From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Fri, 9 Jun 2006 08:10:35 -0700 Message-Id: <200606090810.35322.corey_s@qwest.net> From: "Corey" To: 9fans@cse.psu.edu Subject: Re: [9fans] gcc on plan9 User-Agent: KMail/1.9.3 References: <55be250fa871ed9b484f3725566b80f7@quanstro.net> In-Reply-To: <55be250fa871ed9b484f3725566b80f7@quanstro.net> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Topicbox-Message-UUID: 647a946c-ead1-11e9-9d60-3106f5b1d025 On Friday 09 June 2006 06:15, quanstro@quanstro.net wrote: > On Fri Jun 9 07:41:19 CDT 2006, corey_s@qwest.net wrote: > > Again, my mistake - I accidently crossed-wires by mentioning c99; when the > > point I was trying make was to draw the parralel/similarity between C and libc, > > and Obj-C and GNUstep ( or FoundationKit, or whatever ). > > regardless, my question remains the same. can you name a specific c99 libc > bit that is missing and explain why it could make plan 9 better? > I am certainly not qualified to raise any contention with what portions of the C99 standard have been purposefully left out of Plan 9's libc. Plan 9 has been around for a long time now, if something obvious or necessary were missing from its libc, I can only imagine that it would have already been remedied. Am I correct in interpreting your question as an assertion that, if Plan 9's libc is functionaly complete, then it logically follows that an Objective-C runtime library such as GNUstep would not bring anything useful to the table? > ii'm not yet convinced that objective c would be a bad idea, but neither have you > built any case for it. > I won't be able to convince, with certainty, even _myself_ that obj-c would be anything more than at best redundant on Plan 9 until I have actually evaluated obj-c _on_ Plan 9. That's the whole point of this potentialy inane little experiment; I'm dabbling. To put it simply, I'm thinking chocolate (Plan 9) and peanut butter (Objective-C) just might make a tasty treat, but I won't know until I try it. Have you yourself already written a non-trivial 9p service in objective-c on Plan 9 and compared that to an equivalent plain-c 9p service? Educated speculation is of course necessary to help prevent rushing off into a fool's mission, but there comes a point where actual applied experience is required in order to turn that speculation into something more concrete. > what specific objective c properties would be beneficial? > Richer exception handling, richer string handling, garbage collection, categories, protocols, introspection, dynamic dispatch, dynamic typing, heck, dynamic everything, unit testing, steptalk. Also: Familiarity. Comfort. ... which in my humble opinion, _within_reason_, can be just as practical/valid as sheer technical merit. All I'm flirting with here, is the concept of a personal Plan 9 installation which is ever so slightly more abstracted from the bare metal. While perhaps you and likely others here might disagree that there _is_ any such "bare metal" exposure, or that this exposure is a feature ( it certainly is ), I myself have a notion that a slightly more "higher level" development environment in Plan 9 might make for interesting project. Others may share this opinion, and I think it would be... fun - to play around and see where it goes. Think of it more like "Plan Commons" rather than Plan 9. I don't want to encroach on Plan 9's clean-room identity - I'm not suggesting that Plan 9 compromise its current focus. If these sorts of questionable/scandalous undertakings are inappropriate to discuss or brain-storm here on the 9fans list, please let me know - I don't want to disturb the peace. > why don't you try using the system a bit. write some code. maybe read the a > aleph papers. i'd be interested in what you thought objective c could offer then. > I definitely look forward to doing so. Cheers! Corey