TMS470 ARM ABI Migration - Texas Instruments

[Table] IAmA: We are Gentoo Developers. AMA/AUA

Verified? (This bot cannot verify AMAs just yet)
Date: 2012-04-22
Link to submission (Has self-text)
Link to my post
Questions Answers
How do you feel about "newbies" adopting Gentoo? As long as you can read a bit and use the tools available to you like equery and genkernel so you don't ask stupid questions, go for it...Sadly, most newbies are (unfortunately) not very capable of that. Though TBH, even I wasn't when I did my first Gentoo install many years ago.
Why is BSD superior to Linux? Is Gentoo/FreeBSD your attempt to close this gap? For one, we have the ACCEPT_LICENSE option. If you put ACCEPT_LICENSE="@FSF_APPROVED" in your make.conf, portage will only let you install software that uses a FSF approved license, so you can be like Stallman if you want. However if you're like me and want to have a good computing experience, you may want adobe-flash and the binary nvidia driver and so on and IMO, you should be able to have that. I'd love the day when lightspark or swfdec or gnash can display all the flash elements i may use on the web and/or the web goes HTML5, but until that happens, i'll stick with adobe-flash, unfortunately.
I believe BSD is superior to linux because of the facts that they're trying to keep things as close to old-school UNIX/POSIX to a point. I don't need 15 options in my rm, I only really need 4 or 5, BSD rm (for example) only has said 4 or 5. Also, BSD is more lax on the licenses they allow in the core system, which is why ZFS is available natively in FreeBSD.
Finally, perhaps the biggest reason I think BSD might be superior to linux is that the kernel and the userland are developed in tandem by pretty much the same people. For example, if a new IO layer is added to the kernel, the userland tools can be switched over to that new IO layer the same day and have the old code removed since you basically can't run the FreeBSD 9 kernel on the FreeBSD 8 userland. With linux, it has to wait for the userland people to catch up, and keep the old code for a while for compatibility. With the BSD way, development can move much faster and get to users much faster.
As far as Gentoo/FreeBSD. IMO, the biggest problem with BSD is that ports and packages aren't quite as good as they may seem (neither is pkgsrc) and BSD init is not great. Gentoo/FreeBSD replaces both of these things with Portage and OpenRC (respectively) and that's a GREAT idea IMO. The only problem with it currently is that it doesn't have enough manpower. The official guides and stage3's are woefully out-of-date and Iit's only available as a unstable/testing mask which isn't horrible, plenty of people run Gentoo Linux ~arch/testing/unstable on their systems, I personally don't.
How often to you recompile your system? Almost never. All the stuff that used to require full system rebuilds on upgrade (GCC and Glibc mainly) has been modified that they don't require a rebuild anymore.
how do you amass the info necessary to keep all the stuff you need for gentoo in a single lifetime? You actually don't need that much. I have a generic make.conf for my desktops/laptops and another generic one for my servers and i just build slightly on top of that and other than that...I know the few commands needed to do stuff (repoman commands, eix-sync, eix, basic emerge usage, basic equery usage) and everything else is documented. I don't remember what easily half the supported FEATURES of portage are, but i can $ man make.conf to look them up when i need to.
Do you think FreeDesktop.org is failing to standardize the essential things we need to have a standard for? Well if they want to push their own packages, i'd say .deb because Ubuntu (and its variants) is and likely will be for a long time coming, the most popular linux distro. Also, rpm distros and Archlinux and Gentoo can extract debs to make them into .spec files and PKGBUILDs and ebuilds, it's just a bit more work for us. It would be lovely if they released tarballs that extracted binary files like the linux binary tarballs for mozilla stuff, since that would make things equal for everyone, but meh.
finally: RPM/DEB, why the hell don't they merge to one compatible format? Yes, and I think it is FAR too influenced by Red Hat. SystemD is already in the draft for the next LSB, RPM is the LSB package manager...There could be a lot more standardization. Granted it could be a lot worse too, but...
My guess is history and established trees. RPM as a tool SUCKS. Yum makes it a lot better. dpkg, however, is a pretty good tool (which apt makes a lot better). Also, whichever distro(s) decide to switch will have to do A LOT of work to replace tools with the alternative, port all the packages over to the new format and so on. A lot of work.
This is all just personal opinion and speculation though pretty much...
