Upgrade to new blogger

I upgraded to the new blogger interface yesterday, which means that I finally have access to labels. Over the week I'll be labeling most of the 300+ posts in the archives. If you are reading this via a feed, it may happen that a whole bunch of old posts lands up in your feedreader. If that happens you can just ignore it as its only the labeling going on.

Chennai Air Show Photographs

I've uploaded a bunch of photographs from the Chennai Air Show that happened at Marina Beach yesterday. The photos are here:

http://www.flickr.com/photos/siddhi/tags/airshow/show/

Code coverage for your Django code

One of the things that I like about Django is that it allows you to run unit tests on the code with relative ease. So I have a bunch of unit tests now, and I was looking to expand it. For that I first needed to know which parts of the code had good coverage and which parts had no coverage. I needed to integrate code coverage with the unit testing framework. In the end, it involved modifying a few files. This is what I did:

1. First, I downloaded Ned Batchelder's coverage.py module.
2. Next, get this script to colorize the coverage output. I saved it as coverage_color.py
3. Put both files somewhere on the python path. I put it at Python24\Lib\site-packages
4. Now we need to modify the django unit test runner to include coverage. Head over to your django\test directory and edit the simple.py file
5. At the top, add the following line:

   import os, coverage, coverage_color

6. Scroll down to the run_tests function. You will see a line like this:

       unittest.TextTestRunner(verbosity=verbosity).run(suite)
       destroy_test_db(old_name, verbosity)

7. Modify it to read like this:

       coverage.start()
       unittest.TextTestRunner(verbosity=verbosity).run(suite)
       coverage.stop()
       if not os.path.exists(settings.COVERAGE_DIR):
           os.makedirs(settings.COVERAGE_DIR)
       for module_string in settings.COVERAGE_MODULES:
           module = __import__(module_string, globals(), locals(), [""])
           f,s,m,mf = coverage.analysis(module)
           fp = file(os.path.join(settings.COVERAGE_DIR, module_string + ".html"), "wb")
           coverage_color.colorize_file(f, outstream=fp, not_covered=mf)
           fp.close()
       coverage.erase()
       destroy_test_db(old_name, verbosity)

8. What that does is to record the coverage when running the tests. It then creates a directory for putting the HTML output, creates the colorized version of the source and dumps it into the output directory. At the end of everything, it cleans up the coverage data
9. We now need to configure the COVERAGE_DIR and COVERAGE_MODULES settings. Open your Django settings.py file and add the following lines:

   COVERAGE_DIR = "scripts/build/coverage" # Where the HTML output should go
   COVERAGE_MODULES = ["apps.catalyst.views", "apps.catalyst.models"] # The modules that you want colorized

10. Save and run your Django unit tests. After running, you will have a html file for each module in the specified directory with the colorized coverage output

Check out the output: The lines in red were not executed by any of the unit tests. Looks like there is a function that I'm not testing at all!

Mobile Monday Chennai

Update: See here for the latest update.

Varun Krishnan is organising the first Mobile Monday event in Chennai. Mobile Monday is something like Barcamp in the sense that its a tech gathering, except that the event is only 3 hours long and the topics revolve around mobility.

Mobile Monday, as its name states, is supposed to be on a Monday, usually after work. The upcoming Chennai event, however, will be on a Sunday (some people call this Mobile Monday on a Sunday or MMS). Mobile Monday is also supposed to be once a month, usually the first Monday of the month. The Chennai edition will be once a quarter instead. If there is demand, the frequency will be increased.

So here are the details

Date: April 22nd, Sunday (2 weeks from today) Update: The date has been changed to sometime in May
Time: 10am to 1pm
Venue: Mobile Worx office

The schedule is three tech talks, 20 minutes each and three product demos, 20 minutes each followed by lunch.

I've signed up for a talk on Location Based Services. I'll be talking about a couple of projects that I did while I was in Singapore.

If you are interested in giving a talk, head over to the Speakers wiki page and add your name there.

Barcamp Bangalore 3 Roundup

I was in Bangalore on Saturday and Sunday attending the third Barcamp to be held there. Here are bits from the event that I found interesting. This is a long post, you have been warned :) Note that the following is not in chronological order.

