|
|
|
Posted
23 days
ago
by
t...@mirbsd.org (MirOS Developer tg)
The MirBSD Korn Shell R45 has been released today, and R44 has been named the new stable/bugfix-only series. (That’s version 45.1, not 0.45, dear Homebrew/MacOSX packagers.) Packagers rejoice: the -DMKSH_GCC55009 dance is no
... [More]
longer needed,
and even the run-time check for integer division is gone. Why? Because I
realised one cannot use signed integers in C, at all, and rewrote the mksh(1)
arithmetics code to use unsigned integers only. Special thanks to the people
from musl libc and, to some lesser
amount, Natureshadow for providing me with ideas what algorithms to replace
some functionality with (signed shell arithmetic is, of course, still usable,
it is just emulated using unsigned C integers now).
The following entertainment…
tg@blau:~ $ echo foo >/bar\ baz
/bin/mksh: can't create /bar baz: Permission denied
1|tg@blau:~ $ doch
tg@blau:~ $ cat /bar\ baz
foo
… was provided by Tonnerre Lombard; like Swedish, German has
got a number of words that cannot be expressed in English so I feel
not up to the task of explaining this to people who don’t know the
German word “doch”, just rest assured it calls the last input line
(be careful, this is literally a line, so don’t use backslash-newline
sequences) using sudo(8). [Less]
|
|
Posted
23 days
ago
by
t...@mirbsd.org (MirOS Developer tg)
The MirBSD Korn Shell R45 has been released today, and R44 has been named the new stable/bugfix-only series. (That’s version 45.1, not 0.45, dear Homebrew/MacOSX packagers.) Packagers rejoice: the -DMKSH_GCC55009 dance is no
... [More]
longer needed,
and even the run-time check for integer division is gone. Why? Because I
realised one cannot use signed integers in C, at all, and rewrote the mksh(1)
arithmetics code to use unsigned integers only. Special thanks to the people
from musl libc and, to some lesser
amount, Natureshadow for providing me with ideas what algorithms to replace
some functionality with (signed shell arithmetic is, of course, still usable,
it is just emulated using unsigned C integers now).
The following entertainment…
tg@blau:~ $ echo foo >/bar\ baz
/bin/mksh: can't create /bar baz: Permission denied
1|tg@blau:~ $ doch
tg@blau:~ $ cat /bar\ baz
foo
… was provided by Tonnerre Lombard; like Swedish, German has
got a number of words that cannot be expressed in English so I feel
not up to the task of explaining this to people who don’t know the
German word “doch”, just rest assured it calls the last input line
(be careful, this is literally a line, so don’t use backslash-newline
sequences) using sudo(8). [Less]
|
|
Posted
about 1 month
ago
by
bsie...@mirbsd.org (MirOS Developer bsiegert)
I uploaded a full bulk build of binary packages for MirBSD/i386 corresponding to the pkgsrc-2013Q1 release. About 7,000 binary packages are available in this build, including the pkgin package manager that makes installing binary
... [More]
packages as easy as apt.
See the pkgsrc page for
instructions on how to install pkgsrc for MirBSD.
Build
logs are available on S3. [Less]
|
|
Posted
about 1 month
ago
by
bsie...@mirbsd.org (MirOS Developer bsiegert)
I uploaded a full bulk build of binary packages for MirBSD/i386 corresponding to the pkgsrc-2013Q1 release. About 7,000 binary packages are available in this build, including the pkgin package manager that makes installing binary
... [More]
packages as easy as apt.
See the pkgsrc page for
instructions on how to install pkgsrc for MirBSD.
Build
logs are available on S3. [Less]
|
|
Posted
about 1 month
ago
by
t...@mirbsd.org (MirOS Developer tg)
Since a while… I am a proud On the other hand… I should probably put up my own, local, list of found caches, considering what happened to me on “Open”caching. And maybe write intros for people new to geocaching
... [More]
, since it’d be
virtually no work now had I done it initially. (And for fanfiction
readers! I wish I’d kept a list of read fics, not just of
these I currently read and/or are currently unfinished.) [Less]
|
|
Posted
about 1 month
ago
by
bsie...@mirbsd.org (MirOS Developer bsiegert)
On Saturday March 23, this year's pkgsrc conference (pkgsrccon 2013) took place in Berlin. Julian Fagir organized it with unending energy, even though pkgsrc is not the primary focus of his NetBSD work. He just took matters in his
... [More]
hands because no one else stepped forward. A big
thanks for that!
