Legacy on ¬ just serendipity 🍀

Arch Linux on macbook installation notes

serendipity@perrotta.dev (Thiago Perrotta) — Sat, 11 Feb 2017 18:36:02 -0300

♠ Imported from thiagowfx@’s gist.

Arch Linux Installation Notes (Macbook, UEFI, LVM ON LUKS, swap, btrfs root and home):

References #

Create an installation disk (USB Flash Drive): http://zanshin.net/2015/02/05/arch-linux-on-a-macbook-pro-part-1-creating-a-usb-installer/
Partitioning (EFI boot partition + LVM on LUKS partition): https://gist.github.com/jasonwryan/4618490
- Naming conventions: http://www.pavelkogan.com/2014/05/23/luks-full-disk-encryption/
- Support docs: https://wiki.archlinux.org/index.php/Dm-crypt/Encrypting_an_entire_system#LVM_on_LUKS
Bootloader: https://wiki.archlinux.org/index.php/systemd-boot
Official Installation Guide: https://wiki.archlinux.org/index.php/Installation_guide

Post-installation #

— § —

Reply via email

#dev #legacy

Linux goodness in your Chromebook

serendipity@perrotta.dev (Thiago Perrotta) — Thu, 01 Sep 2016 16:29:07 -0200

♠ Chromebooks are excellent for testing and playing with Linux userland[1] stuff. And, even better, (almost) every change you make to it can be reset back to its factory state — even in the worst case of completely wiping ChromeOS from your computer. As long as you don’t mess up with its firmware, you can do whatever you want and still be safe. Most models are cheap, battery life is amazing for a cheap Linux laptop (8+ hours, depending on the device and on usage)[2], Linux support is great (well, it ships with a Linux OS, right?) — even for most touchscreen models [citation needed] — and it is simple (as in KISS).

Now, the previous paragraph sounded a little repetitive, but I’d like to focus on two of my previous points here:

You can do whatever you want and still remain safe. Yes. Indeed, you can reset your chromebook to its factory state with just a few clicks — this is still hard to do with Windows and OS X, as far as I know. I think Windows 10 is introducing a feature to ease this process, however I am not sure about it yet. This is super important (and convenient!) for people playing with a ChromeOS device — like me. My usual way to reset the system back to “factory state” was, usually, to install a Linux distro with the btrfs filesystem so I could create a snapshot upon the installation was finished, and then revert back to it at anytime I wanted. However, this solution is a little cumbersome. The second usual solution would be to simply use a virtual machine, but I don’t need to reiterate that this is slow and limited in many ways.
Simple. This is really important if you don’t want to have headaches in the long run. For starters, simplicity is always a trade-off that involves performance: it is hard to be simple with a core i7 CPU, so if you choose the KISS path, you are deliberately sacrificing some of your performance. Once you accept that, we can move forward: Chromebooks are very simple devices. They’re almost directly comparable to a tablet with a keyboard. Sometimes they feel like a full-featured arduino.

These were some small highlights about my decision to get one. Now here’s the good stuff:

There be dragons #

Disclaimer/Warning: these will potentially void the warranty of your device, and they are only recommended if you know what you’re doing. Be a good user and research before messing up with your system. I will provide upstream links with documentation as much as possible.

Enable developer mode — it’s a must; however, it weakens the security of your device. Beware and research! This can be totally reverted with a single keystroke (SPACE) at the login screen that will appear in every subsequent boot of your system.
dev_install — to get fdisk, python, tmux, nano, emerge (for binaries), and so on. These are just a few developer tools, they won’t scale and you won’t get whatever package you want in your vanilla ChromeOS, however these might help a lot with simple tasks. For example, you can format a SD card or an USB Flash Drive directly from ChromeOS upon running this tool — otherwise, you’d usually have to use another computer, with another OS to do it. It uses just a little space (~100M+ or so), and it can be totally reverted with devinstall --uninstall
Switch to the beta channel if you want to be slightly more into the bleeding edge — but not too much as in the dev channel. In fact, at the time of this post, select devices in beta channel have support for Android apps in ChromeOS.
Open a new shell (crosh) with Ctrl + Alt + T. The default username is ‘chronos’, without any password.
User-writable locations are: /usr/local (in particular, /usr/local/bin is in your PATH env variable out-of-the-box (echo $PATH)), /tmp and /home/chronos/user/Downloads. There might be more, but these are enough.
You can also insert removable media (USB flash drives, SD cards), which would show up on /media. ChromeOS can read/write filesystems formatted as ext{2,3,4} and FAT{,32} at least.

These are the basics. However, you will soon find out that you cannot do too much with ChromeOS, even with these modifications. In this case, there are three options, and all of them involve getting a Linux distribution in some way:

(RECOMMENDED) Install a Linux distro on either a SD card or an USB Flash Drive and dual-boot from it. It is really easy to dual-boot, you just press C-d to boot from the internal eMMC / SSD, or C-u to boot from external storage. There’s no need to play with grub, EFI, or any (other) fancy bootloader. If you change your mind later on, just format your SD card: there is no need to change anything on ChromeOS.
(NOT RECOMMENDED BY ME) Completely wipe ChromeOS by installing another Linux distribution on its place. Although it’s easy to revert this process through recovery media, it completely defeats the purpose of getting a Chromebook in the first place, especially now that Android apps are starting to become available on them (for touchscreen devices, of course).
(PROBABLY WOULDN’T RECOMMEND, BUT YOU MAY LIKE IT) Install a chroot alongside your ChromeOS, such as crouton. I tried it however I hated it. It is very buggy and opinionated, somewhat similar to oh-my-zsh: both of them promise to deliver a good framework and they’re kinda large open source projects, however they do lots of automatic things for you, and you end up not knowing what is happening to your system. It opposes the Arch Linux philosophy, being user-friendly rather than user-centered. Some people like it, I won’t argue this, it can work for them; but not for me. An alternative would be to install a chroot manually, without the help of crouton.

I’ve chosen the first option, installing Arch Linux ARM on my Chromebook. So far, so good: it’s working very well. Storage is easily shared between the two OSes (by using the SD card). There’s more to get out of my system, and I intend to share new findings in this blog. In particular, the most important one: to find an easy way (not crouton!) of running whatever Linux-compatible application I want from within Chrome OS, without dual-booting. I have a few ideas (well, it’s just a matter of creating a KISS chroot), I just want to polish them a little more before trying them out.

Footnotes #

I wouldn’t say the same for kernel stuff though. It is not trivial to install a bootloader (such as grub) or to boot a custom kernel in a Chromebook. It’s possible indeed, it’s just not as nice or as easy as an average PC or Mac laptop from today, especially if you have the intention of keeping the upstream OS (Chrome OS) in the device.
People probably would cite a Macbook, a Surface Pro, or something else similar here. For starters, did you see how much they cost, in comparison to an average Chromebook? Such a comparison wouldn’t be fair.

— § —

Reply via email

#dev #legacy

Installing Haiku from an existing linux

serendipity@perrotta.dev (Thiago Perrotta) — Fri, 03 Apr 2015 17:20:19 -0300

What’s Haiku in the first place? #

It is an operating system, period. What might cause a small surprise is that it is not based either on Linux or on BSD — and yet it is (probably) runnable in your modern computer/laptop!

What I most like on it is that the system is integrated with the GUI, so the end user gets a nice experience. This is not so common as it sounds: most Linux distributions are simply a collection of programs and utilities put together in one place, but they are not necessarily integrated — however, you can integrate them. This makes all the difference between user-friendly and user-centered paradigms¹.

Anyway, Haiku is simple, so for me this post is more a hobby than something useful that I will use in the future; however, I find that knowing about more and more about different operating systems has its own advantages.

For more, see https://www.haiku-os.org/about.

Installing Haiku #

Note: most of those instructions come from the Haiku installation guide. Also, thanks David Couzelis for kindly giving me some advice and pointing me out to them.

Get Haiku. You can do it either from the official download page, or get nightly releases from the nightly builds page. I personally recommend the nightly releases; I first installed the latest non-nightly one, but later on I discovered that is was very old (circa 2012): it didn’t even have a package manager.
Which image should you download? In this post, I am assuming we’ll install Haiku directly to a (real) disk, so I’m downloading the raw image. Actually, the anyboot image can also be downloaded — and this is the one you will get if you opted for a non-nightly version; however, the anyboot image will be converted to a raw one in step 3 before we proceed. So, skip the next step if you downloaded the raw image.
Got your anyboot image? Now, convert it to raw (run this from a bash compatible shell):

$ dd if=haiku-anyboot.image of=haiku.raw bs=1M skip=$(expr $(od -j 454 -N4 -i -A n haiku-anyboot.image) / 2048)
$ dd if=/dev/zero of=haiku.raw bs=1 seek=506 count=4 conv=notrunc

Update (2022): These days dd supports status=progress to display the image writing progress.

Now that we got a raw image, we are writing it directly to a disk partition. First things first: find 3GB or more of free space in your disk. Then create room for a partition in there (for example, with fdisk or gparted). Got it? Now create a partition in there. I’ll assume the partition is /dev/sda42. Please change 42 for the appropriate number in your case.
Copy the raw image to the partition (this should be done as ROOT); WARNING: double check the partition and the disk number, otherwise you might lose data.

$ dd if=haiku.raw of=/dev/sda42 bs=1M conv=notrunc

Make the installation bootable: you’ll need to compile and run the makebootabletiny program, which can be downloaded from the Haiku installation guide. It is a simple C program, so:

$ gcc makebootabletiny.c -o makebootabletiny
% ./makebootabletiny /dev/sda42  # this one: as ROOT

Make your bootloader know about Haiku. If you’re using grub2z, you can add something such as the following lines to /etc/grub.d/40_custom:

menuentry "Haiku OS" {
    set root(hd0,42)
    chainloader +1
}

Then run as root: grub-mkconfig -o /boot/grub/grub.cfg

Done! Now you should be able to boot into Haiku.

Now what? #

This post is not a review of Haiku, so I’m stopping here. However, if I write a review about Haiku, I’ll do that from Haiku 🙂.

I’m just leaving this here: https://www.haiku-os.org/slideshows/haiku-1

By the way, Arch Linux is user-centered, which means that you’re supposed to integrate the system as you wish. If you don’t wish that then you’re screwed anyway, so go away 🙂 ↩︎

— § —

Reply via email

#dev #legacy

[FAQFlow #3]: Hello, SaltStack (for Arch?)

serendipity@perrotta.dev (Thiago Perrotta) — Thu, 26 Mar 2015 19:48:15 -0300

♠ Salt is what you want when you have to manage several servers. It looks like a good option for open-source servers (Unix systems), so I’m giving it a try. A common alternative is the community version of Puppet.

Motivation #

Historically, I’ve always used Arch on my notebook because it’s the machine I use, customize, and play with the most. Arch is perfect for it—I’m always with the latest software and always tinkering.

But my desktop? I rarely use it compared to my notebook. I’ve tried installing Arch on it, but always wiped it days later because I tend to over-customize Arch boxes and didn’t want to duplicate work across machines.

What distro is good for “non-mainstream” personal use but that you won’t use much? Each has problems:

Funtoo/Gentoo: Too many compilation times. You don’t want that for a machine you don’t use often.
Ubuntu/openSUSE: Okay, but Ubuntu is often out-of-date; openSUSE’s zypper packages are bloated.
Fedora: Changes releases too often. You don’t want to reinstall.
Alpine/grml: Great for other purposes, not “my-quiet-desktop”.

Maybe the problem is me. I got too strict about my personal environments. Then I thought: the problem is lack of automation. If you don’t want to touch machines you don’t use often, automate them. Upgrades, maintenance, etc. SaltStack to the rescue!

SaltStack Hello World #

We begin with hello world. For network environments? A ping, of course.

I’m installing Salt on my Arch notebook (master) and my Arch desktop (minion). Salt uses a master-minions hierarchy. My notebook is the master; my desktop is the minion.

Salt is cross-platform, so you can use different distros. You need to care about your init program; systemd eases everything for me:

$ systemctl enable salt-master # (in master)
$ systemctl enable salt-minion # (in minion)

Salt is now set up. After installing, you have to connect your machines. This is well documented (RTFM). I connected them on the same LAN network and they continue to recognize each other across reboots.

