From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail1-relais-roc.national.inria.fr (mail1-relais-roc.national.inria.fr [192.134.164.82]) by walapai.inria.fr (8.13.6/8.13.6) with ESMTP id pB79r8TX017686 for ; Wed, 7 Dec 2011 10:53:08 +0100 X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: Ah8BAOM2307ZSMDdi2dsb2JhbABDqlIiAQEBCgsLBxIFIoFyAQEEATo/BQsLISUPAQQNGyETFIdzArVmizQEmiGFMYc9 X-IronPort-AV: E=Sophos;i="4.71,313,1320620400"; d="scan'208";a="134301277" Received: from fmmailgate01.web.de ([217.72.192.221]) by mail1-smtp-roc.national.inria.fr with ESMTP; 07 Dec 2011 10:53:03 +0100 Received: from moweb002.kundenserver.de (moweb002.kundenserver.de [172.19.20.108]) by fmmailgate01.web.de (Postfix) with ESMTP id 9D6371A434B6F for ; Wed, 7 Dec 2011 10:53:02 +0100 (CET) Received: from frosties.localnet ([95.208.118.96]) by smtp.web.de (mrweb002) with ESMTPA (Nemesis) id 0MD8RQ-1RZHQv2Wz6-00GbtY; Wed, 07 Dec 2011 10:53:02 +0100 Received: from mrvn by frosties.localnet with local (Exim 4.77) (envelope-from ) id 1RYEBR-0007zh-45; Wed, 07 Dec 2011 10:53:01 +0100 From: Goswin von Brederlow To: Ashish Agarwal Cc: Paolo Donadeo , OCaml mailing list References: <4EDE33A0.6070004@gmail.com> <4EDE568C.9040803@lexifi.com> Date: Wed, 07 Dec 2011 10:53:01 +0100 In-Reply-To: (Ashish Agarwal's message of "Tue, 6 Dec 2011 20:48:26 -0500") Message-ID: <87liqooghu.fsf@frosties.localnet> User-Agent: Gnus/5.110009 (No Gnus v0.9) XEmacs/21.4.22 (linux, no MULE) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Provags-ID: V02:K0:4weRs5TqNiAMW8ifmEvucMfW7bVdywHZTL8pKkpNx9e quD7eTJSfzvIhyM3rC8CkKusr+UasSvgEWPfWNrrq3rjUrlY5L k76O0iVAslCUWmtGkUxZwB1aqGZQD2+B+/OpnUIGt8+moMk4tc OeznlAHEmA/6viSrji0ksYqiqr8XJTBKBTI5DioL3yDjYrMEWN 8e7rKYkgRnDzeA+4n4zWA== Subject: Re: [Caml-list] Some comments on recent discussions Ashish Agarwal writes: > A "standard library" does not imply "big" or that it is part of the standard > distribution. Both Batteries and Core would make fine standard libraries. > Neither is very big and both are independent of the standard distribution. But > having 5 different standard libraries is annoying precisely because the library > covers just basic concepts. We don't need 5 APIs for string functions. They are > all easy to learn independently but a pain if you have to know them all. A > beginner won't bother taking the time to do a google search and find which one > to use; before that they'll use another language. We used to have 100 seperate little modules all over the place. Now we have 2 contenders for a "standard library". I think we are going the right way there. Lets leave that allone for at least a year or two. And we do (initially) need 5 APIs for string functions and we need to test them all and decide which one fits best. And there can be more than one that should be kept. E.g. one API with exceptions and one with option types. But then all modules in the "standard library" should provide those APIs we want to keep consistently across all modules. Lets agree that we eventually want one standard library. Can we also agree that the standard library can be external to the compiler, like Batteries or Core is? > I also think the debate about github, ocamlforge, etc is fruitless. People have > different preferences and who knows what new tools will be developed a couple > years from now. Instead of arguing about which one is best, we should accept > that we will not agree on this. And why do we need to? We can aggregate the > information from multiple sources and present it in a uniform way on a > community website. ACK. Totaly different fruit. Even talking about a standard library is imho fruitless (at this point in time) and OT to the original problem. MfG Goswin