The flight from Zurich to Berlin was uneventful. It was my first
flight to TXL airport (I normally arrive at SXF), and arriving there is
incredibly quick and convenient compared to the latter. The terminal is
very small, and it takes just five minutes to go from the plane to a
bus to the city.
Now for the conference itself: we started at 12pm on Saturday with a
program of talks but no fixed schedule. Due to this, the conference
took a long time (we finished only at 9pm or so) but on the other hand,
it allowed for lots of interesting and fruitful discussion. At no point
did we have to cut a question short because of a lack of time. Overall,
I think that this was an excellent choice and made the conference more
useful and productive.
We were about 21 people – mostly pkgsrc developers (of course)
but also a Debian Developer (Ralf Treinen, who presented his work on
Mancoosi), a FreeBSD dev and some interested users. I won't give an
exhaustive recollection of all talks here but simply comment on a few
ones that I found particularly interesting.
The most important theme of the conference was virtualization and
cloud computing. Jonathan Perkin and Filip Hajny gave a talk about
their company's product, SmartOS, and how it uses pkgsrc. SmartOS is a
"cloud OS" based on OpenSolaris. It boots from a read-only medium (such
as a CD) into a lean system that only does the administration of all
the zones that it runs. All useful work happens in zones, which are a
sort of lightweight VM solution specific to OpenSolaris. The zone
images include access to a very complete set of pkgsrc packages for
things such as a compiler. They can also run other OSes (NetBSD!) by
setting up a zone that runs KVM. Joyent runs a large public cloud with
SmartOS, where customers purchase virtual machines by the hour. This is
similar to Amazon EC2 but with a focus on high performance.
Hubert Feyrer gave another talk about a similar theme. He described
the use of Ansible for provisioning and setting up VMs. Ansible can
automatically create VMs on EC2, gather the necessary information (such
as the IP address) and do various setup tasks without further user
interaction. This was all very impressive, even though the live demo
failed. This was for two reasons: Somebody deleted the sudo package for
amd64 from the NetBSD ftp server (boo), and the i386 VM failed to come
up, the kernel paniced on startup. Joerg speculated that this was due
to _some_ machines in their DC not having PAE enabled, while the i386
kernel uses PAE. This was interesting, as I had noticed the very same
problem when I set up the netbsd-386-bsiegert continuous builder for
Go.
Amitai Schlair alias Schmonz came out in a passionate defense of the
venerable pkglint. He put
the source on github and started refactoring the code
and adding tests. He calls this approach TED for "Test Eventually
Development" ;) and advocated a similar approach for the pkgsrc
infrastructure: Every time a developer takes five minutes to understand
a part of the infrastructure (when making a change, for instance), he
or she should write a test for it. This is a very pragmatic and doable
approach, in my opinion, and we should all do this.
I gave a slightly amended version of the "Go on NetBSD" talk I had
given at FOSDEM 2013. There were a lot of valuable questions and
discussion, both about the language and about how to package software
written in it.
Aleksej Saushev ended the day with a talk that was not in the program
about the Google Code-In and the problems that developers and
particularly new contributors face. If pkgsrc can get more contributors,
it gets more fixes, which in turn makes it more useful to users. More
usefulness leads to more users, leading to more contributors. We should
do more to get into this virtuous circle. There are about five
different mechanisms to build and/or deploy packages in pkgsrc: build
directly with "make package", pkg_chk, pkg_comp, the old bulk build
scripts and pbulk. The basic frustration that should be overcome is the
following: you want to upgrade a set of packages, the old ones are
removed, new ones are rebuilt, and the build fails. Rolling back is
difficult in general. pbulk could be a valuable solution to this, but
its standard config is heavily tailored for a different use case, and
its _two_ separate pieces of documentation are contradictory,
incomplete and confusing. So the talk contained a call for action to
fix those minor annoyances and generally document things better, which
makes it easier for everybody.
