Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2014-07-19

| Channels | #marpa index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:57 jeffreykegler joined #marpa
00:59 jeffreykegler re http://irclog.perlgeek.de/marpa/2014-07-18#i_9044170 -- btw, I took this as a positive for Lua, rather than a negative.
01:01 jeffreykegler My reasoning goes like this -- if C is your favorite language, you are probably a good judge in these matters -- at least you're in the company of folks like Linus Torvalds, Ken Thompson (pre-golang) and RMS (who might rank C behind LISP/Scheme).
01:02 jeffreykegler So, therefore, if a large number of C programmers have adopted Lua as their despised last-resort 2nd language, this is high praise.
01:03 jeffreykegler My study of Lua has reconciled me to a lot of its features --\
01:04 jeffreykegler My current biggest dislike -- arrays are 1-based.
01:05 jeffreykegler Actually, it's a bit more complicated than that -- sort of the opposite of Perl.  In Perl, as Larry put it, you can make your arrays 1-based, "but don't do that".
01:06 jeffreykegler In Lua, you can make your arrays 0-based, if you insist, but the language is set up so that that is not the road to happiness.
01:13 jeffreykegler ronsavage: re http://irclog.perlgeek.de/marpa/2014-07-18#i_9048498 -- I attempted to start listening to this, but it sounds like it was recorded underwater -- it's very hard to listen to.  Did you notice the same?
01:20 ronsavage I didn't listen. The start was so hard to hear, and I'd just watched the one by the designer of Lua.
01:32 jeffreykegler Here, at least, are the slides: http://www.slideshare.net/jgrahamc/lua-the-worlds-most-infuriating-language
01:33 jeffreykegler Too bad the talk isn't listenable, but some of what was said is evident from the slides.
01:36 ronsavage Thanx.
01:41 ronsavage As I read it (http://www.lua.org/manual/5.2/manual.html#2.1) a Lua hash key can't point to the value nil. Surely that restriction will be a PITA?
01:44 jeffreykegler It's a feature. :-)
01:46 jeffreykegler The problem with "undef" is that a non-existent value is undef, but you also have explicit undef's.  This is something we have gotten used to in Perl, but it is in fact a PITA, and repeatedly shows up in some nasty way.
01:47 jeffreykegler Lua's nil is completely consistent, once you get the hang of it, and has a lot of advantages.
01:48 jeffreykegler For example, you can use nil safely as a error return, precisely because you can reasonably be assured that nil is not being used as some kind of special value.
01:49 jeffreykegler This is not the case with Perl's undef, which you might decide means "use the default", but then when you need an "error -- no value", you find yourself stuck.
01:50 jeffreykegler The fact that you can't really store nil's anywhere discourages you from using them for too many things.
01:56 ronsavage I'm not entirely convinced. I see it as a quality-of-code matter. I people abuse the usage of a feature, that doesn't mean the feature is at fault. Just having features is a good thing.
01:56 ronsavage Now calling function off_to_lunch()!
14:35 shadowpaste joined #marpa
14:59 jeffreykegler joined #marpa
23:32 ronsavage joined #marpa
23:34 ronsavage I people -> If people. And by abuse I mean assigning multiple meanings to a single value. It's a classic design fault.....

| Channels | #marpa index | Today | | Search | Google Search | Plain-Text | summary