Mitechie.com

I had a pretty depressing conversation on Halloween. I was at a part with the head of a local community college's computer dept. I guess they've integrated the IT/CS-ish sides of things. I had wanted to ask how one goes about teaching a class as it's something I've thought might be a fun/interesting side project to do at some point. I know a couple of topics well enough to teach these days.

I found out they only offer one Linux course in their whole program. Then we got talking back-end and she mentioned that they're moving off of Linux for computers/servers the college runs on. What?! Why in the world would you do that? She made the comment that they can't find/keep Linux admins. So it's easier for them to move things to Windows servers and have the people on hand to run them.

This leads me to the chicken and the egg thing though. I mean, if you offer one Linux course, how do you train Linux admins? I only had one/two in my degree days, but it was enough to make me follow through with learning a lot more on my own. With so little exposure, it's not surprising they can't find the people. It's also not like this school is in an area where a ton of professionals would pick to live. Flint, MI isn't exactly my first choice to be.

It's definitely sad to see that even somewhere you might consider a "linux friendly zone", the education world, is a place where it's not always welcome.

We had another fun MiDevelopers meeting tonight. Our usual turnout of 7ish people. The new location worked pretty nicely, except that the wireless didn't work. Good food/drink though. It did allow me to test out the JoikuSpot application on my E71, which turns the phone into a wireless AP. It looks like Intrepid's version of Network Manager will do the ad-hoc connection ok now and a couple of guys were able to use my phone as an internet connection. Very cool to know.

I have to apologize for the iPhone hating all night. I tend to rant very well and had many opportunities to do so. When you feel strongly about something, can't help it right?

We had a nice influx of Ubuntu users so there was a mini-Intrepid release party going on. Half the group had upgraded and we were testing out/showing off some things. A couple of guys had never used Gnome Do so we had to set them straight.

On the programming side, we had a new developer show up. Matt joined us and was from the Farmington-ish area. Another PHP developer in the group, but also has aspirations of checking out this "Python" thing. We were all more than happy to explain to him how lovely life will be after giving some time to Python education.

Adam brought in his new netbook toy to show off. Very sharp looking Apire, I think it was. We also spent some time debating on how best to design a Tomboy replacement. Something that could exist on the web, on the desktop, and have access to via cell phone. The key though is to try to mimic the interaction and ease of use Gnome Do + Tomboy provided. A tough call. The conversation definitely has me itching to play with some code though. Some really good ideas tossed about.

Sorry for the late notice on this month's meeting. I think we'll hold the next one at the same location and bump up the date to be the Thurs before the holiday. So mark it on your calendars now. Have to block off that time to drink and talk code/linux.

It's pretty sad that Jorge got it blogged out there first, but I finally have in my copy of my Intro to Bazaar article I wrote for Python Magazine. I started editing for them a while back and it's been a really fun experience. I learn a lot reviewing articles, it increases my time spend playing with Python, and every once in a while you get to do something like write an article. I'd like to thank 'lifeless' in irc for giving it a look over so I didn't make a fool of myself along with the Python Magazine team for letting me work on my path to those riches I keep hearing writers talk about.

Wait, that's a running author joke? Oops.

The follow up, coming out in the November edition, is one I'm a bit more proud of as it has a bit more information. Make sure you get your Python Magazine subscriptions in order. There are some great authors putting out some awesome material.

I ran into this post talking about how he's always uploading files for people to see. It really hit home. In the past few days I've been uploading some custom tweaks to the Gwibber ui along with my irssi config for people to see. I launch my ftp client, upload the files, wheee. The idea in that blog post is that you could just have it things prewired to just launch any file you wanted and then return a link to that file.

That's cool and all, but I have a couple places I upload things. I have a different spot for work/home things. I also was reviewing an article on Amazon Web Services for PyMag and I got thinking, how could would it be to have the script be able to upload to S3 as a publicly available flie, and then get a link back to it.

So I started working on a new app I'm calling sendoff. It's a lot more complicated than that bash script, but it allows you to define multiple server locations, have sane defaults, and there's more to come. A lot more. I need to figure out how to display progress information, get hooks for S3 and FTP uploads, etc. Right now it only supports scp via the paramiko library (very cool btw).

