Mozilla Calendar Project Blog Its about time for some news…

The turnout on Tuesday wasn’t good. So, we’re throwing another test day this week, on Tuesday, July 31st. We could really use your help.

The former prototype event dialog has been promoted to be the new standard event and task dialog for Lightning and Sunbird. The developers are interested in feedback and bug reports of the test community to improve the new dialog. Also, the new Lighting “today pane” has landed, and we need your help testing it and the mail/calendar switch.

You can find all the information on our Test Day wiki page. We look forward to seeing you in #calendar-qa and Litmus on Tuesday, July 31st.

Happy Testing!

Next week, we have a regularly scheduled test day. Our alarms seem to have regressed a bit in the last few weeks, so we want to take a critical look at those and determine what specific bugs are there and how to reproduce those bugs. This will give development the direction they need to get them fixed.

Also, several user interface improvements have landed in Lightning, including the new mail/calendar switch and the “today pane”. Join us to check out the new features and find any regressions these massive landings have caused.

As usual you can find all the information on our Test Day wiki page. We look forward to seeing you in #calendar-qa and/or Litmus on Tuesday.

Happy Testing!

We had a good test day on Tuesday. Thanks to everyone that dropped by and helped with the testing. We managed to narrow down the “remote calendar problems” to two basic issues:

• ICS/WebDav: It seems that most of these problems stemmed from incompatible Limit/LimitExcept statements and a change to our code base that causes writes to fail if authentication is required for PUTs but not GETs. This is best summed up in bug 387559
• ICS/WebDav and CalDav: There are also proxy issues affecting both clients for these server types. The proxy configuration seems to have been recently broken, with various reports stating that it worked in 0.3.1 and that it did not work in 0.3.1. We need further clarification on this issue. This issue is summed up in bug 373439

We were able to track down the first issue to a specific check-in, and that is now being investigated. However, we need to do the same with the proxy issue. If you have a proxy configuration please try out various Lightning nightly builds and see if you can determine the build where the proxy configuration stopped working.

Things We Could Use Help With

• Determine the date the proxy configuration stopped working as mentioned above.
• Look at the list of these new remote calendar bugs, and see if we can duplicate some against others (leave us comments in the bug about your ideas).
• Give us suggestions on how we can improve calendar QA

Thanks very much to everyone that turned out for the test day. And congratulations to Alan Schwartz who won the prize for most creative test.

Happy Testing!!

We have had a startling number of “my events have disappeared!” bugs in the wake of the 0.5 release. Bugs like 381573, 373439, 386197, and 386734. We must figure out what is going on here. Are these all the same issue? Are they several different issues? So, we’re going to have a Remote Calendar test day on Tuesday.

This test day is going to be different because we’re encouraging everyone to get creative with their testing. See if you can help us find out why these events are disappearing. Take a look at Litmus, and see what we tested there, then think about ways to go beyond that. On Tuesday, we’ll be giving an award for the most creative and effective test, so be sure to add your test to the Test Day Wiki page.

We’ve been doing a great job as a QA team, but I think this issue gives us an opportunity to think about how we can improve the quality of Lightning and Sunbird. I don’t believe that the only solution is to quit our day jobs and work on this around the clock (although that might work).

I think we can find ways to work smarter, to do more with the time and the people we have. Maybe we should look into automating some of the testing. Maybe we should look at the weak parts of our Litmus test suite and fix those areas. Maybe we need to delve deeper into performance and memory leak profiling.

We have started this wiki page for brainstorming ways to improve every aspect of calendar QA. We’re going to take your ideas, prioritize them based on what we need the most before 0.7 and put them onto our revised Calendar QA ToDo list.

We will see you on Tuesday. Get creative, bring us your best tests, and prepare to discover where all those events have gone.

The test day on Tuesday was absolutely incredible. The response was awesome.

• 30 people took part

332 tests were run

• 90% coverage in litmus on all tier 1 platforms (Windows, Linux, Mac Intel, Mac PPC)
• 9 non-duplicate bugs found.

First place and the prize of the test day went to ssitter with over a hundred tests run in addition to serving as a moderator on the channel. Ray swept in with a close second at 59 tests and several good bugs. Andreas was on his heels with 54 tests. While these guys may have been in the lead, everyone’s contribution helped us achieve 90% test coverage, and that’s what I’m so excited about.

Several good bugs were found, but none of them were serious enough to block the release. We are still testing, of course, but so far everything looks good. The team is currently in the process of cutting the release builds which will include a localized Lightning and a universal Mac build of Lightning. No code changes are going in, so once we have the release build built, we will do a localization and functional test day. If we don’t find any blockers, we’ll release that as 0.5.

