The behavior of gcc has changed significantly since version 3.x. In 3.x, -O3 has been shown to lead to marginally faster execution times over -O2, but this is no longer the case with gcc 4.x. Compiling all your packages with -O3 will result in larger binaries that require more memory, and will significantly increase the odds of compilation failure or unexpected program behavior (including errors). The downsides outweigh the benefits; remember the principle of diminishing returns. Using -O3 is not recommended for gcc 4.x.
Some GNOME hackers have discussed what form GNOME '3.0' would take, such as radically changing its user model or taking advantage of new technologies. However, the changes in this roadmap are more incremental, designed to fit within the basically stable UI and APIs we guarantee within the 2.x series. For more on the radical changes that could be in a GNOME 3.0, see the long-term ideas at ThreePointZero. And remember, even then, the GNOME 3 APIs would be available in addition to the existing GNOME 2 APIs, so there is no risk that today's applications would break in the future.
Thanks for the links. They explain debugging, but they don't answer what I was getting at: If you know there exists a certain function in a library, how do you find out what it does (and how it does it), and how do you assert that your new implementation does the exact same thing.
Milan, Italy Norilsk, Russia Pittsburgh, Pennsylvania Mexico City, Mexico Dakar, Senegal Sumgayit, Azerbaijan Linfen, China La Oroya, Peru Cubatao Valley, Brazil Kabwe, Zambia
Exactly. You don't have to ship it with your product, you don't have to offer a download, but, if customers ask for the source, you have to give it to them (or point to the original source, if you didn't change anything). That and clarifying on distributing, under which license you ship, is the minimum for GPL (as i understood it, IANAL).
Someone should write a alsadump program that saves everything that should go through the sound card to a.wav file. And the same for video. That should bring discussions like this to an earlier end. (and maybe lead them directly to "trusted computing"/DRMed hardware)
Besides the obvious question "who watches the watchers", I always ask myself on these privacy topics:
Where does the money come from to pay people to sit around the whole day and watch other people? There must be an enormous amount of data accumulating...
Someone should write one or two conformity tests, so every time one says "OOXML is nice", one can point to that (failing) test until Microsoft fixes the bugs.
For dark web browsing, install the Web developer toolbar. Select Disable->Page colors. In the Firefox preferences: Content->Colors, you can set the default background and text colors.
Slashdot Mirror
-O3 is slower than -O2 or -Os in gcc-4
The behavior of gcc has changed significantly since version 3.x. In 3.x, -O3 has been shown to lead to marginally faster execution times over -O2, but this is no longer the case with gcc 4.x. Compiling all your packages with -O3 will result in larger binaries that require more memory, and will significantly increase the odds of compilation failure or unexpected program behavior (including errors). The downsides outweigh the benefits; remember the principle of diminishing returns. Using -O3 is not recommended for gcc 4.x.
http://www.gentoo.org/doc/en/gcc-optimization.xml
From the GNOME website:
Some GNOME hackers have discussed what form GNOME '3.0' would take, such as radically changing its user model or taking advantage of new technologies. However, the changes in this roadmap are more incremental, designed to fit within the basically stable UI and APIs we guarantee within the 2.x series. For more on the radical changes that could be in a GNOME 3.0, see the long-term ideas at ThreePointZero. And remember, even then, the GNOME 3 APIs would be available in addition to the existing GNOME 2 APIs, so there is no risk that today's applications would break in the future.
=> Further see http://live.gnome.org/ThreePointZero
I liked that idea. Maybe it's just a version bump to reflect the progress they're making.
First dogs and apes, now elephants? They are building a zoo up there!!
Thanks for the links. They explain debugging, but they don't answer what I was getting at: If you know there exists a certain function in a library, how do you find out what it does (and how it does it), and how do you assert that your new implementation does the exact same thing.
I didn't get this at first, so here. Its from space balls
How does your usual reverse engineering work flow look like? (How do you start, short note on tools, do you use (unit) tests)
Milan, Italy
Norilsk, Russia
Pittsburgh, Pennsylvania
Mexico City, Mexico
Dakar, Senegal
Sumgayit, Azerbaijan
Linfen, China
La Oroya, Peru
Cubatao Valley, Brazil
Kabwe, Zambia
Very wise, moderate me "Overrated" on a post that hasn't been rated! Thanks, I didn't like my Karma anyway. Next time adjust your preferences...
I like it when the model after real desktop work flows.
Just like this Proof-of-concept desktop environment (ok, might be a little offtopic)
Imagine a (beowulf) cluster of those
It was a joke.
ideally on your website
Exactly! They should just create a data structure and search algorithm with O(1) in all use cases.
Linux has that, it's calledExactly. You don't have to ship it with your product, you don't have to offer a download, but, if customers ask for the source, you have to give it to them (or point to the original source, if you didn't change anything). That and clarifying on distributing, under which license you ship, is the minimum for GPL (as i understood it, IANAL).
Someone should write a alsadump program that saves everything that should go through the sound card to a .wav file. And the same for video.
That should bring discussions like this to an earlier end. (and maybe lead them directly to "trusted computing"/DRMed hardware)
Besides the obvious question "who watches the watchers", I always ask myself on these privacy topics:
Where does the money come from to pay people to sit around the whole day and watch other people? There must be an enormous amount of data accumulating...
It has graphs and screenshots! It *is* a study!
Finally the name JavaScript makes sense
How much is that in real money? :-P
The subtitle will be [citation needed] ;-)
Someone should write one or two conformity tests, so every time one says "OOXML is nice", one can point to that (failing) test until Microsoft fixes the bugs.
It worked for web standards, why not here too?
Perfect example for this is the NetworkManager ... seen dozens of times as cool feature in Ubuntu, actually a project of RedHat.
Epiphany has it:
"'hostname' could not be found.
Check that you are connected to the internet, and that the address is correct.
If this page used to exist, you may find an archived version:
* in the Google Cache
* in the Internet Archive"
For dark web browsing, install the Web developer toolbar. Select Disable->Page colors.
In the Firefox preferences: Content->Colors, you can set the default background and text colors.
https://addons.mozilla.org/en-US/firefox/addon/60
PS: If you use GNOME, Firefox will reuse the GNOME color scheme, so set it to e.g. "High contrast inverse".