Python Conference

I met a few guys from the BangPypers groups - Anush, Anand, Kushal, Swaroop and Sidu. Apart from the Python conversations, I also asked if anyone knew what was happening on the planned Python conference. No one had much information, so I'm not really sure whats happening. Is the Bangalore python conference going to happen?

The Knowledge Foundation

The Knowledge Foundation is the name of the group that is behind various unconferences (plus Proto.in) in Chennai. We had a discussion on how to spread the unconference movement to other cities and other non-tech settings. While Barcamps have been a big success in India, there has not been much activity outside of a few tech events like Barcamps and Mobile Mondays. TKF has been working on that in Chennai where apart from Barcamp we've had Blogcamp, Wikicamp, Proto.in and Rooftop Film Festival (that reminds me - I have to upload the RTFF photos tomorrow). Coming up are another edition of Proto, Mobile Monday and Shoot. Quite a few people were enthusiastic about having small gatherings on other such topics. It would be great to see more unconferences on a whole range of topics in various cities.

Silver Catalyst

I talked to a few thoughtworkers about Silver Catalyst - Amir, Sidu and Vivek. Amir had recorded my session and had also interviewed me at the second barcamp bangalore back in December. I also met Umesh and a couple of other guys who had also attended my session in the previous barcamp. It was interesting to see how the tool had progressed in the three months between the barcamps.

Sidu took some time out to point me towards Mingle, which is a tool developed by the new Studios division of Thoughtworks, scheduled to release in June. I wanted to discuss Silver Catalyst a bit more with Vivek, but I lost him in the crowd somewhere.

Organisational Culture

There were three sessions that I thought touched on the same topic of culture from different angles. The first was by Savita Kini on Products vs Services in Indian companies. The second was by Satish about whether soft skills training was required. The third was by Shourya on the Bambi code jam initiative at GE.

Savita pointed out a number of challenges and opportunities facing Indian companies. Shourya made a good point in the discussion asking why we think that products are superior to services. Some of the points raised were profit margin and scalability. I'm with Shourya here that I'm not convinced that products are always better than service. Profit margin depends on the product. The XBox for instance is sold at a loss. Mobile phones have very thin profit margins. Scalability is also not a great reason as it all depends on the market. If you put out a product in a small market, there may not be much scope for scalability.

In the end, I think the biggest reason that product companies are not as prevalent is because of the organisational culture. Large companies are by nature risk averse. Since there are a lot of profitable service companies, we tend to follow them and not take the risk of doing something new. One of the cool thinks about Barcamps is that it gets a lot of product startups together.

That ties in with the second talk about soft skills training in IT companies. Soft skills are certainly required. The problem arises when the training is imposed rather than sought. Sukumar made a good point that knowledge can never be given, it can only be taken. I dont think a one day training can build teams for instance.

Satish had a good point which, to me, got to the heart of the matter. Fishermen work in teams because they know that if they don't, all of them will have trouble with dinner. The result of success and failure is right in front of them. They don't need to take a training class to realise that. The problem in large organisations is that there is a disconnect between the work and the outcome. Developers who do the work cannot see the outcome of their work, either in the form of customer satisfaction or bottom line impact. Thus they associate with the closest outcome that they can see - the salary - and start working for the money instead of the company.

A startup culture is like the fisherman team. Everyone is aware of pretty much everything and they know the outcome of their work. The question is can we get a startup like culture in a medium or big company?

Which brings us to the third talk by Shourya on Bambi, a codejam initiative taken in GE. He explained how he got the idea from the Yahoo Hackday in the first Barcamp Bangalore, and implemented it in the form of a codejam at GE. He explained how there was trouble convincing the management about holding it. Finally when it was held, it produced one IP plus a number of other cool, but not so useful projects. The second time around, the management actually asked for it to be held and it was held with some changes in the format.

A codejam, 20% time to work on your own project and other such initiatives are 'cool' initiatives. The problem is that they do not have direct ROI consequences. However, the actual dynamics are a lot more complex. Such initiatives for instance may (or may not) help attract smarter programmers, or reduce attrition, or build up a specific company culture. It may fail if the current programmers are not the type to take part in such events. So the success depends on both the current culture as well as the future culture.

