levlaz

In 2010, when I was living in Maryland I purchased the new B.o.B album that had just come out on iTunes and was listening to it while I walked two miles down Cherry Hill Road to target. I don't remember exactly why I walked. Perhaps my car was in the repair store. At target, I bought a new bicycle for the first time in many years. Walked down means I had to ride it up. This didn't last very long and I had an embarrassing walk of shame since I was not able to make it up the hill.

Back then, and even today, I know nothing about bikes. I am pretty sure I bought the worst possible bike for the occasion. It was slow, clunky, and felt like it would fall apart at any moment.

It's worth noting, that at the time I was working at the National Naval Medical Center which was under 10 miles from my apartment. 10 miles in beltway traffic can quickly turn into a 90 minute commute. The only other option was taking the train, which was in an inconvenient U shape. The train station was a few miles away and the bus to take you there was slow, also in some inconvenient letter shape, and overall the commute time was not much better.

Riding a bike to the train station was a viable option, and I ended up doing just that a number of times. The best part about this is that there was a bike trail directly next to my apartment that took you up to the University of Maryland campus and the metro station. This trail was beautiful, and there was even a creepy swamp straight out of a horror movie that would be filled with Silent Hill esque fog in the mornings.

The best memory that I have of this bike was the time that my Ford Focus broke down for good. I loved that car. It was the firs thing I bought with my first military pay check. I got it with around 7 miles on it brand new in 2007. I drove all over the east coast and the midwest in that egg shaped hot red car. I blew the speakers out listening to house music that Gerald introduced me to. I popped it in third gear one time and chased a woman down Wisconsin Ave in a fit of road rage when she cut me off one day. One of my friends joked to me that the moment I reached 45,000 miles the car would break down.

Damn that person. Literally the day I reached 45,000 miles my clutch went out on the beltway. It was the most frustrating experiences of my life. I somehow made it back to my apartment. This was one of the most memorable moments of that old bike. I rode it, in the middle of the winter, over ice, to a Honda dealership.

Let's make one thing clear. When you show up to a dealership on a bicycle in the middle of the winter, you just made the day of whoever is lucky enough to come talk to you first because there is no way that you are leaving there without a car. I got a Honda Civic. Also brand new. No clutch this time. That car, named Chester, is still around. My dad drives it these days.

In 2011, when I was preparing to leave Maryland and move back to Ohio I sold the bike to a University of Maryland engineering student for a fraction of what I paid for it. I remember watching him ride away into the sunset. That was the last time I rode a bicycle.

Maryland, Ohio, and New Jersey where I spent most of the last decade are not really big bicycle towns. San Francisco on the other hand is full of bike lanes, bike shares, and every morning you can see hundreds of cyclists commuting to work like a herd of gazelles down Market Street.

I remember one of the doctors that I worked with biked to work every day. Unlike the folks around here who do it unpretentiously, it was an entire event for him. He would wear the whole tight clothes getup, take a shower before he started to work, and then change into his uniform. Must be nice, who has time for that?

Ever since I moved here, I have been wanting to get a bicycle. A few weeks ago I asked my twitter followers to recommend a bike shop. My good friend, and co-worker, Tad made me an offer I couldn't refuse. Rather than recommending a bike shop or a bike model he gave me an old bike instead. Tonight, I finally got a chance to go pick it up and take it for a spin.

It was amazing.

By far, this is the best bicycle that I have ever ridden on. It has huge wheels. It takes very little effort to pick up speed. It's fast. Most of all, its fun. I felt like a kid again riding on that thing.

We rode to Golden Gate park to watch the awesome photosynthesis light show at the Conservatory of Flowers. I learned about The Wiggle and rode home from the Haight to SoMa. There was something truly amazing and freeing about biking home tonight. I saw the city in a whole different light.

Besides living in constant fear of my front wheel or seat being stolen, I cannot wait to take this for a spin all over the city. The first thing I want to do is finally make my way all around Golden Gate Park. That place is huge and walking around would take an entire day. I am too lazy for that. Naturally I am going to join the flock of tourists and take a ride over the golden gate bridge one of these days as well.

My last bike offered me so many great memories that I have not really thought about until now. I can't wait to see what adventures this new bike will have in store for me.

I want to give a public, heartfelt, humongous THANK YOU to Tad. He really made my day.

It seems like every year or so after some scandal or another dies down the developers at Ubuntu get bored and make another one.

The latest is this amazing new feature that uses motd ("Message of the Day") to show random information to the user. Most striking is what seems to be an advertisement for HBO's Silicon Valley.

Why? Who thought this was going to be a good idea and why do we find ourselves in this boat every other year?

