On 26/04/2009, at 12:19 PM, Richard Schoen wrote:
As I am one of those lazy vendors you are talking about I have to
disagree with your summary :-)
That's OK. I never expect anyone to agree with me initially--funny
thing is that after a while they usually come around.
PASE, Qshell, Java, PHP etc... really enhance the "i" by allowing us
to
do more than just RPG and CL applications.
That's just marketing fluff and only works while you have a need for
OS/400 applications. If that need ceases, perhaps because you've built
a business system around PASE applications, then why would you keep
Unix-y things running on a non-Unix box? Especially when most sites
have Unix and WinDOS systems too.
Java yes, PHP maybe but most of what you might do with that can be
done in Rexx which we've had for years with no real interest from OS/
400 programmers, QSHELL is a quick fix to handle stream files in a
more Unix-y fashion primarily to assist Unix vendors, PASE exists
solely to assist Unix vendors. I only use QSH because IBM did not
provide a complete suite of Native tools to deal with stream files
(e.g., scan for text). I only use PASE because QSH does not provide a
full suite of expected Unix tools (e.g., locate differences in stream
files).
I can tell you that we've been able to meld Pase, Qshell, RPG and CL
applications together to provide solutions for our customers that we
would have had to job out to Windows servers without the tools.
String and sticky-tape. Great, you solved a business problem but
you've made the solution less dependent on OS/400 and as a result
easier to move off the system. At some point someone will notice and
ask why is this running in OS/400 when it could just as easily run on
other hardware we already own (for which they perceive a greater
business need).
The funniest thing about this is you solved the problem in a Unix
fashion--this fed into that and massaged with something else (e.g.,
find results fed into grep, result into sed and then into awk equals
business problem solved--I used to think that was neat but I've grown
up).
I believe melding the technologies is key to promoting and keeping the
iSeries viable as opposed to your opinion that these are driving nails
into the iSeries and OS/400-i5/OS coffin.
Nah, it's viable anyway if people would simply activate their brain
and think about things. It's viable because it can interoperate with
other systems. It's viable because it is a better system. Its main
feature is that it is different from other systems (which at their
core are pretty much identical). However its main detraction is that
it is different from other systems. Hmm, how to solve that dilemma?
Be very wary of relying on PASE applications. You are helping drive
nails in your own coffin.
This comment is pure FUD.
Watch this space and remember you heard it here first. I said similar
things 15 years ago regarding the lack of services revenue provided by
OS/400 and that it was a contributing factor in it's decline. It seems
Gerstner formed a similar opinion and directed IBM accordingly. Seems
I was right about that and I expect I'll be proven right about this.
I think OS/400 and its hardware-du-jour is the best business system on
the planet but until those holding the purse-strings use their brain
instead of following the masses OS/400 will remain marginalised. I
dislike pointing out the downside but if you and others blindly
install PASE applications and cobble solutions together from various
AIX components you will find that you've taken the easy path and a
price will be exacted from you in future.
I saw the effect supporting Win3.1 applications had on OS/2 and I can
see the same thing happening here. It initially appears to be a great
idea but the expectation that vendors will create native versions of
their applications to take advantage of native features (e.g., multi-
threading, more RAM, etc. in the case of OS/2) is naive. Most of them
are simply too lazy (or, if I'm being generous, think the short-term
cost is prohibitive) and consider the application they provide is
"good enough" and since "we" seem to accept that they are given
justification.
It's OK if you don't embrace new technology, but these technologies
will
help insure that the iSeries remains viable.
Hah! I co-wrote one of the first Java Redbooks. I'll use new
technologies where they are appropriate. Using an AIX/Unix application
is appropriate on a Unix box. It's not appropriate on OS/400.
For example, why does PHP require its own web server? Why is that
running in PASE? Why can't it use the IBM Apache server (as more than
just a conduit)? Hmm?
I still say that John would be better served spending time learning a
new technology rather than trying to re-write a complex application
such
as SFTP to RPG. That's just unproductive.
I agree. I never said he should rewrite in RPG. That's simply a stupid
waste of time. However, making a Native port of sFtp (or any other
application) by running the existing C source code through the ILE C
compiler is not a waste of time. I've already done that for a few open-
source utilities and added OS/400 extensions in the process. For
example, adding CCSID support to GZIP so you don't lose the CCSID of
the original file when you unzip it. That's the proper way to do things.
Anyway, it doesn't matter. I'm trying to stay out of these arguments.
I'll say my piece and if you don't agree that's fine. Just watch how
things unfold.
Regards,
Simon Coulter.
--------------------------------------------------------------------
FlyByNight Software OS/400, i5/OS Technical Specialists
http://www.flybynight.com.au/
Phone: +61 2 6657 8251 Mobile: +61 0411 091 400 /"\
Fax: +61 2 6657 8251 \ /
X
ASCII Ribbon campaign against HTML E-Mail / \
--------------------------------------------------------------------
midrange.com
