Date: Sat, 22 May 2004 16:40:44 -0400 From: The Anarcat <anarcat@anarcat.ath.cx> To: core@freebsd.org Cc: timh@tjhawkins.com Subject: LibH pronounced dead, need for a new leadership Message-ID: <20040522204044.GB48382@shall.anarcat.ath.cx> In-Reply-To: <4CA59E9E-AC2D-11D8-871A-000A95AD47EE@queasyweasel.com> References: <00d701c43c85$0ca36bd0$6501a8c0@yourw92p4bhlzg> <20040518171751.GA58065@shall.anarcat.ath.cx> <hf1xlhpk17.fsf@valbanera.sdsc.edu> <4CA59E9E-AC2D-11D8-871A-000A95AD47EE@queasyweasel.com>
next in thread | previous in thread | raw e-mail | index | archive | help
--hHWLQfXTYDoKhP50 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable Right. I totally agree with you here Jordan. Maybe the key would be to try to port successful tools from the linux side. I don't know. One thing I know is that I have completely departed from developpement on libh in particular and FreeBSD in general, because of a lack of time, mainly, but also because of the frustrating experience of maintaining a low-end (and here I mean 64M of ram :) server with FreeBSD, compared to other systems. You mentionned Debian, well that's one thing: I discovered APT. I feel that working on a package management system is completely futile once the open source world has such a great tool already tried and true. They are making APT work with other packaging formats (RPM, for now), and I don't see why this wouldn't apply to our side. APT could work with the current FreeBSD package scheme, which itself could be improved upon to match the capabilities of other packaging systems, as being proven by the libpkg experiment. Building an installer is always a bit of the "re-inventing the wheel" hell. It's already been done, it's never quite right, it's extenuating and no-one wants to do it anyways. Maybe claiming the death of libh will bring life to other projects. I'm thinking of libpkg which kinda sprung to life after all, without libh mind you, and the binary updates stuff. Both projects are very exciting and I'm really considering the binary updates stuff for the maintenance of the FreeBSD boxen I have left. Maybe all the BSDs need to get together to solve this problem. I can't believe we can't collectively solve this problem. What I see is multiple people working on similar or related projects each on their sides: pkgsrc on NetBSD, libpkg on FreeBSD, a graphical installer (i don't recall the name) on OpenBSD. We need to focus all those projects together. Getting libh out of the way for that is a great move, actually, and might actually open up the gates for the next step. I don't nor can lead this one. Someone will have to take care of that, some people who have their feet deep into the BSDs and are socially networked in the community. That is why I CC'd core. Hear our call! FreeBSD needs a new installer! Let's wake this daemon! I don't know where or how to announce that, maybe you guys could point me in the right direction, maybe you guys want to do it yourselves. :) Anyways, there's surely a debate to open, although I'm not sure it'll bring anything. This is a highly bikeshed-susceptible material, and what it needs is work, not opinions or debates. Long life FreeBSD, A. On Sam mai 22, 2004 at 01:19:16 -0700, Jordan Hubbard wrote: > libh never lived up to its promise, that much is certainly true. >=20 > The pity of all this is that there's never been the right sort of=20 > project management or even group consensus necessary to come up with=20 > any credible replacement strategy. Libh was never the cause of=20 > FreeBSD's lack of a good replacement installer, merely a symptom of the= =20 > general lack of desire to deal with this problem space. People just=20 > keep doing (barely) evolutionary hacking on the existing tools and=20 > somehow perceiving this as sufficient and that's just a shame since=20 > it's limited FreeBSD's success (trying, as it has, to be "the=20 > mainstream BSD"). Meanwhile, the SuSE and Red Hats of the world=20 > continue to march on and offer installation and package management=20 > experiences that are at least an order of magnitude more extensive than= =20 > FreeBSD's. Even Debian, who one could argue is more of a reasonable=20 > comparison to FreeBSD given that it's all been volunteer driven rather=20 > than subsidized by a corporate sugar-daddy, offers a better=20 > installation experience. Not an order of magnitude better, but they've= =20 > at least put some effort into this area and appear to be continuing to=20 > do so. >=20 > If libh is finally pronounced dead and buried without ceremony, I hope=20 > that this is announced a little more widely and with at least some=20 > attempt to provoke discussion concerning "if not that, then what? =20 > Nothing? Really??" >=20 > - Jordan >=20 >=20 > >>I feel that libh has taken too long the role of a failed promise for=20 > >>the > >>FreeBSD community and rightly deserves to die now.=20 >=20 --hHWLQfXTYDoKhP50 Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (FreeBSD) iD8DBQFAr7rLttcWHAnWiGcRAstxAKCgbeM+Tu/+znhVT3SZ65FBhRUN0wCfc1U0 0gSElD5hLQ4S9HK70mM/RyI= =ZZdE -----END PGP SIGNATURE----- --hHWLQfXTYDoKhP50--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040522204044.GB48382>