Date: Sat, 13 Dec 1997 23:59:57 -0800 From: Amancio Hasty <hasty@rah.star-gate.com> To: Luigi Rizzo <luigi@labinfo.iet.unipi.it> Cc: joerg_wunsch@uriah.heep.sax.de, freebsd-hackers@FreeBSD.ORG Subject: Re: Proposed code merge, objections? Message-ID: <199712140759.XAA10171@rah.star-gate.com> In-Reply-To: Your message of "Sun, 14 Dec 1997 06:53:01 %2B0100." <199712140553.GAA01161@labinfo.iet.unipi.it>
next in thread | previous in thread | raw e-mail | index | archive | help
This is an old problem that we are currently stuck with -- that is the need to have a 3.0 release which aint' going to happen anytime soon well at least according to the last thread that addressed such issue. New kernel features should go in -current. If we open the floodgate to 2.2-stable we may risk de-stabilizing that release;additionally, I don't enjoy too much "#if FreeBSD xxxx " or "#ifdef <system_feature> on the kernel. The Bt848 driver was entirely developed on -current the main bugs found across the board where specific to PCI chipsets --- natoma vs triton . Granted the test population was greater because the driver compiles on both 2.x systems and 3.0 -current systems. One issue that I see is that many are probably scare of 3.0 -current and are not willing upgrading to it. Perhaps we should try to generate a 3.0-release without SMP support or rather make it obviouse that is not yet ready and a little difficult to enable. Amancio > > > I've been told there is no official policy on this, so I want some > > > feedback. I am considering currently a merge of the alog driver > > > (Industrial Computer Source AIO8-P) into -stable. > > > > -stable is not meant to accumulate new features from -current. Unless > > there are strong reasons, new drivers should not migrate there. This > > _is_ official policy. > > So to put things differently, what is the policy to move features > from -current to -stable ? Unfortunately we don't have a port > category for kernel enhancements, maybe we should add one. > > Lot of people prefer to run one version of the OS across different > systems, and this is generally a RELEASE or -stable, and it is a > bit too much to ask people to upgrade to -current to try out some > (for them) interesting feature. > > For things like drivers etc we can gain a lot from testing on different > hardware. I have had a lot of useful feedback from 2.2. users on the sound > driver, infinitely more than what I got from -current users when the > driver was integrated there. I think the same happened for the Bt848 > driver. > > While it is ok to leave individuals (like me) put up a set of patches > for a kernel enhancement, it is annoying for end users to follow the > announcements on the list and hunt for such patches. It would be much > more convenient to get them through the standard means (cvsup etc.) > > Cheers > Luigi > -----------------------------+-------------------------------------- > Luigi Rizzo | Dip. di Ingegneria dell'Informazione > email: luigi@iet.unipi.it | Universita' di Pisa > tel: +39-50-568533 | via Diotisalvi 2, 56126 PISA (Italy) > fax: +39-50-568522 | http://www.iet.unipi.it/~luigi/ > _____________________________|______________________________________ >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199712140759.XAA10171>