Engineer, Solver of problems, Wisconsin Badger
After trying to sort out the best way to facilitate asynchronous communication for a group project, I was reminded of P2. This talk by Pete Davies at WordCamp SF this past year tells a very compelling story about why you should use P2 on your project. You can use it for free on WordPress.com or download it to run on hosted WordPress. It took all of a few minutes to sign up for a free WordPress.com site and switch to the P2 theme. After that it was all work.
Occasionally I need to send sensitive files to someone on the internet. I took a tip I once heard from Merlin Mann and built a script to automate this as much as possible. It requires Hazel and a cloud sharing service that syncs to your Mac. I’m using Dropbox.
The basic idea is to use Dropbox as a place to share files safely and automate the process with Hazel. We’re going to rely on long, obscure filenames along with randomly generated passwords and short availability windows to protoct our files. We can automate all of this through a couple Hazel commands and an Applescript.
The Hazel commands are set up as follows. The first one looks for any file or folder older than 7 days that is a zip archive and deletes it. This enforces our short availability window. You could make the window anything you like. The second command looks for anything that’s not a zip archive and runs the following Applescript on it.
We generate a random password with openssl rand -base64 32 — this means we’ll get 32 random base64 characters. We also create a SHA1 hash of the file to append to the filename. After that, just zip the file with the random password and put the password on the clipboard. You can do more, like generating a notification when the script is done. I have a TextExander snippet that lets you fill in the filename and grabs the password off the clipboard.
Ideally, you wouldn’t send the password along with the link to document. Try to use two different methods to notify someone where the file is located and what the password is — even if it’s just two different email accounts on a separate server.
For now this won’t work with directories because the SHA1 hash of a directory is blank. An easy way to fix that would be to zip the directory first and then get the hash and rename the zip file.
MG Siegler recently wrote about his choice to archive 50,000+ emails one night. And how he used that to decide to archive everyone once a week from now on.
A week ago, I came home after a long night of drinking and wanted to vomit. It wasn’t the whiskey. It was the email.
The way I deal with email is only slightly different from MG’s. The ammount of email I get is extremely tame, but I still think the workflow could help a lot of people. It’s something I’ve worked into after hours of listening to Merlin Mann, in various formats, speak about email and time management.
I use my email in very tight conjunction with Things. It could be any “to do” list style app. OmniFocus is another good one. I like Things for its simplicity.
There’s one rule that I have with email, only read it once. That’s not the perfect way to say it. Of course I read emails multiple times, but when I read and email there are three things that can happen.
When it’s something I don’t have time for, I’ll drop a link to the email in the notes field of Things if I’m going to need to reference it again. This obviously requires you have an active to do list that you keep up with. It might seem like you’re just pushing the problem out of email and into another list, but ideally your to do list is more organized that your email. If you read the same email six times before you do something about it, you were wasting your time the first five times. With a GTD type approach, you can quickly look at your list and know what needs to get done and what things you can currently do based on contexts.
This works for me with a small volume of email. I imagine it gets even better the more email you get. The more email you get, the more time you waste re-reading everything.1
MG writes about how relieving it can be to have the weight of 50,000 emails moved out of your inbox, to a place that you only ever see them when/if you need to. It’s funny because I’m the same way, though I don’t clear my RSS reader every night, it’s something I do regularly. And those red Push Notifications. I don’t think this is uncommon though.
It’s essentially out-of-sight, out-of-mind. I should have known this would be the case since I’m also obsessed with clearing my RSS reader every night (even though I barely use it anymore) and am a slave to clearing red Push Notification dots on the iPhone/iPad.
I only slightly disagree with the final point:
Read most of it. Respond to some of it. Keep all of it. But hide it. Then forget about it. And repeat. And repeat. And repeat.
I don’t keep all my email. There’s a lot of email that I know I’ll never need again. I delete them. Maybe you never get email like that.
This is a follow up to my post, Set up VirtualBox for Web Development, where I describe how to configure a VirtualBox VM with two NICs so that you can develop on a local VM wherever you happen to be. I’m going to describe how to enhance that with a shared folder between your guest and host operating systems so that changes can be immediately reflected with a need to “upload” them to the virtual server.