I'm of the opinion Red Hat is becoming poisonous to the rest of the Linux world. Do you agree, and do you think there's an escape route if they start pulling an Oracle? I agree that they're beginning to become poisonous and push their own agenda to the rest of the linux world. However, when the shit really starts to hit the fan, I think much like with MySQL and OpenOffice, people will fork heavily Red Hat controlled software (udev, glibc (not really RH controlled anymore), etc) and try to rein it back to sanity.
Part of the problem with that will be...It was clear to EVERYONE that Oracle was going to be evil the day the Sun acquisition was completed. Some people outside RH think SystemD is a good idea. Some people outside RH think moving everything to /us is a good idea. Some people outside RH think removing SunRPC from Glibc was a good idea.
so i am merging -e world for no reason right now? Probably, pretty much. As a archtester, I had to do it yesterday to test the new Glibc, but as a user, you almost never need to emerge -e world. i've not NEEDED to recently.
With a new version of gcc, i can inherit the small improvements and slowdowns that it provides by remerging world. Since glibc is a library, all you need to do is reboot to get the changes.
There is no such thing with glibc? And yes, you can get the improvements and slowdowns a new GCC provides by remerging world, but in Gentoo, there is no NEED to remerge world. Nothing will break or go wrong at all if you don't do it.
What are some things you like and dislike about Gentoo? Have you worked on other platforms besides Gentoo? If so, how do they compare in your opinion? Thanks. I'm going to do this as a list because i'm a guy and as we know, guys like to listify everything. :P.
The customizability. I can choose package features, I can have a ton of choice over package versions and I can choose CFLAGS.
The robustness of the tools. Portage (Gentoo's package manager) really feels like a good product, OpenRC (Gentoo's init system) really feels like a good product, eselect feels like a good product (or at least as good as it can be)
The fact that Gentoo is rolling-release.
The status of Gentoo "stable". I really like that Gentoo "stable" is stable enough to use on a production server, but not crazy out-of-date like Debian Stable and CentOS releases quickly become.
The simplicity and power of ebuilds.
Last (and least IMO): The performance improvements. Compiling your software to sane CFLAGS and EXACTLY your CPU microarchitecture will make it perform a bit better than software that's compiled for generic amd64 or i686/i386. IMO, if you're using Gentoo primarily for this reason, you're using it for the wrong reason.
The compile times. They're not nearly as bad as you may have heard in the past, but it would be nice IMO if PyPy (not going with something like firefox or libreoffice as there are firefox-bin and libreoffice-bin packages in the Gentoo tree) took 2 seconds to install as opposed to 15 minutes to install. However I do realize that giving this much customisability is not feasible in binary form and i'd rather "eat" the compile times and have the flexibility than not have it.
As far as other distros i've used: i've played around with most distros. The ones i've primarily got the most experience with are Debian, Ubuntu and Archlinux so those are the 3 i'll compare to Gentoo since if I went over all the distros i've ever used, this comment would be longer than Twilight and probably a better love story as well.
Ubuntu: Difficult comparison as they're both aiming at different markets. Ubuntu is aiming at the new Linux usenew computer user market and I think they hit that quite well. It's got a lot of helpers and if you're a power user, Ubuntu will get in your way more often that not, but for your dad or your grandma or someone similar, Ubuntu is pretty much the perfect choice IMO. (Ubuntu or Linux Mint)
Debian: IMO, Debian USED TO be great. Debian 6 feels less stable than any previous Debian release while still using older software. Debian Testing is becoming a bit too unstable lately and it doesn't get security updates quickly which, especially recently with big holes found in OpenSSL and Samba, is kind of not acceptable and Debian Unstable...Has a tendency of being a bit too unstable to really be a good choice.
Archlinux. Oh boy. I'm going to try to be as charitable and as nice as I can. Arch is bleeding edge and follows upstream which isn't a horrible policy. Except generally, upstreams suck. Also, I kind of like the automated tools in Gentoo for config file merging and such which you don't really get in Arch.
I used Gentoo when I first got into Linux. It took me 1-2 weeks to get everything set up, flags set and things to stop breaking. I would recommend Gentoo for someone experienced because it does not take 1-2 weeks to get everything set up. Even on a crappy 2.2GhZ single-core Celeron I have, it took less than 12 hours to get from a stage3 to a full GNOME system and I had everything setup the way I wanted.