I volunteered to be JLO (Java Language Owner) at CircleCI and I am currently working on getting a sample Spring Framework project running on CircleCI 2.0. I made a simple app bootstrapped with the Spring Initializer. I included Spring Security for the first time and I decided to try out WebJars for static Javascript libraries such as bootstrap. I am using Thymeleaf for templating.The app does not actually do anything yet but I ran into a pretty strange issue today that I wanted to write up here. My home page is pretty straightforward.

<!DOCTYPE html>
<html xmlns:th="http://www.thymeleaf.org">
<head>
    <title>CircleCI Spring Demo</title>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />

    <link rel="stylesheet" th:href="@{/webjars/bootstrap/3.3.7/css/bootstrap.min.css}" />

    <link rel="stylesheet" th:href="@{/css/style.css}" href="../static/css/style.css" />
</head>
<body>

    <nav class="navbar">
        <div class="container">
            <div class="navbar-header">
                <a class="navbar-brand" href="#">CircleCI Demo Spring</a>
            </div>
            <div id="navbar" class="collapse navbar-collapse">
                <ul class="nav navbar-nav">
                    <li class="active"><a href="#">Home</a></li>
                    <li><a href="#">About</a></li>
                </ul>
            </div>
        </div>
    </nav>

    <div class="container">
        <h1> CircleCI Spring Demo </h1>
    </div>

    <script th:src="@{/webjars/bootstrap/3.3.7/js/bootstrap.min.js}"></script>
</body>
</html>

However, when I tried to load up the app with mvn spring-boot:run none of the styles showed up and console showed the following error message:

Resource interpreted as Stylesheet but transferred with MIME type text/html

It turns out, that a default spring-security config will basically block any request unless you whitelist it. The MIME type is a red herring since what is actually happening is that my spring-security config is redirecting all unauthenticated users to my login page (which is login.html) instead of serving up the stylesheet from the /webjars directory. The solution is to update my security configuration to whitelist anything that comes from /webjars

package com.circleci.demojavaspring;

import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.context.annotation.Configuration;
import org.springframework.security.config.annotation.authentication.builders.AuthenticationManagerBuilder;
import org.springframework.security.config.annotation.web.builders.HttpSecurity;
import org.springframework.security.config.annotation.web.configuration.WebSecurityConfigurerAdapter;
import org.springframework.security.config.annotation.web.configuration.EnableWebSecurity;

@Configuration
@EnableWebSecurity
public class WebSecurityConfig extends WebSecurityConfigurerAdapter {
    @Override
    protected void configure(HttpSecurity http) throws Exception {
        http
            .authorizeRequests()
                .antMatchers("/", "/home", "/webjars/**").permitAll()
                .anyRequest().authenticated()
                .and()
            .formLogin()
                .loginPage("/login")
                .permitAll()
                .and()
            .logout()
                .permitAll();
    }
}

Now, the styles load as expected.

Netbeans is a great open source Java IDE. For some reason it is missing from the current stable repository on debian. In order to get it installed as a regular desktop application in Debian Jessie (using GNOME) you should do the following:

1. JDK 8 is required in order to use netbeans. The default-jdk package on Jessie installs jdk7. First you must enable debian backportsand then you You can install it with sudo apt install -t jessie-backports openjdk-8-jdk
2. Download the latest version from the releases page. There are a couple different flavors. I usually choose the one that contains everything. This will download a bash installer script.
3. Open up a terminal and navigate to wherever you downloaded the script from Step 2. Execute the script with sh netbeans*.sh
4. This will run some pre-flight checks and then fire up an installation wizard that will guide you through the rest of the process.
5. Once Netbeans has been installed you can launch it by clicking on the icon that should now be on your desktop.

Autotools is probably the most overwhelming piece of software that I have encountered. Just when I think that I get how it works, something goes wrong and I spend hours digging through man pages and info docs trying to figure out what is going on.

I wish I had found the Anjuta docs that describe how the "magic" behind their project wizards actually work earlier. Like most IDE's Anjuta takes care of doing a lot of heavy lifting in the background. Unlike most IDE's they have excellent documentation (in addition to the source code of course) on how everything actually works. This is extremely valuable and I am grateful to the folks from the Anjuta project who took the time to describe how all of this works from doing everything as a single line gcc command all the way to clicking buttons via the new project wizard.

If you are new to autotools like me, check out this doc. If you dont care about auto tools but want to see what excellent documentation looks like, check out this doc as well.

gtk-doc is a library that helps extract code documentation. When you create a new project with Anjuta it asks if you wish to include gkt-doc. Unfortunately, on Debian stable there seems to be a bug because the autoconf configuration is looking for the wrong version of gtk-doc.