It's already very cool though. I can now upload by default to my home server with

sendoff filename.jpg

and I can do it to work with

sendoff -h work filename.jpg filename2.jpg

So far it's been fun. My first time creating an easy_install-able package from something. So take a peek at it if you want, let me know if you see anything in particular. It's a lot of firsts in there so I know it's a mess and I'm working on cleaning it up bit by bit. Looks like I've found a better project for MiDeveloper meetings for a while.

What a weekend. I started off heading down to PyOhio. What a great little conference day. The place was packed and will probably need a larger venue next year. My only regret was that I didn't grab a drink before I left the hotel since nothing really opened up for a bit.

The talks were really quality stuff. I learned a couple of things that I've been really needing to play with, creating generators and decorators specifically. I loved the lightning talks while lunch was going on. There were some pitches, but also some interesting things. Mark Ramm did some great work representing the Michigan crowd by not only wearing his TG2 hat, but also channeling some Kevin for the Paver talk. All in all, it was definitely worth the drive and I can't wait for next year.

A couple of things that could be better though. The first was my fault. I should have stayed the night of the conference as well as the night before. We ended up bailing a bit before the end in order to hit the highway and get home before 11pm. Part of that is Michigan's lovely construction traffic though. The other thing I had wanted to do was get to know/meet some of the guys around. Since the talks were go go go, and lightning talks at lunch, I never really got a chance for that social interaction. It was much more classroom-ish. I think it'd be good to try to include some social time or something next year. This might be tied more with the first point where I should have stayed around longer and maybe tried to do dinner with some guys.

Make sure you mark it on the calendar when it comes around next year.

So I'm only just over two weeks late with this. I need to review the last MiDevelopers meeting. We met up at the Hamlin Pub which worked out great. We had wireless, power, and some liquid refreshment available to us. It was the first meeting and turnout was more along what I had expected initially. There were several people that had wanted to come that had conflicts in the end, but we got six of us together for the evening.

I spent most of the time trying to get the bzr-avahi plugin going so I could share my Django project with Craig who came. After finally giving up and just going the old fashioned "bzr serve" route we managed to get my repo over to Craigs EEEPC. The fun really got going then. Installing Postgres on a EEEPC seems funny, but it works! However, setting up virtuanenv on there, not so well. So no Django hacking today.

So the event wasn't a big hacking success. Several guys showed off some things, and there was a lot of discussion. So I think everyone had a good time. We ended up bailing about 10pm before the music got on stage. It gets just a bit loud.

Final vertict, it's a work in progress. We're definitely going to have another one. The last Thurs of this month is the 31st. I'd like to have this one somewhere around the Royal Oak area. Maybe over by Woodward if there's something. So let me know where there's a place that has

1. Room for about 5-10
2. Wireless
3. Beverages
4. Power outlets somewhere in there

I'm down to the last chapter of my second Django book, Learning Website Development with Django (very good by the way, a nice compliment to the official book). I've been pushing at getting it done because I have a few other Python related books pre-ordered right now. The book on SqlAlchemy is on the way so have to finish this one up soon. The other reason for crashing through it is that with my new MiDevelopers meeting, I want to get a project up and going so that I might work on it with some guys on Thursday. So I decided to rework my old spam thing. Some screenshots of the old thing are here. Basically, the idea is I quarrentine email messages marked as spam. In case of a mistake, I pull them all into a fulltext db and if a user goes asks where a missing email is, I can quickly check it, and if found, release it to the end user.

Currently the messages are parsed into the db via a Python script, and the web front end is PHP. It seemed like a good project to move to using all Python by porting it over to Django. This time I'm working on getting unit tests behind it, model sharing between the cron scripts, a new web front end, and hopefully I can implement the feature I've been wanting that will allow users to check their own messages by logging in with their imap accounts.

It's fun stuff because it's a bunch of new things. Django, using bzr full time for the whole project, and I think I'm going to try using Dojo for the Javascript on this one. So if you're going to come on Thurs and are interesting in some Python stop by and see me and we can work on this a bit. Who knows, maybe it can grow into something useful for other mail admins. After all, I hate the stupid spam folder. Ever try to scroll through the 5000 spam in your Google folder? It'd be much easier to say "I didn't see something on Tues I was expecting", and then go try to find it.

