Here we are for another awesome Democamp! Here are the 3 musketeers cracking jokes at the audience. A special word has to be said for David Crow, as mentioned from Michael McDerment, who organized the very first BarCamp here and who is single-handedly responsible for putting Toronto on the Web 2.0 map, so-to-speak. Bravo muchachos. That said, there is an upcoming BarCampEarth which is basically a series of BarCamps around the world held everywhere within 2 days, from August 26-28. Sadly, there is no BarCampEarth Toronto, and since this is a community thing, something should be done. It can’t be the same people organizing everytime, so step up people and give a hand in organizing this event! Without any further ado, here are the demos:

Wild Apricot
The first demo was about Wild Apricot(WA) and the goal was to get known (and of course for feedback). A rather tasty fruit and a funny name, WA proposes an easy to use and easy to configure web-app targeting associations and non-profits. It makes setting up a website for organizing events as painless as possible; amongst the list of features supported, there are registrations, optional individual payments (if the event organized isn’t free) and much more. The company’s been developing web-apps for 6 years now and their mantra is usability, and it shows! Their website says it’s 5 mins from the moment you sign up to the moment you use it, I say its 3!

It is build in .NET, sports WYSIWYG editing, and using some more AJAX sauce to make the application more intuitive. An event can literally be published with 2 mouse clicks and a few key strokes.
The presenter wore the company colours (quite a sight and deserves a picture) and generously offered the attendance free apricots.

It was a good presentation, the web-application seemed to be very polished and this looks to be a great solution for less technically abled people. A nice way to begin this democamp.

Jobloft.com

Next, we had the pleasure to hear from the guys behind Jobloft.com, a startup founded by young and enterprising Ryerson students. Also built in .NET, their web-app uses the Google maps API and other technologies in combination with job postings to create job hunting 2.0 for young people.

Focussing on 3 specific sectors (retail, food and hospitality), they aim to be THE place for teenagers and young adults to look for jobs. Special care has been taken in designing the interface, borrowing features here and there (the left menu when searching inspired from Amazon) to create a much simplified interface as compared to the likes of Monster.ca and Workopolis, which cater to a much larger variety of professions.

The founders took around 8 months, from September to May to bring this beast from paper to release. And right now, they have 36 signed employers offering jobs, over a variety of cities in Canada. They provide features such as SMS messaging, RSS feeds and plan to offer community-building features in the future.

The startup is apparently receiving some praises from employers and appears to be a simpler, more effective job posting service than TheBigOnes at this targetted demographic. One example of that is as follows: a job hunter enters a home address then selects a job. Using the Google API, the distance to the workplace from home is displayed, as well as the route to get there; teenagers would, in my opinion, choose a job based out of convienience rather than make a career out of it.

Now my question is: What is the growth potential? Is this section of the job market always expanding? What is the business model? In any case, kudos to those guys, having started the company with seed funding from their relatives, on a tight budget.

FileMobile

FileMobile‘s introductory description doesn’t do the app any justice. The presenter first explained that it was a nice way of sending your media from your camera phone to your blog. It turns out from the demo that it’s much more than this; it’s actually a full-on multimedia sharing application.

One part of it is a media management program. As far as file input is concerned, it allows file uploads from computers or directly from the cellphones as well as on the fly webcam feed recording. Built in flash with FlashCom (now Flash Media Server), it allows all sorts of video broadcasts. Sporting features like picture re-sizing, simulteneous operations (resizing an image while uploading files etc), it is a decent media management software, but online.

The second part of this app is its ‘sharing‘ component. We were treated with live demos(this one features Fabian picking his nose). The app can create entries automatically in the blog engine of your choice, automatically detecting them. In addition, community features are offered, allowing fellow FileMobile members to browse your stuff, if you set them to be public. We had the opportunity to take a peek at this feature, and we got confirmation that there IS a use for all the camera-phone to FileMobile account transmission; the first 3 cameraphone pictures seen were boobs.

For Mac and Linux users out there fearing compatibility problems, the app is meant to work on your system. Boob hunting has never been made easier.

Languify

Originally an experiment from John Phillip Green at Nuvvo, this framework has been developped by a fellow U of T and Greg Wilson student, Nicolaas Handojo.

They demo’ing for 3 reasons:

1. Usefulness: To find projects this could be used in.
2. To raise awareness of its existence.
3. For feedback.

In a nutshell, it is a mechanism to get any UI translated in any number of languages (spoken, not programming languages). Taking input from any .yml, .NET .xml files or in .csv‘s, the framework can substitute the UI’s language through an easy to use interface. Through the same interface, the text-to-be-translated can be viewed and translated on the fly. The translations can be stored in a database or as a flat file, in the above mentioned input formats. Languify was built using Ruby on Rails.
This tool might prove to be useful for applications like Jobloft.com which hasn’t been built with internationalization in mind.