I don't recommend Gentoo to people anymore unless they want to learn. Sabayon Linux (if they must have a gentoo system) or Debian/Ubuntu is what I normally recommend now. The trick to getting things setup the right way the first time is to do like: emerge -pv gnome, look at all the packages it's about to emerge, use equery u to look up any USE flags you're unsure about and put the ones you want to set in make.conf or /etc/portage/package.use, then emerge gnome. That way you get the USE flags pretty correct first-time. Also, many USE flags have been standardized for the past few years.
Why would you recommend someone using Gentoo atll? Basically, Gentoo is still the very best choice if you want full control over your system. You just don't get that with Debian or Ubuntu.
let me emphasize: there is no perfect distro. Gentoo is not better than Ubuntu, and Ubuntu is not better than Gentoo. Exactly. There also isn't even a perfect distro "for you" (or me or anyone). However there MAY be a "closest to perfect" distro for said person at this current moment.
I kind of like the automated tools in Gentoo for config file merging and such which you don't really get in Arch. Not really. Someone ported etc-update to Arch, check AUR. That kind of doesn't count IMO. It should be a seperate project or in pacman-contrib or pacman itself or something.
Speaking of Arch, what do you think of their kind of continuous upgrading versus well defined release of something like Ubuntu? Gentoo is rolling-release as well and I prefer a rolling-release distro personally. I hate that except for bugfixes and security updates, the "stable" repos of non-rolling-release distros are basically locked till the next release, 6 months later.
How is the userbase for Gentoo. Is it growing or shrinking? It feels like it's staying about the same. The stereotypical ricers are leaving Gentoo for other distros and new people (who are generally a bit saner) are coming in.
The stereotypical ricers are leaving Gentoo. You mean the kids who would tell everyone that Gentoo is the greatest distro ever and all others suck? And use CFLAGS="-funroll-all-loops -ffast-matf -fomgz-sofast"
Link to funroll-loops.info Just found this. Is this your website? :P. Nah, I could only get the .biz (not true, that's probably a porn site, don't go there...). There were some stupid users, but every OS/distro has it's share of stupid users.
What type of packages get accepted into the tree? Pretty much anything that at least one developer is willing to maintain and that Gentoo can legally distribute an ebuild for.
What are your thoughts on systemd? The basic idea is good in theory, but the implementation has some problems. Maybe one day it will work well enough on my systems to e.g. not freeze at shutdown; until then, I will stick with openrc. I should also say that to their credit, systemd upstream developers have written good specifications for some of systemd's interfaces, making it quite possible to implement them for other rc systems.
How healthy is Gentoo's community? Will the distribution still be around 10 years in the future? Gentoo's community works well, but it's not without defects. Although epic bikeshedding is probably the most visible issue, I would say that the biggest real problem is ambitions overreaching available resources in some important areas. That said, I am quite certain that the community is strong enough for the distro to survive for as long as desktop computers are relevant.
If you could change one thing about Linux in general, what would it be? A security architecture that isolates a non-root user's applications from each other except for well-defined, well-audited interfaces. Desktop Linux needs to learn from Android's example (in Android, every application has its own uid that it runs as).
How do you feel about your unending 'support' from /g/? Honestly don't bother me.
Would you agree that Gentoo is easier to use than Ubuntu? If you know what you're doing with linux, I would say Gentoo probably is easier than Ubuntu. All the automagic stuff Ubuntu does will just get in your way if you're an advanced user. Also, debian's/ubuntu's packaging format is fairly arcane and not that easy to work with. The tools (dpkg, apt-get, synaptic, etc) are all fairly good, but the format is kind of a pain to work with.
If you're new to linux, Gentoo will be VERY difficult to use. You need to have a decent knowledge of how things work on linux and your hardware and such to use Gentoo well. Granted, the Gentoo Handbook ( Link to www.gentoo.org ) is pretty easy to follow even if you don't know what you're doing as long as you have good reading comprehension skills, but for a novice to linux, I still wouldn't recommend Gentoo over Ubuntu.
For a novice, probably not. The learning curve for portage is pretty steep.
For a developer, or an experiences Linux user, it is much easier to customize at a low level, and therefore easier to use.