/home/levlaz/git/librefocus/configure: line 13072: syntax error near unexpected token `1.0'
/home/levlaz/git/librefocus/configure: line 13072: `GTK_DOC_CHECK(1.0)'

On Debian stable, the version of GTK doc that comes with thegtk-doc-tools package is 1.21. In order to resolve this error you need to update configure.ac to use the newer version of gtk-doc as shown below:

GTK_DOC_CHECK([1.21])

Then you need to regenerate the entire project and everything should work as expected.

Anjuta is an excellent IDE specifically when it comes to writing applications for GNOME. On Debian stable, there seems to be a bug having to do with a missing dependency. When you create a project for the first time using the new project wizard and then try to execute it; Anjuta will complain that you must have libtool installed. I already have libtool installed, but it is looking specifically for some tools found in the libtool-bin package. Installing this package resolves the issue.

sudo apt-get install libtool-bin

The other day Aosheng send me an article to read from the verge. When I tried to read it, it took about 5 minutes to load because of the 15 various JavaScript things that were running in addition to ads loading in the background. Firefox was unhappy, and even when I tried to turn on "Reader View" (which strips out all of the junk) it took another minute to load. I've been on a UNIX binge lately so I figured there had to be a clever hack to make my own reader view in a terminal. This is where pandoc comes to the rescue. I've written about this tool in the past discussing how to easily convert Markdown to PDF. It turns out that pandoc also supports arbitrary URL arguments which means that you can convert HTML files on the fly without having to download them first. This means that we can take an arbitrary URL, pass it into pandoc, and spit out plain text. Furthermore, we can pipe this into less to get a nice pager for longer documents. The full string is shown below:

pandoc -f html -t plain
https://www.theverge.com/2017/5/4/15547314/edward-snowden-cory-doctorow-nypl-talk-walkaway
| less

In the example above, -f specifies the input filetype, in this case HTML. -t specifies the conversion filetype, in this case plain text. Pandoc supports a ton of different formats, you can read the man page for more info.

The next logical step is to make a script like my wordpress mutt poster to make this even easier. You could make a simple program called reader and put it in /usr/local/bin/reader. The contents of this script are:

1
2
3
4
5
6

#!/bin/bash
# Terminal Reader Mode using Pandoc and Less

url="$1"

pandoc -f html -t plain $url | less

You can then use this  by typing reader $URL.

Soemtimes you are hanging out in your terminal and you just want to be able to post something to your blog quickly. I was pretty inspired by Derek Siver's OpenBSD post [1] where he really embraces the unix philosophy of having one tool to do a job correctly and putting together various small tools like this to come up with a solution for the problem that you are trying to solve with your computer. Wordpress with Jetpack makes it dead simple to post to your blog via email [2], even if you do not have a mail server configured. I was able to write a three line bash script to "automate" creating a new post from my command line.

1
2
3
4
5
6
7

#!/bin/bash
# Bash Utility to Post to Wordpress using Mutt

subject="$1"
WP_ADDRESS=

mutt -s "${subject}" $WP_ADDRESS

I saved this file in /usr/local/bin/wp and whenever I am inspired to fire off some quick thoughts to this blog I can run wp "Blog Post Title" which dumps me into a vim buffer that once I complete is sent off via mutt to wordpress. [1] https://sivers.org/openbsd [2] https://jetpack.com/support/post-by-email/#examples

The Debian Policy Manual dictates that all packages should come with documentation. In order to save space in the debian archive these documents need to be compressed with gzip. There are a ton of these files floating around in the /usr/share/doc directory. Recently I wanted to read some of the documentation. If you try to open the file with cat it spits out binary gibberish. You can of course unzip the file as you normally would and open it up that way, but it turns out there is an easier way. Using zcat you can read the contents of compressed files just like you would with cat.

zcat is identical to gunzip -c. (On some systems, zcat may be installed as gzcat to preserve the original link to compress.) zcat uncompresses either a list of files on the command line or its standard input and writes the uncompressed data on standard output. zcat will uncompress files that have the correct magic number whether they have a .gz suffix or not. GZIP(1) man page.