Looks like some people need to update their rules. Guess I have a different interpretation of "Users".

I've been trying find more and more places to use a distributed version control system in my daily work. I eventually want to move all of my personal svn repos to bzr or hg, but for now I wanted to see if I could use something with work's svn repositories. I've heard that bzr has this bzr-svn thing that allows you to work with a svn repo. This seemed ideal. I mean I can't convert stuff for work. I don't want to try to svn co something, add a .bzr or .hg dir to it and try to use both systems on a set of code. This bzr-svn seems ideal. You point bzr at a svn repo and it'll check it out. You can work on it, commit, work offline, everything. When you're ready you bzr push and it commits to the svn repo as one nice changeset. How sweet is this?

So the couple of tricks go like this. When you checkout a svn repo you might need to prefix the url with svn+. This gives bzr a hint that this is a svn repo. The first checkout is VERY slow, but it's got to transform all of the changesets so projects with more history will take longer. It makes sense and is a one time deal so it's not too painful. Making changes from there is just normal bzr usage. You've got a bzr repo in your lap. Once you commit your changes and you're ready to submit the changes to the original svn repository you just run bzr push. The first time you run it you'll have to provide the url you used to check it out, after that it'll do something like 'using saved url xxxx'.

So now I can managing feature branches locally. It's really a cool idea if you have an existing svn repo and you don't want to have to hand out accounts for creating/merging branches and such. Definitely check it out if you're stuck working with existing subversion repositories.

I've written about my issues with Zend Studio and how it's forced me to move on to try out Komodo as my full time editor/IDE. I did eventually (one week after submitting my support ticket) get an answer that was exactly what I posted in the ticket, which was recreate your profile. I never got any support or even mention of my request for some method of debugging the startup of ZS so I could figure out which files/part of my profile was corrupt, allowing me to salvage the rest of it. So I basically told Zend in the ticket, thanks for the editor since 2003, but I'm done. kthxbye

So I've been trying out Komodo since then, which has been a couple of weeks now and I think it's time to move on. There are a couple of big issues that I'm not happy with, but overall the editor is pretty nice. Definitely give it a try if you haven't.

The first big thing is autocomplete. We don't have namespaces in PHP yet, so I have a bunch of classes with lovely long names and I just live on autocomplete. Komodo does an ok job of autocompleting after some special character. In Python for instance, doing os. will get me a list of methods in the os module. However, typing out a constant I had defined at the top of my script would not autocomplete. I've checked the preference for "complete as you type", but it won't work. It also fails on inherited classes it seems. In my script I'm working on now I inherit from the CommandLineApp and Komodo fails to autocomplete any method in the base class there. It just means I have to go load up that class and look through all it's methods to find what I wanted. It wouldn't be so bad if I could have gotten the "Go To Definition" to load up the CommandLineApp code since I easy_installed it. No go on that front as well. I did add the path to the .egg to my Python path in Komodo, but I don't know if it can open eggs or not.

The next thing is a small one, but one that drives me nuts. I could not find any way to put the code tab, the one that lists the classes in currently open tabs and lists their methods for easy jumping around, to the right side of the editor. I like to have my project files on the left, editor center, and my function list on the right.

Finally, the third straw was using the project pane for doing any file manipulations. Creating new folders didn't seem to actually create them. Copy/paste a file and renaming it didn't work either. I ended up with a renamed original file. What was very strange was that I copied/pasted the file to a different folder. I did finally get a new folder to work by using "Add new live folder" whatever that means. Sorry, but while I love my command line, and I can do a cp && mv to get the desired result, it would be nicer/faster for me to be able to do it in the IDE. I'm sure I'm just missing something here. I mean, these simple tasks must work and the options are there to do them. Oh well.

So next up I think I'll try to get eclipse and use PDT with Pydev. See if that works for me. Pydev I liked so that part should be ok. I haven't tried PDT in a while though so we'll see.