[Chicago] Python Hype

Fri Apr 29 17:45:08 EDT 2016

The "Gartner Hype Curve" is enterprisey, unscientific management gobbledy-goop.  The kool-aide at Deloitte must be strong.

Let us look at what is being "graphed".

The x axis is time but without any metric on the axis.  The most you can say is that points to the right come after points on the left.  Imagine buying a bond where the issuer promises to pay you in the future, but refuses to specify how far in the future.

They y axis is the most squishy and enterprisey.  It is labelled "Expectations".  This is not a metric (measurable) quantity.  It is impossible to compare one set of expectations to another set for exactly the same reason that a set of strings is not comparable to a set of integers.  But wait, you may say, if we are talking about the same technology across time aren't we talking about the same set of expectations?  No, the graph is implicitly saying that set of expectations is a function of time, i.e. the set of expectations changes with time.  If the expectations are changing then whatever metric function you are applying to the set is either itself changing or is extremely simple.  Either way the output is essentially meaningless.

Let's substitute 'expectation of a financial return' for expectation.  This would give us a graph that would essentially be a stock price chart.  The Gartner curve does not look like the graph of any successful technology company - not Apple (APPL), Microsoft (MSFT), Oracle(ORCL) or Intel (INTC).  The closest stock chart I've seen to a Gartner curve is Textura's (TXTR) and I don't think that the event points on the hype curve have any relation to the expectation of financial return for that company.

Maybe you can sort of see the curve occurring in stock charts.  Maybe Apple is overlapping hype curves of Mac, IPod, and IPhone.  This is the province of chartists who look for patterns like "head and shoulders" and "support/resistance levels" in stock prices.  These patterns occur in a random walk as well.  I won't be winning anything on Quantopian.com by automating the recognition of either chart patterns or the hype curve.

The Python Foundation will never issue dividend checks or provide the founders with a payoff when they get acquired in a buyout even though a lot of people have invested in Python.  So we don't even have money as a proxy for the y axis.

So we can't measure where we are on the curve, either on the x or y axis, and it has no predictive capability except as a sequence of 'events'.  There is no science in applying it.  As a paradigm it cannot be proven wrong nor can it be proven right.

It does posit that the next to final step is third generation tools and libraries.  I think we have that in Python for web (1:WSGI, 2:Zope 3:Django/Flask), graphing (1:Tkinter Canvas, 2:Matplotlib, 3:Bokeh) and many other fields.  The final prediction of the curve is 20 to 30% adoption.  I don't think that any modern language will ever get there for the same reason that NBC will never again have 30% market share.  There is too much competition today.  Even in the days of Cobol/Fortran domination there were plenty of RPG and 360ASM jobs.  And the curve calls the final time the Plateau of Productivity.  We are certainly productive.  So in answer to your question, that is where we are.  And it doesn't mean anything that the hype is rising.

Phil Robare

From: Chicago [mailto:chicago-bounces+proba=allstate.com at python.org] On Behalf Of Brian Ray
Sent: Thursday, April 28, 2016 9:35 AM
To: The Chicago Python Users Group <chicago at python.org>
Subject: [Chicago] Python Hype

I am working on a talk regarding "Python Hype"  https://www.linkedin.com/pulse/python-hype-brian-ray<https://urldefense.proofpoint.com/v2/url?u=https-3A__www.linkedin.com_pulse_python-2Dhype-2Dbrian-2Dray&d=CwMFaQ&c=gtIjdLs6LnStUpy9cTOW9w&r=VXIryE9UwJGlNMLzgMzDT4_t2NMrZf6alSphHwSEwC0&m=BWZTPsddtmLmZLnvndpsWWmDDTjPCE94pHKONtGd50E&s=j82bEB-JdLR-EvcZWzyGzmuQoLlk-Qupi0BIc4kJn6Y&e=>

Would love some feedback. As you are my key support group, I'm looking for your opinions, thoughts, comments, concerns....

Also, expect a Survey soon to follow.

Thanks!

--
Brian Ray
@brianray
(773) 669-7717
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/chicago/attachments/20160429/b78df908/attachment.html>