Now the best part. From my master (notebook):

# root@arch /home/thiago # salt '*' test.ping
<$MYHOSTNAME>:
  True

Everything is properly set up. Now what? I don’t know—I’ll discover!

But I tried automating a system upgrade:

root@arch /home/thiago # salt '*' pkg.install refresh=True sysupgrade=True feh
<$MYHOSTNAME>:
  ----------
  feh:
    ----------
    new:
      2.12-1
    old:

  giflib:
    ----------
    new:
      5.1.1-1
    old:

  imlib2:
    ----------
    new:
      1.4.6-3
    old:

  libexif:
    ----------
    new:
      0.6.21-2
    old:

  libid3tag:
    ----------
    new:
      0.15.1b-8
    old:

Really nice! I can install or remove any package. This is all about automation. I’ll see in the next days if I keep using Salt.

Footnotes #

Why Salt? I just picked it first. I could use other technologies. I don’t recommend it right now—just testing it. The choice was pure serendipity.

— § —

Reply via email

#dev #legacy

Recording screencasts on Linux

serendipity@perrotta.dev (Thiago Perrotta) — Sat, 21 Mar 2015 21:03:48 -0300

♠ One-line self-documenting functions, put this into your shell init file:

# ffmpeg x11 screencasting
# it will just record video: no audio
# dimensions of your screen are automatically calculated
_ffmpeg-screencast() {
 ffmpeg -f x11grab -s $(xdpyinfo | awk '/dimensions:/ {print $2}') -r 25 -i :0.0 -qscale 0 $HOME/ffmpeg-out-$(date '+%Y-%m-%d-%H-%M-%S').mkv
}

# ffmpeg x11 screencasting with audio
# it supersedes the previous function
# source of the audio is determined with `pavucontrol`: open it then go to the "recording" tab. You can select between the microphone or your computer as audio sources
_ffmpeg-screencast-audio() {
 ffmpeg -f alsa -i pulse -acodec pcm_s16le -f x11grab -s $(xdpyinfo | awk '/dimensions:/ {print $2}') -r 25 -i :0.0 -qscale 0 $HOME/ffmpeg-out-$(date '+%Y-%m-%d-%H-%M-%S').mkv
}

Observations #

If you want to get this automatically, you might be interested in my dotfiles repo. In particular, inspect the .alias file, where those functions are located (at least at the time of this post; things change).
Audio recording works best with Pulse Audio (citation needed). I haven’t tested with alsa only.
Pavucontrol is just a fancy interface for configuring pulseaudio, but there are others.
Tweak those parameters…ffmpeg is a really powerful tool.
I tested this on the i3 window manager, and it works beautifully. I tested some GUI screencasting programs on it before and the results were poor.
Source of this script: a merge of several ones on the web.

Oh yes, and you can stop the screen capturing with the ‘q’ key or with a classical Control-C.

Edit: thanks Victor for a small improvement with awk in the above functions.

— § —

Reply via email

#dev #legacy

★ [FAQFlow #2]: Android Upgrading/Rooting Notes

serendipity@perrotta.dev (Thiago Perrotta) — Wed, 11 Mar 2015 22:41:06 -0300

♠ I successfully upgraded my Android phone from Jelly Bean (4.1.2) to Kit Kat (4.4.2). Here are notes about my upgrade process.

Key steps #

Backup your data before trying anything! Make sure it’s available elsewhere—cloud or computer.
Make screenshots of installed apps: press Volume Down + Power buttons together for a while. Save these screenshots!
Did you backup your data? (Really!)
Is your bootloader unlocked? If not, unlock it first. See the XDA Developer Forums.
Is your device rooted? Maybe root it first. See the XDA Developer Forums.
What do you want to do?
- Upgrade Android version? Search for “OTA upgrade”
- Customize it? Search for “custom ROM”
Get a decent recovery. TWRP (Team Win Recovery Project) is one of the best. CWM is another option. A recovery is useful for flashing ZIPs, wiping cache, installing custom ROMs, and other maintenance.
Reboot into bootloader with adb: adb reboot-bootloader
Or reboot into recovery: adb reboot recovery
While in fastboot, you cannot use adb.
Get updated adb: ftp://ftp.mozilla.org/pub/labs/android-tools/
Upgrade from Windows or Linux: Usually you can do both. Windows is common among power Android developers. Use what’s available and recommended by your tutorial.
Don’t install blindly: Read, read, and read before anything! Check opinions of other users.
Need support? Use:
- IRC (Freenode: #android-root channel)
- Facebook Groups
- Google+ Groups
- Reddit
- Search for groups about your phone model first
- XDA Forums (technical/advanced place—don’t ask newbie questions unless there’s a specific section)
Debloating is a good term to search for after installing/upgrading your device.
Google Play isn’t the only option. See F-Droid (contains FOSS software).
Only upgrade with a charger connected.
Don’t install untested or poorly tested ROMs. It’s better not to install anything than install something that doesn’t work well. Use your head and inspect context carefully.

These items sum up key points. If you want to complement this post, feel free to leave a comment. Happy hacking!

Footnotes #

Device: Motorola Razr I XT890, with Intel Atom (x86) CPU and 1GB RAM
XDA Developer Forums: http://forum.xda-developers.com/

— § —

Reply via email

#bestof #dev #legacy

Running Ogre3d locally (without installing it)

serendipity@perrotta.dev (Thiago Perrotta) — Mon, 09 Mar 2015 03:05:25 -0300

♠ Simple build systems, huh? Hello there. In this post we will run the “Basic 1” example from the Ogre3d wiki without installing ogre in the system. For that, I’ll assume you’re using an Unix system. First things first: grab the latest ogre version. At the time of this post, it is the 1.9 release, and you can find its repository on Bitbucket. Note: you want to download the 1.9 version using the tag download feature from Bitbucket; do not simply download the master branch. After that, extract it, create a build directory inside its directory, cd into it, and do a

cmake .. -DCMAKE_INSTALL_PREFIX=$HOME/.lib -DOGRE_INSTALL_SAMPLES=TRUE -DOGRE_INSTALL_DOCS=TRUE -DOGRE_INSTALL_SAMPLES_SOURCE=TRUE -DCMAKE_BUILD_TYPE=Release

This is a pretty much standard cmake build. If your system is missing some of its dependencies, you will be warned and you should install them. Then:

make OgreDoc install

Now it is a good moment to read xkcd. The usual difference here is that we don’t do a normal make install. Why? Because we don’t want our stuff ending up on /usr/local. Instead, we want it somewhere where we have full access as a normal (non-root) user – in this example, $HOME/.lib/ogre.

Basic 1 #

Now, head to the basic 1 page of the ogre wiki and download its example. You should try to build it, but take care to change all ogre3d references from /usr/local to your new install directory ($HOME/.lib/ogre). Yeah, that’s it. Note: if the basic 1 download doesn’t come with a CMakeLists.txt file, you can download it from the Ogre CMake wiki page (clean ogre cmake project). You can use it verbatim; just put it in the same directory as the basic 1 files and – again – change any references to /usr/local to your new install directory. Also, copy the dist/ directory too (if it is missing from the Basic 1 project), I’ll add a small observation: in the case cmake complains about pkg-config, try adding this line to the top of your CMakeLists.txt (but after the project version):

set(ENV{PKG_CONFIG_PATH} "$ENV{HOME}/.lib/ogre/lib/pkgconfig:$ENV{PKG_CONFIG_PATH}")

Credits: the build style I adopted here was derived from the Ogre Arch Linux package. Thanks Sven for your excellent packaging skills! :) Update: in your plugins.cfg , do not use “$HOME” as your PluginFolder. This seems to cause issues when you run your ogre app. Instead, expand $HOME manually (in my case, to ‘/home/thiago’). This is not much portable, I know, but there must be some workaround for this I haven’t discovered yet.

— § —

Reply via email

#dev #legacy

Mulheres na computação, cadê vocês?

serendipity@perrotta.dev (Thiago Perrotta) — Sun, 08 Mar 2015 16:26:42 -0300

♠ 08 de março. Você sabe que dia é hoje. Muitos dizem que é um dia especial. Já eu, pessoalmente, em geral sou indiferente para qualquer comemoração do tipo, porque eu acredito que todos os 365,25 dias do ano são igualmente especiais. Não é necessário tirar um único dia do ano (menos de 1%!!!) para dizer que as mulheres são especiais¹.

Mas bem, hoje resolvi aproveitar a deixa para fazer uma pequena divulgação. Como você pode ler no título, eu restringi o post a “mulheres na computação”.

Muito bem, pois é que existe um grupo bem bacana na comunidade do Arch Linux chamado Arch Women, que no fundo é uma organização sem fins lucrativos que visa incentivar a participação de (mais) mulheres no mundo do Arch e de open source em geral. Mais do que isso, no fundo é um grupo que acredita na igualdade de gêneros e que quer enviar qualquer tipo de preconceito para /dev/null.

A organização é bem bacana, de vez em quando são organizadas algumas aulas pelo IRC, também existem projetos de “estágio”, digamos assim (um exemplo típico é o Google Summer of Code) e, bem, é uma comunidade open source, então as possibilidades de interação são inúmeras.

Adicionalmente, existem alguns encontros (encontro talvez não seja uma palavra boa. Vamos dizer: reuniões) todo início do mês no IRC, onde são realizadas algumas discussões sobre as coisas que eu falei aí em cima.

Ah sim, esqueci de falar duas coisas: 1) não é uma comunidade só para mulheres; na verdade, é uma comunidade que incentiva a participação de mulheres na tecnologia. Então, existem vários rapazes por lá também. Se você quiser ajudar, seja bem-vindo. 2) já falei de IRC umas duas vezes mas não disse onde: é no irc.freenode.net, no canal #archlinux-women. É, o idioma é em inglês.

Enfim, essa história de “mulheres na tecnologia” (computação, whatever) é frequentemente tratada como um tabu, mas não precisa(va) ser assim. E eu aproveito a deixa desse post para incentivar mais as pessoas a criar organizações que incentivem a inclusão de mulheres no mundo digital. Não precisa ser só a Arch Women, afinal.

Até mais ver!

afinal, elas são especiais sempre, certo? ;) ↩︎

— § —

Reply via email

#legacy #serenity

Getting started with Ogre3d

serendipity@perrotta.dev (Thiago Perrotta) — Thu, 05 Mar 2015 16:46:49 -0300

♠ I struggled a little to get started with Ogre3d graphics, so here are a few notes. In this post we will be able to run some basic examples from the framework.

Installation #

Get a distribution with a nice packaging system and install ogre from it. I can cite two: Arch Linux family (pacman-based) or Gentoo family (portage-based). The point is: DON’T install the ogre package from deb-based distros (Debian and Ubuntu).

They have a few problems:

1.1 Packages are usually outdated #

In the case of ogre, this isn’t a super problem. Its current version is 1.9, and Ubuntu offers 1.8. You can live with that.

1.2 Some files are missing from the packages #

I compared the upstream ogre tree (from Bitbucket) with files installed on Ubuntu (dpkg-query -L libogre-1.8-dev), and some files were missing. One example: SDK Trays.

1.3 Some files are installed in non-standard locations #

For example, x86_64-linux-gnu being used as a directory. This is non-standard. And what about /usr/lib/ogre? This folder doesn’t even exist on Ubuntu, but should!

Worse yet: Ogre documentation assumes the default installation directory is /usr/local. This assumes you don’t use a package manager and used make install after compiling, which requires sudo rights. This is too restrictive. Everything that ends up in /usr/local (programs installed manually) eventually becomes a mess.

Disclaimer: I’m not trying to convince you not to use Ubuntu for ogre3d development. I’m just stating I had an awful experience. If you can handle it, please comment how you did that!

I’m using Arch for ogre3d development now; if forced to use Ubuntu (work, lab), I’ll have to install ogre locally on the system.