One of the advantages to working with virtual machines for development is having a sandbox to throw stuff in. Not having to install PHP or MySQL on your local machine is nice, and if something goes wrong, just wipe it and start over (or boot from a snapshot). But wouldn’t it be nice if you could save your files and have the changes instantly reflected on the VM without going through an app like Transmit? I’ve been working with CodeKit recently, and one of the nice features is that it will refresh the browser for you automatically when you save your changes, but if you have to “upload” the files to a VM every time you save, this feature isn’t quite as useful. So, let’s fix that.
A couple of things came together in an interesting way leading up to this post. I’ve been working with CodeKit for a few weeks and it had started to become obvious that my workflow was a little flawed. Uploading to the VM after every save was getting old, even with Dropsend from TextMate. I knew about shared folders between VirtualBox host and guest, but gave up after briefly looking into it because I didn’t know how to install Guest Additions via the command line.
Then, someone showed me Vagrant. It’s an awesome app for automatically provisioning “lightweight, reproducible, and portable development environments.” I had a problem with it though; my MacBook is old and slow. Part of the automatic process is using Chef to essentially set up all the apps you need to run your environment – PHP, Apache, MySQL, etc. Everything went smoothly until Chef started doing its thing, then the CPU would jump to 100% and everything would lock up. I could limit this to 75% of the the host’s CPU if I wanted, but that wasn’t the real issue. I never really gave it a chance to finish, but it took long enough that it would be completely impractical for me to wait that long every time I needed to provision a VM for a new project. So, back to plain VirtualBox.
As I mentioned, there is a way to set up a shared folder between the guest and host – Vagrant does this automatically for you, which is the main feature I was interested in anyway. Once you have the guest OS configured the way you want it, it’s only a few steps to create the shared folder:
First, install Guest Tools.
sudo apt-get install dkms
sudo apt-get install build-essential
sudo apt-get install linux-headers-$(uname -r)
sudo reboot
After the machine reboots, go to the “Devices” menu and click “Install Guest Additions”. This will essentially add Guest Additions as a CD-Rom, just like if you were using a desktop OS, but you have to mount it manually. Then, run the Guest Additions setup.
cd /media
sudo mkdir cdrom
sudo mount /dev/cdrom /media/cdrom
cd /media/cdrom
sudo ./VBoxLinuxAdditions-x86.run
If you haven’t already, you’ll need to shut down the guest to add a shared folder in the VirtualBox settings. From VirtualBox, click on the VM you want to set up, go to “Settings”, “Shared Folders”, and add a new share. Then, mount it with the following, where “yourshare” is the name you gave the shared folder.
mkdir /path/to/mountdir
sudo mount -t vboxsf yourshare /path/to/mountdir
Now you can just save your files into that folder and they will be shared on the guest as well. Eventually, when I get a new machine, I’ll try Vagrant again. This works excellent for me right now though.
Instead of mounting the shared folder directly to the path you need it at, you may want to use a symbolic link to take advantage of the auto-mount feature of VirtualBox. When auto-mount is turned on, the folder will be mount in /media as the name you set it up as in VirtualBox, but it will have a prefix of ‘sf_’ in front of it. To create the symbolic link, you would run the following command.
ln -s /media/sf_yourshare /path/to/mount
I’ve used many different development styles through the years. For a while I was a big fan of using Coda to develop live on the server. Recently I’m using more of Textmate and Transmit1. And, while I still like the how fast I can move while writing changes straight to the server, I don’t necessarily want to be doing anything that could potentially break a client’s website — even for a short time. The solution to this is pretty simple; run a virtual machine that acts just like the remote server. You can make your changes and when you’re sure it’s right, you can make one push up to the live site. This probably seems pretty obvious, but I’m going to talk about how I have my VM configured so that it works even on a strange network.
We want a virtual machine that has a static address that will work even if we pick up and decide to work out of a coffee shop for a day or, say, a warehouse in Northern California. It needs to act just like a production server, which means it also needs to be connected to the web.
/etc/network/interfaces. Make sure you leave eth0 as a dynamic interface and set eth1 as a static interface for your VirtualBox network. ScreenshotSomething I like to do, just to make this a little easier to work with is to set that address to a local domain in my hosts file. Something like local.dev seems to be popular, but you can name it whatever you want. Obviously if you name it google.com you’ll need to start using Bing or something. 😉
I realize this was a pretty quick overview of the process, so if there’s anything that was unclear, let me know in the comments or shoot me an email.