So here I am, doing my job, standing in front of a whiteboard asking questions posed through a translator about undocumented things I might or might not know something about when out comes this piece of international brilliance that takes three fucking hours. I am not sure if I’m pitting the customer, the translator, the system, or even myself, but something was terribly terribly wrong.
(loosely retold with most repeated parts skipped to avoid boredom)
(also it might seem confusing because a) It’s confusing b) I had to remove names and specific terms to avoid getting fired)
Customer (Engineering Manager): <About 10 minutes of monotone Japanese>
Translator: They have a concern.
Me (trying to look worried): What is < >-san’s concern?
T: <About two minutes of groveling Japanese>
C: <About five minutes of monotone Japanese>
T: nods turns to me They want to know how to do more with feature X when
interfacing with device A.
Me (trying to look understading): Ah, I see! Would < >-san like a better explanation of how to use feature X?
T: <About two minutes of groveling Japanese>
C & T: <Monotone ten minute discussion in Japanese with some numbered boxes getting drawn and erased on the white board>
T: < >-san understands how to use feature X. They want to know how to do more with feature X with interfacing with device like device A.
Me ( ): Like device A? Is < >-san anticipating a change in device A specifications? And what do you mean by “do more with”?
C & T: <More discussion and whiteboard stuff>
T: < >-san does not anticipate a change in device A specifications. However, when device A specifications were sent in by the manufacturer they were non-standard and unexpected. You even called it “weird” device A. < >-san requests information on how to use feature X with a more “weird” device B.
Me: Ah. What are the differences between device A and device B
T (Strangely already possessing this information): No no no, you do not understand! < >-san proposes a hypothetical device B that is more, as you say, “weird” and non-standard than device A. They are worried about interoperability. There is no actual device B, but they would like to know how to use your feature X with such a device if it is ever created.
Me: :eek: Can < >-san give an example of some non-standard or “weird” specifications that we could expect from such a device B that they want to use with our feature X?
C & T: <Twenty minutes of discussion and whiteboard stuff, resulting in two drawings>
T: < >-san cannot provide a specific example because device B does not really exist, but what if device B was like this or like this points at drawings
At this point I look at the drawings and kind of start to get the picture as to what they’re asking. Note that it has been roughly an hour and a half since we started on this question. Device A outputs data. It’s predecessors outputted data in the right order, but Device A decided to output data in a strange fixed pattern (a linear function of one variable) to avoid interference. We had to change Feature X to reorder the data and added ability to customize what the reordering function coefficient is. Their description of Device B was two examples of a non-linear polynomial reordering.
Me: Ah. Well if such a device B is built and < >-san chooses to work with it, we can modify feature X interface to allow for a custom polynomial reordering function. This will create additional parameters to define the polynomial.
C & T (Drawing, etc.)
T: What if device B is even “more weird?”
C (draws something non-polynomial. smiles. sits down)
Me: :eek: Like that? (point at the non-polynomial drawing)
C (In english): No, in general.
Me (Thinking: AHA, you bastard! You speak English. I knew it!): Well in general we cannot promise to support abstract real time reordering so anything more weird than device A will have to be considered when and if it is ever encountered.
C & T (Lively discussion in Japanese, CEM has a smug look, yeah he showed me! :rolleyes: )
T: < >-san understands your strategy for supporting the examples given. They are concerned about interoperability since flexibility has to be added on later on and is not built into the system from the get-go. I do not know enough about the subject to understand if they are happy about this understanding, but this issue is for now closed.
And people ask me why I don’t try to learn Japanese. Among many reasons, one of the key ones is that using an interpreter significantly buffers my homicidal drive. Sometimes I want to smack my co-workers for using terms like “weird device” or “garbage customer section” around Japanese customers. They become fixed proper nouns and then I have to sit there writing a document entitled “Garbage Customer Section: Handling and Best Practices”. Any attempts to rename things or refer to them using other terms results in confusion that can eat days if not weeks of engineering time (I kid you not).
I need a drink.