Building with Arch #

pacman -Syu ogre

Now see if you can run the most basic pre-example from Ogre. Download it. Files include:

ogreapp
- CMakeLists.txt
- BaseApplication.h
- BaseApplication.cpp
- TutorialApplication.h
- TutorialApplication.cpp
dist/ folder

There’s a CMakeLists.txt, so the build should be straightforward. Create your build directory, cd into it, cmake .. and make. Copy the OgreApp executable into the dist folder manually.

Note: The build system could be better. Just do it.

There’s a warning: see the plugins.cfg config file. Change /usr/local/lib/OGRE to /usr/lib/OGRE.

Basic Example 1 #

Get it from Ogre Basic Tutorial 1. Uncompress it, move its files to a ‘basic1’ directory, and copy CMakeLists.txt and dist/ from the previous example. Copy and paste the code from the wiki to TutorialApplication::createScene in TutorialApplication.cpp. Build as usual, and copy the OgreApp executable to the dist/bin directory. Done!

PS: I’ll likely write another post about Ogre.

Footnotes #

The problem isn’t cmake (I had experience with it). The problem is the documentation. Ogre’s wiki is scattered. There should be one entry point for “getting started with ogre”, but there are several. A typical beginner struggles to decide which one to begin with. Documentation outside ogre3d.org is poor—forums, Reddit, blogs are either inactive, old, or poor.
Bias: I love both Arch and Gentoo.
PPAs? Please don’t use them.

— § —

Reply via email

#dev #legacy

Documented dotfiles

serendipity@perrotta.dev (Thiago Perrotta) — Thu, 05 Mar 2015 04:06:31 -0300

♠ I just documented my dotfiles to make them easier to understand. Check them out on GitHub.

— § —

Reply via email

#dev #legacy

[FAQFlow #1]: XMonad

serendipity@perrotta.dev (Thiago Perrotta) — Wed, 04 Mar 2015 21:45:47 -0300

♠ WTF is a FAQFlow? See the footnotes at the end of this post.

XMonad is a window manager for Unix systems written in Haskell. This is an introductory post on how to use it.

Getting started #

First, install XMonad. You can either install it manually or use your system’s package manager.

If you’re using X11 with a display manager, proceed normally. If you don’t use a DM, put this at the end of your ~/.xinitrc:

exec xmonad

Now run startx.

Common questions #

1. XMonad is lightweight but GHC (Haskell compiler) is way too big (>400MB+). Do I really need it? #

Yes, if you really want to get the most out of xmonad. This is part of the game. If you already use Haskell, this won’t be a problem. Otherwise, read about GHC in “The Architecture of Open Source Applications”—there’s a good reason it’s that big.

2. I just started xmonad. But there is nothing on my screen—it’s blank all over the place #

This is normal. XMonad is lightweight! There’s nothing fancy getting in your way. If you think there should be lots of widgets and desktop icons, maybe you should start with an easier tiling window manager first. i3 is a good choice.

3. How do I open a terminal? #

Mod+Shift+Enter. Mod is usually Alt, but can be changed to the Windows key later.

Bootstrapping #

Now you have a terminal. What’s next?

How to change workspaces? #

Yes, xmonad has workspaces: Mod-1, Mod-2, etc.

How to change layouts? #

Mod-Space. It cycles through the three default available layouts.

I don’t like xterm! It’s way too ugly! Can I change the default terminal emulator? #

Yes! Create a config file at ~/.xmonad/xmonad.hs:

mkdir ~/.xmonad
cd ~/.xmonad
vim xmonad.hs

Add this:

import XMonad

main = xmonad default Config
  {
  -- set your terminal. Examples: konsole, gnome-terminal, urxvt
  terminal = "urxvt"
  }

Lines beginning with -- are comments. Change “urxvt” to your favorite terminal emulator and run:

xmonad --recompile

This makes GHC compile your xmonad.hs. Do ls now and you’ll see the compiled results (assuming compilation succeeded).

Now restart xmonad:

xmonad --restart

Close your current terminal and open a new one (Mod-Shift-Enter). Done—now you have a pretty terminal!

This post will have a continuation.

Footnotes #

FAQFlow: A new way to document and teach introductory skills or tutorials. The word is a portmanteau of “FAQ” (frequently asked questions) and “workflow”.
Customizing xterm: You can customize ~/.Xresources to get a nice-looking xterm, but most people either don’t know what a Xresources file is or don’t bother creating one.

— § —

Reply via email

#dev #legacy

Thoughts on Instagram

serendipity@perrotta.dev (Thiago Perrotta) — Mon, 02 Mar 2015 05:18:09 -0300

♠ Note: this is a brief article which exposes my opinion about a free web service.

For a long time I refrained myself from creating an Instagram account for some reasons. Those reasons are not much different for the delayed birth of my Facebook profile. Let’s talk a little about them and discuss if they were or not valid:

A. Too mainstream: yeah, Instagram is pop. One of the top downloads of smartphones app stores nowadays. So what, you just stop using a service just because it is mainstream, and period? No, no, no. But…for every mainstream service, caution is needed.

A.1 Privacy: Most accounts from people I currently follow on Instagram are public by default (neverthless, this is the default setting of Instagram). Not cool, dude, not cool. This is a short essay and you probably know what I would talk here, so let’s be brief : most of those people are either not concerned or don’t know what the problem of this deliberate practice might cause them a day.

A.2 Did you read its terms of service? Photos are stored forever in there (in principle), even after you delete them. And what one person – for example, either a complete stranger or the company behind Instagram itself – could to with them? So many things. The most common underrated example is: selling them – plus information associated with them – to other companies, so they can (for example) better target ads to you. This is not simply a matter of ads: it also creates a bigger and bigger bubblein which you get stuck.

I’m just posting random photos in there, so I don’t care. But hey, look at other people profiles. They post photos about themselves all the time. $$$$oooooo many data…one more thing, remember: Instagram was created exactly for this purpose: you are encouraged to use it just from your phone (not from the web), and there is a big “shoot photo” button. While you can upload photos to it manually (instead of shooting and submitting them there directly), this is not encouraged by design. So hey, a big database of public photos of Bob and Alice, available 24/7.

Remember: when the service is free, YOU are the product (author needed).

So, is maintreamness a problem? That depends on you.

**B.**A social network represents what most of its users make from it. And…what most (I repeat: most) users use Instagram for? For bragging. That’s it. You might try to argue it is different, but it is not. One could say this is human nature, but if it is, then I find it pretty sad. I always had this suspicion, but only after creating an Instagram account I could confirm it. So what? Pretty simple: I’ll just delete the app from my smartphone. I won’t delete my account because I may need it in the future (don’t forget: people post data; data might become valuable anytime later on). However, if you find bragging cool (I’m not judging, just stating that I don’t like this), then Instagram is probably the perfect social network for you. It is even better than Facebook (actually, it supersedes it; you can automatically make every new photo of yours appear on Facebook automatically, yay!) in this sense.

C. Design: despite the previous item, I have to say something: I appreciate the engineering and the design of the service. Engineering? Well, I previously said the service is consumed by many people, millions of them, and yet it has a very good response time. Design? Yeah, the app and the service are very well written and conceived. UX is amazing, the app is pretty much intuitive and easy for newcomers to learn it. And it follows the Unix philosophy as a bonus: “do one thing and do it well” ==> photo sharing. ATM I don’t know about the team, but I would probably be very happy if I worked in a successful service and ecosystem like it.

Conclusion : of course I generalized a few things¹, but I believe they are not far from the truth. This was a brief overview about Instagram, there are no right or wrong conclusions: you think what you want and use it as you wish. Have a nice day.

Just bragging? Oh no. There are some artists and brands out there. They win real money from their visibility on the $ocial network. ↩︎

— § —

Reply via email

#dev #legacy #serenity

Account deletion request

serendipity@perrotta.dev (Thiago Perrotta) — Sun, 01 Mar 2015 14:15:20 -0300

♠ There is a fact that is not usually much publicized about accounts in the web: in general, if a given web service doesn’t have a delete button, you can instead request manually the deletion of your account, by e-mailing the administration of the service. Usually the e-mail you want to contact is localized somewhere on the ‘contact’ or ‘about’ page of the service, or even in the homepage. If not, you can try to make educated guesses:

Through the last months I occasionally deleted some of my accounts and I confirm that this method usually works all the time. (Bias: )Well, at least if you subscribed to services with good admin/policies in the first place… But really, even when my account couldn’t be deleted for…er…bad design in the database, I got nice replies from real people (not robots :-P (yet)). There are a few exceptions for this. One is when it isn’t an advantage for the service to have you gone. Example? Facebook. You can only deactivate your account; there is no option to permanently delete it. And don’t you even try to contact Zuckerberg…other exceptions I have yet to experience. If you have any good (or bad) experience about account deletion requests, don’t hesitate to share it in the comments. Have a nice day.

— § —

Reply via email

#legacy

Thoughts and notes on deploying Wikimedia

serendipity@perrotta.dev (Thiago Perrotta) — Mon, 23 Feb 2015 14:23:33 -0300

♠ Wikimedia is the most full-featured wiki software out there and powers Wikipedia. Let me talk about what you may not know yet.

I tried to deploy it on Debian. While Debian tries to set everything up for you, pulling dependencies like Apache and MySQL, this doesn’t necessarily reflect what you want.

First, Apache creates a new process for every connection, which doesn’t scale (see the c10k problem). While probably the most used web server nowadays, I prefer something more lightweight. Two good options are nginx and lighttpd.

Second, PostgreSQL is better than MySQL. This is not something I can fully argue about, but most people I know agree on that.

Conclusion: if you use Debian, either you’ll struggle to get the packages and software you want working together, or a sub-optimal solution will be deployed. I prefer a KISS operating system. For familiarity, I chose Arch Linux.

Note: Be aware that some people discuss using Arch as a server, mainly for stability reasons. The point is: choose an OS that doesn’t get in your way.

Installation #

Get php: pacman -S php
Get a CGI program for php. I like fastcgi: pacman -S fcgi (alternative: php-fpm)
Get a database server. PostgreSQL: pacman -S postgresql
Get a web server. Lighttpd: pacman -S lighttpd
Get wikimedia and optional dependencies: pacman -S wikimedia php-gd php-intl php-pgsql imagemagick

Now it’s a matter of configuration. I advocate using “/srv/http” instead of “/var/www” for storing web files. However, wikimedia gets installed to “/usr/share/webapps/wikimedia”, so you have to either create a symlink or use lighttpd’s mod_alias to create an alias to this path. I prefer the last option so wikimedia gets placed in a specific directory—for example, “/wiki”.

Consult the Arch Wiki or lighttpd doc pages for configuration details. After configuring lighttpd with fastcgi and wikimedia, configure php (enable extensions and fix cgi path info to 1) and PostgreSQL.

After everything is properly configured, start the services. Arch uses systemd:

% systemctl enable lighttpd
% systemctl enable postgresql

That’s it.

Footnotes #

c10k problem: http://www.kegel.com/c10k.html
Arch Wiki — Server: https://wiki.archlinux.org/index.php/Server
Arch Linux forum discussion: https://bbs.archlinux.org/viewtopic.php?id=162434
Reddit discussion: https://www.reddit.com/r/archlinux/comments/1wqpdw/do_you_use_arch_for_production_servers/

— § —

Reply via email

#dev #legacy #selfhosted

ksnctf

serendipity@perrotta.dev (Thiago Perrotta) — Sat, 21 Feb 2015 21:38:13 -0300

♠ CTF: http://ksnctf.sweetduet.info/

Write-ups: https://manhhomienbienthuy.github.io/

— § —

Reply via email

#dev #legacy #security

From orgmode to asciidoc (OR: ditching emacs)

serendipity@perrotta.dev (Thiago Perrotta) — Fri, 20 Feb 2015 14:42:16 -0300

♠ Hello people, I’m back. The portal will open in 3…2…1... OK, here are some updates. After taming the beast for two long years, I’m now slowly ditching emacs from my life.

