From owner-cvs-all Fri Apr 17 09:39:11 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id JAA25252 for cvs-all-outgoing; Fri, 17 Apr 1998 09:39:11 -0700 (PDT) (envelope-from owner-cvs-all@FreeBSD.ORG) Received: from ns1.yes.no (ns1.yes.no [195.119.24.10]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id QAA25232 for ; Fri, 17 Apr 1998 16:39:07 GMT (envelope-from eivind@bitbox.follo.net) Received: from bitbox.follo.net (bitbox.follo.net [194.198.43.36]) by ns1.yes.no (8.8.7/8.8.7) with ESMTP id QAA06828; Fri, 17 Apr 1998 16:39:04 GMT Received: (from eivind@localhost) by bitbox.follo.net (8.8.8/8.8.6) id SAA08646; Fri, 17 Apr 1998 18:39:03 +0200 (MET DST) Message-ID: <19980417183902.45662@follo.net> Date: Fri, 17 Apr 1998 18:39:02 +0200 From: Eivind Eklund To: Matthew Hunt , Jun Kuriyama Cc: cvs-committers@FreeBSD.ORG Subject: Re: cvs commit: CVSROOT avail References: <199804171557.IAA05893@freefall.freebsd.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.89.1i In-Reply-To: <199804171557.IAA05893@freefall.freebsd.org>; from Satoshi Asami on Fri, Apr 17, 1998 at 08:57:28AM -0700 Sender: owner-cvs-all@FreeBSD.ORG Precedence: bulk On Fri, Apr 17, 1998 at 08:57:28AM -0700, Satoshi Asami wrote: > asami 1998/04/17 08:57:28 PDT > > Modified files: > . avail > Log: > It is my second pleasure in two days, this time welcoming > Mr. Matthew Hunt ("mph") to our esteemed (and overworked) rank of > ports committers. Welcome aboard, Matthew and Jun, and thanks for going with FreeBSD Spacetours. (All of the below goes for you too, Jun - just replace Matthew's login with your own). Your CVS-meister will be Peter Wemm - take any repository-related problems with him. Patching the hull yourself (AKA writing directly to the repository) is strictly not allowed without prior permission; this include all cases. In case of emergency, Cc: to root@freebsd.org. Be aware of the severe implications of running a blow-torch ('cvs import') in the ship - it might sometimes be necessary, but we really recommend using Joergs new blowtorch with safety features (the 'easy-import' script on freefall). Besides Peter Wemm, there are other officers you'll get to know: Satoshi Asami A lean, mean ports-machine. This guy owns the docking machinery (bsd.port.mk) which you are not allowed to touch without permission, and have ultimate authority over all modifcations to docking setups (ports). Touch them when he's declared a hands-off, and you'll get your commit privileges suspended faster than you can say 'ports-freeze'. To avoid him having to throw too many dinner-plates through the hall (fixing up trivial mistakes), it will be appreciated if you check that your new docks work with autopilots (packages correctly), apply to the correct class of ships (fetch correctly), and pass the standard tests (compile from scratch and pass portlint). Jordan K Hubbard The usual takeoff-manager (release-engineer). When you forget to merge good, wholesome changes from -current (HEAD) to -stable (at the moment RELENG_2_2 - not some magic number), it is Jordan who is most likely to get the extra work. So remember to merge - he's overworked enough as it is. Jordan is also the guy who is constantly claiming that he's not forgetting to set up forward files for new committers, so you'd probably be well of logging into hub.freebsd.org and setting /var/forward/mph to point to where you want your mail to land. There's probably some letters from your sweetheart waiting for you on hub already.. Bruce Evans Perfection police. When you do a commit that could have been done better, Bruce will be there. Even though it can be extremely annoying to be reminded of your own imperfections at all the in-opportune moments, Bruce is a good guy too. Garrett Wollman If you speak some obscure network internals, and you think FreeBSD has a lithp somewhere, then Garret is likely a good person to know about - he's responsible for the overall state of FreeBSD networking internals. John D. Polstra John is the guy behind our great 'shoot pieces of FreeBSD