https://bugs.documentfoundation.org/show_bug.cgi?id=172799

--- Comment #18 from ady <[email protected]> ---
A teenager decides to ride his bike. He noticed the tyres (aka tires) are
deflated, so he pumps up them. While riding, he sees a big truck is stuck under
a bridge, and people are gathered around discussing how to remove the truck
without further damaging the bridge. The teenager suggests to slightly deflate
the truck's tyres.

The next week the teenager visits CERN. For the trip to it, he takes with him
an apple. At CERN, several physicists, engineers and technicians are discussing
a specific problem in the collider. Even if the teenager has heard about an
apple fallen because of the force of gravity, the chances he can really opine
about the problem with the collider with any relevant idea would be extremely
close to zero.

There is no real explicitly kind way of proving someone else that he does not
have the necessary knowledge and experience. Eyal does not have the necessary
knowledge and experience regarding spreadsheet Functions and formula errors. I
do not want that to sound subjective or personal or mean. It is not my
intention. There is no simple way of expressing that in written, somewhat
concise language; at least not with my language skills.

OTOH, saying to someone that you are happy to see him go away, is not kind. I
do not have the language knowledge in order to potentially interpret that
phrase as meant with kindness, unless my life is being threatened by a mob (or
by one individual) and you are worried that if I stay I would get hurt.

As for Ben having the required knowledge, I do not know. Assuming that he does
have it (which is not limited to having the coding skills, FWIW), I do have my
doubts regarding the required experienced. Compatibility with other spreadsheet
tools regarding spreadsheet Functions is not a mere optional; it is a very
important factor, a goal. Even when a Function returns "the wrong" result, it
matters, because after decades of using those functions users already take
advantage of those specific "wrong" results in order to obtain the actual
result they are seeking.

Being able to modify the LAMBDA() function in an incompatible way speaks about
coding skills, but not about experience. You need to keep contact with those
that take care of ODF future versions. You need to be mindful of the problems
that incompatibilities bring. You need to be aware that modifying a spreadsheet
Function in Calc, whichever the bug report that triggers that modification,
will have negative consequences for users that have already been using that
Function when it was behaving differently. Additionally, it speaks about QA and
reviewing patches, since an experienced and knowledgeable developer should have
raised concerns about someone modifying a Function in incompatible ways; the
patch should had been paused and not committed, until relevant discussions were
held – and my educated guess would be that it would had been rejected.

There are bug reports for Functions such as TYPE(), COUNTIF(), COUNTIFS()
XLOOKUP(), XMATCH(), DATEVALUE()... There are important enhancement requests,
for instance for tracing Precedents and Dependents (awaiting for years, with
tens of users subscribed to the bug report, and who knows how many more that
are interested but seeing nothing done they just move on to other spreadsheet
tools). In some cases the problem is against ODF, and/or weighting the current
behavior of the Function against the negative consequences of matching it to
ODF (and in some cases, developers with adequate knowledge and decades of
experience with Calc have opposed the "correction" to match a Function
according to ODF).

In the past I would had suggested to look for a mentor, with knowledge and
experience in Calc's development. Within TDF, or within current associated
organizations, there is no longer such developer available.

So, the matter of correcting or improving Calc's Functions is not that simple.
And even when it is, Calc is still lagging.

As for this tdf 172799, there is no one-to-one correlation between error codes
and specific Functions. The error code being displayed is not always the actual
problem with the specific formula. Someone with enough knowledge and experience
building workbooks with complex formulas would know that. Going back to the
start of this rant, a teenager that thinks that an apple falling from a tree
would be enough to answer all the questions about the Standard Model for CERN's
tests cannot even start to grasp the complexities involved.

-- 
You are receiving this mail because:
You are the assignee for the bug.

Reply via email to