wtf ....maven2 surefire 2.4 b0rken
Jesse Kuhnert
It looks like someone forgot to apply some of "Alex"'s (read: sarcasm) new changes to surefire booter to handle passing Properties objects in :
java.lang.IllegalArgumentException: Unknown parameter type: java.util.Properties
at org.apache.maven.surefire.booter.SurefireBooter.constructParamObjects(SurefireBooter.java:767)
I didn't want to say anything in the users list but it ~was~ supporting testng 5.5 - so I hope that was just a misunderstanding and not a new feature we should expect.
wtf .....
---------------------------------------------------------------------
Posted via Jive Forums
http://forums.opensymphony.com/thread.jspa?threadID=87388&messageID=154245#154245
Brett Porter
this elsewhere. Another developer partially deployed the snapshot.
I've deployed the whole thing - hopefully that will correct the issues
(update with -U - please let me know if it resolves problems).
But as I pointed out the other day, this stuff is still under
development and may break, which is why it's not in the production
repository.
- Brett
Jesse Kuhnert
I was addressing it at whoever the generally slimy group of people involved in the new changes are. I'm assuming that means you as well.
>
> I'm not sure who you are addressing this to, but I
> was just seeing
> this elsewhere. Another developer partially deployed
> the snapshot.
>
> I've deployed the whole thing - hopefully that will
> correct the issues
> (update with -U - please let me know if it resolves
> problems).
>
> But as I pointed out the other day, this stuff is
> still under
> development and may break, which is why it's not in
> the production
> repository.
>
> - Brett
>
---------------------------------------------------------------------
Posted via Jive Forums
http://forums.opensymphony.com/thread.jspa?threadID=87388&messageID=154247#154247
Brett Porter
Well, that was uncalled for. You can put your conspiracy theories away, thanks.
The situation as it stands:
- trunk is now deployed to the development repository, whereas
previously your branch was.
- trunk has never contained your changes. They will get applied
gradually, but I have to try and understand them first.
- trunk does now include Alex's changes which as far as I can tell
enable TestNG 5.5 usage and expose the extra functionality
We're all aware that more work is still needed. Obviously some people
have been successfully using your code and those builds got broken by
the other deployment which I'm sorry about. I'll deploy that under a
different version so that people can pin to it temporarily.
As I talked to you about online last week, I generally felt the
changes you made were too complicated and I'm not happy with using
internal APIs of TestNG that may break again. You seemed generally
disinterested with having anything to do with it going forward, so I
was going to run with it from there.
- Brett
Jesse Kuhnert
I'm sure people can be understanding that there will still be more work to do, but first things have to actually f-ing work don't they? As they stand now nothing is actually working. Was this one of the changes you felt good about ?
---------------------------------------------------------------------
Posted via Jive Forums
http://forums.opensymphony.com/thread.jspa?threadID=87388&messageID=154261#154261
Cédric Beust ♔
--
Cedric
Jesse Kuhnert
Ah well.. Don't worry - I won't post here again - but all of your general "technical / moral " purity is pretty much down the drain in my book. It obviously extends only so far as it serves your own best interests.
> > Posted via Jive Forums
> >
> >
> http://forums.opensymphony.com/thread.jspa?threadID=87
> 388&messageID=154261#154261
> >
> > >
> >
>
>
> --
> CÃ(c)dric
>
> >
---------------------------------------------------------------------
Posted via Jive Forums
http://forums.opensymphony.com/thread.jspa?threadID=87388&messageID=154276#154276
Cosmin Marginean
lot, but rarely (if ever) on this group.
We are all aware that your technical skills are confirmed and that
your opinion is relevant and to be analyzed, Jesse, but you must
understand that but what you are doing here is considered
flame-war-starting attitude and personal offence.
It is a pitty for someone to see this kind of behavior on an
open-source project development group. This discourages both the
development team and the other people having access to this group.
The problem is however that this kind of behavior was allowed in the
first place, and I have Cedric to blame here. However, this doesn't
mean that we should act like teenagers and continue promoting this
kind of attitude just because someone did it before.