It all depends on what you want to do. Roughly speaking, in Ubuntu, easy things are easy, but the hard things are sometimes nearly impossible. In Gentoo, easy things are usually not easy at all, but hard things tend to be pretty doable due to the distro's amazing customizability. If you are an experienced, full-time developer or server administrator, Gentoo can definitely save you some time and sanity. If you just need a simple Linux desktop to play around with or to do your computer science homework, installing and maintaining Ubuntu will take much, much less work.
How did you became a Gentoo dev? I started helping out and proxy maintaining packages and archtesting for x86, I was then asked to become a dev so I could actually maintain said packages myself. I got a mentor, took the 2 necessary quizzes and submitted more ebuilds to show that I could write ebuilds. I then met with my recruiter a few times and then I became a dev. :)
I have been a user of Gentoo since about 2004.
In the last few years, I started reading through ebuilds and eclasses, reading mailing lists, hanging out in IRC, and basically turned into a knowledge sink for all things Gentoo.
I started to work on bugs related to the www-client/chromium web browser. I took the "ebuild quiz" and became an official tester for the chromium team.
The chromium team leader suggested I become an official dev; that process took several months, concluding last August.
I've been using Gentoo probably since 2002 or 2003. I started contributing to Sunrise (the "official" user-maintained overlay) around 2006. I had been using Gnome for a long time; Gnome 3.0 was going to come out in the spring of 2011, and of course I wanted to take a look at the beta versions. So I enabled the Gnome overlay on my Gentoo desktop machine to install a Gnome 3 pre-release, started filing bug reports about it (and there were a lot of bugs!), and after hearing that Gentoo's Gnome team was understaffed, I volunteered to help out with Gnome 3 support. For some time I simply emailed patches to the Gnome team members, and having proved myself, was given commit access to the overlay maybe a month or so later. I did a lot of work on Gnome 3.0 and 3.2 in the overlay, hung out in the IRC channels, took the quizzes, and was granted full developer status in October 2011.
How do you get a mentor? Ass kissing on IRC or some formal channel where you can make requests? FORMAL ass-kissing on IRC.
Really though, my mentor was one of the people who asked me to become a dev so there was already a relationship there. :)
In which country do you reside, and what is your educational background? I reside in the US and I went to university for 2 weeks and then dropped out. Been working in pretty good jobs ever since.
Born and raised in Detroit, MI, USA. I have a Bachelor of Science from the University of Michigan-Dearborn, majoring in Computer and Information science.
Born in Russia, reside in the US. Have undergraduate degrees in mathematics and computer science; dropped out of a mathematics PhD program.
I reside in the US and I went to university for 2 weeks and then dropped out. Been working in pretty good jobs ever since. How did you get your job, though? It's one thing to know the right programming languages, but did you have to demonstrate projects that you had worked on, prior to getting the job? What did you do? Personal app development as a hobbyist? Working on collaborations? Like... what did you give as "work experience", in order to get "good jobs"? I did have to demonstrate projects that i'd worked on, but at that point, i'd been coding for easily ~6 years so I had plenty of stuff I could show.
Why did you drop out and did you just learn how to code by yourself? Well I was coding when I was a kid (like 12). And I dropped out because in 2 weeks i'd learned nothing and realized that it was basically a waste of my time.
I really want to learn how to code, where would you suggest me start? Honestly, i'd suggest starting with C, a lot of people will disagree with me, but I feel that learning C somehow unlocks understanding of coding syntax really helps you understand most languages used, like Python, Perl, Ruby, PHP, C++, Java, etc.
To learn, i'd recommend starting with K&R (The C Programming Language by Kernighan and Ritchie) which I recently re-read. It's REALLY a great book whether you're a beginner at C/programming, intermediate at it or advanced.
What do you do in an average week or so? Read and attempt to solve bug reports, test and add new versions of packages (or entirely new packages), debug encountered problems and write patches to fix them, answer user questions, talk to my fellow developers, move packags between visibility levels (or ask others to do so), and in the time that's left, write code for a few long-term projects that I am interested in.
Is developing volunteer, or like a paid job? Gentoo development is 100% volunteer work that we do in our spare time. However, quite a few other Linux distros do make money (generally from commercial support contracts) and are therefore able to pay their developers.
Do you work on concepts or writing it all, or both? A concept that's not backed by a working implementation isn't worth much. In a volunteer development community, if you have a bright idea, it's your responsibility to ensure that idea turns into code that someone can use.