_# archers, don't struggle yourselves!_
sudo pacman -Rnsc $(pacman -Qdt emacs)
_# *ntoo users, don't lose opportunity cost regarding precious compiling time!_
sudo emerge -vC emacs && sudo emerge -avuDN @world && sudo emerge --depclean -a

Emacs doesn’t simply go out and we are done here. No, no, no. Rehab, guys. Org-mode was a fellow companion, now what? I even moved out of wordpress just to blog with org files¹. Orgmode is the most amazing to-do / organizing / planning / tracking tool I’ve ever used, however it is completely tied up to emacs. The only thing we can do about that is to get all the knowledge and discipline that orgmode teached to us, and then move on.

And I moved on.

I’m now slowly converting my text formatting habits to Asciidoc. It is an oldie (2002+) thing, but people are still using it, and very well, by the way. Asciidoc is not a replacement for markdown², but rather a full-featured text formatting tool — markdown is KISS. I also found a nice³ tool to blog from an asciidoc file directly to Wordpress; and it is called…blogpost, you are right. It was created by the same author of asciidoc, so you can probably conclude it is reliable. OK, goodbye for now, till next time.

Footnotes #

now you realize this was useless, since I moved back to Wordpress. But at least I learned some Hakyll and Haskell in the process. ↩︎
neither the other way around. ↩︎
not sure if it is really nice, but if you can see this footnote, then I can guarantee you the tool still works. Update : actually, the tool works indeed, but it doesn’t play nice with footnotes. ↩︎

— § —

Reply via email

#dev #legacy

Games: index of

serendipity@perrotta.dev (Thiago Perrotta) — Thu, 12 Feb 2015 14:03:36 -0300

♠ I am a huge fan of games in general, although nowadays I’m more of a casual player. Today I realized I never wrote a blog post about them.

Social networks of games never really existed or worked at all. Steam is the closest to one today, but in practice you become stuck in their platform. What would one expect from a social network of games?

See what your friends have been playing → serendipity + you can play together
Track the games you have played → gamification + organization

In this sense, Kongregate satisfies those properties, although it has the same problems as Steam: its network comprises only games on their platform. To be successful, a potential new social network should separate metadata from games—track only information about games, not host the games themselves. Unfortunately, I don’t know one today, so I decided to create my own list of games. I will track comments about my gaming styles, preferences, and games I’ve played from now on.

— § —

Reply via email

#legacy #serenity

C++ man pages

serendipity@perrotta.dev (Thiago Perrotta) — Thu, 15 Jan 2015 13:00:00 -0300

♠ Today I stumbled upon across a project called cppman¹, which I believe has a very suggestive name indeed. Manual pages for the C programming language are almost always available on Unix systems. Don’t you remember the signature of the qsort function? No problem, a man 3 qsort will refresh your memory. What about printf? What about stdlib.h? Man won’t disappoint you. However, what about C++ man pages? I first discovered that they weren’t available by default about three years ago. Okay, so what do I do? A simple google search will probably lead you to Stack Overflow². At the aforementioned time, there weren’t really useful answers in there. Okay, some of them pointed out to libstd++ docs, but using them wasn’t much convenient. The best documentation sources you could find were cplusplus.com and cppreference wiki³. You can live with these on-line references, but you will quickly get bored with them. Now, back to cppman, the boredom killer and time saver.

cppman — quick reference #

Installing it #

If you’re an user of Arch Linux, it is available in the AUR at the time I’m writing this post.
Otherwise, you can install it using python’s pip – ya know, pip install cppman.

It doesn’t really matter the method you choose, just make sure to add it to your PATH.

Using it #

cppman sort
cppman map

Don’t cry, it is real. You can optionally cache all of its man pages, but I don’t find this strictly necessary for me. The default option (no cache) downloads manual pages on-demand. Oh, it uses vim as its default pager, so you can search for keywords with /. Also, hyperlinks are everywhere, so you can use K or C-] to follow them. To go back, use C-t. By the way, it is compatible with man, so if you decide to cache its man pages, you can view them later with man (instead of cppman). Happy ~~hacking~~ reading!⁴

Footnotes #

cppman — you don’t need to read the whole article; if you’re in a hurry, just open the project page and forget me! ↩︎
As always, huh? Stack Overflow thread. ↩︎
The last one is by far much better maintained. ↩︎
Wait, since when reading manual pages is a source of happiness? ↩︎

— § —

Reply via email

#dev #legacy

The Eudyptula Challenge

serendipity@perrotta.dev (Thiago Perrotta) — Wed, 07 Jan 2015 12:54:40 -0300

♠ Update: I stopped at task 06 on February. Little seems to have stopped responding e-mails. Which is sad, because I was liking those tasks =/

I am participating in the Eudyptula Challenge, which is not something exactly new, but it is as able to get you out of your comfort zone as if it were¹.

It puts you into the mind of a Linux Kernel developer, where you are required to write and submit patches and to complete some tasks.

Subscribing to the challenge is a mini challenge itself²: you must write to little (the superuser behind the thing) an e-mail in plain text. This means: NO HTML. This looks like a silly requirement, but it’s the way how the (real) Linux developers work, and it is actually slightly fancier than you probably imagine. Little himself recommends that the participants don’t use either the Gmail web interface or Outlook.

I intend to document here my own experiences with Eudyptula³.

Setting up #

The first thing I went after was a decent and updated Linux distro; doing the challenge within a Linux rather than anything else offers a better experience. Zero effort here, as I’ve already had Arch Linux⁴ installed.

Next: a good text editor. Wait…why not two? I am using vim to compose e-mails, and emacs to write code.

What’s next? Oh yes…a decent sendmail program. It is not mandatory to have one of those, but it eases a lot the process of writing e-mail in plain text. I grabbed msmtp plus its MTA. Writing a config file for it is pretty straightforward.

And now I was ready: I subscribed to the challenge.

The task #01 #

Hours later, little⁵ acknowledged my e-mail and sent me the instructions for the first task.

I should write my first module for the Linux Kernel.

A good start. I didn’t even know how to begin, oh no!

The task was to write a module that would write ‘Hello World’ to the debug level log of the running kernel when loaded.

First, I had to install some dependencies to be able to compile a couple of C programs. Actually, the base-devel group from Arch already contained most of the things I would need. linux-headers and bc were also relevant.

My first module, written in C⁶, looked like this in the end:

#include <linux/init.h>;
#include <linux/kernel.h>;
#include <linux/module.h>;

MODULE_AUTHOR("Thiago Perrotta");
MODULE_DESCRIPTION("A gentle Hello World module");
MODULE_LICENSE("GPL");

static int __init hello_init(void)
{
  printk(KERN_DEBUG "Hello world!\n");
  return 0;
}

static void __exit hello_cleanup(void)
{
  printk(KERN_DEBUG "Goodbye World!\n");
}

module_init(hello_init);
module_exit(hello_cleanup);

I didn’t get it right in my first trial, of course. In particular, I wrote ‘Hello World’ to the INFO log level, using KERN_INFO, instead of KERN_DEBUG.

And I also needed a Makefile to be able to compile it:

obj-m += hello.o
KERNEL ?= /lib/modules/$(shell uname -r)/build

all:
    make -C $(KERNEL) M=$(PWD) modules

clean:
    make -C $(KERNEL) M=$(PWD) clean

The Makefile wasn’t right in my first trial, either. Another requirement of this task was that this file should provide an environment variable, as an alternative location for the build directory of the kernel, opposed to being a hard coded value. I provided one from the beginning, but the way I wrote it was slightly wrong.

Now, this code didn’t just pop out of my head out of nothing. Although I know how to program in C, I never did kernel programming before. I had to look up for some documentation about how to write a module. It wasn’t super straightforward, because I found many documentation sources about writing modules for the Linux 2.6.x version, which is outdated now⁷, and the modern version of writing modules is a bit different from it.

Also, I was careful not to accidentally find a solution for this task directly. Of course, some people have already done it, and they could have written something about their experiences, as I am doing right now.

Anyways, after writing those two files and testing that they really worked as expected, I loaded my freshly built module. And yay, a simple dmesg would reveal my hello message out there. It was really cool to see it worked:

thiago@archpad ~//01 % sudo modinfo hello.ko
filename:       /home/thiago//01/hello.ko
license:        GPL
description:    A gentle Hello World module
author:         Thiago Perrotta
depends:
vermagic:       3.17.6-1-ARCH SMP preempt mod_unload modversions

After that, I’ve sent everything I should to little, and today I got its reply, about the second task. So, let’s go…

What have I learned with this task? #

how to write a Makefile to compile a kernel module
how to execute shell commands within a Makefile
how to use an environment variable within a Makefile
how to write e-mails in plain text and send them with msmtp
how to send attachments without base64 encoding⁸
how to load, unload and get information about modules
modules are so dynamic and easy to load, huh? Like USB devices. I would never imagine that.

It is also a excuse for me to write something here :-) ↩︎
Pun intended. ↩︎
I can’t write this in one shot. Really. Eudi←yptull←a. EUDYPTULA. ↩︎
The distro that will conquer the moon. ↩︎
Which turned out to be a bot. ↩︎
Damn, everything will be written in C from now on. That’s what we get with kernel programming, I guess. ↩︎
Only lazy or paranoid Linux sysadmins will tell you otherwise 😉 ↩︎
This turned out to be the hardest thing of this task, yet it wasn’t even part of it. ↩︎

— § —

Reply via email

#dev #legacy

★ 2014 - Small review of a really nice year

serendipity@perrotta.dev (Thiago Perrotta) — Wed, 24 Dec 2014 04:51:52 -0300

♠ Disclaimer: I haven’t reviewed this post. It was written in one shot. I hope it is understandable.

This is possibly the last post from me on this platform. However, I’ve said that a thousand times before. I’ve tried many platforms: Google Sites, Jekyll, nanoc, and Octopress.

Last year I wrote a similar post. Now, here I will register some small changes, tidbits, and bits I’ve learned and discovered this year. I won’t create links—just pure text.

The actual post #

Detailing everything would be tough, so I present this as lists, trying to be chronologically organized and reasonably brief.

Campus Party #

This was my first great event. In particular:

I participated in a hackathon for the first time. It was from Mozilla, and my code is on GitHub in the “Riddlefox” repository. It’s a riddle game for Firefox OS.
I participated in my first Capture the Flag (CTF) competition, without knowing it was a CTF. Promoted by CERT.br, I won a Kindle Paperwhite 2, ending in second place. Today, the Kindle is one of the devices I use most.
I opened my eyes to communities thanks to open source communities at this event. Today, I’m an active member of the Arch Linux community.
I opened my eyes to web development. I’m not a super fan yet, but it’s on my planning list.

Arch Linux #

I was searching for a community to settle on for two years. The Arch Linux community is perfect for me. I learned:

Communities
MediaWiki syntax/style
IRC for open source projects
The importance of the command line
The importance of “do it yourself” and not feeding help vampires
Sometimes user-centered > user-friendly
The feeling of knowing intelligent people and becoming more intelligent with them
Tiling window managers are awesome (currently using i3)
Mailing lists

Programming languages #

One of my goals this year. Languages I learned include, in order: emacs lisp, cmake, golang, javascript/node.js, ruby, haskell, Matlab, and R.

The most important things are the paradigms:

Haskell: functional programming
Golang: systems programming
Ruby: dynamic programming

My goal for next year is more depth-first than breadth-first. I want to improve my knowledge in some of these languages, probably digging deeper on Haskell.

MOOCs #

I became more interested in them month by month. Coursera is my preferred platform, though I learned to like edx. The best MOOC this year was on Functional Programming using Haskell, conducted by Erik Meijer.

Presentations #

LaTeX + beamer became my de facto standard. For collaborative editing, Google Docs/Slides continues to be the best. LyX is my preferred tool.

Books and reading #

This year was great. My Goodreads profile tells me I read 50 books this year, almost double from last year (26 books). Most were technical.

I can’t live without Goodreads anymore.
You don’t need to buy books on Amazon for a Kindle. I’ve never bought one. Epub and mobi are the best formats.
I can’t live without Calibre anymore.

Operating systems #

I learned more about FreeBSD, grml (a nice distro for system administrators), and other Linux distros. Now I try to say Unix and *nix instead of just Linux.

Docker and containers are simply amazing. They will be the future. I also played with filesystems like btrfs and xfs.