The demo showcased a choice of languages consisting of Japanese, German and English, so for those who might ask, yes unicode is supported. Hurray for 2-byte characters.

Michael McDermot

Last but not least, Michael McDerment from Freshbooks.com came to do a presentation on how to measure the success of a web-based service. Bending the no-powerpoint rules, he used html slides to discuss the problems web businesses face concerning sales.

He adapts the sales funnel model to web services, the top being visits, followed by trial then ending by purchase. If visits are 100% of the incoming traffic, trials go down to 10% of those people followed by only 1% that actually purchase the service. And THAT’s considered good.

The classical problem is this: How to increase sales? Some splurge more on marketting, generating more traffic.

What Michael suggests is to first use a more sophisticated funnel model. The following one is more realistic, with current web services now offering trials and free limited usage:

Starting at 100% with visits, 10% at trial, 9% at login to trial account, 5% as active users and 1% at purchase.

Notice how he added the ‘login to trial account’ phase. The solution he proposes is to analyze where people are lost in the funnel model and attempts to minimize this number. As far as FreshBooks is concerned, he found that he lost 60% of the people that signed up for a trial to the actual trial login process. Only 40% bothered to go in and fill all the forms before being able to log in. A thoughtful change to the sign-up process for trials fixed that problem for Freshbooks. The moral of the story is this: expand the funnel model, find the ‘limiting factor’ and swiftly fix it. In his case, it was a usability problem.

We only got to the actual theme of the presentation in the last few minutes; Google Analytics, while good at providing high level data is not so good at providing more detailed info. Amongst the data to watch are: login/logout rates for a single user, which page are being accessed more frequently and of course finding and eliminating limiting factors. Another way to test the service is to use the classical usability tests: using random people to go through a task, asking questions etc…

Conclusion

This was yet another exciting democamp, with very polished applications. The next democamp might not take place at the end of August because it coincides with BarCampEarth, but more details are to be posted. Until then, see you later!

This is the next iteration of DemoCamp (Read this first!).

Demo Schedule

We will have 5 demos:

1. WildApricot
2. JobLoft.com – Google maps-powered job board
3. Filemobile Power tools for bloggers, letting users videoblog and moblog with ease.
4. Languify A not-yet-released tool from the Nuvvo team & NicolaasHandojo to manage translating your user interface to new languages.
5. Mike McDerment (FreshBooks.com) – How to Measure the Success of Your Web Service

The world belongs to early risers. We do everything we can to delay sleeping, yet we can’t get enough of it. Yet, sometimes we do manage to get to bed early, but we fail to catch the Z’s because we are so anxious to sleep. Being a recent grad, the unregular sleeping schedules are still fresh in my memory, with the odd eating hours as well as the sleepless nights browsing the internet (“research”).
“Sleep is for the weak”, that’s what some actually believe. Well guys, I think you are very wrong. Not only is sleep very important, but due to our newfangled social and professional obligations (damn Edison and whoever invented the candle) we are also expected to stay awake after dusk and still rise a couple of hours after dawn.

I have found a few excellent articles on the subject of sleep, notably on how to become an early riser, its sequel, possibly how to annihilate sleep anxiety and finally, a wonderful article about everything you always wanted to know about sleep but were afraid to ask. The latter goes into hypothetical evolutionary explanations of our sleeping patterns WRT circadian rhythms, and how to make them fit our societal requirements, all written in a funny, pleasant style.

I do not know what to blame it on; sometimes when bedtime comes i’m “in the zone” while coding away, therefore I push back my sleeping hours, otherwise I’m often reading really interesting articles (notice the “s” after “article”) and want to get through till the end, and yet other times… I’m just losing time doing totally unproductive things (yet so satisfying).

The 3 most useful rules I find to waking up everyday at the same time are as follows (but not limited to):

1. To always get up at the same time. No matter when you go to bed, wake up at the exact same time everyday. That way you’re setting your biological clock to automatically be prepared to wake up at a thus-defined time. (i.e no sleeping in, much to my dismay)
2. If rule #1 is observed, go to bed whenever you feel tired. Let your body determine the amount of sleep you need! If your body is set to wake up at 8:30am everyday, you should be pretty lethargic at around midnight or so. Sometimes, 11 could be too late, especially after a pretty heavy workout!
3. Developing a going-to-bed ritual. Mine entails brushing my teeth and cleaning up the smorgasbord of Firefox open browsers (which sometimes actually leads me to NOT sleep!). This will cue your body to go to bed.