What got you into Linux/Gentoo in the first place? Middle-click paste. Blew my mind when I first learned about it. That middle mouse button got me into Unix, and then Linux. I started out by running Red Hat, switched to Slackware, and at some point realized that I was compiling so many packages by hand, that I might as well install a proper source-based distro.
In your opinion what makes emerge a better packaging/distribution system over others like dpkg and rpm? From the developer perspective, writing and testing a new ebuild is much easier than creating a new rpm/deb package. First, Gentoo's eclasses are a big help with avoiding boilerplate. Second, you can just drop your ebuild into a portage tree, there are none of the intermediate steps that are needed with debs and rpms; it's the same sort of developer-time advantage that writing in a scripting language (edit-test-repeat) has over using a compiled language (edit-compile-test-repeat). From the user perspective, USE flags are certainly a big advantage.
What is your favorite ice cream? The plain (and I mean really plain, not vanilla-flavored) Soviet-style ice cream in little paper cups that I remember from my childhood.
What desktop environment do you think is the best? And what did you think of the Gnome 3 redesign? Gnome, of course. I honestly love almost all parts of Gnome 3, modulo a few warts such as the power off menu item that's invisible by default. (And most of those warts can be fixed with extensions.) These days, Gnome 2 to me feels quaint and archaic. That said, Gnome 3's interface is certainly a big change, and if you already have Gnome 2 in your muscle memory, it will probably take you 1-2 weeks to figure out how to be equally productive in Gnome 3.
If you had to name one application that you live by (bash, chrome, etc.), what would it be? Git. Of Linus Torvalds's two creations, the Linux kernel and git, git may well be the greater.
I have worked as a developer on proprietary OSes, and I'm getting my feet wet developing for FreeBSD these days. For keywording, if a developer doesn't have access to a x86 box (say all his boxes are amd64), he can add a new package to the tree just keyworded for amd64 and request that the x86 team test it on x86 and add the x86 keyword if everything's OK.
Does Gentoo's x86 team do work that remains specific to the Gentoo distro, or does their work mainly go into the linux kernel itself, so as to be available to all distros on the next kernel upgrade? If there is Gentoo specific stuff, can you give an example? When developing for x86, do you guys do things differently when developing for virtualized x86 than for bare metal x86? Is there a difference, basically? Perhaps I should clarify what i mean by the fact that i'm an x86 arch team member. Gentoo supports arm, amd64, alpha, hppa, ia64, m68k, mips, ppc, ppc64, s390, sh, sparc and x86 architectures and a few prefix arches and alt project arches, though prefix and alt aren't relevant right now) The primary work of the x86 team is stabilising packages and keywording packages for the x86 architecture. Basically, the arch teams all do mainly Gentoo-specific stuff. Some Gentoo devs also contribute to other projects, but some don't. Generally though, when we patch around a bug or something, we try to get the patch accepted upstream for the next release (or we try to get it accepted before we have the ebuild apply the patch).
Thanks for the reply. Good to know that arch-specific bugfixes are patched upstream as well. Well not even arch-specific bugfixes, any patches we apply in ebuilds for any reason we try to put upstream or use a solution suggested by upstream if it works.
General opinion on Sabayon? I see nothing particularly wrong with it. The main developer is a Gentoo developer and contributes back to Gentoo and other than that...It's really a decision you have to make for yourself.
Do you want a distro that's 20 minute install like Ubuntu and you're at a desktop, but is binary first with the powecustomisability as a secondary thing or do you want a distro that will take longer to get setup, but it'll be set up exactly how you want it after install and it'll always be exactly how/what you want?
I've never used Sabayon, but I can say that its developer community includes some awesome people who have helped to identify and fix bugs in the packages that I care about. I wish them the best of luck.
In my mind that defeats the point of 'stable'. Devs should just use the unstable packages that they wanted. Most distros, I think, allow you to specify stable and unstable packages. I'd be surprised if arch didn't. Arch doesn't. Well...You can fall back to a older package by installing that .pkg.tar.gz manually or editing the PKGBUILD back to the older version and building it, but Arch effectively has ONE set of repos, not necessarily a stable/unstable branch.
There are in fact testing repos in Arch. Well...Yes there are, but they're quite small and VERY few packages go in them. Generally only new big suites (new major KDE version, new major GNOME version) or packages in core (kernel, glibc, zlib, etc) go in testing at all or sometimes the results of soname rebuilds.
