Re: Need support for the IBM support

Hi,

You made a few good points but I would like to differ with you in some.

companies usually are not interested in finding bugs.
This certainly is not true. IBM does spend a lot of money in fixing
bugs for customers, and it is generally more costly to fix a bug once
it has reached the field. IBM does spend a lot of money in finding
bugs, so that they can fix them pre-release. This makes good business
sense for IBM for two reasons:
1 - a bug in the field mars IBM's image to some extent
2 - Since the cost of fixing it pre-release is lesser, it means more
profit for IBM

If you have to fix a bug, this binds resources that could otherwise be
used to develop flashy new features for the next release to make marketing
The key to success in any UNIX distribution is the quality of support a
company gives. The major factor large customers (for eg. Boeing,
Nestle) use UNIX and not linux is because of the support available for
it (which includes bug fixes), inspite of the fact that Linux has a
richer set of features than UNIX.

a) you have a support contract
Usually IBM is only bound to give support if you have a support
contract. But, even if you don't have a support contract, IBM usually
gives you some sort of help. Keep the customer happy is IBM's policy.

b) you are ready to upgrade to the latest bleeding edge version
Not necessarily. IBM still supports 53E, 53H, 53D, 52Q, 52M, 52L and
even though 51G has come to the end of its life-cycle - critical
patches are still being dropped into 51G.

c) you are persistent enough to wade through layers of support personell
whose job it is to keep the bugfix requests from the developers
This could be a problem in certain parts of the world.

d) you can pester them enough to actually reproduce the bug and
Reproduction of a bug is usually a major hurdle developers, and support
personnel face. In many cases, customers use proprietary software in
their AIX boxes which IBM does not have access to the source code of,
sometimes even the binaries of. Without the actual recreation scenario
(a step-by-step set of steps) and only a vague description it is really
difficult to recreate a problem and to fix it. Even if a fix is arrived
at, it is difficult to test it.

e) the next release is too far away to say "Wait for release N+1, I'm
sure the problem goes away in this release".
You need not wait for the next release. PTFs (patches) are available
from fixcentral for all problem fixed. These PTFs are made available in
the general release of the next level, but if a bug has been fixed,
you'll find the PTF in fixcentral.


Thanks and regards,
Rajbir Bhattacharjee



Laurenz Albe wrote:
Piotr Wyderski <wyderskiREMOVE@xxxxxxxxxxxxxx> wrote:
No, I give up. I don't want and have never expected any support
from IBM; I just want to report them a bug. If IBM can't simply
provide a freely-accessible bug reporting web page, then let them
find their bugs themselves.

I understand your feelings.

Contrary to what most programmers or open source developers expect,
companies usually are not interested in finding bugs.

If you have to fix a bug, this binds resources that could otherwise be
used to develop flashy new features for the next release to make marketing
happy.

In my experience, a software company will only fix a bug if
a) you have a support contract
b) you are ready to upgrade to the latest bleeding edge version
c) you are persistent enough to wade through layers of support personell
whose job it is to keep the bugfix requests from the developers
d) you can pester them enough to actually reproduce the bug and
e) the next release is too far away to say "Wait for release N+1, I'm
sure the problem goes away in this release".

So IBM is not worse than others in this regard.

Yours,
Laurenz Albe

.

Relevant Pages

  • Re: How to address fixes in OS/2 to IBM ?
    ... It isn't really a bug. ... It's something IBM has already told us they won't fix. ... The variable that is causing the error is "fTaskPtr" being non initialized, the rest of the initialization of GRADD variables might not be strictly necessary but does not hurt either. ...
    (comp.os.os2.bugs)
  • Re: Third party Maintenance of z/OS R4 after March 2007
    ... MVS/XA system that was in another country. ... by "support". ... I understand that IBM means they will not fix new problems, ...
    (bit.listserv.ibm-main)
  • Re: 11.5 Stability ?
    ... and new features then it's fine to try at FC1. ... One good point that may not be obvious for the customers: When IBM finds a bug ... released (at least when you look back at Informix' history). ...
    (comp.databases.informix)
  • Re: Need support for the IBM support
    ... how do I submit a bug report to IBM? ... all our attempts to report it via on-line ... support have failed. ... If IBM Poland refuses to open a PMR because you are using an evaluation ...
    (comp.unix.aix)
  • Re: IBM PL/I support is still available
    ... I know because I have been in contact with IBM Support Germany and the PL/I ... And it's not a compiler bug, it's a runtime library bug - hepwm20 is the ...
    (comp.lang.pl1)