In preparation for my job starting on the 31st, I have put it in myself to wake up early everyday (early for me is 8:30 am!). Ever since monday, I’ve succeeded on waking up at the same time and I’m very proud of it, even though its only 4 days now! I can’t wait to start =)

These are the steps I took in enabling the ESS 1869 sound chip to work in Ubuntu.

These fixes will actually work on any GNU/Linux distribution provided you have alsa. This is a fix for the ESS Audiodrive 18XX series of sound chips, not just for the 1869. Those chips were low cost alternatives to full fledged sound cards such as the Sound Blaster 16, both running on an ISA bus and with comparable features: Full-Duplex sound, 44.1kHz 16-bit audio sampling, FM synthesis and Plug-and-play(pnp). Mind you, pnp was a great feature at the time! No longer did you need to manually configure the settings, the drivers would automatically detect and configure the hardware… at least in theory!

Manual configurations were still required to guarantee smooth operation, pnp hence obtaining the redefinition of “Plug and Pray”. In my case, this is exactly what happened; ISA pnp is a little bit flakey and to make the sound chip work, I had to specify the drivers as well as the card settings.

A driver is a piece of software used to allow the kernel (the operating system) to interact with some hardware. modprobe is a tool used to add or remove modules from the kernel (the real operating system) and we will use it to load the driver. sudo (super-user do) is a program used execute other programs with super-user priviledges (similar to an administrator for windows users), alternatively, one could use the su command to put a terminal in super-user mode.

Step 1

Go to /etc/modprobe.d/ and create a file name soundcard there. ( requires super user priviledges)
Copy and paste this in the file:

alias sound-slot-0 snd-card-0
alias snd-card-0 snd-es18xx
options snd-es18xx enable=1 isapnp=0 port=0×220 mpu_port=0×388 fm_port=0×330 irq=5 dma1=1 dma2=5

The “options” line should contain the settings of your sound card, which may or may not be the settings above.

Step 2

To test if the settings are correct, try these commands on a terminal:

sudo modprobe snd-es18xx
sudo /etc/init.d/alsa-utils restart

Now play a sound file. If it works, awesome, if not, you may have to change the sound device settings to use Alsa. If it still doesn’t work, maybe you have to change the sound card settings.
Step 3

For a permanent fix, you will want to load that driver module everytime your computer starts.
Edit the file /etc/modules and add the following line to the bottom:

snd-es18xx

and voilà, you’re pretty much set.

Ubuntu, as I had assumed,would appear to be more lightweight than a Microsoft counterpart (as in being less bloated; potentially much less demanding on hardware). Eager to take Ubuntu for a spin on a less powerful machine, I got the opportunity to install the much-hyped distribution on a Compaq Armada 3500. The purpose of this post is to help the poor people having to go through what I have been through!

The laptop i configured as such:

• Pentium II 350 MHz
• 128 MB EDO-RAM (EDO!)
• 6 GB hard drive
• Chips F69000 Video card
• ES1869 Sound system
• 8X speed CD-ROM drive
• D-Link DWL-650+ PCMCIA card
• 3COM 10BaseT ethernet PCMCIA adaptor

Eager to boot the live CD, I watched the spinning poppy-seed bagel for around 30 mins after X windows initialized. I realized that there might not be enough memory to contain the live CD, so I scoured the net to find a distribution with a smaller footprint. My instincts lead me to something running XFCE and lo and behold, I found Xubuntu.

Seeing the memory requirements were 128MB, I happily downloaded the iso and I was back in the business. Or so I thought… Taking 10 mins to log on the live CD and 3 mins for each installation step at the minimum, (and also locking up each time at step 5 of 6), I grew really tired of trying (I tried at least 3 times!). I then realized 128MB was not enough despite the published requirements.

Enter Xubuntu – alternative installer. It is the text-only installer for Xubuntu. This made installation a snap. After happily booting into XFCE, I noticed X.org was not running XFCE at the correct resolution and did not want to change no matter what (running at 800×600 vs 1024×768. On an LCD this is extremely annoying). The sound card was not working and the wireless network card wasn’t connecting either. The only redeeming factor was that the ethernet adaptor was working flawlessly.

Keep in mind that I’m installing Ubuntu on older hardware. Recent hardware doesn’t have these problems.
Coming soon: The next post will explain how to fix the problems.