[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Ousting bad people

From: David Kastrup
Subject: Re: Ousting bad people
Date: Mon, 03 May 2010 18:10:10 +0200
User-agent: Gnus/5.13 (Gnus v5.13) Emacs/24.0.50 (gnu/linux)

Werner LEMBERG <address@hidden> writes:

> On the other hand, you obviously have too much time, David, to write
> such long replies

I wish I did.

So you suggest I'd rather _not_ make point-to-point replies to negative
reviews and hope that eventually somebody else figures out that the
points are not valid with respect to the reviewed code?  Fat chance.

But anyway, ignoring a negative review basically _was_ my second attempt
after commenting the original code did not help _anything_ at all (and
nothing Han-Wen wrote would suggest he even looked at the comments I
wrote to address his concerns): I completely junked the whole
verification and moved its functionality elsewhere, in a different
language, because the first reply made clear that there was no chance
that I could get the C++ code doing the task in the most straightforward
and efficient way matched to the task at hand accepted before blowing my

So I scrapped the whole thing (pretty annoyed because it was already
working perfectly and documented extensively _on_ request) rather than
bothering to argue about it, and it helped exactly zilch.

After three iterations, one of them a complete change of strategy,
nothing in the reviews suggests that anything but the very first version
was even being looked at.

Do you really think I have the time to spare to do three times the
necessary work without people even noticing?

No, I haven't.  I also don't have the time to rant and have everybody
spit on me.  But I was pretty much out of other options.

I know _no_ other project that makes it so absurd and hard to be
permitted to contribute.  There are double standards for insiders and
outsiders, and it is hell for outsiders to get anything but trivial code
accepted.  Most other people just go away, or they behave themselves
long enough (or don't venture beyond trivial code long enough) to become

I prefer getting work _done_, but it does not seem like this option is
available to me within this project.

I'll likely stop using the Rietveld process since it restricts reviews
mostly to inside people with Google accounts, and is mostly not making
people actually try out the code.

Instead I'll just post patches on the list again.  git is tailored to
this workflow and it's dead easy to let it process patch series with
"git am", in a different branch or otherwise.  Perhaps that will lead to
a more diversified response.

David Kastrup

reply via email to

[Prev in Thread] Current Thread [Next in Thread]