From mboxrd@z Thu Jan 1 00:00:00 1970 X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me X-Spam-Level: X-Spam-Status: No, score=-1.9 required=5.0 tests=BAYES_00 autolearn=unavailable autolearn_force=no version=3.4.4 Path: eternal-september.org!reader01.eternal-september.org!reader02.eternal-september.org!.POSTED!not-for-mail From: Paul Rubin Newsgroups: comp.lang.ada Subject: Re: How to get Ada to ?cross the chasm?? Date: Thu, 10 May 2018 13:00:24 -0700 Organization: A noiseless patient Spider Message-ID: <87k1sb1dt3.fsf@nightsong.com> References: <322f9b26-01de-4753-bb50-6ef2f3d993d8@googlegroups.com> <87a7th9pd1.fsf@nightsong.com> <87h8no1nli.fsf@nightsong.com> <874ljo1hvy.fsf@nightsong.com> <87vac4z2lh.fsf@nightsong.com> <87lgcszjdn.fsf@nightsong.com> <87sh6z1kkg.fsf@nightsong.com> Mime-Version: 1.0 Content-Type: text/plain Injection-Info: reader02.eternal-september.org; posting-host="8ee8a87c586ad688bbbd7b3afaf6ee4e"; logging-data="25060"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX19f/fMWyWUWMD2z+IPIc77u" User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/25.3 (gnu/linux) Cancel-Lock: sha1:aOLfjm/jxZbSmlWp+TvOd0ljr5k= sha1:yN6dKOCwpqSvhABkaOo4QT1P02k= Xref: reader02.eternal-september.org comp.lang.ada:52208 Date: 2018-05-10T13:00:24-07:00 List-Id: Dennis Lee Bieber writes: > Then Ada has been doomed since 1980: > "Designed by committee" (so was COBOL and it lasted) I don't see that as a problem: Haskell and Scheme were also designed by committees and PL geeks are very fond of them. > "Too large" There's some confusion here, I wrote that >> if "crossing the chasm" means getting everyone to read the 1000 page >>(or whatever) ARM, then Ada is doomed. That doesn't mean the language's largeness implies doom all by itself. C++ is large but it's mainstream. The issue is whether you have to read a 1000 page document to use the language at all, or if it's sufficient to start with a subset and expand outward from there. Most C++ users start with a subset, and I suspect not very many know the full language. I myself certainly don't. I would have thought starting with a subset of Ada is also legitimate, but if it's not, then Ada is probably doomed. > "Compiler is too slow" (something wrong with one's software > development practices when the compiler speed is of more concern than the > lifetime of a correct application) That was in the discussion of how to add more phrases to the CBSG, which is to say, how to change its configuration in a deployment environment. Randy's answer was to just include the Ada compiler as part of the deployment, and add phrases by modifying and recompiling the Ada code. I think Ada programmers are used to the idea of the development environment and the deployment environment being two different things. A big slow development environment might be tolerable, but if the Ada compiler is part of the deployment as well, then its size and speed have to be considered again. The program takes 7 seconds to recompile on a fast x86 box, and 1.5 minutes on an ARM server that I just tested. Since anything post-deployment is "run time", that number has to be considered as part of the runtime performance, which makes GNAT (with this implementation approach) look pretty terrible compared to something like Forth or Picolisp. Anyway, one of the claimed advantages of Ada over C++ is that Ada's module system makes compilation faster, so perhaps you're trying to have it both ways.