Reporting bugs to the net (was Re: Another bug in DEC Fortran) 
Author Message
 Reporting bugs to the net (was Re: Another bug in DEC Fortran)

|> Thanks for giving me yet another opportunity to ask that if you
|> think you've found a problem with a compiler (Digital's or anyone else's)
|> that, before you post to the net, you report the problem to the vendor's
|> support organization.  (At the very least make sure you're running the current
|> version!)  If you don't get satisfaction from the vendor then I suppose it's
|> reasonable to post here - just be aware that you may not get an authoritative
|> response.  If the problem is fixed in the current product release, you may not
|> get much sympathy, either.

In particular, I think the following bears repeating, loudly and often:

|> General newsgroups should not be used as a substitute for reporting
|> problems to a vendor's support organization.

Well said!!

Thanks, Steve.

----------------------------------------------- Roger Glover
XXXX\ \ / \ /XXX  \ / \ X   \ /\\\  X///X /\\\  Cray Research, Inc.
/ \ / \/ /\ / \ / \X /\ X  \  / \   X\  \ X     DISCLAIMER HAIKU:
//X/  X\\\X //X/  X \ X X\\   / \   X/X \ X \\\   CRI may not
/ \   X///X / \/  X//XX X  \  / \   X  \\ X   \   Share these opinions of mine
/ \   X   X /\\\/ X   X X///X /XXX/ X///X /XXX/   This is not my fault



Sat, 14 Jun 1997 20:11:01 GMT  
 Reporting bugs to the net (was Re: Another bug in DEC Fortran)

: In particular, I think the following bears repeating, loudly and often:
: |> General newsgroups should not be used as a substitute for reporting
: |> problems to a vendor's support organization.
: Well said!!

Actually, the title of the post clearly indicated that it was about a bug.
Anyone who is not interested can pass it.  I am not particularly interested
in the details, but I am interested in the fact that the post was made.  We
are going to be buying PC compilers shortly, and the number of posts about
this compiler tells me that we don't want it.  Obviously, Microsoft falls
into the same category for non-UNIX Intel.  We haven't seen many bug posts
about SunSoft Intel f77, so we're favorably inclined towards that one.  The
point is simply that these posts do convey useful information, and I'm not
inclined to discourage them as long as their title indicates the contents.




Tue, 17 Jun 1997 05:55:11 GMT  
 Reporting bugs to the net (was Re: Another bug in DEC Fortran)

Quote:
>: In particular, I think the following bears repeating, loudly and often:
>: |> General newsgroups should not be used as a substitute for reporting
>: |> problems to a vendor's support organization.
>: Well said!!
>Actually, the title of the post clearly indicated that it was about a bug.
>Anyone who is not interested can pass it.  I am not particularly interested
>in the details, but I am interested in the fact that the post was made.  We
>are going to be buying PC compilers shortly, and the number of posts about
>this compiler tells me that we don't want it

[EDIT]

I'd take the bug reports with a big grain of salt.  Lots of people are ready
to blame the compiler when it's something else.  I sure wouldn't base a big
software purchase on Usenet bug reports alone. (Not that you said you would...)
A question in the form of "is this a bug or am I doing something wrong?" would
seem fair game to me.

There is a need for some place to report real, suspected, or imagined bugs,
but I'm not sure the lang groups are it.



Tue, 17 Jun 1997 00:55:59 GMT  
 Reporting bugs to the net (was Re: Another bug in DEC Fortran)

Quote:
(Mike Boucher) writes:

>Actually, the title of the post clearly indicated that it was about a bug.
>Anyone who is not interested can pass it.  I am not particularly interested
>in the details, but I am interested in the fact that the post was made.  We
>are going to be buying PC compilers shortly, and the number of posts about
>this compiler tells me that we don't want it.  Obviously, Microsoft falls
>into the same category for non-UNIX Intel.  We haven't seen many bug posts
>about SunSoft Intel f77, so we're favorably inclined towards that one.  The
>point is simply that these posts do convey useful information, and I'm not
>inclined to discourage them as long as their title indicates the contents.

I don't think you can really tell much of anything based on the number
of "bug" reports.  My experience is that most such reports end up being
user errors or misunderstandings of the language.  The other thing to
understand is that some compilers are much more widely used than others -
meaning that you're statistically likely to see more postings about them.

