[Tutor] Python3 : Yes --- Python2 : No ?

[David L Neil]

On 21/09/19 4:50 AM, Albert-Jan Roskam wrote:
> On 15 Sep 2019 09:48, Alan Gauld via Tutor <tutor at python.org> wrote:
> On 15/09/2019 07:17, Hendrickx Marcel via Tutor wrote:
>> Link with the info =>
>> https://www.python.org/doc/sunset-python-2/
>> News for me...
> It really shouldn't be a surprise. As the link says its
> been announced several times and was always
> the intent.
> ==>> Some bankers pretend to be surprised:
> https://www.techrepublic.com/google-amp/article/jpmorgans-athena-has-35-million-lines-of-python-code-and-wont-be-updated-to-python-3-in-time/


My first temptation is to comment that business decisions are often made 
by managers and board members who have little understanding of computers 
and technology - which leads into *two* reasons why the same (type) 
demand that their specialist staff have five years' experience in some 
facility which has only been around for two!
(read Dilbert cartoons for similar)

Banking is no exception to this. To gain a good idea of banking 
management's understanding of technology (and finance!) ask yourself the 
'who' and the 'how' of the Global Financial Crisis?
(sorry, my sympathies require a microscope to see - and my advice, 
should your bank be acting similarly, is to consider moving elsewhere!)


That said, it is very difficult to put either a cost or a benefit on 
'technical upgrades'. Which is why the text-books say "there are no 
technical projects".

The business benefit of changing from Python2 to 3 has been dubious in 
the present, and difficult to comprehend in the future tense. There is 
no direct benefit in the sense of marginal income. Even on the technical 
side there have been pythonista feeling forced/quite comfortable staying 
with Python2 because the libraries they wish to employ are not (yet) 
available in Python3. I was reviewing a system yesterday for a PoC, and 
discovered (to my horror) that they have yet to complete their Py3 
upgrade. Plus the old saw: "if it ain't broke, don't fix it".

Thus, the technical viewpoint and criteria for assessment are a long way 
from those of our 'business' colleagues!


Further to Alan's dates: (*ten* years already!) I've just finished 
reading (a somewhat irreverent article, laced with British sardonic 
humor - which many other cultures just don't 'get') this week's "The 
Register":

<<<
At the turn of the century, panicked organisations found themselves 
calling in COBOL coders, often those responsible for the decades-old 
systems in the first place, to deal with the impact of those crucial 
extra digits.

That some today reckon the event was a bit of a damp squib, manufactured 
by highly paid contractors to add a Ferrari or two to the garage, is 
testament to the monumental efforts by those involved at the time. 
You're welcome.
 >>>

They are talking about a time when a *FORTY* year old language (and its 
code-base) was in need of checking and amelioration - a need which had 
been identified for about the same length of time as the language had 
been in-use (certainly, my education included an observation that 
two-digit year-dates were going to become an ever-greater problem).

	if this_year - birth_year > minimum_age: # allow...

	if 19 - 91 > 18: 	# ok
	else: 			# no, you're not old-enough to vote!
				# even though (you think) you're 28

Oops! Whereas:

	if 2019 - 1991 > 18:	# welcome to 21st century democracy!

So, this problem of 'delay' is nothing new!


WebRef:
COBOL: Five little letters that if put on a CV would ensure stable 
income for many a greybeard coder
EVALUATE COBOL-AGE WHEN 60 CONTINUE
By Richard Speed 16 Sep 2019 at 09:09
https://www.theregister.co.uk/2019/09/16/cobol_at_60/


Disclaimers:
1. This is no apologia, I have (almost) only used Py3 for the last five 
years (and probably longer) - that said, am keenly awaiting the release 
of CentOS8 (next week, dv), which upgrade will under-pin the migration 
of a long-running Py2 system. Once again: a project which is very hard 
to justify from the stand-point of 'conducting our business' and only 
comes under the 'technical debt'/don't store-up trouble for the future, 
headings!

2. I was PM on a Y2K Project, which became the first significant (and 
real-time) operation to enter the new century, and by 'living or dying' 
act as a bellwether for organisations world-wide. Somewhere, I have 
(multiple) certificates from organisations such as Microsoft, because we 
found issues before they did themselves - so it's not only bankers then! 
I've heard many criticise Y2K as 'overblown' or 'manufactured into a 
legend', however we 'sweated bullets' to make it look easy - and some 
who didn't paid a significant price, up-to, or is that down-to, the 
eventual collapse of whole companies.

3. I do have a gray/grey beard, but no, I don't drive a Ferrari - in 
fact, if you own a car, chances are it is newer and more 'flash' than mine!

-- 
Regards =dn