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 <no.email@nospam.invalid>
Newsgroups: comp.lang.ada
Subject: Re: meaningfully/compellingly "advertising" Ada on StackOverflow
Date: Thu, 17 May 2018 21:37:22 -0700
Organization: A noiseless patient Spider
Message-ID: <87bmddshod.fsf@nightsong.com>
References: <6420bab2-0aef-4d36-b978-525e4de45e7e@googlegroups.com>
	<aed23075-1271-4e2c-b5f2-8da705e1a73e@googlegroups.com>
	<a86dc533-a8f5-4d1c-9b4c-11d7963c6d15@googlegroups.com>
	<pdkrvk$n53$1@franka.jacob-sparre.dk>
	<1559505943.548291689.457576.laguest-archeia.com@nntp.aioe.org>
	<87fu2psqpj.fsf@nightsong.com>
	<300fd624-72a0-4171-90dc-49ad5df4fa21@googlegroups.com>
Mime-Version: 1.0
Content-Type: text/plain
Injection-Info: h2725194.stratoserver.net;
 posting-host="f0affb4512ce5ff754742fac3a6698cf";
	logging-data="24622"; mail-complaints-to="abuse@eternal-september.org";
	posting-account="U2FsdGVkX18gUlb03G6baPyLVMKPSajm"
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/25.3 (gnu/linux)
Cancel-Lock: sha1:Gcx+nASQsqfgkEuhCl+nNINLG3M=
	sha1:uoOA8k+2A/1f2KvH2VBX/9bKbSg=
Xref: reader02.eternal-september.org comp.lang.ada:52421
Date: 2018-05-17T21:37:22-07:00
List-Id: <comp.lang.ada>

Lucretia <laguest9000@googlemail.com> writes:
> You don't get the point I made. That "benchmark" is to see how fast
> different languages go using the same algorithm, without using any
> safety features. If you hobble Ada by including them, you will make
> people think it's really slow. It's not about the safety features. Do
> you get it now?

No.  The Python and Haskell programs are slow because Python and Haskell
also have safety features.  The data signals the right conclusion: C is
dangerous and fast, while Python and Haskell are safe and slow.  A
completely standard cost-benefit tradeoff.  As a user I appreciate being
presented with this info so I can make an informed choice.

Ada's reason for existence revolves around safety.  The benchmarks with
the safety checks off are useless from my perspective as a C programmer.
If I want to write fast dangerous code I already have C for that.  If I
look at Ada at all, it's for the purpose of writing safe code.

So I care primarily about the speed of the safe version (in part so I
can compare it with Python and Haskell).  The unsafe version is just C
with bloated syntax and I'm content to keep using C instead.

I like Dan'l Miller's suggestion of having two Ada benchmarks, giving
both the safe and unsafe measurements.  Then people can decide what's
relevant to their own requirements.  Giving just the unsafe measurements
is uninformative or maybe even misleading.