You also have to have a good understanding of just which compiler it is
being discussed, what the nature of the problem is and how relevant
that is to you.  I must admit I'm a bit confused at the last part of
your posting quoted above,  Which compiler are you referring to?  The
title was from a note reporting a bug in DEC fortran for OpenVMS VAX -
probably the first bug report against that compiler I've seen here in
months, despite it being one of the most widely used Fortran compilers
in the world.  Also, the bug had a fix released for it a year ago and
involved a rarely used extension.  Now how does this relate to your
wanting to buy a PC compiler?  I don't know anything about the SunSoft
PC compiler - I don't recall anyone here commenting on it.  (I didn't
even know they made one.)

I would like to clarify my position regarding posting bugs to the
newsgroup.  What prompted my initial outburst was having received
two mail messages, over the span of three days, from different people
saying "Here's a bug report against DEC Fortran which I just posted to
comp.lang.fortran."  My response was "Why?"  While I'd prefer that
people report problems to our service organization (which is able to
provide correction kits if necessary), I understand that some people,
especially at universities, don't have ready access to vendor support
and hence turn to the newsgroup.  That's fine.  What I don't understand
is why people who already know how to contact the vendor go ahead and
post anyway.  In darker moments I envision their perhaps doing so for
the pleasure of being able to show off "in front of the class".  Nearly
all of the true bug reports I see (which are a small fraction of the
claimed bugs) are obscure and difficult to characterize, and hence of
little interest overall.  And nearly all of them already have corrections
available from the vendor.

I'm glad to get bug reports wherever I can find them, but I don't think
it's right to turn this newsgroup into a transcendant bug list.


DEC Fortran Development           WWW:  http://www.digital.com/info/slionel.html
Digital Equipment Corporation     CompuServe: 75263,3001
110 Spit Brook Road, ZKO2-3/N30
Nashua, NH 03062-2698             "Free advice is worth every cent"




Tue, 17 Jun 1997 09:54:28 GMT  
 Reporting bugs to the net (was Re: Another bug in DEC Fortran)

Quote:

> I would like to clarify my position regarding posting bugs to the
> newsgroup.  What prompted my initial outburst was having received
> two mail messages, over the span of three days, from different people
> saying "Here's a bug report against DEC Fortran which I just posted to
> comp.lang.fortran."  My response was "Why?"  While I'd prefer that
> people report problems to our service organization (which is able to
> provide correction kits if necessary), I understand that some people,
> especially at universities, don't have ready access to vendor support
> and hence turn to the newsgroup.  That's fine.  What I don't understand
> is why people who already know how to contact the vendor go ahead and
> post anyway.  In darker moments I envision their perhaps doing so for
> the pleasure of being able to show off "in front of the class".  Nearly
> all of the true bug reports I see (which are a small fraction of the
> claimed bugs) are obscure and difficult to characterize, and hence of
> little interest overall.  And nearly all of them already have corrections
> available from the vendor.

> I'm glad to get bug reports wherever I can find them, but I don't think
> it's right to turn this newsgroup into a transcendant bug list.

1)  Of course bugs should be reported to the vendor.  This can be done
    either formal via official channels or via more unofficial channels
    such as email to someone in the development team. This first should
    always be done, but it can be a tremendous task to get through
    all the desks (and a large fraction of the desks are at the customers,
    so this is not vendor bashing).

2)  But I do think that it is appropriate to ALSO post to a relevant
    news-group/mailing-list about the bug, if one consider the bug to
    be so relevant that other people could be bothered by the same bug.
    And the fact that the bug is fixed in a new version is not necesarry
    relevant, because a lot of sites do not upgrade immediately for
    various good reasons. (and if you want an example from your own
    backyard: we are still using DEC FORTRAN 6.1 on most of our machines
    because 6.2 needs VMS 6.1 and if we upgrade to VMS 6.1, then we will need
    to install DECnet/OSI, because we use PSI software today and that will
    not work on VMS 6.1, and we have not been espacially eager to install
    DECnet/OSI, but 2 week and then 7-9-13)

                                                                   Arne

Arne Vajh?j                             local DECNET:  KO::ARNE
Computer Department                     PSI:           PSI%238310013040::ARNE

                WWW URL: http://www.hhs.dk/~arne/arne.html



Wed, 18 Jun 1997 20:50:09 GMT  
 Reporting bugs to the net (was Re: Another bug in DEC Fortran)

Quote:
(Arne Vajhoej) writes:
>    (and if you want an example from your own
>    backyard: we are still using DEC FORTRAN 6.1 on most of our machines
>    because 6.2 needs VMS 6.1

That's not true for VAX, you could be running a VMS version as old as
5.4.  On Alpha, the documentation says that OpenVMS AXP 1.5 is the
required minimum version, though if you're running 1.5 you may sometimes
need to add a switch to the command line in order to have the compiler
not use a new math routine - see the release notes for details.