Games #

Not a great year for games. The only one worth playing was Tom Clancy’s Splinter Cell: Conviction. Kongregate is slowly dying.

Music #

Spotify was my best discovery of the year. I’m trying to listen to different music genres. I also discovered blockify for blocking audio ads on Spotify (though I don’t recommend it—Spotify is a great service offered for free).

The podcasts I most like is Bad Voltage (presented by Jono Bacon, Jeremy Garcia, Bryan Lunduke and Stuart Landridge).

Computer graphics #

I created a game (repository “cg-sokoban”) and developed a ray-tracer. OpenGL wasn’t the most exciting thing I learned, but I got interested in game engines and image processing/computer vision.

Vim and Emacs #

The epic battle continues. I switched to vim for most daily text editing needs this year. However, I’m not done with emacs. My personal rule: if your editing task takes less than an hour, vim is better; otherwise emacs is better.

Hackathons #

I participated in my first hackathon with a team; my team got best place in a specific category (HackPuc).

Android #

I learned tricks to optimize it. I have the same device (Motorola Razr I) for almost 2 years, and it’s still fast. Root is necessary for maintenance.

Privacy #

Privacy becomes a concern more and more. I became an adopter of OpenPGP and I’m trying to move away from the Gmail web client. I’m using msmtp for sending emails, but I still need a more complete stack.

Facebook is becoming deprecated. My news feed is blocked from the web. Most of my contacts are boring, narcissistic, and futile.

Twitter: Better as a news feed than Facebook, though I never interacted with most people I follow.
Tumblr: Now using it for sharing links instead of Facebook.
Flickr: My preferred place for photos. Flickr is the best place for sharing photos, though this may change.
Reddit and Hacker News: Still the best sources of serendipity I know.
RSS feeds: Using Inoreader now. XKCD is probably the best RSS feed.

Podcasts #

This year was huge for podcasts. I have 47 podcast feeds, a ridiculous number. I listen regularly to ~15 of them. I’m adapting to listening at 2x speed. Some I listen at 2.5x.

Languages #

One of the best things I did was write in this blog in English. This improved much the way I think in English. What’s next? I’m thinking about studying French or German.

Orgmode #

The best TODO/agenda/organizer I discovered. Both for discipline and structuring your thoughts and plans.

The future starts with you #

This was probably the largest post I written in this blog. I’m pretty satisfied. I covered most of the things I thought. I haven’t talked about everything—programming contests, hackergroup, bookmarks, dotfiles, repositories, what I hated and loved to learn, professors, and the GPA system.

In this year, more than halfway through my college degree, several things became clear about my next steps. The next year I’ll probably study more computer graphics, computer networks, programming languages, and algorithms.

What is the future of this place? I don’t know. However, I tried to stop several times. This means something. I’m not 100% comfortable with using Wordpress. I’m getting bored of this format/style.

For all of this, I’m migrating to https://thiagowfx.github.io/archive. I’ll change the way I post too. If you follow by RSS, keep it in your feed list for a while until the new blog gets its own feed. The best way to get in touch with me is by following me on Twitter (@thiagowfx).

I wish you a good New Year. Let’s move on.

Observation: I’m on vacation. I’m open to study groups or business. If you want to study something with me or want me to work for you, this is a nice moment.

— § —

Reply via email

#bestof #legacy #meta

Tmux: a gentle introduction

serendipity@perrotta.dev (Thiago Perrotta) — Sun, 14 Dec 2014 03:00:38 -0300

♠ There are plenty of tutorials and insights about tmux on the web. This post isn’t another tutorial. Instead, I’m listing four purposes of why you should use tmux. I’m not here to convince you to use any tool; just read and create your own judgment about it.

Okay, tmux #

Observation: You should probably install and open tmux to follow this mini tutorial. After executing tmux, you’ll probably notice nothing fancy happens—it just feels like opening a new terminal.

Purpose #1: Terminal multiplexing #

Tmux is a terminal multiplexer. If you are used to opening several windows of your favorite terminal emulator (terminator, gnome-terminal, xfce4-terminal, konsole), you should probably stop that now.

Observation: <prefix> denotes your tmux prefix key. The default is C-b, but it can be changed—for example, vim users tend to dislike this setting because C-b is frequent in vim (it scrolls one page up, like Page Up). C-a is a common alternative.

Here’s how you open a new tab (window in tmux terminology) on tmux: <prefix> c. No mouse; no need to play with two windows. To switch between windows: <prefix> n for next and <prefix> p for previous.

Now for multiplexing: to create a new horizontal split, use <prefix> ". To create a new vertical split, use <prefix> %. To switch between splits: <prefix> o. This subset of keystrokes is sufficient to boost your terminal productivity.

Purpose #2: Pair programming #

Suppose you want to share a terminal session with a friend. What is a terminal session? Think of it as an analog to desktop sharing in the GUI world. However, it is more than simple sharing—the observer has full control over the terminal, just like you.

Therefore, a terminal session is suitable for:

Pair programming
Remote support (from a technical person)

What do you do? Open tmux. Tell your friend to SSH into your machine. They run tmux attach-session. This assumes you only have one active session. That’s it. To quit, execute exit or type Ctrl+D.

Purpose #3: Highly configurable #

Tmux is very configurable and can be scripted to do almost anything concerning shell organization. Look at other people’s .tmux.conf files, merge them, and create your own reasonable configuration.

Purpose #4: IDE/programming workspace #

Tmux can be used as an “IDE” or programming workspace. This becomes really easy with tmuxinator. The idea is to center all your workflow into one place: tmux. No need to swap windows.

Conclusion #

Stop hitting alt-tab and using the mouse. Use tmux—but only if it would boost your productivity. Otherwise, stick to your habits.

Observation: What about GNU screen? I haven’t tried it. Most sources I read and most people I know prefer tmux, finding it more modern and powerful.

Bonus: Detaching from a terminal, then re-attaching to it later is useful for monitoring servers, though there are better solutions for that.

— § —

Reply via email

#dev #legacy

Using proxy servers on Linux

serendipity@perrotta.dev (Thiago Perrotta) — Tue, 09 Dec 2014 02:48:45 -0300

♠ There are some times when you need to connect your PC to an access point which requires you to route your connections through a web proxy. No panic! It is really simple to config your linux desktop (or server, for that matter) to do it. You just have to modify your http_proxy and https_proxy environment variables.

For example:

export http_proxy=http://1.2.3.4:3128
export https_proxy=http://1.2.3.4:3128
export HTTP_PROXY=http://1.2.3.4:3128
export HTTPS_PROXY=http://1.2.3.4:3128

After that, just execute your favorite web browser then test your new settings. Beware! You have to execute your web browser through the command-line, in the same shell that you have exported those environment variables. For example:

$ export HTTP_PROXY=http://1.2.3.4:3128 && chromium &

Starting the .desktop application of your web browser, for example, wouldn’t work, unless you store those environment variables in a system wide directory (good recommendations include /etc/profile and /etc/profile.d/_).

However, from my experience, this is limited in the sense that it would be the default proxy of your system, which isn’t what you probably want – the common use case is: connect to some AP, switch to its proxy, and after a while disconnect from it and throw away its proxy.

Simple, isn’t it? If you want to automate this task a little more, please refer to the Arch Wiki – from where this post was originally inspired.

Also, I needed to apply this setup in a real world situation and I can confirm it works really well.

— § —

Reply via email

#dev #legacy #security

Raspberry Pi Overview - Part #0

serendipity@perrotta.dev (Thiago Perrotta) — Sat, 06 Dec 2014 01:11:54 -0300

♠ Hello people,

In the next days I’m gonna begin a new mini series here in the style of my last one, about FreeBSD.

Motivation #

I always wanted to have an opportunity to test a Raspberry Pi. I still remember my enthusiasm when I first heard about it (probably by a podcast or a Linux BBS). Okay, those small devices tend to attract people anyways, however the Pi always got a special place in my head. Although other projects such as Arduino, Galileo and Beaglebone ask for some attention, none of them is comparable to a Pi in the sense of upstreamness.

Now, after a little bit of serendipity and other factors, me and some friends won two Raspberry Pis on a recent hackathon. This was a complete surprise (I believe almost all prizes and/or competitions of my life always were one) and I thought it was (is!) the perfect opportunity to review one of them. So, let’s do it, shall we?

Hello, Pi #

What are the first impressions after you see a Raspberry Pi? Well, you’ll probably say: “whoa, it is very small for a computer!”. Maybe you won’t get too impressed in our modern days of modern smartphones and phablets, but let’s forget them for a moment.

— § —

Reply via email

#dev #legacy

M-x review-this-blog

serendipity@perrotta.dev (Thiago Perrotta) — Fri, 28 Nov 2014 15:19:08 -0300

♠ Dear reader, this place reached more than a year of existence, and this is a milestone I never dreamed of. I always thought that I would only blog for a few months. However, here I still am.

I always emphasize that this is a personal blog, therefore its posts are not necessarily optimized to please you—they are optimized to be clear, reflecting my thoughts as precisely as possible. Yet here are almost 10k views.

Views are not super important because I don’t win anything with ads (on purpose), but they of course motivate me to continue writing here. Now, as you are with me, I decided it would be a good idea to listen to you. There is a Google Docs form where you can say things you like and dislike on this blog, and also suggest future topics.

If you have some time, please respond. Thanks!

— § —

Reply via email

#legacy #meta

Matlab debootstrapping #1

serendipity@perrotta.dev (Thiago Perrotta) — Tue, 18 Nov 2014 21:45:00 -0300

♠ Alternative title: “Matlab, second impressions”. I am trying to learn Matlab super fast for prototyping purposes. Fast learning isn’t my preferred style for approaching a new programming language. However, this is necessary at the moment.

Disclaimer: This is not a tutorial. Just annotations and notes. Information is not necessarily correct, so consume it at your own risk.

Matlab Paradise #

Everything is optimized for matrices and vectors. Matlab means “matrix laboratory”.
A number is treated as a 1x1 matrix. This is really interesting and has nice implications and syntactic sugar.
Matrix manipulation: A = [1,2,3] is a row vector; B = [1;2;3] is a column vector; C = [1,2;3,4] mixes them together.
Slicing: Python style, but with different syntax. 1:10 means range(1,10) and 1:2:5 means range(1,5,2) but Matlab is inclusive (5 is present).
Indexing vectors: A(1,2), A(2,1). Beware: indexes begin from 1, not zero.
Performance: slow, slow, slow. See benchmarks on Julia’s homepage.
matlab -nodesktop for command-line prompt. Self-contained, and you can still browse help with doc lu or help lu.
Introspection #1: whos. Know everything about your variables.
Workspace view: Really nice. Save variables with save out.mat. Load them with load out.mat.
Some Unix goodness is available: ls, cd, pwd, ls -l.
However, mkdir and rmdir work differently.
Hello, world!: fprintf('string here') or display('hello, world!').
Strings are enclosed with single quotes, not double. Why?
Functions: C/C++/Java style. f(a,b,c,d,e).
Some functions have more than one return value: ret1, ret2 = f(a,b,c).
Clear screen: C-l or clc.
Beware! clear deletes your variables. clear var is more specific. clear all is even more destructive.
Auto-completion with TAB. No C-Space.
Need help? lu( then F1. Hints? C-F1.
Matlab hotkeys seem optimized for emacs users. See C-y (yank/paste), C-k (kill/copy), C-a and C-e (navigation), C-s (search), C-_ (undo).
open file.m to open files without the fancy GUI.
delete file.m. Not rm.
Comments with percent sign (%). Why not hashes?
Run a Matlab script: cd into its directory then type file<CR> in the prompt.
version! R2014a == 8.3.0.x
C-c for long computations.
rand(2,2) if you ran out of creativity for numbers.
v = [1 2] equals v = [1,2]. Who needs commas?
Element-wise operators are preceded with a dot. Example: .*.
Love intelligent prompts: C-p, M-p, C-n, M-n, <UP>, <DOWN>.
Miss your shell? system('echo Hello world!').
File handling: fid = fopen('file.txt','w'); fprintf(fid,'...'); fclose(fid);

Introspection #

which lu.m, which lu
edit lu.m, edit lu
help <command> spits output
doc <command> opens help browser
whos shows your variables and symbols and functions
type <command>. Like Unix’s cat.
Try double-clicking on a variable in your workspace
path

Internals: Functions, variables, types, control flow #

sum(a), min(a), max(a), median(a), var(a), prod(a), mean(a), size(a), mode(a), std(a)
false is 0. true is 1. Case sensitive!
isvector(a), isempty(a)
Not is a tilde ~, not ! (C-style)
Matrix indexing: Column 2: a(:,2); Line 2: a(2,:).
inv(a) for inverse
find (a < 3) filter
2^3 == 8
Quick matrix generation: zeros, ones, eye, magic, rand
Math: cos(pi), sin(pi), tan(pi), tanh(pi)
More math: rad2deg(pi) == 180, deg2rad
det(m), rank(m), trace(m), a,b = lu(m)
if, elseif, else, end. No parenthesis required.
for k=1:10; display('hey'); end
while
switch b; case 1; display('one'); end. No colon after cases.