How do you feel about Arch? I answered this question elsewhere in these comments. I'm not a fan. The biggest problem IMO is that they follow upstream very closely and upstreams are generally wrong.
Can you elaborate with examples where upstream has been wrong? GNOME 3.0 being considered "stable"
Glibc dropping SunRPC.
Glibc moving ia64 support into ports.
Python 3 being considered the primary python implementation, even though VERY FEW 3rd-party modules(?) support it.
Udev forcing /usr to exist before it starts itself.
(proposed, not done...Yet) GNOME considering SystemD as a dependency.
Now that udev and systemd have been merged, are there any people working on an upgrade path from openrc to systemd? There are many people work on that actually, but OpenRC will still be an option with udev for quite a while and i'm fairly sure (but don't quote me) that after udev requires systemD, Gentoo will maintain a patch that prevents udev from forcing systemD or someone will fork the version previous or something.
The merge of udev sources into the systemd source tree changes nothing in the short- and mid-term. Please read the official announcement for the merge. You (and we in Gentoo) will still be able build udev independently of systemd just as easily as before; the only difference is that to get udev sources, you will need to download systemd-$VERSION.tar.xz instead of udev-$VERSION.tar.bz2.
What do you personally consider to be the 3 most important pros for using Gentoo? Technically, not philosophically. :) Conversely, what do you consider to be the 3 greatest weaknesses(cons) of Gentoo? Bonus points for your ideas on how to fix/address them! The compilation. This is probably the biggest killer of not being able to put Gentoo on low-end machines (without a binhost) and even on nice machines, sure, i'd rather pull Firefox in 10 seconds as opposed to 10 minutes. However, due to the fact that pretty much EVERY Gentoo system is different from every other Gentoo system, there's no great way to fix this that's financially viable. The fact that portage supports BINHOSTs, but there are only a few public BINHOSTs still out there and (AFAIK), they're not terribly up-to-date and don't contain all packages. Public BINHOSTs would help for people putting Gentoo on low-end boxes like I said...
How many hours per week do you dedicate to Gentoo? Probably 10 to 20. That would include helping people in IRC.
Probably about 10-25 (not to upstage floppym)...Archtesting can be time consuming.
On a side note, what do you think about the whole Funtoo thing? I started using Gentoo probably 5 or 6 years ago, took a break about 2 years ago, and just started back up with Funtoo. I absolutely love the little tweaks (like the build USE flag and boot.conf), and I'll probably be sticking with it (benefiting still, of course, from awesome Gentoo devs like yourself). But yeah, what's your general take on Funtoo? Have you used it? I've answered this question in other places in this thread...Too lazy to link though. Ctrl+F is your friend. :)
Why do you suppose gentoo isn't more popular? what hurdles does gentoo have to overcome for world dominance? The setup can be a bit overwhelming. It's not just {Enter}, {Enter}, {Enter}, reboot, done. I kind of like this because it's a barrier to entry for people with no reading comprehension skills and it lets you customize before you're in the installed system.
The compile times. Mainly because people's perception is still stuck in 2004 when it took 18 hours to get from a stage 3 to a desktop on a 2GhZ Pentium 4 (and like a 3 hour compile for Firefox). It's 2012. We have multi-core machines. Almost no single thing takes an hour to build even on a Core 2 Duo and with a quad-core or core i7 or whatnot...
Mostly related to #2. It's people's misconception that Gentoo is really hard to install or that it takes forever to compile stuff...It really doesn't.
I think most people can do the install now that genkernel is the recommended method to build a kernel, so that's not so hard anymore. As for the misconceptions...It's really hard to bust misconceptions, but I hope this AMA is helping/will help and i'm going to be going on FLOSS Weekly in about a month to talk about Gentoo and i'll hopefully bust some misconceptions for their audience as well.
One of the reasons I love Linux is that the desktop I built five years ago performs just as well as it did back then; I'm not going to upgrade to a Core i7 merely for Gentoo. The thing is though, that's not strictly true. Firefox has gotten more bloated, Chrome has gotten more bloated, GNOME has become a lot slower (on old hardware at least) with the advent of GNOME3. KDE4 is (IIRC) a fair bit slower than KDE3.
I mean if all you use is Fluxbox/Awesome and Midori as your browser and such, it won't be much worse than it was 5 years ago, but not many people do that.