My take-home message – and my next project idea – is the following:
each time that I do a MirBSD bulk build using pbulk, I have to do a lot
of painful steps to set up the right build environment on all my
machines. This time, I will try to automate this process with Ansible,
making up the recipes as I go along, and then (more importantly)
publish these recipes for others to use and to share. [Less]
|
|
Posted
3 months
ago
by
t...@mirbsd.org (MirOS Developer tg)
*winkt Ventilator*
Natürlich mit MirKaffee (enthält Milch, Kakao, Kaffee, Rohrzucker)!
|
|
Posted
3 months
ago
by
t...@mirbsd.org (MirOS Developer tg)
I’ve been doing too much lately, which has led to reduced performance and enjoyment. Also I’ve not been able to work the full hours of my dayjob, reducing what I had on my overtime account. I’ll be taking a step back and try to
... [More]
un-load. This is my notice, I’m not explicit on
where, and I’m not cancelling anything special (not even those mentioned
in the next paragraphs).
I’m disappointed with Google/Nianticproject Ingress. It’s frustrating
(nothing lasts; also read this
posting), buggy, battery-draining, sometimes too time-consuming
(especially with only GPRS) and I don’t get warm with the Android 2.3
based Cyanogenmod on the borrowed device. Using it without a big
screen device having the Intel map next to you is futile. I could go into
detail but won’t. I won’t stop playing, as it’s a good excuse to go outside
and combines somewhat with geocaching (unless you’re trying to actually
play Ingress, in which case you’ll just be walking/cycling/driving between
portals at maximum speed). And there’s that connection with Liferay…
Fun is important in securing volunteer work; bugs and other
random happenings (example) can drain the fun.
To end on a positive note, I’m absolutely, totally happy with mksh user and distributor feedback,
including the bug reports and feature requests, how well almost all
people deal with feature rejection, and the speed of integration of
mksh(1) updates lately. The only thing I’m unhappy wrt. mksh is my
own lack of speed regarding implementing the cool new things I’ve
been, as an mksh user, waiting for because I want and even
need them for some cool programs written in mksh I would
love to write, so I can use them.
I’ve got roughly 350 mails in my INBOX (all read, but most of them
being action items; some due… before this weekend, evilly enough, the
one I’m thinking of is GnuPG/MIME encrypted, which means extra effort
to read it). Just so you know. (And a couple of other things that
really could use some fixing, which I can, in theory, do. And lots of
requests for spending real life time with.)
I’m still reachable via eMail and IRC (mostly), will respond, will
try to persuade my employer to send me to CLT 2013 next month… just,
don’t deadline me right now. I’m not taking a VACation either (though
I probably should, had I money). [Less]
|
|
Posted
3 months
ago
by
t...@mirbsd.org (MirOS Developer tg)
On Planet Debian, Vincent Bernat wrote: The drawback of this approach is that if you rebuild configure from the released tarball, you don’t have the git tree and the version will be a date. Just don’t do that. Excuse
... [More]
me‽
This is totally inacceptable. Regenerating files like aclocal.m4 and
Makefile.in (for automake), configure (for autoconf), and the likes is
one of the absolute duties of a software package.
Things will break sooner or later if people do not do
that. Additionally, generated files must be remakable
from the distfile, so do not break this!
May I suggest, constructively, an alternative? (People – rightfully,
I must admit – complain I’m “just” ranting too much.)
When making
a release from git, write the “git describe” output into a file. Then,
use that file instead of trying to run the git executable if .git/. is
not a directory (“test -d .git/.”). Do not call git, because,
in packages, it’s either not installed or/and also undesired.
Couldn’t comment on your blog, but felt strongly enough about this I
took the effort of writing a full post of my own.
(But thanks for the book recommendation.) [Less]
|
|
Posted
3 months
ago
by
t...@mirbsd.org (MirOS Developer tg)
The MirBSD Korn Shell must-have bugfix releases R42b (for the current formal release series) and R41c (for the last bugfix-only stable series) have been issued. Debian testing/unstable users get the fixes through their package
... [More]
management, versioned 40.9.20120630-7 targetting wheezy (hopefully).
Debian experimental users will receive a 42b-1 upload, closer to
what other distro packages provide, soonish, targetting jessie.
Since we lack anything better (that, unless Freshmeat, is actually
usable), by request of our packagers, new mksh(1) releases will be
announced on our miros-mksh@ mailing
list starting now. [Less]
|
Copyright
©
2013
Black Duck Software, Inc.
and its contributors, Some Rights Reserved. Unless otherwise marked, this work is licensed under a
Creative Commons Attribution 3.0 Unported License
. Ohloh
®
and the Ohloh logo are trademarks of
Black Duck Software, Inc.
in the United States and/or other jurisdictions. All other trademarks are the property of their respective holders.