Symbolic math library/toolbox #

syms x y z shorthand for x = sym('x')
x = sym('x', [2 1]) for a symbolic column vector
f = x + y^2 symbolic function
subs(f, [x y], [1 2]) evaluates the function
gradient(f), hessian(f), ccode(f) magic functions!

Fancy stuff #

diary on, diary off, help diary to record terminal sessions

— § —

Reply via email

#dev #legacy

★ i3 window manager

serendipity@perrotta.dev (Thiago Perrotta) — Mon, 17 Nov 2014 02:35:49 -0300

♠ I’m a huge fan of the i3 tiling window manager. It was the first ‘simple’ window manager I tried, and it is still my favorite one. It eases the management of your windows, is optimized for keyboard control, and maximizes all the space you have available, thus being a nice option for laptop users. i3 is packaged for most Linux and BSD distributions. You just have to install it through your package manager, then enjoy simplicity and agility. In this post I’ll share a couple of configurations and tweaks I use for my day-to-day computing and programming tasks on i3. First thing: the statusbar. My preferred one is i3blocks. A statusbar should be easy to config and tweak, and should not be bloated. Only the information you really care about should be displayed on it. Colors are also welcome, as they are a good design practice. Essential blocks include the battery of your laptop, the system load, the current date and time, and whether you are connected to the web or not. I also like to have a block which displays my current network speed (download/upload) and another one for system memory and swap. Optional things include volume and the currently playing song. Second thing: the background of your desktop. You’ll rarely see it, but it is nice to have something other than a black background fill color. feh is the best wallpaper setter for i3. A simple

feh --bg-scale '/home/thiago/Pictures/wallpaper.jpg'

sets your background image, and creates a $HOME/.fehbg shell script, which you can run to reload your wallpaper later on. Third thing: a i3lock which blurs the current view of your desktop is very nice. I like i3lock-wrapper from i3-extras repository by ashinkarov, available on GitHub. There are other interesting things there. j4-make-config is great for managing i3-themes, in the case you dislike the default one. It is available on http://www.j4tools.org/. Workspace management should be simple, so I don’t use names for them. Just numbers. When you use names, it becomes annoying when a given window is not in the proper workspace. While you can force them to go its proper place in your i3 config file, this process is more automatic than I need and it keeps getting in my way, so I decided it is better not to give names to my workspaces. If you use dual monitors, xrandr is your best friend. I like to have the following aliases in my shell rc file:

alias xrandr-t-hdmi-connect="xrandr --output LVDS1 --primary --output HDMI1 --auto --left-of LVDS1"
alias xrandr-t-hdmi-disconnect="xrandr --output HDMI1 --off"
alias xrandr-t-hdmi-mirror="xrandr --output LVDS1 --primary --auto --output HDMI1 --auto --same-as LVDS1"

arandr also helps when you need something a bit more advanced. Those set of configs makes me happy on i3. There are several more i3 customizations you can do, and I’ll leave a couple of references here. Other ones were already mentioned through the post.

Having a good window manager, the next step is to achieve a good shell environment (urxvt + tmux + zsh). More on that later.

— § —

Reply via email

#bestof #dev #legacy

Vim shortcuts everywhere

serendipity@perrotta.dev (Thiago Perrotta) — Thu, 13 Nov 2014 12:03:07 -0300

♠ Recently I’ve been switching my mind and workflow from emacs keybindings to vim ones, so there is nothing better than being able to use those shortcut keys on as many applications as possible. Currently ones I’m trying include:

vim itself
emacs with evil-mode
firefox with vimperator (there’s an Arch package for it)
chromium with vimium
luakit (KISS web browsing)
tmux
i3 window manager

Edit: forgot this one:

zathura (PDF/PS/DJVU viewer)

And I’m pretty sure there is plenty more to find out.

— § —

Reply via email

#dev #legacy #vim

btrfs trouble, trouble, and troubleshooting

serendipity@perrotta.dev (Thiago Perrotta) — Thu, 06 Nov 2014 00:01:43 -0300

♠ The last weekend I lost more than 8 hours trying to recover from a serious bug that reached the btrfs filesystem in recent days. This bug was relatively specific: it would only affect Linux users running a specific kernel version (3.17.1, if I recall correctly), a specific version of the btrfs-progs package, and systems with read-only snapshots — yeah, I’ve been using the snapshot feature of the btrfs filesystem extensively.

First Symptom #

After a pacman -Syu, my entire filesystem became read-only. I couldn’t save anything. So I rebooted it.

Then I found myself directly in the initramfs rescue prompt. I’ve always had nightmares with this prompt. I wouldn’t have access to all my files until I got rid of it.

A few Google searches alerted me that this was a severe/critical bug. However, fortunately (or unfortunately…), other beta and enthusiastic people also had it. The Arch Linux bug tracker and its discussion forums contained mentions about it. They served as an initial “what am I supposed to do now, besides screaming?”.

The Recovery Process #

I had backups, but they were a few weeks old. During this process, I fell in love with the grml Linux distribution. It is amazing for recovery processes and inspired the creators of Arch (the other inspiration is CRUX, which I still haven’t tried). It’s based on Debian/testing.

From grml, I logged into Hexchat (an IRC client) in the #btrfs channel, and within less than five minutes I got help and orientation from a guy who had the same problems. He told me what I was supposed to do, and it really worked! Now I have my system back.

Why It Took 8+ Hours #

It wasn’t that straightforward. Mainly because I was downloading several distros and dd’ing them onto USB Flash Drives from a grml virtual machine inside the Windows 8 desktop from my parents’ house. I did this several times until I realized (well, the #btrfs guy realized) that I needed a recent Linux kernel to do the recovery process.

What Linux distribution has a recent kernel? Arch, of course. The actual recovery process was done inside an Arch live medium.

Grml taught me several things to expect from a recovery distribution designed for system administrators. Unfortunately, its kernel was relatively old at the time (it’s based on Debian, so I can’t blame it).

The Takeaway #

Although losing several hours, I:

Discovered a nice Linux distribution
Was convinced (again) that IRC is still awesome
Could post on the Arch forums confirming the solution I used
Had fun in the process (YES. I love this sysadmin thing.)

Moral of the story: BACKUP YOUR SYSTEMS!!!! And sleep well.

I’m now using ext4 on my root partition and btrfs on my home partition. This means I can hibernate my system now — I missed that for a while.

References #

— § —

Reply via email

#dev #legacy

My last *nix setup

serendipity@perrotta.dev (Thiago Perrotta) — Tue, 21 Oct 2014 12:19:04 -0300

♠ I’ve just completed my objective with this blog. Should I start a new one? After looping through a BSD distro and dozens of Linux distributions, I’ve finally decided to settle down and stay away from new test drives. This time should be definitive—I really do not intend to replace my system anymore. The point is now I have a stable configuration and a few preferences that I would like to share with you.

The Setup #

I don’t intend this to take too long, so I’ll be brief:

Filesystem: btrfs instead of ext4. Snapshots are awesome, and I also can benefit from its SSD and compression performance.
Partition scheme: split your root (/) and home (/home) partitions. Always. Unless you’re in a hurry, or unless you do not intend to create a non-root user. Since I’m using btrfs, I also split my 300-500MB ext4 /boot partition. My preferred size for root is (30 ± 15)GB. /home takes up the rest. No LVM.
UEFI: I don’t use it now, but I would probably do that the next time I buy a new computer. Just take care of your ESP partition and disable secure boot.
Bootloader: syslinux! GRUB2 is way too automatic. Syslinux is simpler and faster. If you’re in a hurry, stick to GRUB2.
Linux Distro: Arch Linux. Period. If you’re in a hurry, use a derivative of Arch. If you’re even more in a hurry, use a derivative of Debian or Ubuntu. Linux Mint is my preferred choice, but Xubuntu or Lubuntu are good lightweight options.
BSD Distro: FreeBSD. I love its documentation and traditional Unix feeling. But it was the only BSD distro I’ve tested up to today.
Tiling window manager: i3. KISS, KISS, KISS. And elegant. With i3blocks, please.
Stacked window manager: Cinnamon.
Full-featured file manager: dolphin (or nemo, but only if you are using Cinnamon).
Lightweight file manager: thunar.
Terminal/text editor: vim.
GUI text editor: emacs.
Shell: zsh. With grml configurations, auto-completion and syntax highlighting. It’s perfect. Oh-my-zsh is a common alternative.
Terminal emulator: urxvt (rxvt-unicode). With tabs.
Monospaced bitmap font: terminus.
Monospaced scalable font: Source Code Pro (from Adobe).
System font: either Bitter, Cantarell, Ubuntu or Droid Sans.
Plug-in manager for emacs: el-get.
Plug-in manager for vim: pathogen.
Display manager: lightdm.

Conclusion #

This is the best Linux/BSD setup I’ve ever dreamed of, and I’m really happy. The original purpose of this blog has been achieved. Through 2 years, what I most posted here were concerns about Linux. The purpose was to share my own ignorance and growing maturity towards this awesome subject area.

Now my focus turns to Programming languages. Today I feel the same insecurity towards programming languages that I used to feel two years ago about Linux. I want to be very good at a few programming languages, and at least know a couple of them (maybe 5-10). I want to be part of a programming language community and evangelize programming.

This blog will still exist. I will probably keep posting here with the same rate I post nowadays, but I will probably talk less about Linux and more about other things. However, when I get into programming deeply, I might create a new space for that purpose.

If you are still reading here, thank you. I really appreciate the support of the people that express interest in this blog. I also hope that others could learn something during all this time.

— § —

Reply via email

#dev #legacy

From Emacs to Vim – Chapter #2 – Philosophy

serendipity@perrotta.dev (Thiago Perrotta) — Sat, 18 Oct 2014 23:47:21 -0300

♠ Previous post. Oh man, up to this point I am slightly disoriented about emacs shortcuts. It has been probably one month since I am using vim almost exclusively. I just use emacs to blog here and to use orgmode. Yes, I still know so many of Emacs shortcuts and keybindings that I can even recite them to you while I am dream. However, I am not practicing them so much as before, so they will probably fade away some time.

But I have used emacs for two years, so I won’t forget the basic commands, terminologies, functions and shortcuts, so no problem at all. Okay, so I’ve just finished reading Practical Vim. I super encourage you to buy the book and support its author, because this was (is!) a fantastic book! I also thanks Jason Ryan, forums and wiki admin from the Arch Linux community, for recommending the book. It is one of the reasons that made me commit myself to learn (and stick to) vim.

Now, in this post I’ll talk a little about the differences between emacs and vim. This is supposed to be an informal post, with information coming out of my head on-the-fly, so don’t expect much consistency or formality here. I’ll list a couple of topics/categories and tell you what text editor I most like in each of them.

IDE vs text editor #

Vim is not an IDE. I’ll repeat that. Vim is not an IDE. Vim is just a text editor. Don’t even try to make it a IDE. If you want a solid text editor to behave like an IDE, use and stick to Emacs. Emacs is not just a text editor; it is a full-featured text editor. Theoretically, it might be anything you want. There are no limits about the customizations you can do to extend emacs. The limit is the sky (actually, the memory of your current hardware).