DEC Fortran Development           WWW:  http://www.digital.com/info/slionel.html
Digital Equipment Corporation     CompuServe: 75263,3001
110 Spit Brook Road, ZKO2-3/N30
Nashua, NH 03062-2698             "Free advice is worth every cent"




Thu, 19 Jun 1997 00:07:54 GMT  
 Reporting bugs to the net (was Re: Another bug in DEC Fortran)

Quote:
: (Mike Boucher) writes:

[munch]
: I must admit I'm a bit confused at the last part of
: your posting quoted above,  Which compiler are you referring to?  The
: title was from a note reporting a bug in DEC Fortran for OpenVMS VAX -
: probably the first bug report against that compiler I've seen here in
: months, despite it being one of the most widely used Fortran compilers
: in the world.  Also, the bug had a fix released for it a year ago and
: involved a rarely used extension.  Now how does this relate to your
: wanting to buy a PC compiler?
I was simply making the point that bug reports that I find on the net help
us to make buying decisions, in our case for a PC compiler later this year.
Someone wrote to ask that bug reports not be posted, and I was writing to
respectfully disagree.  If the title of the posting indicates that it is
a bug report then people who don't want to hear about it can avoid it.

At the risk of introducing more irrelevance than I have already (inadvertantly)
done, I should say that we like SunSoft's compiler a lot.  I say that because
you mention here and others have mentioned in email that my posting seemed
to say that SunSoft's compiler was not good and I think it is.

: I'm glad to get bug reports wherever I can find them, but I don't think
: it's right to turn this newsgroup into a transcendant bug list.
I agree, but I don't think it's gotten out of hand.




Sat, 21 Jun 1997 02:12:33 GMT  
 Reporting bugs to the net (was Re: Another bug in DEC Fortran)

 > I am not particularly interested
 > in the details, but I am interested in the fact that the post was made.  We
 > are going to be buying PC compilers shortly, and the number of posts about
 > this compiler tells me that we don't want it.  

This seems a remarkably poor way to make the decision in view of the
selection of data concerned.  Is there any evidence that the number of
bug reports posted to usenet is correlated with whatever features of
the compilers concerned are of interest any more than personal
testaments are correlated with medical efficacy?



Sun, 22 Jun 1997 01:55:44 GMT  
 Reporting bugs to the net (was Re: Another bug in DEC Fortran)


:  > I am not particularly interested
:  > in the details, but I am interested in the fact that the post was made.  We
:  > are going to be buying PC compilers shortly, and the number of posts about
:  > this compiler tells me that we don't want it.  

: This seems a remarkably poor way to make the decision in view of the
: selection of data concerned.  Is there any evidence that the number of
: bug reports posted to usenet is correlated with whatever features of
: the compilers concerned are of interest any more than personal
: testaments are correlated with medical efficacy?

IMHO
It is a useful counterballance to reading the adverts, hype articles in the
trade papers, handouts coming from the sellers of the product.  Also useful
counterballance to the statements of the true believers...

Of course something "Consumer Reports"(USA) or "Which"(UK) might give
better data.  So would a scientific survey...

It is best of course to do a thorough test for several years before
buying the product.... but that is not as feasible as letting several
people try it independently for a week or to and watching what they
say.

Note: this message is going to several different groups and I'm
not saying that your language/group has more bugs, hype, or faith than any
other.

I could have saved weeks of development time if I had known
that some of the problems I was trying to fix were not my bugs at all but
were in the OS/Compiler/Hardware.

--

Disclaimer::=`CSUSB may or may not agree with this message`.
Copyright(1994)::=`Copy as long as you include this copyright and signature`.



Wed, 25 Jun 1997 04:24:26 GMT  
 
 [ 11 post ] 

 Relevant Pages 

1. BUGS, BUGS, BUGS, BUGS, C4 BUGS

2. strange problem with DEC fortran !!! Bug or feature

3. Bug in DEC Fortran V6.2-508

4. BUG REPORT: Color bugs in tk4.0???

5. I can't find my bug report in the Scriptics Bug DB.

6. expr bug report not a bug

7. Bug reporting for Intel Fortran

8. g77 non-bug I reported (actually in NR-Fortran )

9. PopupMenu bug, menu bug, apple bug (worm)?

10. Tcl8.0p1 error reporting bug report and patches

11. DEC Fortran 90 Price Reduction, Combined Packaging with DEC Fortran

12. System Bug, Or Am I Really Dense?

 

 
Powered by phpBB® Forum Software