Ports system quality and trolling

[Jerry]

On Sun, 28 Aug 2011 14:43:14 -0700
Doug Barton articulated:

> On 8/28/2011 1:54 PM, Michal Varga wrote:
> > On Sun, 2011-08-28 at 15:30 -0400, Sahil Tandon wrote:
> > 
> > [...]
> >> Criminal?  Indifference?  This sort of troll-ish hyperbole is
> >> decidedly unhelpful. 
> 
> FWIW, I agree with Sahil that this post of Jerry's was over the top,
> as several of his have been of late. To use the word "criminal" in
> this context is sufficient all on its own. To accuse people who spend
> an enormous amount of their own free time trying to make this thing
> work of being indifferent is just plain rude.

Specifically, I said "borders on criminal". Get your facts straight.
Second, I was specifically referring to the act, not the individual. In
addition, "Spending enormous amounts of time" != "valid excuse". A mass
murderer can spend enormous amounts of time planning his crime. Does
that absolve him from the actual crime?

> > To contribute my random few cents to the debate (without actually
> > contributing anything of worth, so you don't really have to read
> > it):
> > 
> > Replies like these already made me discard like 20 of my own emails
> > in the past, mid-write, exactly because of this expected outcome -
> > accusations of trolling, because, why not, that's really what it's
> > all about, right.
> 
> Well, no. :)  Personally I find this message of yours to be well
> thought out and well stated. You raise some valid concerns without
> making personal attacks. That kind of feedback is always welcome.
> 
> > So to say for myself - I do not know Jerry, but I definitely share
> > his sentiments and even find his tone quite funnily (is that a
> > word?) appropriate, as the ports quality, over the last year, went
> > totally, horribly, down the drain.
> > 
> > On some of my desktop setups, I keep about 900-1000 installed ports
> > (and there are some ~200-300 for servers in general). There already
> > seems not to be a single week, even once, without some MAJOR
> > breakage that always takes hours (sometimes days) to track down and
> > fix by my own ...
> 
> FWIW, my experience has not been even close to yours, although I do
> find broken things occasionally.

I have 1,112+/- ports installed on any given day on one of my machines.
Your statement, without knowing your exact configuration, applications,
etcetera is not a valid counter statement. It stands to reason that an
increase in the number of ports used would show a corresponding
increase in the number of port failures.

> > And I know that every time I'd start writing a mail about it, my
> > tone would be exactly the same as Jerry chose. With the expected
> > result of "Zomg stop trolling",
> 
> First, if you find something broken, please report it; in a calm,
> factual manner; ASAP. That will help us fix it ASAP and help avoid
> other users having to share the same frustration. Second, if you have
> concerns about the direction that things are heading in a more
> general way, feel free to express them as you have here. You may find
> that people agree with you. :)

And then again, you may not.

> > or for a change, the ever popular megahit "Patches welcome"
> 
> Sometimes that *is* the correct answer though. There is only so much
> that the existing pool of volunteers can do. If we don't get new
> people who are willing to get their hands dirty, the project dies.

A quick death is often the best course of action.
 
> > On a weekly basis, again and again, there are port
> > updates being introduced with what seems to be absolutely no testing
> > whatsoever, some breakages take multiple takes on fixing by their
> > respective port maintainers,
> 
> While I'm certainly not going to say that mistakes never get made,
> with very nearly 23,000 ports, and a nearly infinite number of
> possible OPTIONS combinations, shaking out all of the corner cases
> can be very difficult for even the most dedicated of maintainers.
> But, see below.
> 
> > new versions of major dependencies get introduced only to be rolled
> > back few days later;
> 
> If you're talking about the recent ruby update, an enormous amount of
> work went into that prior to the trigger being pulled in an effort to
> make it as smooth as possible. It's unfortunate that in spite of that
> effort there were still some "issues" that were only discovered after
> users rushed to perform the upgrade. In this case backing out the
> change was the responsible course of action.

If this was an isolated issue, I would certainly agree with you.
However, it has become the norm and not the exception for this sort of
behavior.
 
> > So, was really Jerry's tone so trollish?
> 
> Yes. There is a world of difference between expressing concern about
> the issue (as you have done) and attacking people on a personal level.

The root problem is group mentality. Attempting to convey a message
that runs counter to the "group" is like making a statement against the
Pope, or the towel head god, or what ever. People immediately become
paranoid and defensive. They act like a group of school girls who have
to have their egos messaged on a daily basis.

> Now, how do we fix this? It has been suggested numerous times that one
> solution to this problem would be a "stable" ports tree. The idea
> being that after changes have had a chance to shake out for a while
> in the head of the ports tree they get merged back to a stable
> branch. This needs to happen, yesterday.

What really needs to happen, besides that, is that FreeBSD realize that
there are wireless-N devices in existence, have been for over 5 years,
and actually develop real, fully functional drivers for them. Opps,
sorry, I forgot to blame the manufacturers and, or Microsoft for that
failure. I'll skip the part about getting the device to operate without
having to waste time attempting to get it manually configure just to
make a simple network connect. That would be assuming that there was a
driver available for the device to begin with.

> The other thing that will help between now and then is to manage your
> change windows a little more conservatively. Except for
> security-related updates there is almost always zero reason to
> upgrade to new versions of things immediately after they hit the
> ports tree. With all due respect to those involved, one of the
> reasons the ruby thing was such a mess was that users jumped in and
> started upgrading stuff without knowing what they were doing, or why.
> Personally as soon as the notice about the upcoming change went out I
> put the knob in make.conf to keep my systems at 1.8 to make sure I
> wasn't affected.

I have no problem with that. No one should be forced to update their
system if they choose not to. However, to carry your statement to its
logical conclusion, you should issue a warning that attempting to
update your system carries dire risks since the updates have not been
properly tested. For the record, users knew exactly why the were
updating "ruby", they wanted to. If it was not to be used, then why
release it? What they did not know was that it was going to bite them
in the ass, like so many other updates (cups+gnutls) have lately. If it
had been failing on a few obscure programs, then I could probably say it
was an unfortunate oversight. When it starts failing on major
applications used by a large number of FreeBSD users, then it should be
labels what it is, incompetency. Opps, did I hurt someone's feeling?
Well, you screwed up my system and wasted hours of my valuable time, so
now we are even.

> So all this is not to say, "You're wrong!" but instead to try and
> constructively address the concerns you've raised, and hopefully
> provide some suggestions on how to address them.

That and $1.50 (give or take) will get you a cup of coffee.