How (not) to deal with our own inconsistencies

One of the hardest goals to achieve in life is self consistency: as we have mentioned before, our tastes are constantly in a state of flow that implies that tomorrow we will not like the same things that are like today, at least not exactly, or even if we do it might be for different reasons. The unavoidable consequence of our volatility is that we end up contradicting ourselves, and that is not a comfortable situation to be.

There are many possible ways to deal with the inconsistencies, but one of the most common ones is the denial: "I have always liked it, how can you think otherwise?"Depending on how obstinate each part is, the conversation can escalate our just die out with the condescending acquiescence of the more flexible side. Even more frequent, but also more problematic for the losing side, is admitting our inconsistency, for instance showing agreement with the proposal even if we have not been able to carry it out. And it is problematic because either admission is negative: either we did not have the strength of character to pull it through or we did not have time (i.e. good planning skills) to do it. 

Photo: Erich Ferdinand

When dealing with this kind of situations at work, I try to deflect the attention away from the culprit and focus instead on whatever is wrong. There is no point in burning anyone at the stake (and risking their future cooperation on even their departure) when the situation requires some joint effort to be solved. And that is precisely what I did (or tried to do) this morning: someone was talking about a problem they had had with the instrument and Martin, with an excellent engineering sense, asked which version of the software they were using because, unfortunately, the software is still evolving and sometimes newer versions can contain not only fixes but also new errors. Anyone who is even mildly aware of the complexity of the software should know how important it is to know exactly which version was at fault, but inexplicably our programmers did not have an answer, and I could not hold myself.

Over many years and across several different space missions I have come to reverently love accurate test reports: in development environments it is not uncommon to find yourself without a test plan, even if some test has to be done, so in many cases the only real alternative is to forge a plan on the fly. This is precisely where adequate and precise records are hugely important. If the test succeeds, the record will set the baseline on how to reproduce the success, so that next time the test can actually have a plan: repeat the previous run. If the test fails, the record will become the measuring stick to determine when the problem is solved. Without a test log it is possible that something that "used to work" does not work anymore and we will never know if some new error has been introduced or if, by some reason, we are doing something differently that explains the error. Conversely, we might not be able to reproduce an error that was happening, so the development team would have a harder time to address it.

With this personal history and the situation we were in I limited my comment to say that I was strongly disappointed that, in spite of my previous recommendations to keep public logs in our wiki page, we were once again facing a problem that we were unable to trace. The remark was obviously addressed at the two developers that were doing the test, but I was very careful not to point to anyone in particular. In fact, one of the developers took it the right way, accepted their part of the fault and excused themselves for their sloppiness. However the other one counter that they had perfectly usable personal logs and that I had no right to complain. My response could not be other than pointing out that personal logs are not the same as public logs as I requested, so in spite of their fury they were obviously on the wrong. Besides, even the personal logs were not accurate enough if they were not able to say which version of the software they were using. To prove my point, I went to the public logs that I had written around that time and pointed out how I had indeed recorded the version of the software the day before and the day of the failure. Sadly, they were so utterly outraged that I doubt that my demonstration held any convincing value in their view. They just protested again that my attack had gotten personal and I was just being a jerk, at which point the project manager budged in and cut the conversation short.

In all honesty, I have to admit that I spoke with the clear intent to hurt. I did not want to shame anyone in front of the team but I wanted them to feel my disappointment, and there was a reason for it. I had made similar suggestions several times before, always in much milder terms, something to the tune of  "it would be convenient if everybody kept track of their activities in the wiki pages so that the rest can follow along". The second time I probably emphasized "very convenient" and the third one went with "essential", but none of those manage to effect a change; they still kept doing their test on the fly and just trying to remember how the did them the previous time. The funny thing is that every time I mentioned it, every one of the developers agreed that it would be great. That is where the inconsistencies of the title show up: they all agreed with the idea but they did not act on it. That is why this time I wanted to let them feel the heat.

Probably one part of the visceral reaction was driven precisely by this inconsistency. The outraged developer was in fact furious with themselves for failing to meet their own standards. They knew I was right to complain and they just wished they had thought about it before the test. But they had not so, in the opinion of a humble coffee shop psychologist, they projected their self anger (for failing) against me (for point out their failure). My hope is that the encounter today has been painful enough to them that they will change their habits, but this is far from a given. I will probably go to talk to them in person as soon as the home-office situation allows it, but not so much to apologize as to let them see why I had to do what I did. I can totally accept everybody's inconsistencies, because I am the first one to suffer from many, but that would not prevent me to try to make the best out of our joint effort. I will keep you posted with my progress. Have a nice evening.