Date: Sat, 30 Sep 2017 12:30:14 -0700 (PDT) From: Don Lewis <truckman@FreeBSD.org> To: rkoberman@gmail.com Cc: freebsd-ports@freebsd.org Subject: Re: Status of portupgrade and portmaster? Message-ID: <201709301930.v8UJUERE025933@gw.catspoiler.org> In-Reply-To: <CAN6yY1ukz_gn3Ny4J52qoq0KjGmvDxLEZrBenXPA7o-YC%2BHSyg@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 30 Sep, Kevin Oberman wrote: > John did state that he would continue to support synth. I can't say if he > has continued to make contributions. In any case, only poudriere is > available for maintaining ports in HEAD and I, for one, feel that it is > simply unacceptable as it make FreeBSD unusable for those of us with only > "small" systems where the weight poudriere simply can't be justified. (I > have no system with other than SATA disk drives and, for my current needs, > 1 TB of SATA on my development system and .5TB on my production system is > adequate. Both systems are physically constrained in expansion capability, > though otherwise easily meet my requirements. 1 TB should be plenty-o-room for poudriere in most cases. The machine I use for building packages only has a mirrored pair of 1TB Western Digital Green drives that were purchased years ago for another project where they were eventually replaced, so I just happened to have them handy when I put my package builder together. I use that box to build a set of about 1800 ports for FreeBSD 10 i386, FreeBSD 11 amd64, FreeBSD 11 i386, and FreeBSD 12 amd64. There are some of the larger ports in that set, like chromium, firefox, thunderbird, openoffice-4, openoffice-devel, and libreoffice. I also run the other supported release / x86 combinations when I'm doing port testing. %zpool list NAME SIZE ALLOC FREE EXPANDSZ FRAG CAP DEDUP HEALTH ALTROOT zroot 888G 608G 280G - 64% 68% 1.00x ONLINE - The biggest consumer of space is actually my collection of VM images that get used when this machine isn't building packages. My biggest constraint is CPU cycles. I/O is generally not a problem because I was able to max out RAM in this machine and use tmpfs for most things in poudriere. Centralizing port building like this allows me to continue to use some really ancient and slow hardware, like my 2003 vintage laptop that only has a 160 GB drive that is split between both Windows and FreeBSD and 1 GB of RAM. It's still perfectly adequate for light use running a browser and editing documents with one of the office products, but trying to build those ports on it is totally out of the question. I also have a Via C3 machine with only 256 MB of RAM that I use as a lightweight server and I maintain using the packages produced by poudriere. This also allows me to avoid bogging down my daily desktop machine with port builds. It's somewhat more modern, but a big batch of port builds would probably make it really laggy for a long period of time. That said, if you only have one machine, synth is probably a better choice. The situation on 12.0 should be fixable by someone with the proper skillset, but it illustrates the problem of synth being the only real consumer the ADA toolchain (which John also maintained) on FreeBSD. Basically we've got an important tool that had a single point of failure, and even without the the politics, we'd be in the same situation if John had been run over by a bus. By contrast, poudriere is mostly shell scripts, which makes me shudder for totally different reasons. Another issue is that synth is only available on x86 because that what the toolchain limits it to, so that leaves the our other architectures out in the cold.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201709301930.v8UJUERE025933>