Re: [E-devel] Fwd: About release, snapshots and openBSD port

[Carsten H. (T. R. <ra...@ra...>]

On Sun, 13 Nov 2011 02:26:00 +0100 Stefan Schmidt <st...@da...>
said:

> Hello.
> 
> I have not joined any of these flame wares before as I don't think to
> change anything significant but only start to hurt peoples feelings
> for each other. But I had to join here as it started to look like a
> witch hunt on raster here. Please take a moment when reading this.
> Thanks for your time.
> 
> On Sat, 2011-11-12 at 13:43, Youness Alaoui wrote:
> >
> > you don't seem to read what I wrote, you ignore the facts.. the facts are
> > not the commits or how bad they were, the facts were your attitude and
> > condescending bullshit. but yeah,you don't seem to be able to acknowledge
> > that, since you're perfect and everyone else is wrong.
> 
> Great, personal insults are getting us really forward here. This is
> one of the social skills you are calling here for. Discussing with
> others without getting into personal insults. Given this mail and the
> long rant where you "behaved like a dick" (citation from you) are
> letting me wonder if you are able to call others for things you do not
> handle very well on your own. Something to thing about.
> 
> > Thank you Vincent and Gustavo for sharing your concerns about this, and
> > it's too sad that the new contributor has become another victim of raster's
> > poor social skills. That's what I wanted to avoid, that's what I wanted
> > raster to understand, and I was hoping for him to reply with something like
> > "sorry if I offended you, that wasn't my purpose" and that's it, the guy
> > stays with us, but I guess raster has too much pride and is too
> > self-centered to recognize his own faults.
> 
> And you wanted that to happen by forcing him into a corner?
> 
> That is almost always the best recipe to get the opposite of what you
> wanted. Forcing people trigger over reactions from them. Self
> protection, naturally for humans. Changing the behaviour of people is
> a long and exhausting  process. Nothing you can do by sending of
> several mails. And before people even accept what they here from
> others they need to respect them. Respect them for their doings and
> ideas they have come up with over time. Again nothing you can achieve
> in some weeks.
> 
> > I think I will follow Vincent's advice and not reply to this thread
> > anymore, raster clearly showed he has no comprehension of what people are
> > trying to tell him here, so this is just an endless drama with no possible
> > resolution.
> 
> Black and white thinking all around. Sadly we live in a grey world.
> Nothing is only black or only white. Lets have a look at what problems
> we have here and what possible solutions we can come up with. (That
> what we should aim for in the end, a solution bringing the project
> forward).
> 
> Raster is stressed out. Short on time and running at the edge of what
> is possible for him all the time. Thats a fact and on of the biggest
> problems here. Stress calls out on people making hard decisions and
> one of this is being brusque to others. I have observed this a lot at
> myself when being in stressful times. Family and friends had the
> pleasure to get me in such a mood. And even after I recognized this
> at myself (the first step, you know), it is very hard to change at
> all. Again, behaviour changes are the hardest.
> 
> The work part of raster stress we can't influence much. He has to
> handle this on its own. And I personally hope that he realizes how
> near he comes a actual burnout if he keeps going like this for more
> months.
> 
> But now to the things we can change. You and Gustavo are trying to
> change this project in a direction that should be more welcoming for
> developers and users. Making the community grow. I welcome this move,
> but doing something like this can not happen by bringing everything
> down that happened so far. Raster brought this all to the point what
> we have today. Motivating people on the road and de-motivating people
> on the road. Again, very natural as we don't live in flower-power
> land. :)
> 
> So to change to bring in change to this community you need to earn the
> respect of the other developers here before steering the way forward.
> Bluntly speaking nobody wants to accept orders from people he does not
> respect or being paid by. And even the last part may be very hard
> sometimes. ;)
> 
> For the matters at hand the following could be done:
> 
> o Tarballs: Everyone seem to speak about daily tarballs. What I read
>   from the openbsd guys have not been daily tarballs but tarballs for
>   an alpha or rc to check if everything is fine beofre the actual
>   release. Such tarballs are fine and have already been acknowledged
>   and done before. They will even get some QA. And QA is something
>   that differs from daily tarballs, like your script or a simple make
>   disctheck, will produce. SOLUTION: Wait for the alpha and rc
>   tarballs.

yup. exactly my point. :)

> o OpenBSD patches: Vincent asked raster to have a look at the patches.
>   He did look and pointed out what was wrong. Agreed, a bit to blunt

thanks stefan! yes. i am blunt. i don't sugarcoat things and tiptoe. but i
wasn't rude. and i'm not apologizing for being blunt and factual as now the
witch-hunt-crew are thinking i must. 

