ps enhencements (posix syntax, and more)

[Garance A Drosihn]

At 2:31 AM +0200 6/7/04, Cyrille Lefevre wrote:
>Garance A Drosihn <drosih at rpi.edu> wrote:
>>At 12:03 PM +0200 4/27/04, Cyrille Lefevre wrote:
>>>here is a description of the last PR#64803 updates :
>>
>>The latest info I see in PR #65803 does not match some
>>things that you describe in the rest of this message.
>>The following comments are based on what I have looked
>>at in the updates in the PR.  I have not had much sleep,
>>so this message may be a little confusing in parts.
>>
>>>*** the kernel part has been reworked and validated in
>>>      the last patch set.   ...OpenBSD -k ...
>>
>>I haven't looked at what you have for -k, but I did try
>>what you had for KERN_PROC_SESSION and it didn't seem to
>  >work.

Please do not take my private messages and reply to them in
public.  If I thought we had to hash out every little detail
in public, I would have sent my messages to the mailing list...

>  > I probably should, but I fear that before I have the
>  > time to read & understand & test & install those
>  > updates, you will just have rewritten them all over
>  > again.  This is a bit frustrating...
>
>I didn't understand why you want to reinvent the wheel ?

There is a key point that you are overlooking.  I was already
working on my own large updates to `ps' before you sent any
messages to any mailing lists.  I was discussing those publicly
on the mailing lists.  The major changes that I committed in
April was just the "safe part" of the larger work I was doing.
It was the parts of my larger change that I felt were safe to
MFC into 4.x-stable.  At the time I was pushing those in to
5.x-current so I could have them adequately tested in time
to MFC them before 4.10-release shipped.  I did manage to do
that.  I then hit end-of-semester here at RPI, at which point
I have zero free time.  None.

The remaining changes were things that I doubt I will ever
MFC (just because there are too many differences between `ps'
in 4.x vs 5.x).  Right near the end of the public testing
of that first set of changes, you showed up with your update,
wishing that someone would pick up the update.  You were
probably not on the mailing list where my earlier discussion
had been going on (freebsd-standards), so you missed that I
was already working on `ps'.

I am not "reinventing the wheel" after you wrote your update.
I am not doing that any more (or any less) than you were.  We
just both happened to start working on this at about the same
time.  Things like this just happen from time-to-time...

I am continuing with updates I was already working on before
you presented your huge update.  You were quite enthusiastic
about your changes, so initially I put my work on hold and
asked you various issues I saw in your updates.  I sent multiple
messages.  I got no answers.  After a few weeks of waiting, I
finally had some free time again so I decided to go back to the
work I was already doing.  I did that because I tried various
parts of your update and THEY DID NOT WORK.  Thus, it is much
*LESS* work for me to continue with the updates that I already
understood (because I am writing them...), than to figure out
what all 4,000 lines of your update was doing, and all the
side-effects of that update.

My updates do not address everything that your massive update
addresses, but then your massive update does not cover some
of the things I have in the pipelines.  No matter how we slice
it, it will take work to combine the two streams.  If I am the
one doing the commits, then I need to understand the code I am
committing.  The biggest mistakes I have made have happened
when I committed someone else's update because "it looks OK",
without really understanding what it did.  I do not intend to
make that mistake again.  It will take me a fair amount of
time to understand all that is done by your update -- and I am
not going to commit any of it until I am sure that I understand
it.  If my name is on the commit, then I will be the person
responsible for it.

I am still interested in looking over your changes, so I will
check the PR.  Right now I am actually focused on newsyslog,
but I'll look at these when I get back to looking at `ps'.

-- 
Garance Alistair Drosehn            =   gad at gilead.netel.rpi.edu
Senior Systems Programmer           or  gad at freebsd.org
Rensselaer Polytechnic Institute    or  drosih at rpi.edu