Also, if all you use is Fluxbox/Awesome and Midori, the compile times won't really be bad at all even on a old Pentium 4 so...
When is Linux going to implement Threefish? When you write the patch.
As someone who'd one day like to use Gentoo, what is, in your opinion, the best way to go about learning Linux and actually becoming fluent with it? I would honestly say learn slowly. If you have a system that can handle VM's well, maybe try Gentoo in a VM where you can't bork anything, but if not or if you don't want to just jump into the deep end...
I'd recommend using Debian, spending as much time tinkering on the command-line as you can, have fun playing around...Maybe try building your own kernel on Debian which is one of the hardest parts of the Gentoo install.
Also, read through the Gentoo Handbook ( Link to www.gentoo.org ) and see if you feel overwhelmed by anything in it...If so, google and learn more about the command or the section (like "partitioning with fdisk" for example would be a good Google search), if not or when you don't anymore, give installing it a try. It's actually quite fun...Unless stuff breaks. :P.
If stuff does break though, try not to get frustrated. You'll have it fixed and almost anything you might screw up in Gentoo can be fixed without needing a reinstall.
I think you need to immerse yourself to become "fluent". Get a working Linux system, and do not use anything else. If you break it, work on it until you fix it.
I recommend keeping a spare Windows machine around for when you really hose your Linux system. As you learn, that will happen less frequently.
Last updated: 2012-04-26 09:06 UTC
This post was generated by a robot! Send all complaints to epsy.
submitted by tabledresser to tabled [link] [comments]

Q&A #1  BINARY OPTIONS Binary Options 100% ITM strategy from 1000 to 20000 in 6 min live trading How to trade Binary Options for beginners - Binary Options 101 BINARY OPTIONS TRADING TUTORIAL #2 BINARY OPTIONS TUTORIAL - YouTube

We have implemented ARM V5T support in our DBT and demonstrate execution rates of up to 1148 MIPS for the SPEC CPU 2006 benchmarks compiled for ARM/THUMB, achieving on average 192%, and up to 323%, of the speed of QEMU, which has been subject to intensive manual performance tuning and requires significant low-level effort for retargeting. SPEC has chosen not to allow source code changes for the CPU2006 suite, except under very limited circumstances. By restricting source code changes, SPEC separates the activity of porting benchmarks, which has a goal of being performance neutral, from the activity of using the benchmarks, where the goal is getting the best score possible. An LLVM-based hybrid binary translation system. SIES 12, Karlsruhe, Germany, June 20-22, 2012. Google Scholar; Hazelwood K, Klauser A. A dynamic binary instrumentation engine for the ARM architecture. In Proceedings 2006 International Conference on Compilers, Architecture and Synthesis for Embedded Systems. CASES'06, Atlanta, Georgia, 2006; pp Pin [Hazelwood and Klauser 2006] and Valgrind provide dynamic binary instrumenters for ARM. DELI [Desoli et al. 2002] exposes client API for fine-grained control over the guest application SPEC2006 Benchmarks Build gem5 with a cache protocol cd gem5/ scons build/ALPHA_MOESI_CMP_directory/gem5.fast . Note that you can build with other protocols in gem5/build_opts but only under ALPHA ISA because the released benchmarks were compiled for ALPHA ISA

[index] [11115] [11867] [17733] [18813] [20660] [29558] [5958] [28601] [15787] [17337]

Q&A #1 BINARY OPTIONS

Chapter 1 - Introduction to binary options trading: brokers, how it works, example of trade Chapter 2 - Bid/offer levels from the brokers: what it means in terms of probabilities to end up in the ... Hi! This is my binary options video blog. I will show you how I earn money on binary options with simple binary options strategy BINARY OPTIONS TRADING Easiest 400$ for half an hour - Duration: 5:07. Ana Trader 602 views. 5:07. You NEED to try THIS STRATEGY 2020 +1220$ - Duration: 6:02. Ana Trader 1,103 views. BINARY OPTIONS TRADING TUTORIAL Hello everyone!:) My name is Anastasia, but it's too hard to pronounce, that's why you may call me just ANA. I'm a pro trader for more than 2 years already and this ... 04:28 USDJPY binary options trading signals review 06:30 How to get binary options pro signals It's one of the most accurate binary options strategies. The best aspect of this strategy is signal ...

Flag Counter