>   maybe. He did mention better options though. Something that people
>   like to ignore in this thread. (BSD specific malloc changes in
>   mempool instead on every file using it, not changing API/ABI without
>   discussing it here). SOLUTION: To calm this down you or Vincent or
>   someone else can keep working with them to gte the changes in. That
>   involves understanding why so much changes are needed and bringing
>   it up here to discuss about a solution. Uninteresting work like

yup. which is why on the email list i was saying i hadn't seen any patches -
but am open to them. on irc vincent asked me to look at them. 

>   reviewing patches from the ml and putting them into svn. But it
>   helps to balance the load. Raster is nobody  who calls others for
>   doing things for him. He waits until he comes to it and does them
>   alone. To me that looks like he lost his faith in this because it
>   did not work out well in many cases. Sure, that is something he
>   needs to improve. That is nothing that stops others from stepping up
>   and doing it without being asked for though. Mike, Vincent, Cedric
>   and others are reviewing patches here on the ml. That takes of load
>   from raster. He did not ask them to do it.
>   For your tarball script you could do the same. Why must it happen on
>   the main machine? You can host them yourself and when they are
>   really becoming popular they can get moved. Not enough capabilities
>   for hosting? I bet e.fr or others can help out there.

sure - any of us can run the script - still have to upload to e.org, make dir,
do announce etc. - well announce could be skipped but in the past thats what we
have done. but if u don't at least to make distcheck and you don't run and test
and check to see that what u just tarred up works, then the quality is no
better than an svn checkout - which is what we have now already. and yes - i'm
stressed and busy and i'll instantly push back on anything that looks like
"some more work" unless i see a gain proportional to the work involved, or
greater. :)

> Phew, long mail. To long actually. The main point here is that it does
> not help to forcing raster into a corner here. He is the main driver
> of the project and if people like to expand it need to be done in a
> way that do not offend the people that are already working on it.
> People earn respect and faith of others in their work by actual
> doings. There must not be a leader who is always right and delegates
> work into his hierarchy of minions. You have a pet peeve topic? You
> want to improve it? Don't ask for permission, do it and improve the
> situation gradually.

thanks stefan. you actually read the content at hand. something others are not
doing. yes - i could be less blunt. if armani, vincent, kakaroto, gustavo
simply said "you're a bit too blunt" i'd agree - i am blunt. if me being blunt
offends someone - then i'm not going today sorry and pretend i've been rude.

all over the world someone will get offended at almost anything you say. if
people read the actual log it'll be clear - i'm blunt - but not rude. well not
rude until someone decides to actually directly attack me. i will generally
respond in-kind.

let's look at the core - this started with 2 things:

1. me saying no to daily tarballs as they were no better than a randokm svn
checkout
2. me giving a review on patches with questions (then answered) and pointers as
to how to do it better. the review was blunt.

now let's discuss offence / feeling bad / de-motivation when patches are
reviewed badly. maybe that's the point? maybe , if the code in question is
bad, you SHOULD feel bad. it's a lesson to learn that hopefully ensures you
don't make that same mistake again. i've made my own mistakes and feel bad even
about code that's in EFL now - but i don't go get de-motivated. i've had others
review my code before and been made to "feel bad". i've learned from it. the
reason why reviews are blunt and tough is... that that patch will become
maintenance baggage for those who follow you long after you've gone. it will
eventually create work if it creates problems. breaking efl api's in itself is
very bad for packagers to do - and yes - i'm very annoyed at that being done.
we can't stop it - that's the nature of open source, but it's inconsiderate to
go doing that as we end up with the support burden as a result of the api
breaking. yes - i'm not going to smile and be all love and roses when i see
that kind of thing put in front of me. but i was not rude or condescending - i
stayed factual other than my "that's bad. really bad" level 
comments. :|

-- 
------------- Codito, ergo sum - "I code, therefore I am" --------------
The Rasterman (Carsten Haitzler)    ra...@ra...