By default, this will put all of the output into your terminal window, which is fine for most files. The other place where this can come in handy is when you are trying to look through compressed log files. In this case, having to scroll around the terminal may not be a great option. You can pipe the output of zcat into other programs such as less in order to be able to page through long files. For example, if I wanted to read the first 10 lines of a compressed log file, I could do so with the following command:

levlaz@debvm:/var/log$ sudo zcat syslog.2.gz | head -n 10

The output of this command would look like this:

May  2 22:27:43 debvm rsyslogd: [origin software="rsyslogd" swVersion="8.4.2" x-pid="585" x-info="http://www.rsyslog.com"] start
May  2 22:27:43 debvm kernel: [    0.000000] ACPI: LAPIC_NMI (acpi_id[0x1a] high edge lint[0x1])
May  2 22:27:43 debvm kernel: [    0.000000] ACPI: LAPIC_NMI (acpi_id[0x1b] high edge lint[0x1])
May  2 22:27:43 debvm kernel: [    0.000000] ACPI: LAPIC_NMI (acpi_id[0x1c] high edge lint[0x1])
May  2 22:27:43 debvm kernel: [    0.000000] ACPI: LAPIC_NMI (acpi_id[0x1d] high edge lint[0x1])
May  2 22:27:43 debvm kernel: [    0.000000] ACPI: LAPIC_NMI (acpi_id[0x1e] high edge lint[0x1])
May  2 22:27:43 debvm kernel: [    0.000000] ACPI: LAPIC_NMI (acpi_id[0x1f] high edge lint[0x1])
May  2 22:27:43 debvm kernel: [    0.000000] ACPI: LAPIC_NMI (acpi_id[0x20] high edge lint[0x1])
May  2 22:27:43 debvm kernel: [    0.000000] ACPI: LAPIC_NMI (acpi_id[0x21] high edge lint[0x1])
May  2 22:27:43 debvm kernel: [    0.000000] ACPI: LAPIC_NMI (acpi_id[0x22] high edge lint[0x1])

Many new and existing Debian users want to help make the distribution  better but do not quite know where to begin. Debian comes with a very handy package called how-can-i-help which tells you after each apt invocation the current bugs that are associated with packages on your system. The "Work-Needing and Perspective Packages" (WNPP) listing is a bit overwhelming for new contributors. What better way to figure out what packages need your help than by seeing a list of them each time you use apt.

The first time you run apt after installing this package it will likely spit out a long list of packages that need your help. Each subsequent time it will only show new packages or changes. In order to see the master list again you can use the how-can-i-help --old command to see all packages that need your help. I think this is a great way to get engaged with the software that you rely on each day.

Although getting started with Debian development is not trivial, this lowers the barrier a bit and provides some clear direction on what to work on since the list includes packages that you are using every day.

There is no official debian package for the nextcloud client. There have been a handful of RFP bugs reported but it looks like no one has taken this on yet. I want to get more involved with debian packaging so this might be a great first package to maintain. For the time being, the owncloud client is still backwards compatible with nextcloud. Unfortunately, the version that ships with Debian stable (8, jessie at the time of writing) is a bit old. When I tried to connect to my nextcloud instance it complained that my password was incorrect. Luckily, there is a slightly newer version available in jessie-backports  which has no trouble connecting to nextcloud. The steps to get a working version of owncloud-client to work with the latest stable version of Nextcloud are as follows:

1. If you have not already, enable jessie-backports
   1. Open up /etc/apt/sources.list
   2. Append deb http://ftp.debian.org/debian jessie-backports main to that file.
2. Run sudo apt-get update
3. Install the latest version of owncloud-client with sudo apt-get install -t jessie-backports owncloud-client

You should now be able to connect to nextcloud without any issues.

If you spend any time mucking around config files in Linux you are likely to run into some syntax errors sooner or later. Recently I was setting up cgit on Debian 8 and was banging my head against the wall for a few minutes trying to figure out why apache was so unhappy.

Symptoms

The key issue was when I restarted apache2 like I normally would after adding a new configuration it spat out an angry message at me.

root@nuc:/etc/apache2# sudo service apache2 restart
Job for apache2.service failed. See 'systemctl status apache2.service' and 'journalctl -xn' for details.

Troubleshooting

The first place that I would look is the error logs. However, in this particular case they were not very helpful.

root@nuc:/etc/apache2# tail -f /var/log/apache2/error.log
[Mon May 01 21:00:11.922943 2017] [mpm_prefork:notice] [pid 20454] AH00169: caught SIGTERM, shutting down

Next, I read the error message per the suggestion from the restart command. This was also not very helpful.

root@nuc:/etc/apache2# systemctl status apache2.service
● apache2.service - LSB: Apache2 web server
 Loaded: loaded (/etc/init.d/apache2)
 Drop-In: /lib/systemd/system/apache2.service.d
 └─forking.conf
 Active: failed (Result: exit-code) since Mon 2017-05-01 21:05:58 PDT; 1min 45s ago
 Process: 20746 ExecStop=/etc/init.d/apache2 stop (code=exited, status=0/SUCCESS)
 Process: 20697 ExecReload=/etc/init.d/apache2 reload (code=exited, status=1/FAILURE)
 Process: 20920 ExecStart=/etc/init.d/apache2 start (code=exited, status=1/FAILURE)

