Hi Jeff (Zeffie),
As you've been dragging my name up a few times in your recent postings I'll
add my $.02 to this unecessary discussion.
Personally I see it like this: Sure, /etc/dovecot.conf certainly could have
been marked config/noreplace. BUT: If *you* (i.e.: Zeffie) distribute
software that modifies the system - ANY system for that matter - then you
have to know exactly what you're doing and which loose ends you might run
into.
When you provide updates that don't take the stock configuration into account
as it is present on that server, then I see this as a lack of testing.
Especially as you could have adressed any adverse side effects which the
absence of config/noreplace might cause with just a few lines of extra
scripting in your RPM.
If something worked before (like SSL for Dovecot) and your update breaks it,
then blaming someone else for this oversight is certainly not the right way
to go.
To quote your own words from [cobalt-users]:
> updates should be tested on devel boxes as needed...
In that case:
dovecot-1.0.12-1.0.Z1.centos4.Zeffie.com.BQ.i386.rpm
"This breaks the dovecot SSL connection on port 995"
= I'm sorry, in that case I'd rate his as: FAIL.
> It's my understanding that Hsiao was paid by turbo Linux to build it for
> them.. As far as I know Brian just made the cd and I'm not aware of
> anything major or on the same scale that Michael has done...
Oh, I certainly haven't contributed as much code to the BlueQuartz SVN as
Hisao or Brian have. There is absolutely no doubt about it. Especially as I
asked for and got commit access to the SVN only like a couple of months ago.
Who contributed what to BlueQuartz is easily verified by checking the SVN logs
at this URL: http://bluequartz.org/trac/log/ as it shows a brief summary of
the changes, when the update was submitted and who checked the code change
in.
That you talk Brian's contributions down to "Brian [...] just made the cd" is
easily proven as utterly wrong by checking the SVN logs at the URL above.
Brian's name shows quite frequently there, does it not? You said "as far as
you know", which *could* be an indication about how little you may know about
the state of the code in question or the project in general.
Either that, or it is what most people will perceive it for: A quite
transparent attempt at lessening the contributions of individual developers
to elevate your own contributions to the BlueQuartz SVN. Which are like ...
zero? Correct me please if I'm wrong on that. And no, it's not because "we
wouldn't let you". It's because you didn't ask. Instead of suggesting
changes, offering coding time or asking what you could contribute you emailed
threats to developers offlist whilst trying to play us out against each
other - both off list and on the list. This kind of strategy is
called "divide and conquer" and its not working well in a close knit
community such as this.
From your posting [coba-e:12205]:
> Also..
> These updates are not intended for anyone selling BQ servers or operating
> systems. If you would like to use these updates for servers you rent out
> or in any "paid for" BQ product you should contact me for a site or system
> license... This means if you sell/lease servers running BQ and you want to
> update them, You should contact me first...
>
> If you just want to install this on your own server or if you have less
> then 5 servers where you sell webhosting, it's free...
Code that is in the BlueQuartz SVN or in the BlueQuartz.org YUM repository has
no such strings attached. It is released under Sun BSD which gives all users
the greatest deal of leeway what they can do with it and how they use it.
Code that is in the BlueQuartz YUM repository has been and is written with
updateability to 5106R/5200R (BlueQuartz on CentOS5) in mind to make sure
that those future updates won't cause conflicts. So I don't really see why
anyone using BlueQuartz to build and sell servers or who rents out servers or
who uses BlueQuartz in any commercial fashion would use the updates you
rolled up and anounced here. Unless they were misled or willingly want to
take easily forseeable and quite avoidable risks.
According to the rules of this mailing list (see
http://bluequartz.org/ml/index.html ) "All postings [on this list] must be
free of copyright restrictions that limit distribution". Your statement as
shown above could be viewed as a violation of these rules as you *are* indeed
limiting the distribution.
Furthermore the rules of this list state: "Advertisements are not appropriate.
This includes, but is not limited to, announcements of new personal products
and free trials by those who stand to gain from such announcements.", which
your anouncement also may easily run into conflict with.
Lets look at your reply to Brian in [coba-e:12197]:
> Last I checked this was the BlueQuartz list and if you feel your work is
> perfect, I would suggest you fork off.
Instead of pointing Brian - whose contributions to the community are without
any question or doubt very well known - towards the door, you should take
note that you are violating a third rule of this mailing list, which
states: "Personal attacks such as name calling and personal insults will not
be tolerated. Comments that are intended only to enrage the recipient or
others rather than contribute to thoughtful discussion are prohibited. " If I
had to judge this, then I'd rate telling Brian to "fork off" after all he's
done for the project is certainly neither thoughtful, nor productive, but
quite an insult.
Jeff, you can't on one hand hide behind the rules of this mailing list when
someone disagrees with you and then willingly violate other parts of said
rules as if they don't apply to you.
Could you please read http://bluequartz.org/ml/index.html again and make sure
that your future postings on this list are in accordance with it?
Thank you!
--
With best regards,
Michael Stauber