all over the world' technology, AKA cvsup. If you need to cvsup directly from freefall (and this can be useful at times, though it shouldn't be a regular habit unless you need it) you should ask John to set this up for you. ... and some important privates: Steve Price Steve is Problem Report-meister. and will be assigning you additional trouble sometimes. He tries to fix it all himself, but should really be less reluctant to use us others. Brian Somers Official maintainer of /usr/bin/ppp and LPD. He's ten times as efficient as god on changing PPP, so if you intend to help out there, be prepared to feel quite inadequate as he end up doing everything. Mark Murray Mark is the guy in the black hat, the one always one step ahead of NSA. If you can't read this: Vs lbh qb nal jbex gung vf pelcgb-eryngrq - cyrnfr fraq zr (znexz@serrofq.bet) n pbcl bs lbhe cngpurf sbe vapyhfvba vagb gur Vagreangvbany Pelcgb Ercbfvgbel. Orggre fgvyy, nfx zr sbe na nppbhag ba gung znpuvar naq pbzzvg gur qvssf gurer nf jryy. Guvf nccyvrf gb nyy pbqr va fep/rObarf, frp/frpher naq fep/pelcgb. then you shouldn't be touching src/eBones, sec/secure and src/crypto. (Anybody that feel left out - speak up. I've just included the people that I thought were likely to need their roles explained - no significance attached). For your convenience, we repeat some of the magic tricks of using CVS with SSH below: 1. Always use ssh public key authentication with agents to work against freefall. This is (a) secure and (b) extremely convenient. Details below. 2. Set your CVS_RSH=ssh 3. Store your local CVS-repository in /home/ncvs, as freefall hasn't got a /home/FreeBSD alias, so you can't store the repository anywhere intuitive. (Well, OpenBSD doesn't implement /home/OpenBSD, either - they're worse...) 4. Set alias lcvs='cvs -d /home/ncvs' 5. Set alias fcvs='cvs -d mph@freefall.freebsd.org:/home/ncvs' Now, use lcvs to checkout/diff/update locally, and when you want to commit to freefall, use fcvs commit in the same directory. Remember: Doing a 'lcvs diff' before you commit is usually a good idea. 6. (Vary depending on preferences only) echo 'diff -u' > ~/.cvsrc echo 'update -Pd' >> ~/.cvsrc will give you unified diffs and pruned updates with directories. You'll probably want these two options. Now, for the ssh-setup (if you haven't done this already): 1.1 Put 'ssh-agent sh ~/.xsession-real' in ~/.xsession after having copied your old .xsession to .xsession-real. 1.2 Generate an ssh private/public key pair by running 'ssh-keygen' 1.3 Copy the public part of the key (~/.ssh/identity.pub) to freefall:/home/mph/.ssh/authorized_keys 1.4 Check that /home/mph and /home/mph/.ssh is owned by 'mph', and does NOT have group write access. (Otherwise, it Just Won't Work). Now, when you want to get ready for authentication, run 1.5 'ssh-add' to add your private key to the authentication agent. Take particular note of the neat X-display if you run 'ssh-add < /dev/null' - I have this in my startup. (I am in a fairly secure environment, and I lock my display any time I leave my machine). After having stored your key in the authentication-agent (step 1.3 above), just ssh to the places you've added your identity.pub to - you'll be authenticaten through your agent, no passwords. The same goes for secure remote `cvs commit's. Oh, and when you fix the fact that somebody else have delivered a report of a leak in the ship or a suggestion for an improvement[1], you can run 'edit-pr ' on freefall to close it. It's easy, just like closing a door - you just change 'open' to 'closed'. Relax now - the impending takeoff will be at your hands, but it will _not_ include any mutilated bodies on the ground (AKA world-breakages) or bodies falling from the sky (AKA release-breakages). Everything will be just fine. No committer has ever screwed up. :-) Starship Mate Eivind signing off. [1] Both are known as Problem Reports, or PRs for short - this is because they give the committers extra problems. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message