May 01 21:05:58 nuc apache2[20920]: Starting web server: apache2 failed!
May 01 21:05:58 nuc apache2[20920]: The apache2 configtest failed. ... (warning).
May 01 21:05:58 nuc apache2[20920]: Output of config test was:
May 01 21:05:58 nuc apache2[20920]: apache2: Syntax error on line 219 of /etc/apache2/apache2.conf: Syntax error on line 22 of /etc/a... section
May 01 21:05:58 nuc apache2[20920]: Action 'configtest' failed.
May 01 21:05:58 nuc apache2[20920]: The Apache error log may have more information.
May 01 21:05:58 nuc systemd[1]: apache2.service: control process exited, code=exited status=1
May 01 21:05:58 nuc systemd[1]: Failed to start LSB: Apache2 web server.
May 01 21:05:58 nuc systemd[1]: Unit apache2.service entered failed state.
Hint: Some lines were ellipsized, use -l to show in full.

Inspecting the error message, we see that it is unhappy with line 219 of the main /etc/apache2/apache2.conf file. Looking at that line we can see that it is simply loading all of the other config files in sites-enabled which means that before it even gets to load my new cgit config file it fails.

Help

So now that we have done some basic troubleshooting. It's time to dig into the manual for further information. I know that the config file is failing to load, and knowing my fat fingers it is very likely a config error on my part. Before reading 200 pages of documentation on the apache website we should take a look at the built in help to see if we can find something of value.

root@nuc:/etc/apache2# apache2 -help
Usage: apache2 [-D name] [-d directory] [-f file]
 [-C "directive"] [-c "directive"]
 [-k start|restart|graceful|graceful-stop|stop]
 [-v] [-V] [-h] [-l] [-L] [-t] [-T] [-S] [-X]
Options:
 -D name : define a name for use in <IfDefine name> directives
 -d directory : specify an alternate initial ServerRoot
 -f file : specify an alternate ServerConfigFile
 -C "directive" : process directive before reading config files
 -c "directive" : process directive after reading config files
 -e level : show startup errors of level (see LogLevel)
 -E file : log startup errors to file
 -v : show version number
 -V : show compile settings
 -h : list available command line options (this page)
 -l : list compiled in modules
 -L : list available configuration directives
 -t -D DUMP_VHOSTS : show parsed vhost settings
 -t -D DUMP_RUN_CFG : show parsed run settings
 -S : a synonym for -t -D DUMP_VHOSTS -D DUMP_RUN_CFG
 -t -D DUMP_MODULES : show all loaded modules
 -M : a synonym for -t -D DUMP_MODULES
 -t : run syntax check for config files
 -T : start without DocumentRoot(s) check
 -X : debug mode (only one worker, do not detach)

Success! It turns out we can run a linter on a specific config file using the -t flag.

Solution

root@nuc:/etc/apache2# apache2 -t -f sites-available/git.levlaz.org.conf
apache2: Syntax error on line 22 of /etc/apache2/sites-available/git.levlaz.org.conf: </VirtualHost> without matching <VirtualHost> section

Doh! Such a silly mistake with a missing </VirtualHost> closing bracket. Fixing this syntax error resolved the issue. The main takeaway for me is that the best part about most Linux tools is that they usually give you everything you need in order to succeed. We were able to troubleshoot and resolve this issue without resorting to google and running random commands that stranger posted on the internet 5 years ago.

Debian comes with a very handy utility called update-alternatives that helps to set default tools for various tasks.

It is possible for several programs fulfilling the same or similar functions to be installed on a single system at the same time. For example, many systems have several text editors installed at once. This gives choice to the users of a system, allowing each to use a different editor, if desired, but makes it difficult for a program to make a good choice for an editor to invoke if the user has not specified a particular preference.

On Linode, it seems that the default editor is nano, I prefer to use vim for editing git commits, visudo, and other things that use the default editor which is symbolically linked through /usr/bin/editor. The update-alternatives package basically changes the symbolic links for you. In order to change your default editor, you simply need to run the following command:

sudo update-alternatives --config editor

The output of this command is shown below. You will see a list of all of your editors that you currently have installed and will be asked to make a choice.

There are 3 choices for the alternative editor (providing /usr/bin/editor).

