This blog consists of geeky computer-related content, as well as posts pertaining to music and jazz in particular. I realize that this makes for a rather odd mix. To make things easier, if you are interested in my music posts and don’t care about the computer stuff, you can find these posts here.
Kudos to Christos Manios for coming up with this solution!
The instructions provided by Yubico for integrating two-factor auth with OpenVPN lack a PAM configuration for Debian based systems. The following configuration file (/etc/pam.d/openvpn) worked for me:
auth required pam_yubico.so authfile=/path/to/yubikeys id=22010 debug
auth required pam_unix.so try_first_pass debug shadow nodelay
account required pam_unix.so
While the prerender.io team does not provide explicit directions for getting their service/NodeJS service to work with Loopback, it isn’t hard to do so. At the top of your top level bootstrap JS file, after the “app” variable is declared simply provide the following:
app.use(require('prerender-node').set('prerenderToken', '<your token here>'));
Then, from the prerender.io service dashboard, add a URL to your site manually (this URL needs to be world-accessible) and see if it caches the content correctly. Within my AngularJS controller I needed to provide “window.prerenderReady = false;” and “window.prerenderReady = true;” to indicate when all content was loaded, but this quirk may be particular to my setup.
You can also test the prerendering simply by appending “?_escaped_fragment_=” to the end of a URL. It may look different from your live site, but all text content should display.
I’m convinced that one could create some pretty compelling musical products and services in a similar manner to how one would build a tech startup, and that there are practical applications for musicians in thinking this way. Read more »
I get messages about needing to start a new backup frequently, and because I backup a lot of data this takes forever and is really annoying. I have found the following method for repairing a TM volume:
chflags -R nouchg /Volumes/[yourshare]/[whatever].sparsebundle
hdiutil attach -nomount -noverify -noautofsck /Volumes/[yourshare]/[whatever].sparsebundle
fsck_hfs -drfy /dev/disk3s2
In the sparsebundle file is a plist called “com.apple.TimeMachine.MachineID.plist”. Edit it and remove “RecoveryBackupDeclinedDate” and the corresponding date, and change “VerificationState” to 0.
You can also use DiskWarrior on Time Machine volumes from DiskWarrior’s file menu -> “Rebuild Disk Image”.
I had installed my Loopback app using Yeoman some time ago, and was confused by the fact I was missing server/middleware.json, which Strongloop’s documentation indicated I should have. I discovered that installing a blank template app using slc included this file (and presumably other support for middleware), and eventually figured out that my Yeoman Loopback generator was out-of-date. I had some difficulties updating the generator via “yo” (although I didn’t spend much time working on this), but managed to update it by running:
sudo npm install -g generator-loopback
Assigning a second network interface within the EC2 console and setting up the additional OS configuration is not terribly hard (I used the instructions here), but there was an extra catch to routing my EIP though this interface…
Read more »
I was trying to post to a model in the Loopback Explorer and certain fields referenced IDs from related models, only no matter what format my JSON was in my data was being inserted as an array where each index value corresponded to each letter in the string of text. I discovered that the cause of this was because my relationships weren’t setup correctly. In my case the “model” value in my relationship declaration in my model’s model.json config file wasn’t respecting the proper case sensitivity of the model.
When relationships are correctly established, the model schema displayed in the Explorer will show that the input type is “objectid” rather than simply “object”. With the input type set as “objectid”, you simply have to provide the ID of the related model entry, rather than
Hopefully this makes sense to somebody other than me
I was interested in migrating some servers from the older generation Amazon EC2 m1.small to the newer (and more cost effective) t2.micro instance types, but I had some difficulties in making this jump. Here is what I figured out… Read more »