en --- You received this message because you are subscribed to the Google Groups "Clojure" group. To unsubscribe from this group and stop receiving emails from it, send an email to clojure+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
