If you are reading this chapter, then you really like Watir. In that case, you deserve to know every little detail about it. Read on.
I was thinking about open source, and I have asked myself: "So, what is an open source project, after all?" After a lot of thinking (and a few beers) I think I know the answer. Open source project is code, documentation and community. So simple.
Watir is fairly old project. According to github.com, the first commit containing Ruby code was made in October 2003. That is almost 8 years ago. In internet years, it is almost a century.
I thought the code was first in SVN repository at RubyForge. Bret commented on the early draft of the book: "Actually started in CVS, then was ported to SVN and then Git. Nobody was using SVN in Oct 2003." From RubyForge it moved to SVN repository at OpenQA and now is git repository at GitHub.
License it an important part of every open source project. The very first commit had the license, and as far as I can tell, it was BSD, although it was not explicitly stated.
In January 2005, when the first version of Watir, 1.0.1, was released as a zip file (according to RubyForge) we had some code. Zip file? Really?! Really. RubyGems did not exist back then. And you thought I was joking when I sad it is almost a century in internet years.
Watir 1.4.1 (August 2005) was a big success, downloaded almost 45k times. That was also the first version that was available as both a gem and exe file. Watir is 23rd on the list of top RubyForge downloads, if that means anything these days. Bret commented on the early draft of the book: "Back then Watir was one of the top 5 downloads on Rubyforge. That was before rails."
Watir gem is now hosted at RubyGems.org, all gems have over 55k downloads, and the current version (1.8.1) over 2k downloads. Watir gem is not the only gem included in the Watir project, there is also firewatir, commonwatir, safariwatir, celerity, watir-webdriver and the youngest one, operawatir. There used to be chromewatir gem too, but it is replaced with watir-webdriver.
There are a few frameworks that simplify writing Watir code: taza, watircraft, watirloo, and watirsplash, but only the last one is updated.
There is even one fork of Watir project, Vapir. Ethan (he never said his last name) created a fork in May 2010.
The first significant documentation I could find in repository was made by Bret in April 2004. When I started using Watir (March 2005) it already had pretty good documentation. It had user guide written by Jonathan Kohl and RDoc documentation. RDoc documentation is now hosted at RubyDoc.info that auto-generates it with every new release.
Internet Archive Wayback Machine says we had a web site since February 2005, and around August 2009 we moved the site to http://watir.com/. Paul Rogers registered watir.com domain in April 2005 (according to whois.net). Maybe watir.org would be better for open source project, but it was taken. Alister Scott did a great job with the web site. The plan is that all you need to know will be there, but not overwhelming like wiki. With time we were just adding more and more stuff there, maybe it is the time for a cleanup.
I am also very proud that we do not have www in front of our domain at our web site. I even tried going to www.watir.com and it redirected to watir.com. I was so proud. If you do not know what I am talking about, take a look at http://no-www.org/.
Wiki is an important part of our documentation. We had it in early stages of the project. The first web site had a link to Web Testing with Ruby wiki (still live, but obsolete, from February 2004). I do not remember there was anything useful at the wiki while it was at RubyForge. We started working on the wiki when we moved to OpenQA. Alister did a lot of work on it, and I have moved the user guide there and renamed it to tutorial. We have a lot of documentation there at the moment, but it needs some love. A lot of stuff should be updated.
We have a bug tracker at OpenQA. Previously it was at RubyForge, for firewatir at Google Code, and for safariwatir at GitHub. All of them are now at OpenQA.
Since October 2007 we have Google Custom Search that searches all Watir related sites but nothing else. It was usually embedded or linked from our web page.
Wtr-general mailing list exists from October 2003, about the same time when the first commits were made to the code, but way back then the list was not about Watir as we know it today. It was about a larger topic, web testing in Ruby. We moved wtr-general from RubyForge to OpenQA, and from there to Google Groups.
Wtr-general says Watir project was stared by Chris Morris, Johnatan Kohl, Bret Pettichord and Paul Rogers.
We also have wtr-development mailing list at RubyForge, and it is still there. One of the rare things we did not move around. We have even managed to make Jira bug tracker send mail to wtr-development when any ticket is created or changed. I think it makes the developers more aware of the problems people have.
We had wtr-core group at RubyForge for a while, but that is not active any more.
I see spam at other lists from time to time, but I think we managed to keep our lists clean, polite and productive.
In the past 2,5 years I have been pushing Stack Overflow as "the" place to get Watir support, but had little luck, up until recently. I think in the last few weeks or months, almost every day a new question is posted there. Some people also started participating, and I am really happy for that. We have also tried to create a separate Watir Stack Exchange site, but that failed. I think it was good that we have tried, at least we created some buzz around Stack Overflow/Exchange. There is 256 questions tagged Watir there. (Nice, round number.) We finally have enough active users to get the questions answered.
