Re: Running SQL Workbench console from PASE

On Fri, Oct 9, 2015 at 3:56 PM, Justin Dearing <zippy1981@xxxxxxxxx> wrote:

On Fri, Oct 9, 2015 at 3:20 PM Buck Calabro <kc2hiz@xxxxxxxxx> wrote:

I can't figure out what the business value IBM is getting by taking on
the maintenance / support / documentation / distribution burden of
custom IBM i forks of popular OSS projects, versus an alternative of
modifying PASE to simply compile and run them as-is (after configure,
make, make install)?

THIS A thousand times over, and I think gcc and python fills a lot of the
gaps to make the rest of the OSS ecosystem work, or easily portable.
However, some projects do need attention, and internal resources to make
sure they continue to run on the IBM i. They should follow the example of
Microsoft with projects like jquery, the linux kernel, openssl, and openssh.

[Big spiel about Microsoft's involvement and contributions.]

I think you read Buck's post almost exactly opposite to the way I read
it. To me, it sounds like Buck is saying that IBM is *not* doing a
good thing (for IBM) by increasing its involvement and assuming more
responsibility. Buck is, at the very least, saying that it costs
something for IBM to be doing this. What I believe he is contending is
that IBM, if they're going to expend resources for the purpose of
expanding open source on the i, they should do it by "completing"
PASE. That is, PASE began life as "just enough of an AIX" to get
certain pet Unix projects running (specifically Smalltalk, according
to Frank Soltis). It has expanded and improved since then, so it's
closer to a "full" AIX, and thus closer to a full Unix system, but
isn't all the way there yet. Notably, it doesn't even ship with a
standard C compiler out of the box. I think Buck is proposing that IBM
just commit to making PASE really and truly a full-fledged Unix, with
everything that a Linux or Unix developer could reasonably expect, and
just ship THAT. And then any other open source stuff can run and be
built upon that. Instead of this endless parade of new options to
5733-OPS.

In other words, Buck is arguing for *minimizing* IBM's further
involvement, basically the opposite of what you've described Microsoft
as doing.

(I'm just paraphrasing how I understood it. I am not sure the cost of
"finishing PASE" and then potentially keeping it up to date is smaller
than their current approach.)

John Y.