I prefer vim here. For two years I tried to make emacs the perfect IDE. You can achieve good results with it (yeah, good ones!), however, it is best when you keep it simple. I like this aspect in vim, and now I understand why almost everyone of the Arch Linux community that I know prefer vim rather than emacs.

Vim is KISS. It is kiss in text editing. For example, vim defaults are very good; I can survive without a .vimrc. However, emacs defaults are terrible (from my point of view, and considering that I know it can be better). I simply cannot use emacs for much time without my .emacs file.

Emacs shortcuts mostly involve pressing <Control> or <Meta>, plus a letter or number key. Sometimes you have to do this (which I’ll call a chord) two or three times. For example, to quit emacs you press <C-x C-c>. Vim shortcuts usually involve only a few chars, but you have to switch to normal mode to activate them (most of the time, at least).

Now, which is better? That depends. From my point of view, emacs win in respect to speed. I usually issue a command faster in emacs than in vim. However, vim commands are better to navigate through text than emacs ones, because they are well integrated.

Furthermore, an extensive use of Emacs can lead to RSI (repetitive strain injury). Even Richard Stallman had it once. Vim commands are slightly less aggressive (in respect to the keyboard), so the health of your hands and fingers will benefit from them.

Some people also recommend the mapping of the caps lock key to <Esc> in vim, or to <Control> in Emacs. They claim that the caps lock key is not much used, so it is more useful when mapped to <Esc>/<Ctrl>. Well, I personally never accustomed myself to this convention. It has a natural problem of compatibility: every new computer you use won’t be configured this way. So I don’t find this very portable, and I don’t want to bring a .Xmodmap file everywhere.

So, who wins here? Emacs. However, remember what I’ve said: Emacs is faster, but it will hurt your fingers if you use it for prolonged hours.

Command-line #

Emacs command line wins. Sorry vim, that’s it. Short and clear. It is even more powerful with smex, ido and icomplete. At this point I don’t know any enhancement for the vim command line, but I have to say it is pretty good also. For now, I’m only using wildmenu. But emacs wins, period.

Movement of the cursor through a buffer #

Vim is better. Period. Like I’ve said before, it is very well integrated in this respect. You can switch your cursor to almost anywhere in your buffer with just a couple of keystrokes.

Now, in Emacs I am usually mechanical, and keep repeating myself. For example, if I want to go a few lines above, I’ll probably hit <C-u C-p> a couple of times. In vim, currently I press k a few times too, but this is due to lack of practice. There are faster ways of moving in vim than I am used to.

Buffer management #

The two editors are very decent in this regard. Since I am more used to Emacs, nowadays I find easier to manage emacs buffers; however, the vim way of navigating through them is pretty decent, so I won’t declare a winner in this category.

GUI (graphical user interface) #

It is almost impossible to use emacs for a couple of minutes if you are not using its X version. Its console version is just awful. In contrast, vim is perfect on the console, but I find gvim pretty bad (well, at least, I feel like I am editing the console when I used it: no relevant differences).

So the emacs GUI is better. It is probably even better on OS X, however I cannot confirm that right now.

Books #

Emacs books will usually teach you emacs lisp. It is very hard to learn about emacs without learning what is actually happening under its hood. When you realize that everything you do in emacs is just Emacs Lisp, you really find that very beautiful. A functional language being used for the greater good.

But this is also a problem: if you don’t care about elisp, you will probably get bored. Fast. Very fast.

Now, vim books usually teach you how vim commands are integrated with themselves and with vim itself. At least the books I’ve read up to today. This is better on a practical basis, when you want to quickly become good with it (you see, I already feel fluent in vim with just one month usage – I felt fluent in emacs only about a couple of months, and part of this reason was bad out-of-the-box defaults).

The point is: you don’t need to know about vimscript to actually get work done on vim. Therefore, vim books are better in respect to being pragmatic; buuuuut emacs books are better regarding teaching the inner workings of the editor, and how to program and extend it.

Out-of-the-box, better defaults #

Vim. vim. vim.

Ubiquity #

Vim. vim. vim. And vi. However, “micro emacses” such as mg (the program that Linus Torvalds uses) and zile are very good. But they are very basic, they are just micro text editors with emacs keystrokes; not much features.

One-line answers #

Color themes / colorschemes #

Emacs!

Plug-ins and extensibility #

Emacs is more extensible, but vim plug-ins give you less headache.

Communities #

There are more people using vim than emacs, but the emacs community is still good today. …and my creativity ran away now. Which additional topics do you believe that vim or emacs is better than its counterpart in? Comment.

— § —

Reply via email

#dev #legacy #vim

How do I blog with org2blog?

serendipity@perrotta.dev (Thiago Perrotta) — Sat, 11 Oct 2014 14:26:00 -0300

♠ TL;DR: Video link: Vimeo. This is my blogging workflow with Wordpress, Emacs and org2blog. Example markup:

* **bold**
* _italics_
* `code`

Code:

$ sudo apt-get update && sudo apt-get upgrade

A new section #

Text in the new section.

Another section #

Text in another section. Those are usually the formatting and commands I most use. There are others, of course, but this is usually a fast (and comfortable!) way of writing. You should also notice that Emacs automatically break lines for me! Now, let’s post this post. I’ll post the video link soon.

— § —

Reply via email

#dev #legacy #meta

Living on Ubuntu - Instance #2 - Creating a new package

serendipity@perrotta.dev (Thiago Perrotta) — Sat, 11 Oct 2014 14:14:52 -0300

♠ These are notes from a few weeks ago when I created a sample package for Ubuntu. This article is written both as a guide and a timeline. There is no explicit tutorial here, so get your feet wet by yourself, and make sure to follow the references. Edit: enumeration became a mess here. Sorry for that.

Initial setup #

Introduction; Packaging new software for Ubuntu; Debian Packaging Tutorial
Get a couple of packages for packaging (duh):

sudo aptitude install gnupg pbuilder ubuntu-dev-tools bzr-builddeb apt-file packaging-dev

Make sure you have a PGP key.
Send your PGP key to Ubuntu servers.

gpg --send-keys --keyserver keyserver.ubuntu.com <KEY ID>

Make sure you have a SSH key.

Building environment #

Create a clean build environment:

pbuilder-dist <release> create

At the time of this post, I used ’trusty’ as the release name.

Make sure you have a Launchpad account.
Update your Launchpad information and upload both your SSH and your GPG key into there.
Identify yourself with bzr.

bzr whoami "Bob Dobbs [subgenius@example.com](mailto:subgenius%40example.com)"
bzr launchpad-login subgenius

Creating the package #

Start a new package, and choose its type to create a new template.

bzr dh-make hello 2.7 hello-2.7.tar.gz

Explore and edit the files of your new template. In particular, see the debian directory.
bzr add and commit your changes. This is a DVCS such as git, so this should feel familiar.

Building and testing #

Building the package:

bzr builddeb -- -us -uc

The previous flags tell the system not to use GPG. -nc might also be useful: it tells the system not to start from scratch.

To view the contents of a package:

lesspipe *.deb

To install / remove your new built package:

sudo dpkg -i *.deb # attention to the architecture of your package!
sudo dpkg -r <name-of-your-package>

An alternative is to invoke sudo debi.

Check for errors.

lintian *.dsc *.deb

Optional testing with chroot #

Next (optional; just to test the package in the chroot you have set up before, with pbuilder)

bzr builddeb -S
cd ../build-area
pbuilder-dist trusty build *ubuntu*.dsc

Upload #

Upload your package to Launchpad:

dput ppa:<your-launchpad-username>/<your-ppa-name> *.changes

If you wanted to create a pull request instead:

bzr push lp:~<launchpad-username>/+junk/hello-package

And that’s it! Then just share your PPA with your friends and clients. I hope this guide could serve to you as a quick reference of creating a new Ubuntu (or even Debian) package.

— § —

Reply via email

#dev #legacy

Everyday Podcasting - #2

serendipity@perrotta.dev (Thiago Perrotta) — Fri, 10 Oct 2014 21:24:10 -0300

♠ Continuation of this post. Nothing fancy, really. Just to say that Bad Voltage is the most awesome podcast in the planet. Period. And goodbye.

— § —

Reply via email

#legacy #serenity

Compiling / Installing pacman on Ubuntu

serendipity@perrotta.dev (Thiago Perrotta) — Thu, 09 Oct 2014 23:20:52 -0300

♠ No, this is not the classic game. Here are logs.

git clone git://projects.archlinux.org/pacman.git
cd pacman/
./autogen.sh
./configure LIBS="-lpthread -larchive"

Attention here! ./configure without this parameter didn’t work on my machine. The error was: DSO missing from command line.

make
sudo checkinstall

Attention here! You could simply (and näively) do a sudo make install, however, this wouldn’t create a .deb package. Since we have good tools, we should use them!

Now, you can remove pacman anytime you want with dpkg -r pacman. No need for make uninstall.

Why this post? Because of this forum thread.

— § —

Reply via email

#dev #legacy

From Emacs to Vim - Chapter #1 - Vim plug-ins

serendipity@perrotta.dev (Thiago Perrotta) — Mon, 06 Oct 2014 01:50:00 -0300

♠ This is the continuation of the first post. As you get the feeling and overview of vim, the first thing you’ll probably miss as a probable-soon-to-be-emacs-orphan (actually, not really, I’ve already told you I’m not going to abandon emacs – for example, this post is being written within it, with org2blog – (too many level of parenthesis, please help me!!)) are your customizations and custom plug-ins and extensions. A couple of google searches would bring me to projects like janus and vim bootstrap. Those look nice and I would probably appreciate them if I were a beginner. However, as a super user – if you read this statement as bragging, then you are probably not understanding the purpose of this mini series –, I like to be in control of the personalization and settings of my text editor. Right? So, I’ve chosen to control my own plug-ins. In other words: bottom-up instead of top-down. Now, the two state-of-the-art plug-in managers you’ll find out there to manage your vimscripts are vundle and pathogen. I began using Vundle, but after a few hours I realized that pathogen was KISS and superior. Period. Not that Vundle is bad. It looks (works) okay too; in this case it was simply a matter of personal choice. The way Vundle works is as follows: your .vimrc will be populated with several lines such as:

Plugin 'tpope/vim-surround'
Plugin 'username-from-github/repository-name'
Plugin 'matchit'
Plugin 'vimscript-name-from-vim-dot-org'
Plugin 'git://www.absolute-path-to-my-site-here.com/repository-name.git'

Then, you just have to install those plug-ins with a simple command:

:Plugin Install

Done. It is simple, yes, but I liked more the approach of pathogen: you maintain a git repository with your plug-ins. Each new plug-in is a git submodule within your repository. Yay, this means it is super-easy to set up and share your plug-ins with other people. This is exactly what I’ve done: you can peek out my vim repository here. Now, there is one super good advantage here over emacs. The cost of adding a new plug-in is practically zero. This has two meanings, actually:

vim start-up time is ridiculously fast. Even with a dozen plug-ins, I still don’t notice it. With emacs, if I don’t run it as a daemon, it is a nightmare: a couple of seconds (yeah, a couple of seconds is really a nightmare, not kidding) of start-up time.
I don’t even have to touch my .vimrc to add a new plug-in. I just have to clone a new repository into my $HOME/.vim/bundle folder. Period. With emacs, I have to continuously poke my .emacs file to add, remove and constantly tweak my extensions. Yeah, I’ve been doing that for a while, and it does provoke a lot of stress and anxiety. You already have to tolerate a little bit of RSI with emacs, so you don’t need more problems.