I have started recording Watir podcast in May 2008, and that makes us one of the rare projects that has it's own podcast. It is hard to tell how much impact the podcast has. I like to think that it gives us a human face (well, human voice at least). So far I have recorded 44 episodes. I started using podtrac (for tracking how many times a podcast is downloaded) in December 2009 and since then we had about 300-400 downloads per episode. Episode #26 with Alister (we were talking about moving our web site to watir.com) is interesting. I am not sure why, but it has about three times more unique and total downloads than any other show. I guess I will have to have him on the show more frequently. I am not sure what I will do with the podcast in the future. My current plan is to reduce the size of the episodes to 5-15 minutes. Also, it is interesting that old episodes are still downloaded, even the very first one.
One of the first "real" marketing adventures was free ad at stackoverflow.com. It is hard to measure how successful it was. We did it two times (January 2010 and June 2010). I could not find any numbers on how successful the ad was, but additional exposure could not hurt. The third round is up right now, but I did not have the time to create an ad yet.
We also have #watir IRC channel at freenode, but I never got used to hanging out there. When I have the time, nobody is there. When I have to work I usually shut down mail, messengers and stuff like that to reduce distractions.
People have been blogging about Watir. We have a list of people that mostly blog about Watir at our web site.
I remember seeing articles about Watir in print magazines, I wrote one for a local computer magazine.
I am sure I am not the only one that speaks about Watir at local conferences and meet-ups. We even organized a few conferences specifically about Watir: AWTA in Austin, Texas (2000-2009); Watir Day in San Francisco, California (2011).
If you are just starting using Watir, I would suggest that you start with watir-webdriver gem. I think it is the future of Watir project. Some time in the future we will probably no longer develop other gems, just watir-webdriver. It will probably not happen soon (in the next few months), but it might be sooner than you expect.
If you are already using Watir, you can continue to use the gem(s) you have used so far. The only exception is firewatir gem. It can not drive Firefox 4. Good new is that watir-webdriver gem can. So, if you have to drive Firefox 4, start moving your tests to watir-webdriver. If you are still using Firefox 3, you can use either firewatir or watir-webdriver.
Jari Bakken, creator of watir-webdriver gem, also created celerity gem. If you have a need for headless browser driver, take a look at celerity. There is also headless driver in watir-webdriver, but I did not have the time to play with it so far, so I could not recommend it. My guess is that it is pretty usable, but that is only a guess.
Andreas Tolf Tolfsen, lead developer of operawatir gem says Opera should be supported via watir-webdriver gem in the near future. Until then, if you need to drive Opera, use operawatir, but have in mind that you might need to move to watir-webdriver soon.
Safariwatir was created by Dave Hoover in July 2006. Tom Copeland (of RubyForge fame) took over the project in October 2009. Safariwatir was never really popular, but it was there, waiting for somebody to take care of it. If you need to drive Safari on Mac (safariwatir does not work on Windows) give it a try. There is a rumor that watir-webdriver will be able to drive Safari in the future.
There is only one thing going on regarding documentation of Watir project. This book. If you think it is a good idea to write a book about Watir, please let me know. Sometimes I am wondering if it was a good idea to start writing the book.
We have created watir account at twitter.com (June 2009). It is pretty active. Alister and I post there when there is something significant, like new releases or blog posts. I also used it to link to new Stack Overflow questions, but not any more. It has 242 followers at the moment. We used to display tweets with #watir hash tag at the Watir web site, but it got flooded with a lot of posts in some strange language. At the moment, all posts containing @watir will appear in the sidebar of watir.com.
We had a LinkedIn group for some time (April 2008) with 554 members, but just recently I started posting relevant Watir links there. Usually the same stuff as Twitter account.
Since all the cool kids are at Facebook, so are we. I have created the account a few months ago but I did not have the time to start using it since this February (2011). I wanted facebook.com/watir url, but it is taken, so I took facebook.com/watirproject. Usually the same stuff gets posted there as all other social media we use. 82 likes so far.
I was hoping we would get some interaction at Twitter, LinkedIn and Facebook, but it is pretty quiet so far.
We started receiving donations at watir.com via PayPal since 2009 and via Pledgie since at least January 2010. Andreas suggested we start using Flattr in March 2011. We have received about $1000 in donations.
What would I like to see happening in the (near) future, regarding Watir code, you ask? In no particular order:
- Watirspec is "the" test suite for all Watir implementations. Each gem has a score and you can see what fails and what works.
- Watir-webdriver is the only driver, and it can drive all popular browsers on all popular platforms.
- One (or more) Watir frameworks that is up to date.
- Watir participates in Google Summer of Code.
- I hope we will raise some money so we could fund further development.
I guess every project needs (more) books, updated web site and wiki, and in case of Watir, updated RDoc would be nice too.
Watir community is a topic I think a lot. There a few things I think we should do in the near future:
- Joining Software Freedom Conservancy and moving all of our stuff there (domain, money, trademark...).
- We regularly create Stack Overflow ads, when ever the have a new round.
- It would be great if more people were writing about Watir, especially on their blogs.
- I think meeting regularly (at least once a year) is really important. We could meet at conferences that attract Watir users anyway, like Selenium Conference (may be renamed to Browser Automation Conference in the future). Organizing our own conferences and meetups is also important. Viaqa Watir Day in Zagreb, Croatia will be in June 2011.
- It would be great if I could get some funding. It would really help me with working on documentation, podcast, this book...
\newpage