Selection Path Priority Status
------------------------------------------------------------
 0 /bin/nano 40 auto mode
 1 /bin/nano 40 manual mode
 2 /usr/bin/vim.basic 30 manual mode
* 3 /usr/bin/vim.tiny 10 manual mode

Press enter to keep the current choice[*], or type selection number:

Behind the scenes you can see that all this does it updates the symbolic links.

levlaz@dev:~$ ls -al /usr/bin/editor
lrwxrwxrwx 1 root root 24 Feb 10 20:49 /usr/bin/editor -> /etc/alternatives/editor
levlaz@dev:~$ ls -al /etc/alternatives/editor
lrwxrwxrwx 1 root root 17 Apr 28 18:56 /etc/alternatives/editor -> /usr/bin/vim.tiny

There are many other things that can be configured this way. For more information reading the man page for update-alternatives is worthwhile.

I spent way too much time troubleshooting an issue I was having with sed today. This is a common theme sometimes where I spend upwards of an hour debugging something that is ridiculously obvious.

I was trying to replace a string in a file. This is super simple to do with sed.

sed 's/string/replacement_string/` $file_name

The problem is that this command just spits out the result. If you want to actually save your change you must use the -i flag.

sed -i '/s/string/replacement_string/` $file_name

For more information on how to not suck at Linux like Lev please refer to man $command ;)

TIL while working on this issue that if you dump a file into public/index.html then rails will just serve that up for you. A lot of tutorials out there talk about making a static pages controller, which seems like overkill (unless you have a lot of static pages .. but if thats the case why are you using Rails?)

Wow, I spent so long trying to figure out why the hell karma was not working for me, it turns out its because it was installed globally. For instance. In my projects package.json I had:

"scripts": { "test": "karma start karma.conf.js" } ...

When I ran npm test - it told me sh 1: karma not found Every other possible combination also did the same thing. i.e.

node_modules/karma/bin/karma

./node_modules/karma/bin/karma

node ./node_modules/karma/bin/karma

I could totally execute this myself from the shell, so I had no idea what was wrong. Then I finally stumbled upon this GitHub Issue. After uninstalling karma globally, npm uninstall -g karma I was able to run npm test without any issues. I still have no idea why this works or didn't work. But at this point I just want to go back to writing tests.

Similar to a previous post where I wrote about how to run flask tests without installing your app, another common thing that you might want to be able to do is import your app from some arbitrary script. This is especially useful when running your app with apache mod_wsgi. This module expects the app to be installed globally or at least in the python path. Unless you install the app in a traditional sense this will not be true. The solution is just to inject the path prior to running your import statement like this.

sys.path.insert(0, '/var/www/blog')
from blog import app as application

This import will actually work.

Who knew that flask had a cli? Previously I used to just use manage.py just like Django does it to "do stuff". The CLI is great, but again it follows the theme of kind of wanting you to install your flask app. (I really should do this). So in order to get your app to work you must point the FLASK_APP variable to the actual python file (not your app module). This is true even if you have a true python module. For instance.

export FLASK_APP = blog/blog.py
flask run

Works, while

export FLASK_APP = blog
flask run

Does not. Even though blog consists of:

blog/
  __init__.py
  blog.py

The other day I read a post on Hacker News about premature optimization in web application development. This was an excellent post in its own right, but one thing that jumped out at me was the call to action from the author to follow him on Twitter at the end of his post. I liked that it was subtle, simple, and effective  (I followed him on Twitter). Naturally, I stole this idea for my own blogs. This method uses the Follow Button provided by Twitter. In this post, I will show you how to add your own "Follow Me on Twitter" Call to Action on  your own Wordpress or Ghost blog.

Wordpress

I used the Bottom of Every Post Wordpress Plugin because I found that hacking the main wordpress loop caused my CTA to show up at the very bottom of a page rather than at the end of the post content. In addition, I made a small change to this plugin to make sure that it only appears on single posts rather than on the home page. Edit bottom-of-every-post/bottom_of_every_post.php from the Wordpress Plugins Editor and update the method to be:

if( is_single() && file_exists( $fileName )){

 /* open the text file and read its contents */

 $theFile = fopen( $fileName, "r");
 $msg = fread( $theFile, filesize( $fileName ));
 fclose( $theFile );

 /* detect the old message in code to try and eradicate my name and #
 showing up on strange websites that are run by lazy people */

 if( $msg == "<p>Call for an estimate 724-498-1551<br><a href=\"mailto:corey.salzano@gmail.com\">corey.salzano@gmail.com</a></p>" ){
 $msg = "<p>Thank you for installing the Bottom of every post WordPress plugin. To find out how to change or remove this message, read <a href=\"http://wordpress.org/extend/plugins/bottom-of-every-post/installation/\">the instructions</a>.</p>";
 }

 /* append the text file contents to the end of `the_content` */
 return $content . stripslashes( $msg );
 } else{

 /* if `the_content` belongs to a page or our file is missing
 the result of this filter is no change to `the_content` */

 return $content;
 }

The key here is

is_single()

Next, edit bottom-of-every-post/bottom_of_every_post.txt and add your call to action. Mine looks like this.

