The last two weeks have been intense and this has contributed to my lack of blogging our status.
Whidbey Beta2 continues its inexorable shutdown. We fixed some compiler bugs and a particularly nasty IDE crash but apart from that looked good. Our QA team finished their final test pass and we feel good about the quality of the Beta going out. It is a Beta but we feel its going to be a good one. There was much drama as we found and debated bugs that we wanted fixed. Side story – a year ago or so the 2 nights before we released our first ever CTP drop, we found a bug and we fixed it, promising up and down the chain of command that it was a simple fix. Then the day before we shipped the CTP we saw a problem that was pretty bad – that we thought was a regression from the previous days fix. An unnamed but popular QA blogger who was on point for releasing the CTP heard of the bug, saw me in the hallways and his first words were – “Its been a good run brother”. We then debated new jobs for the next hour until it was found the regression only happened on a build customers would not see. There was definitely a moment last week where we looked at each other and jokingly uttered the famous words “Its been a good run”. But the positives far outweigh the specific bug. Our team did a great job in tightening up our lockdown processes. As a result we became much much more careful about fixes and this is going to serve us well as we shut the product down for RTM.
Another huge thing keeping us busy is RTM planning.We recently had reviews with our release team where we charted out when we thought we could release and deliver a high quality product. That really helped us straightening out where we were w.r.t RTM. As a result our plans for the next couple of weeks are very very targeted to releasing the right product at the right time. A bunch of my work is to drive us setting dates to hit targets. W refer to them as “stakes in the ground” with the theory that sometimes you need to set the goal and then get the team to rally and figure out how to hit that. It works pretty well.
I’m proud of our dev team who are going to take the time to write more checkin tests and run more automation to stop regressions in RTM before they get checked in. This kind of preventive work now will payoff big later. Another impressive moment came when Steve Kruy our dev manager sat down with me and in about 10 minutes commandeered my machine, and excel and figured out a no bs schedule for how long it would take us to fix bugs after accounting for people taking vacations, parental leave, support for internal and external customers of our tools, conferences and what not. It was pretty impressive and really makes you love having an experienced hand at the helm of the dev team.
A number of people in past comments have asked me when we ship…I just cant answer that. Sorry folks, professional restrictions.
Andy Pennell left this week. He will be missed. As will Damon, who also left. It just means we have more window offices to hand out. See my previous posting on jobs 🙂