But one of the main points of Eckel's rather negative analysis of the checked/unchecked distinction is -precisely- that it lends itself to the abuse you mention. That is WHY no language since Java's invention has chosen to repeat the experiment. And sure enough, the problem really is not as bad in Python or Ruby as it is still in Java.
On Jul 31, 12:12 am, Doug <beafd...@gmail.com> wrote: > On Jul 30, 4:51 pm, Zsolt Vasvari <zvasv...@gmail.com> wrote: > > > By the way, the whole JSONException stuff is foobar-ed: > > > There is a way to avoid JSONException to be thrown by using the opt* > > methods for reads, which I am doing now as I know that I created the > > data and it must be there. But, there is no way to avoid catching the > > JSONException on put* operations as it will check the arguments to be > > NULL's and if they are, it throws the JSONException instead of just a > > NPE. > > Right, and this is kind of my point. The gap between checked/ > unchecked exceptions themselves are not the problem. It's the abuse > of the system that's the problem. Wrapping a NPE inside a checked > exception is subverting the purpose of an unchecked NPE, specifically, > to point out a programmer error. > > You said in your last post that you're "Not sure how the system > implements that, it may save it to a file, I don't know, I don't > care." I think this abuse of the exception system has got you > confused about which exceptions you need to catch and which not. > These abuses are giving checked exceptions a bad name. > > Suggesting that checked suggestions should be banned is, to me, kind > of like saying that prohibition of alcohol is a good idea because it > keeps people from doing stupid things after they abuse the substance. > People are always going to do stupid stuff in their software if > they're not disciplined or understand how things can best work. > > Checked exceptions have their place in the world, and it's up to the > engineer to figure out how best to use them. The JSON serialization > components that Android has included seem to violate best practices > here. (Incidentally, the Android JSON stuff comes > fromhttps://github.com/douglascrockford/JSON-java. Douglas Crockford is > supposed to be one of those really smart guys that we pay attention > to. Maybe he should stick to JavaScript?). The offending code in > JSONObject conveniently checks for null on the key and throws a > checked exception, which is bad. It would have been better if it just > threw unchecked NPE, or even better, unchecked > IllegalArgumentException, because it's just writing to a map. But it > was Free Software, so maybe you got what you paid for. :-( > > Sorry you've been troubled by this questionable engineering. At least > you're free to choose another JSON serialization library. > > But I am curious why you're serializing to JSON but not sending or > saving it anywhere, unless I'm misunderstood your previous message? > > Doug -- You received this message because you are subscribed to the Google Groups "Android Developers" group. To post to this group, send email to android-developers@googlegroups.com To unsubscribe from this group, send email to android-developers+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/android-developers?hl=en
