[illumos-Developer] webrev: reimplementation of od

[Joerg Schilling]

"Garrett D'Amore" <garrett at nexenta.com> wrote:

> On Mon, 2010-10-18 at 21:13 +0200, Joerg Schilling wrote:
> > "Garrett D'Amore" <garrett at nexenta.com> wrote:
> > 
> > > > > Assuming I get some good reviews of this, I'll probably integrate it on
> > > > > Monday.  It compiles cleanly, so its simple to just d/l and compile if
> > > > > you want to try it yourself. :-)
> > > > 
> > > > -	Does not include largefile support
> > >
> > > Wrong.  Its related to my next point.  I explicitly support large files
> > > directly, and do not require a compiler option to enable it. 
> > 
> > I do not expect you to "see" the problem like I did within the first 2 minutes
> > and without compiling. I however expect people who like to integrate code to do 
> > tests. If you did test, you did know that the code does not work with large 
> > files as expected.
>
> Your comments are *always* so helpful Joerg.  You love to critique other
> people's work, but never provide adequate detail.  I don't think you
> understood my code correctly, because it was designed for large file
> support.  But whatever.   I'll change it to use the lf64
> -D_LARGEFILE_SOURCE.

The problem is that you sent the reply "Wrong" which signals that you believe 
you know better and that you are not interested in help.

I do not expect everybody to have the experiences in programming that allows me 
to "see" certain problems without running the software, but I expect that 
people test their code. If you did run a simple test with a large file, you did 
know why I wrote that there is no large file support. If you did ask, you 
could know too. As I mentioned above, you send the signal that you don't like 
help. I have been short with my report as I did see several additional 
problems and as I did not like to send a really long list that is not read by 
most people.

You cannot test everything if a program is bejond a certain complexity, you 
still can get a sufficient test coverage by testing the things that are 
expected to cause problems. This is how I get fast results.

Your "od" will create a lot of problems if it is going to replace the current 
implementation as it does not implement everything that is in the POSIX 
standard. I am definitely not able to use your code in case that it does not
support a sufficient amount of the expected functionality, so I would need to 
write my own od replacement. I did already spend enough time with this thread 
that would be sufficient to write a more complete od replacement on my own. I
need to know whether Illumos will be able to step in the Solaris footprints 
and deliver the quality previous Solaris users expect or whether there are 
other interests behind Illumos.

In order to continue the OpenSolaris business, we need more people than we 
currently have. Things will work if there is collaboration. Collaboration means 
talking about plans before coding and it also means not to start a taks that 
others could do better. Collaboration of course also means to value the work of 
others. The announcements for Illumos have been promising. Let us do work in a 
way that is needed in order to implement the announcements. It would be a sad 
experience if people need to start own forks on onnv because expectations on 
quality and collaboration are not met by Illumos.

Jörg

-- 
 EMail:joerg at schily.isdn.cs.tu-berlin.de (home) Jörg Schilling D-13353 Berlin
       js at cs.tu-berlin.de                (uni)  
       joerg.schilling at fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/
 URL:  http://cdrecord.berlios.de/private/ ftp://ftp.berlios.de/pub/schily