[Cross-posted to RPG400-L and OPENSOURCE]
I'll be more inclusive with my quoting than I otherwise would be, so
that the context is visible on both mailing lists.
(Incidentally, I highly recommend that people use "reply all" instead
of just "reply".[1])
On Wed, Jul 5, 2017 at 1:40 PM, Aaron Bartell <aaronbartell@xxxxxxxxx> wrote:
Opinion:
The one area of RPG we have no control over is also one of the things that
is most frustrating; specifically the legacy-laden syntax. Yes yes, RPG is
awesomely stable and integrated, and yes, it has fairly simple modularity.
But it has held on to too many legacy things in its latest 100% freeform
(the chance they had to be rid of some of the legacy stuff was not taken
unfortunately).
Are you saying that you wish IBM had broken backward compatibility?
I've seen this with the move from Python 2 to 3, and it wasn't pretty.
Or do you mean you wish IBM had kept less legacy *functionality*
within the free-form *syntax* when they originally embarked upon
designing free-form? For example, they *did* drop GOTO from free-form
syntax. Are you saying you wished they did more of that?
Steve Sibley's(n1) recent declaration of IBM's continued commitment to ILE
compilers is good to know, but it is important to realize what's being
celebrated as historical evidence of continued success and support; namely
RPG OA and 100% freeform. RPG OA was too little too late to make the
difference necessary for IBM i customers to adequately compete. And I
already commented on 100% freeform above. In short, IBM is keeping RPG
alive, but they've made no moves to show they intend for it to thrive.
n1 - http://krengel.tech/steve-sibley-2017
I fully agree regarding OA. I wonder what it would take to make RPG
*thrive* though. If you mean taking RPG and turning it into something
like Rust or Go, but keeping all the great integration with DB2 (and
the IBM i in general), we're talking some serious, serious development
resources.
Honestly, if you measure "thriving" by ecosystem health and community
participation, I doubt syntax or "language feature" moves are the best
choice. I think it would be more effective to throw their weight
behind building a proper standard library, more like what Java or
Python have. And this could be a collaborative, open-source process
that includes both IBM and the RPG community at large.
The question I ask myself is if I would use RPG for a brand new
application. I always answer "no" based on my above opinions and also the
significantly smaller community.
Very understandable.
Of much more intrigue are things like iSeriesPython (if somebody wanted to
stick with the ILE environment yet have a language with a better syntax and
the ability to pull from a large community repository).
Again, just my opinion.
There is no IBM i news that would make me happier than to learn that
IBM has been in secret talks with Per Gummedal to take over
iSeriesPython, update it to Python 3, and fix a lot of the niggling
issues. There is such a great base to work with there (both in terms
of code and ideas). But I guess IBM would see it as cannibalizing the
effort they already put into their PASE Pythons.
John Y.
[1]For people using Gmail, go ahead and configure the default action
to be "reply all" instead of "reply". I find that instances where
"reply all" is the wrong move consist mostly of saying unflattering
things to a portion of the recipient list that you wish you excluded
(that is, "talking behind someone's back"). So if you engage in this a
lot, well, I guess maybe you should stick to just "reply". But for
most uses, I think "reply all" is what you *meant* to do. And if you
really *do* want to exclude people, it's easy enough to start with the
"reply all" recipient list and cross off the ones you don't want. It's
much more annoying to build a recipient list from scratch, or to
resend a message to someone you meant to include but forgot, isn't it?
midrange.com
