Tag: linux

Google Authenticator and a finished robot

Last Tuesday, we got our robot shipped, or rather, put in a big clear plastic bag with big red zip tie on it. Although we shipped overweight, and missing a few components, this is better than last year on completeness at ship. I have a few pictures in my “Robots” Facebook album, showing off Ursa Major (our robot).

Last night, I upgraded my Debian VM to Squeeze, and set up Google Authenticator on it. The Makefile is kinda terrible, but it works as of now. I just scanned the barcode that loaded the secret key into the authenticator app on my phone, and now I can authenticate with time-based one-time passwords. I got it to compile on Mac OS X (some things are hacked up), but I don’t think it exactly works. I’ll have to dig in further to figure it out though.

Google Authenticator works rather well, although one thing I want is a timeout indicator so I know when my password is going to expire, so I don’t get half-way through typing it and it times out.

That is all.

ScriptBot 2.1 is in the works.

Just a quick post, telling everyone that a new version of ScriptBot is in the works, version 2.1.

So far, we’ve done the first xplugin -> bundled plugin conversion in ScriptBot’s short history, as well as added some little features that make a big impact, such as NickServ identifying, configurable connection port, configuration for whether to join on invite, and more features are of course on the way, in order to make this one of the biggest releases for ScriptBot (aside from 2.0, the massive rewrite).

I hope to also make ScriptBot available in the Fedora repositories, starting with version 2.1, but I’ll need assistance with packaging (see my earlier post on packaging ScriptBot). If anyone is willing to help, that’d be great. Just come on over to #bots on OKSnet. The current roadblock, as mentioned in the post, is packaging a library we use, written in Java, called PircBot.

Why Mac OS X Server can’t be taken seriously.

Mac OS X Server just can’t be taken seriously. Why?

A couple things:

  1. It can’t be run on commodity hardware, it has to be on a Mac.
  2. It can’t be virtualized.
  3. It doesn’t have server virtualization solutions designed to run on it.

To further that, Apple seems to take NO interest in virtualization. I haven’t seen one peep out of them about it. They simply sell desktop-class virtualization products in their store.

So if Apple wants to MAN UP and write some virtualization stuff, then they can be taken seriously.

I mean really. Virtualization is taking the world by storm, and Apple, the company that is supposed to be all hip and such is not participating. Every single product on the current Macintosh line supports hardware virtualization. Apple does nothing with that fact.

It’s sad. Just sad.

Fedora 12 Constantine

Yes, Fedora 12, Constantine has been released.  This release is focused on reinforcing stability, as well as making the desktop easier for new users. I’m hosting a release party on IRC, #Main on theoks.net.

You can read the new features in Fedora 12 on this wiki page: https://fedoraproject.org/wiki/Releases/12/FeatureList

Fedora Activity Day – Fedora Talk (Day 1)

Day One of the Fedora Talk Activity day is over.

Here’s some very quick wrapping up:

  • Icecast streaming is now working pretty decently, however the current test server is in… Germany. (http://publictest16.fedoraproject.org:8000/test.ogg)
  • Improvements on documentation, generally on the web site.
  • Troubleshooting tests are underway.
  • More “ghetto HTML” and such funnies.
  • People thinking my bag o’ chips is popcorn.

Things will continue tomorrow at I don’t know what time. I didn’t hear that part.

The conference has taken place on talk.fedoraproject.org, extension 2001.

Packaging ScriptBot for Fedora, a continuing adventure

You mostly likely know that I’m  a proud contributor to the Fedora Project. I helped out a bit on blogs.fedoraproject.org and I generally also keep the wiki clean. However, I haven’t ever made a package.

So my first packaging adventure is packaging the IRC bot library that ScriptBot uses.

One word: pain.

There’s a few nuances with PircBot. It comes as a jar with source inside it as well as the compiled classes. I extracted the source, made it usable for RPM to build. I scrapped the binaries, and put it into a traditional tarball. So apparently that doesn’t work.

Next, I found RPM wasn’t making a buildroot correctly. So I had to make a hack around that.

Then, it came to the install path. There’s a page on the Fedora Project Wiki apparently. But I didn’t know that! It’s not linked from the frequently visited Packaging pages.

After getting it to build, it went through a scratch build on Koji, then I filed for review on Bugzilla.

That’s when I found the page for Fedora’s Java packaging guidelines. I apparently did need documentation (although all I can put is a link, the documentation isn’t really pullable from the website). So I ended marking the bug for my request as CLOSED DEFERRED.

Expect a part two post when I actually get Pircbot into the repos.  Also likely is a part three when ScriptBot itself gets in.

See the ScriptBot project page at scriptbot.theoks.net.

Things around here might get a lot faster

I was able to compile the memcache PECL extension for my copy of PHP on my Mac. Now I just need a separate machine for memcached. When I get such a thing, OKS Blog would be much better, as queries from the DB would be cached through memcached, and memory is *much* faster than hard drive.

Click on a few ads (seen near the top of the page) to support my crazy venture of a separate memcached machine, and we’ll see how speed improves when we get a memcached machine up.

The memcached machine will be running a stripped Fedora 11 install (as in no GUI), with a 64-bit kernel (because it’s a whole new machine I’m gonna put together). I know that memcached is used in the Fedora Infrastructure with MediaWiki. Possibly, I might have the wordpress-mu install use memcached as well.

For those of you who don’t know what memcached is, memcached is “a high-performance, distributed memory object caching system, generic in nature, but intended for use in speeding up dynamic web applications by alleviating database load.” (from the memcached home page) memcached came from LiveJournal, a project of Danga Interactive. We’ll see if when we get a memcached instance (if we get a memcached instance) things speed up.