Language Selection

English French German Italian Portuguese Spanish

Portage 2.1 Released!

Filed under
Gentoo

The Gentoo Portage Team is proud to announce that version 2.1 of portage has been completed and added to the tree for testing.

New features include:

* FEATURES="confcache" integration; global caching to speed up configure calls,
requires dev-util/confcache
* elog framework and accompanying modules for logging ebuild warnings, errors
and general notices. Collects eerror/ewarn/elog/einfo messages.
* New elog function (should replace einfo in many cases)
* version syntax enhancements allowing multiple suffixes and a new 'cvs'
prefix for denoting "live sources" ebuilds.
* config files as directories enabling more flexible settings management.
* Addition of an register_die_hook method that allows ebuild/eclasses to
register functions to be called for better debugging on errors.
* Addition of pre and post user hookable functions for each ebuild phase, accessible
via portage bashrc. Example would be pre_src_unpack .
* cache refactoring- runtime improvement from 35% -> 65%.
* Intelligently handle and display USE_EXPAND-based IUSE variables.
* FEATURES="parallel-fetch". Download in parallel to compilation.
* Include a "changed or new" USE flag output when --verbose isn't specified.
* Support for splitting out debug information into separate files in
/usr/lib{,32,64}.
* exec subsystem refactoring (now with less bugs!)
* Added sha256 and rmd160 hashes for digests/manifests
* Make --emptytree only apply to ${ROOT} rather than always including /.
* Allow packages to be upgraded that are only depended on via a
"|| ( =cat/pkg-1* =cat/pkg-2* )" construct.
* Ebuild output is no longer cut off early when using PORT_LOGDIR.
* Distfiles indirection- $DISTFILES access goes through a tmp dir to fail
access to files not listed in SRC_URI.
* Emerge now uses --resume to restart itself after portage upgrade.
* Atomic file updates via the new atomic_ofstream class.
* Global updates and fixpackages performance optimizations.
* Tests show that file locking is now more reliable.
* A bash call stack is printed when an ebuild dies in ebuild.sh.
* New rsync option handling by using a generic PORTAGE_RSYNC_EXTRA_OPTS variable
* Manifest2 support that will allow digest-* files to be eliminated from the tree.

Release Notes.

More in Tux Machines

Cloud Foundry (LF) News

‘No Company Is So Important Its Existence Justifies Setting Up a Police State’

You’re talking about very — about specific manifestations, and in some cases in ways that presuppose a weak solution. What is data privacy? The term implies that if a company collects data about you, it should somehow protect that data. But I don’t think that’s the issue. I think the problem is that it collects data about you period. We shouldn’t let them do that. I won’t let them collect data about me. I refuse to use the ones that would know who I am. There are unfortunately some areas where I can’t avoid that. I can’t avoid even for a domestic flight giving the information of who I am. That’s wrong. You shouldn’t have to identify yourself if you’re not crossing a border and having your passport checked. With prescriptions, pharmacies sell the information about who gets what sort of prescription. There are companies that find this out about people. But they don’t get much of a chance to show me ads because I don’t use any sites in a way that lets them know who I am and show ads accordingly. So I think the problem is fundamental. Companies are collecting data about people. We shouldn’t let them do that. The data that is collected will be abused. That’s not an absolute certainty, but it’s a practical, extreme likelihood, which is enough to make collection a problem. A database about people can be misused in four ways. First, the organization that collects the data can misuse the data. Second, rogue employees can misuse the data. Third, unrelated parties can steal the data and misuse it. That happens frequently, too. And fourth, the state can collect the data and do really horrible things with it, like put people in prison camps. Which is what happened famously in World War II in the United States. And the data can also enable, as it did in World War II, Nazis to find Jews to kill. In China, for example, any data can be misused horribly. But in the U.S. also, you’re looking at a CIA torturer being nominated to head the CIA, and we can’t assume that she will be rejected. So when you put this together with the state spying that Snowden told us about, and with the Patriot Act that allows the FBI to take almost any database of personal data without even talking to a court. And what you see is, for companies to have data about you is dangerous. And I’m not interested in discussing the privacy policies that these companies have. First of all, privacy policies are written so that they appear to promise you some sort of respect for privacy, while in fact having such loopholes that the company can do anything at all. But second, the privacy policy of the company doesn’t do anything to stop the FBI from taking all that data every week. Anytime anybody starts collecting some data, if the FBI thinks it’s interesting, it will grab that data. And we also know that the FBI and other such agencies are inclined to label protesters as terrorists. So that way they can use laws that were ostensibly adopted to protect us from terrorists to threaten a much larger number of us than any terrorist could. Read more Also: Numerical Analysis Software Global Market Analysis & Forecast: Analytica, Matlab, GNU Octave, Plotly, FlexPro

Today in Techrights

Android Leftovers