[tex-live] strange discrepancy in running time of etex between TL2015 and TL2017

[jfbu]

Le 2 sept. 2017 à 14:52, jfbu <jfbu at free.fr> a écrit :

> Perhaps Apple does not distribute security patches
> (I think it must do in extreme situations like the bash
> vulnerability of a year or two ago) for Mavericks, i.e.
> for a system released only 5 years ago, but I do not wish
> to contribute to this "programmed obsolescence" whose only
> aim is to bloat software to force people to buy new hardware.


Hi Richard,

(I don't currently subscribe to the tex-live list, hence
I saw your message only on visiting

http://tug.org/pipermail/tex-live/2017-September/date.html

could you please CC to my address if replying?)

of course I wish to thank you and all those contributing
to MacTeX, and the TeXLive team for still supporting
an extended range of older mac os x systems
(I think it is Mojca who provided the time and resources).

I was confronted with the fact that TL2017 darwinlegacy
for Mavericks execute slower than darwing binaries 
from TL2016, TL2015, ...
(my first awakening was with luatex/texlua, where on one occasion
I had found
execution time to be doubled, then later I realized also
pdftex was affected to a lesser but significant extent).
Never ever I would blame people who did all the hard work
and have the expertise I am lacking to do it, but as an
end user I can only note the situation and then try
to find a solution. I definitely would compile myself the
binaries if I was not too afraid of endless problems
with missing dependencies on my system or subtleties
of C compilers etc... 

You reacted in your message to my mentioning of "programmed
obsolescence", but I think it is indeed a vital mechanism
for Apple, as demonstrated by its policies regarding
smartphones and tablets. Not owning any of those I can't
comment more, but I have heard enough interesting real
life experiences from people who do.

You stress that we live in an era of worldwide security
attacks. Then, why should Apple stop providing security
patches for the system it delivered in 2011 or 2012?
Isn't that a breach of trust with its users?

You stress that upgrading is free.
Ok, but not always free of problems.
I have an extensive range of installs via Fink for example.
Upgrading the system will require me upgrading the Fink
binaries. This takes hours. This is lame excuse perhaps,
but I recall hearing that switch to Yosemite was not painless
with the business that the computer appeared to be stalled
for 24 hours when handling /usr/local (I have forgotten the
details and it might not be Yosemite)

Computer hardware and all Apple gadgets like AppleWatch
use precious mineral resources and I am sorry to observe
that zillions of people
have no issue with throwing the thing to the "trash"
frequently and buy the new model,
in an attitude of frenetic consumerism.

I am just the slower guy who knows he owns an extremely
sophisticated product of human engineering. I am grateful
for that but I am afraid
that by upgrading incessantly the main software
I will observe that little by little this hardware will
prove insufficient, be it for memory space or processor
speed or whatever. Yes, this is an attitude of mis-trust
on my part respective to Apple, an attitude
which your more extensive experience indicates might
be mis-guided.

Nevertheless, I stick to my point that when the main
change of a system upgrade by Apple is to unify the
looks of the windows to those in use on smaller screens
like smartphones which I don't own, or to make it easier
to store my private data on "Cloud" owned by Apple or
whomever, I prefer to stay
on the cautious side. Because I gain something with
absolutely no interest to me and I don't know what I may
lose in the process.

Best,

Jean-François