Another problem is that some companies don't really believe in it. Someone was talking about 20% time in their company and how you can work on what you want provided you finish the main work. Since work pretty much takes the whole time, plus a bit more, this is as good as maintaining the status quo. The point of 20% in Google is that you can spend that time even if you are working on something else. They are okay with the fact that the main work will take a bit more time. So that again points to a difference in company culture.

So how can these forces be resolved in managing the current culture or creating a new culture? Thats the big question, but the discussions were interesting.

Proto

There was a discussion on the venue for Proto. A lot of people from Bangalore wanted the next edition of Proto to be held in Bangalore. This was a completely ad-hoc session held under the trees (an ideal unconference session!). [Off-topic digression: When a projector is involved there is a clear speaker/audience divide. Get everyone sitting in a circle and it becomes a discussion with everyone pitching in their views.] Back on-topic, there was a good discussion where it was decided to have the next edition in Chennai with interested people from Bangalore joining the team, and the session after that will be in Bangalore.

Barcamp Bangalore 2 DVDs

Amir from Thoughtworks had recorded a lot of sessions from Barcamp Bangalore 2. It came to 8 DVDs in all after editing. Kesav brought his set of DVDs for us to make a copy. The problem was that it takes a long time to copy all eight. Finally, Kushal volunteered to make copies for us. Thanks Kushal!

Quake deathmatch

A few of us had a Quake deathmatch using OpenQuake on Saturday evening. Jace, Shreyas, myself and a couple of other guys were there. It's been many years since I played Quake so it was great fun. For some inexplicable reason OpenQuake would sometimes stop reading the horizontal axis of the mouse. The mouse would work fine in the menus, but in the game I could only look up and down and couldn't turn left and right. Quiting and rejoining the server fixed the problem, but the frag count went back to zero. It happened thrice in all. Highly frustrating. Absolutely no idea what caused it.

Dinners

A dinner was organised at the end of Saturday. I didn't attend it because I went to meet a few guys at the Forum. At the dinner were ex-thoughtworker and AI hacker :) Ravi Mohan, Bharat Guruprakash, Manoj Govindan, Rajesh Babu and myself. A number of interesting topics were discussed like machine learning, silver catalyst, the ongoing drama at the scrumdevelopment group that has been tragically comical, work at thoughtworks and subex, thoughts on the CSM, agile in India, applying agile to embedded systems, PhD, research, entrepreneurs, dungeons and dragons, board games and so on.

We had to catch our train on Sunday night. Ramesh Babu joined Sukumar and me for the evening. He had hosted Syed, Kiruba and Sukumar on Saturday night. Kiruba and Syed had a dinner appointment so we split up at the station. They returned 5 minutes before the train was scheduled to leave, running and sweating, and then the train left half an hour late :)

Photography

I met Mahesh Shantaram at lunch on Sunday. Mahesh is a documentary photographer based in Bangalore. He has some very nice photographs that you can see online here. Amogh had a session on conservation with stories based on his photographs. He goes into the forest every weekend to take photographs. We later caught up with him at Cafe Coffee Day after the event was over to have a discussion about the Shoot event thats coming up and also if we could join up with him on a trip sometime.

Sessions

A few quick words about the sessions that I attended.

Motvik showed a demo of their WWIGO product. They showed how to use a mobile phone as a webcam and transmit data via bluetooth to the laptop where it was streamed over skype to another person. Inactiv explained their Activ Mobs service, a group SMS service for groups of friends. There was a presentation about Bigswerve by Raj Bala. What they do is to go through the comments at various websites and generate a network based on comment patterns. Comment patterns an example of implicit networks where the network is not formally defined and a comment automatically implies a level of engagement with the site. A few thoughtworkers gave a presentation of Continuous Integration and Cruisecontrol.rb. Sowmya Karmali had a session explaining the Mifos project. Marek Tuszynski had an event on How to organise an open event. Marek was the organiser for AsiaSource and a few other events. I think it started early or I messed up the timing because I really wanted to attend but when I went there the session was just wrapping up. There was a nice session on usability by Hari and Ram.