<p>
 If you made it this far, you should probably follow me on twitter. :) 
 <a class="twitter-follow-button" href="https://twitter.com/levlaz"> Follow @levlaz</a>
</p>

<script>window.twttr = (function(d, s, id) {
 var js, fjs = d.getElementsByTagName(s)[0],
 t = window.twttr || {};
 if (d.getElementById(id)) return t;
 js = d.createElement(s);
 js.id = id;
 js.src = "https://platform.twitter.com/widgets.js";
 fjs.parentNode.insertBefore(js, fjs);

 t._e = [];
 t.ready = function(f) {
 t._e.push(f);
 };

 return t;
}(document, "script", "twitter-wjs"));</script>

You only need to replace your twitter username in the above example and it should work as is. The end result is shown below: {.alignnone .size-full .wp-image-650 width="527" height="89"}

Ghost

I used ghosts code-injection tool to get this to work for every page. I added the following two scripts to the Blog Footer.

<script> 
 article = document.getElementsByClassName('post-content');
 child = document.createElement('p');
 cta = '<p class="follow"> If you read this far, thank you! Follow me on Twitter to stay up to date on what the fuss is all about.<br />';
 link = '<a class="twitter-follow-button" href="https://twitter.com/tralevnet"> Follow @tralevnet</a></p>'
 child.innerHTML = cta + link;
 article[0].appendChild(child);
</script>

<script>
 window.twttr = (function(d, s, id) {
 var js, fjs = d.getElementsByTagName(s)[0],
 t = window.twttr || {};
 if (d.getElementById(id)) return t;
 js = d.createElement(s);
 js.id = id;
 js.src = "https://platform.twitter.com/widgets.js";
 fjs.parentNode.insertBefore(js, fjs);

 t._e = [];
 t.ready = function(f) {
 t._e.push(f);
 };

 return t;
}(document, "script", "twitter-wjs"));
</script>

In the example above you need to change your call to action as well as your twitter username for it to work. I also added a custom style to the Blog Header

<style> 
.follow {
 line-height: 1.5;
 width: 50%;
 padding: 15px;
 border: dashed 1px lightgrey;
 font-size: smaller;
 color: #555;
 font-family: sans-serif;
 font-weight: bold;
 }

 .twitter-follow-button {
 margin-top: 10px;
 }
</style>

The end result looks like this: {.alignnone .size-full .wp-image-659 width="388" height="186"} That's pretty much it. You can go wild with the style to your hearts content. I like this simple CTA and I am excited to see how effective it is on my own blogs.

For some reason there is a whole thread on this seemingly simple tasks. In a bootstrapped Laravel 5.4 instance the following worked for me.

Install Font Awesome with NPM

Import font-awesome in your app.scss file

Copy the fonts to public directory

Adding the following to your elixir config in the gulpfile

My complete gulpfile looks like this:

Run Gulp

If you run gulp you should be able to now start using font-awesome everywhere in your app.

Docker is awesome. Its also quite useful for local development. The following Dockerfile and docker-compose.yml will be helpful if you want to do laravel development inside of docker. I am using Ubuntu as a base, but you can probably use the official PHP image as well. Dockerfile

FROM ubuntu:16.04

RUN apt update
RUN apt install -y php7.0 php7.0-zip php7.0-mbstring phpunit curl php7.0-mysql

RUN curl -sS https://getcomposer.org/installer | php 
RUN mv composer.phar /usr/local/bin/composer

RUN composer global require "laravel/installer"

RUN export PATH=$HOME/.config/composer/vendor/bin:$PATH

docker-compose.yml

version: '2'
services:
  app:
    build: .
    ports:
      - "8000:8000"
    volumes:
      - .:/code
    env_file: .env
    working_dir: /code
    command: bash -c 'php artisan migrate && php artisan serve --host 0.0.0.0'
    depends_on:
      - db
  db:
    image: "mysql:5.7"
    environment:
      - MYSQL_ROOT_PASSWORD=password
      - MYSQL_DATABASE=$your_db
      - MYSQL_USER=$your_db_user
      - MYSQL_PASSWORD=$your_db_password
    volumes:
      - ./data/:/var/lib/mysql
    ports:
      - "3306:3306"