Now, don’t undermine my (our) dear Emacs! Emacs Lisp will (and will always be) superior to vimscript. I have no doubt of this fact and therefore I won’t even discuss it here. Emacs is much more extensible than vim. Okay, then what? You have to find a few plug-ins, right? In Emacs, most plug-ins with decent quality you might find will be available either on MELPA or Marmalade and, of course, on GitHub too. In vim, it looks like the home of scripts is the vim website itself ((1) and (2)) and GitHub. I couldn’t find a central repo or website hosting the majority of its plug-ins, as with emacs. However, there is vim-awesome, which indexes existing ones (this concept is like the ruby toolbox, godoc.org and Emacs paradox extension by Bruce Connor. I won’t discuss the plug-ins I’m using right now, however here is a snapshot of my $HOME/.vim/bundle directory at the time of this post (19 plug-ins, yay):

You Complete Me
ack.vim
ag.vim
ctrlp.vim
delimit Mate
nerdtree
syntastic
tagbar
vim-airline
vim-bufferline
vim-commentary
vim-easytags
vim-fugitive
vim-gitgutter
vim-markdown
vim-misc
vim-repeat
vim-sensible
vim-surround

Most of those have an emacs equivalent. Okay, I’m stopping here for now, until the next post :) Update: next post.

— § —

Reply via email

#dev #legacy #vim

From Emacs to Vim - Chapter #0

serendipity@perrotta.dev (Thiago Perrotta) — Sat, 04 Oct 2014 15:55:56 -0300

♠ Hi there. This will probably become a new mini series. I begin by quoting Vanessa Da Mata:

É só isso Não tem mais jeito Acabou, boa sorte Não tenho o que dizer São só palavras E o que eu sinto Não mudará Tudo o que quer me dar É demais É pesado Não há paz Tudo o que quer de mim Irreais Expectativas Desleais

That’s it There’s no way It’s over, good luck I’ve nothing left to say It’s only words And what I feel Won’t change

[…]

Tudo o que quer me dar / Everything you want to give me É demais / It’s too much É pesado / It’s heavy Não há paz / There’s no peace

Upstream URL with lyrics. Now, a little bit of context. Two or three weeks ago I began learning vim. This is due to a lot of reasons: one of them is that vim is actively used by a lot of members of the Arch Linux community, and I usually am influenced about their tastes — what doesn’t mean that I will comply with everything, it is just a matter of trying new things.

Another reason is that I am actively learning new open source tools, and vim is definitely one of the most widespread and ubiquitous ones. I don’t need to provide statistical sources for that, do you agree? Vi is present everywhere, even on my Android. (Now, almost nobody uses vi anymore, however vi and vim are too much similar that for simple and rapid tasks usually annoying differences won’t even be noticed).

Another reason is that I never became 100% satisfied with Emacs. Hey, I like emacs very much, don’t get me wrong, however there were always some piece of completeness missing in it, so this puzzle never got finished. In particular, orgmode (an emacs task-manager-note-taker-and-everything mode) is a tool that I felt in loved with, and thanks to it I won’t completely remove emacs from my system; you should notice that orgmode is something very specific to emacs.

Although there are some tentatives on vim implementations, as far as I read, they are not very good (at least not compared to a vanilla, emacs orgmode). Okay, those are sufficient for now, but there are more reasons that will reveal themselves along the next posts.

Now, we should go back to the lyrics. I really love how this music is related to emacs; in particular:

Tudo o que quer me dar / Everything you want to give me É demais / It’s too much É pesado / It’s heavy Não há paz / There’s no peace

That’s is right. Emacs always was regarded to as the dinosaur, as the operating system of your computer, as the I-want-to-rule-all-your-workstation program. This has both its advantages and disadvantages, of course. The point with vim is that is tries to Keep it Simple. It will never be everything: it is only your text editor. Nothing more.

Well, although people will always try to write a Twitter Client for it and similar things, I feel it is not natural. No, IT IS NOT!! Now, with emacs, those things are pretty natural. It was created to be extended! (Emacs is described as the extensible text editor.)

Along the last two years, I’ve tested more than a hundred of add-ons and customizable pieces and bits for it: it is so cool how a functional language (Emacs Lisp) can improve a text editor. 2048 game, flappymacs, twitter client, mediawiki client, xkcd browser, RSS reader, calendar application (built-in), calculator (built-in), mail client … Yeah, emacs is everything! Maybe too much for me.

Now, this is the part that makes me feel a little sad: There’s no peace. Yeah, this is a problem. If you browse through my dotfiles repository on Github, more specifically, through its commit history, you’ll see that my _.emacs_ file is one of the files that most changed through the days. I am always editing something of it, and I am never completely 100% satisfied with the outcome.

Because it is to easy to tweak and customize almost every aspect of Emacs, this means it is also possible to customize some aspects that I should probably not even be bothering to (and I should probably not touch those ones). Now, this proved to be a problem…

Planned topics #

PROGRESS Emacs
- PROGRESS introduction, motivations, and so on
- TODO org2blog
- TODO functional programming, REPL and Emacs Lisp
- TODO keybindings, keystrokes and hotkeys paradise (or nightmare)
- TODO more org-mode goodness
- TODO lessons and overall learned discipline
- TODO about the awesome emacs community
- TODO emacs blogs and RSS out there?
- TODO evil-mode
- TODO Emacs on X vs Emacs on Console
- TODO Books and study sources on Emacs
TODO Vim
- TODO how a emacs user feels after being introduced to a modal text editor
- TODO keybindings and maps in different modes
- TODO gvim vs vim
- TODO vim community and resources?
- TODO editing text at the speed of thought, really?
- TODO books and study sources on vim
TODO Conclusions
- TODO Are text editors still relevant today, in a world of full-featured IDEs?

Whoa. I hope you might like this mini series and want to see some (or all!) of the above topics discussed. During it (please notice this comment won’t be repeated further on), feel free to leave out any comments about your specific experiences about vim or emacs. Also, you are encouraged to suggest me additional topics about my experience with emacs (and my soon-to-be-born experience with vim too), and even choose one of the above topics to I write about next. Until the next post! (For now, read this.) Update: next post.

— § —

Reply via email

#dev #legacy #vim

My first ebuild

serendipity@perrotta.dev (Thiago Perrotta) — Sun, 28 Sep 2014 03:24:00 -0300

Ebuilds are not evil

— Larry, the Cow

~~From now on, this will be my overlay repository: https://github.com/thiagowfx/overlay~~

I might change its name in the future, however it will probably remain on GitHub. From my experience with this blog, I realized it would be better to leave a copy of this ebuild here:

# Copyright 1999-2014 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header: $
EAPI="5"

inherit font

DESCRIPTION="A font specially designed for comfortably reading on any computer or device"
HOMEPAGE="http://www.huertatipografica.com/fonts/bitter-ht"
SRC_URI="http://www.fontsquirrel.com/fonts/download/${PN}"

LICENSE="OFL"
SLOT="0"
KEYWORDS="amd64 ~arm ~arm64 ~ppc ~ppc64 x86"
IUSE=""
HDEPEND="app-arch/unzip"

FONT_SUFFIX="otf"
S="${WORKDIR}"

src_unpack() {
  mv "${DISTDIR}/${A}" "${DISTDIR}/${A}.zip"
  unpack "${A}.zip"
}

src_install() {
  FONT_S="${WORKDIR}" font_src_install
}

My conclusion? Writing ebuilds is nice, very nice (yeah, more complex than writing PKGBUILDs, I know, but at least this complexity is justified).

Edit: Thanks Buss for pointing me out a small error (RDEPEND, instead of HDEPEND).

— § —

Reply via email

#dev #legacy

Cool but obscure unix tools

serendipity@perrotta.dev (Thiago Perrotta) — Sat, 20 Sep 2014 15:42:06 -0300

♠ An amazing list! Yet another one which deserves a post just for its usefulness. See http://kkovacs.eu/cool-but-obscure-unix-tools.

— § —

Reply via email

#dev #legacy

Living on Ubuntu - Instance #1

serendipity@perrotta.dev (Thiago Perrotta) — Mon, 15 Sep 2014 23:24:45 -0300

♠ As stated in my last post, I’m now using Ubuntu as my main linux distro. This is temporary and might even hurt a little, however I’m willing to keep going with this goal for a while.

In this post I’m going to document some cleanups and tips/tricks/whatever that I find interesting and, sometimes, important too.

Disclaimer: I am a heavily biased Arch Linux user. Anything related to it in this post is probably not a coincidence. This disclaimer won’t be repeated through other posts of this series. So, you have been warned. This is important to understand some of the motivations behind the changes I’ll list here.

Now, just one thing: one of the reasons I’m using Ubuntu right now is so I won’t spend much time customizing it. So, I am deliberately cutting this list in more than a half. There are other things that I would usually do but I simply won’t.

Enable the root account – by default, the root account is disabled on Ubuntu. This is great so newbies won’t harm the system; however, powerusers don’t like that, this is simply annoying for them. You might do that as follows:

$ sudo passwd root

Then just enter a password for it. Now you might become a superuser by invoking su.

Install aptitude. And never, I mean ever****ever touch apt-get. apt-get is evil and will break your system sooner or later. Top used aptitude commands are probably aptitude install, aptitude upgrade, aptitude remove and aptitude search. The -y flag is useful sometimes, too. So,

$ sudo apt-get install aptitude -y

Then forget apt-get!

Install unity-tweak-tool then change some annoying Unity defaults. What is annoying depends on each user taste.
I like to create a Ramdisk. How? Arch Wiki Ramdisk. This also works for Ubuntu.
Install the classicmenu-indicator package.
I like using a clipboard manager. clipit is a good choice and is well integrated with the Unity panel.
Add/activate workspaces. 4 of them is (usually) okay. Use Ctrl + Alt + directional keys to cycle through them.
Super + D shows the desktop. (Super is usually the key with the windows logo).
Ctrl + Alt + L locks the screen.
Install the nautilus-open-terminal package.
Improve your .bashrc. How? See here. You’ll also find my .bashrc there (search for thiagowfx).

— § —

Reply via email

#dev #legacy

What is taking much space on your Linux/BSD system?

serendipity@perrotta.dev (Thiago Perrotta) — Sat, 13 Sep 2014 21:13:52 -0300

♠ This program is so nice that I had to create a blog post just to tell people about it: ncdu. It is the most awesome and easiest (really easy! I would indicate it to any new Unix user as well) program to detect which folders/files are eating too much disk space on your system. Available for dozens of platforms. TL;DR and tutorial: just install, then type ncdu and wait for the best (worse?). See screenshots.

— § —

Reply via email

#dev #legacy

Legacy on ¬ just serendipity 🍀

Arch Linux on macbook installation notes

References #

Post-installation #

Linux goodness in your Chromebook

There be dragons #

See also #

Footnotes #

Installing Haiku from an existing linux

What’s Haiku in the first place? #

Installing Haiku #

Now what? #

[FAQFlow #3]: Hello, SaltStack (for Arch?)

Motivation #

SaltStack Hello World #

Footnotes #

Recording screencasts on Linux

Observations #

★ [FAQFlow #2]: Android Upgrading/Rooting Notes

Key steps #

Footnotes #

Running Ogre3d locally (without installing it)

Basic 1 #

Mulheres na computação, cadê vocês?

Getting started with Ogre3d

Installation #

1.1 Packages are usually outdated #

1.2 Some files are missing from the packages #

1.3 Some files are installed in non-standard locations #

Building with Arch #

Basic Example 1 #

Footnotes #

Documented dotfiles

[FAQFlow #1]: XMonad

Getting started #

Common questions #

1. XMonad is lightweight but GHC (Haskell compiler) is way too big (>400MB+). Do I really need it? #

2. I just started xmonad. But there is nothing on my screen—it’s blank all over the place #

3. How do I open a terminal? #

Bootstrapping #

How to change workspaces? #

How to change layouts? #

I don’t like xterm! It’s way too ugly! Can I change the default terminal emulator? #

Footnotes #

Thoughts on Instagram

Account deletion request

Thoughts and notes on deploying Wikimedia

Installation #

Footnotes #

ksnctf

From orgmode to asciidoc (OR: ditching emacs)

Footnotes #

Games: index of

Social networks for gaming #

C++ man pages

cppman — quick reference #

Installing it #

Using it #

Footnotes #

The Eudyptula Challenge

Setting up #

The task #01 #

What have I learned with this task? #

★ 2014 - Small review of a really nice year

The actual post #

Campus Party #

Arch Linux #

Programming languages #

MOOCs #

Presentations #

Books and reading #

Operating systems #

Games #

Music #

Computer graphics #

Vim and Emacs #

Hackathons #

Android #

Privacy #

Social networks #