We are getting down to the wire, so keep an eye on this blog, the newsgroup, and Mozilla Zine for the announcement about the next test day. I’m going to try to have it on Tuesday, but it entirely depends on when the release is built.

I’m not sure what a virtual round of applause sounds like, but we should give ourselves a hand. This was a magnificent feat! Thanks to everyone that joined the test day effort and helped out. Thanks also to our moderators who in put in long hours to stay awake and available through the day. We couldn’t do it without you!

Happy Testing!

I’m not sure what happened to this post, but I think there was a glitch between my computer and my chair last week. I distinctly remember writing a post about the test day tomorrow. But, it isn’t posted, and wasn’t deleted. Sounds like “somebody” forgot to click “Publish”, or worse, I wrote it in a dream. ;-S

All questions about my sanity aside, we need a test day. The last of the 0.5 blocking bugs was fixed by Mr. Goold (thanks very much!). We are going to try to spin up a real release (including locales and everything) for both Lightning and Sunbird. But, even if that doesn’t happen by tomorrow, we still need to test the code since these current branch nightlies are going to be the RC.

There will be rewards for this test day! I will announce the winner in this forum after they are announced at the Thursday Calendar QA Chat on May 17.

I know the project has been slipping a bit due to the hectic schedules of the core team, but we are trying very hard to keep it moving forward. The best way for you to help us is to participate in the test days. The regression testing is the last thing we need to do before we release 0.5, so the sooner it’s finished the sooner we release. And that’s exactly why we need your help tomorrow.

Thanks very much. Happy Testing and Good Luck!

Spring is officially here, and we’re going to do some spring cleaning at Calendar QA. We’ll start with a Test Day Tomorrow on May 1st. On Thursday, we’ll hold a QA Work Session to resolve the growing number of “QA Discussion Needed” and “QA Wanted” bugs.

The QA Work Session will take place at 14:00 UTC time in the #calendar-qa IRC channel. The Test Day tomorrow will start at the usual 13:00 UTC, you can always find more information on the test day at our Test Day Wiki.

Every test we run and every bug we verify brings us that much closer to our 0.5 release. Thank you very much for your help. We look forward to seeing you tomorrow in #calendar-qa.

Happy Testing.

I apologize for the late notice, but the Calendar-QA effort needs to get its test day schedule back on track. With only blocking defects going into the code base, now is the time to turn a critical eye toward these recent builds. We need to alert developers to any critical “stop ship” defects we find. And to find them, we need your help at the test day.

Drop by #Calendar-QA and join the bug hunt tomorrow. If you can’t make it tomorrow, we hope to see you at some point before the end of the week (Drop in for our QA Chat).

For tomorrow’s test day we are going to drive it based on the testing recommendations at our Calendar Todo List. I will put together the standard test day wiki page later today. Thanks for your help.

Happy Testing!

Test Day Report

We’ve had two test days since my last post: one on March 20 and the other back on March 6. Both of these were great days. We had hundreds of Litmus Cases run, found several new bugs. Skuribay won the test day on March 6 and Thorn won the test day on March 20. Congratulations to you both!

Function Testing Begins!

As you can see on our new project calendar, Functional Testing for the 0.5 release will begin on Monday, March 26. We need folks to start running through the FFT (Full Functional Tests) in Litmus for both Sunbird and Lightning builds. Please help us cover as many OS’s (windows XP, windows Vista, Mac OS X, Linux flavors etc) as we can on these tests. It is important for us to find as many bugs as we can before the string freeze as possible, so that the localizers have solid builds to work with.

We will be also holding a “Hard Core QA Session”. This will be from noon UTC to 16:00 UTC on Wednesday March 28. We’ll be working on QA Wanted and QA Discussion Needed bugs, to clear those out. Should we finish that, we’ll work on functional testing of some of the newer components and features that have landed.

This is a great chance for folks to drop by and help with some of the core issues of bug triage and deep testing. We’ll be holding several sessions like this in the coming days. Working with the QA Wanted and difficult defects will give you insight into how best to track and find issues, which will help you win test days and make the calendar code more robust.

It is time to ensure that we release the best possible product we can for 0.5, and that begins right now. We need to amplify the QA on Calendar. So, turn up your favorite music, download a nightly, and see how badly you can break it. If you find an unreported issue, drop us a bug.

I’ll see you in #calendar-qa. Happy Testing.