.env file Your .env file is what Laravel uses when it starts up set up various things. The only real thing to change is your DB connection info. A full sample is shown below:

APP_ENV=local
APP_KEY=$your_app_key
APP_DEBUG=true
APP_LOG_LEVEL=debug
APP_URL=http://localhost

DB_CONNECTION=mysql
DB_HOST=db
DB_PORT=3306
DB_DATABASE=$your_db
DB_USERNAME=$your_db_user
DB_PASSWORD=$your_db_password

BROADCAST_DRIVER=log
CACHE_DRIVER=file
SESSION_DRIVER=file
QUEUE_DRIVER=sync

REDIS_HOST=127.0.0.1
REDIS_PASSWORD=null
REDIS_PORT=6379

MAIL_DRIVER=smtp
MAIL_HOST=mailtrap.io
MAIL_PORT=2525
MAIL_USERNAME=null
MAIL_PASSWORD=null
MAIL_ENCRYPTION=null

PUSHER_APP_ID=
PUSHER_KEY=
PUSHER_SECRET=

Gotchas

1. In order to do stuff with the database you should add the following record to your local /etc/hosts file

   # /etc/hosts
   
   127.0.0.1 db
   

2. You should still install npm and run npm install from your local machine so that you can do frontend stuff.

3. Since we define - .:/code as a volume, this means that all of your local changes are immediately visible in the dockerized app.
4. If you need to access the running app or db container you can do so with docker-compose run app bash or docker-compose run db bash

I have been using safari lately, for whatever reason. While going through the second rails course on coursera I came across this handy extension that pretty prints JSON in safari. JSONView I think this might be built into firefox or chrome because I have seen this before, but its super handy if you spend a lot of time looking at json in the browser.

The Flask Docs have a great section for testing. However they assume that you have the app installed with pip install -e . which I almost never do. (Maybe I should start?) I have had trouble with this approach. So one little hack to inject your app into the searchable python path is to do something like this:

export BLOG_PATH=$(pwd) && python tests/blog_tests.py

We use \$(pwd) so that this will "just work" no matter which computer you run it on (i.e. test, dev, other persons dev) Then in your tests/blog_tests.py file you import your flask app like this:

sys.path.insert(0, os.environ.get('BLOG_PATH'))
from blog.blog import app, init_db

You can see the full details here: https://circleci.com/gh/levlaz/blog My biggest issue with installing the app in a traditional sense is that when I run it with apache mod_wsgi it does not seem to actually copy over the right folders and apache is not able to find templates or static files. So that sucks, and I am probably doing something wrong. But this approach works more or less no matter what so sometimes a hack is better than derping around with no solution (as I did for six hours yesterday).

I finally finished reading "Getting Things Done" by David Allen. This has been on my reading list for years so I am glad that I finally got a chance to scratch it off of that list. Overall, it was a good read and I learned how to approach an overwhelming number of tasks with Allen's proven methodology. My biggest takeaways from the book were:

1. Get things out of your head and somewhere where you will look at them later. Big or small, short or tall, write it down.
2. Identify what success, or "done" actually looks like right away.
3. Identify the next step instead of worrying about the scope of a large project.

I've been using Kanboard to manage my day to day work for both professional and personal projects. Before that, when I was using OS X I used a program called OmniFocus which does an amazing job at allowing you to capture items from any context. Using a simple shortcut (Super + Space) it let you get things out of your brain quickly.

No other tool that I know of does this, and its a real shame because the biggest barrier to feeling relaxed about the pile of things you have to do is being able to trust that a specific item is going to be looked at again from any context. When adding a task to an app feels like work (i.e you have to go to a webpage, open an app, etc) then you may not do it.

I attempted to reproduce the magic of OmniFocus with a simple desktop app that I wrote called TaskAdder. When mapped to a keyboard shortcut (Ctrl + Space for me) it lets you add a task to your Kanboard from any context. Using this app for the last few weeks while reading GTD has changed my life.

Overall, the book was great. My only gripe is that it was a bit verbose. Many chapters repeated the same ideas, and the same examples. In addition, although these methods could apply to any human being a lot of the examples and anecdotes that Allen offers come from big wig executives who have secretaries, offices, and enough money to afford his one on one consulting work. My eyes began to roll after the third time that I was reminded to talk to my secretary (which I have never had) about helping me with my workflow.

If you don't like reading verbose books, I still think that looking into the GTD methodology is worth doing. The main website is full of great examples, diagrams, and resources.