I am sure that everyone MIGHT have a personal interest in these types
of projects, however, I think it is pretty obvious that most of the
people here are trying to get things to work for one reason or
another. So I really don't give a shit wether they do it for the
money, for the fame or to masturbate in the evening with the idea that
they are popular open-source ass-lickers.
This is not the first time that I see this kind of behavior and I must
say that I feel insecure about my position and other people's position
in this project while this kind of behavior is allowed (again, I have
Cedric and the other responsible people to blame). I am trully ok with
curse-words, but personal attacks are something totally different.
My bests,
Cosmin
> > C�(c)dric
Hani Suleiman
On Jun 4, 2007, at 4:11 AM, Cosmin Marginean wrote:
> I don't want to sound hypocrite (as Jesse would say), as I swear a
> lot, but rarely (if ever) on this group.
> We are all aware that your technical skills are confirmed and that
> your opinion is relevant and to be analyzed, Jesse, but you must
> understand that but what you are doing here is considered
> flame-war-starting attitude and personal offence.
Actually, the problem is that the above is not entirely correct.
Here's what happened (from my personal perspective, and I'll try to
be polite even though I've always thought that Jesse's an incompetent
idiot, and hopefully his last few posts have convinced everyone else)
- Jesse develops first cut of maven integration, this is very badly
written using lots of internal apis and dumb shit like reflecting
into private methods. This is not a subjective judgement, almost any
halfway competent developer could look at this and see how awful it
is, but nobody ever did (which is perhaps where Cedric is to blame,
he's too trusting).
- Jesse demands some changes to make his maven plugin better, but his
patch is more bad code. Alex starts reviewing the existing code now
and seeing how abysmal it is.
- Alex and Cedric dont accept the patch for tng 5.6
- Alex starts working on maven integration with brett, doing it properly
- Jesse sees this, and throws a temper tantrum (as an aside, I doubt
you'd find anyone who swears more often than I do online, and even I
know that a developer mailing list is not the right forum for it).
Any sane person has nothing to worry about. Cosmin, look at your
contribution for example. Your initial version had a bunch of issues.
It didn't get into svn at all for weeks, then we finally got around
to discussing it, and a bunch of people provided feedback. You
graciously listed to all feedback (the good and the bad), and happily
addressed all the issues. Everyone is happy, no problems at all.
Had you done this Jesse style, you would have cried when you had to
wait a few weeks for your code to get in, and after that, gotten
angry at negative feedback and stormed off in a huff.
Still, all in all, I'm glad this happened. Jesse can go work on other
random stuff away from us, and we'll have less noise on the list and
fewer dumb ideas. Everyone wins!
If there's one lesson to be learnt, its for all contributions to be
reviewed by at least one person (not an issue for the core, but is
likely to become one for external addons like maven/spring/etc), so
badly written/unmaintainable code is fixed and addressed before
anyone gets any delusions of grandeur.
Anyway, this is the last I'll say on this. It's all my personal
opinion, but is all true to the best of my knowledge. Back to your
regularly scheduled programming now!
Alexandru Popescu ☀
>
>
> On Jun 4, 2007, at 4:11 AM, Cosmin Marginean wrote:
>
>
> [snip...]
>
>
> If there's one lesson to be learnt, its for all contributions to be
> reviewed by at least one person (not an issue for the core, but is
> likely to become one for external addons like maven/spring/etc), so
> badly written/unmaintainable code is fixed and addressed before
> anyone gets any delusions of grandeur.
>
I think everybody have found out in a way or another that I've been
reviewing all the code going to the core (and not only) for quite a
while now. And I really believe that my technical arguments were
correct and that my comments weren't offending anyone so far.
./alex
--
.w( the_mindstorm )p.
TestNG co-founder
EclipseTestNG Creator
Cosmin Marginean
>
> On 6/4/07, Hani Suleiman <ha...@formicary.net> wrote:
> >
> >
> > On Jun 4, 2007, at 4:11 AM, Cosmin Marginean wrote:
> >
>
> >
> > [snip...]
> >
>
> >
> > If there's one lesson to be learnt, its for all contributions to be
> > reviewed by at least one person (not an issue for the core, but is
> > likely to become one for external addons like maven/spring/etc), so
> > badly written/unmaintainable code is fixed and addressed before
> > anyone gets any delusions of grandeur.
> >
>
> I think everybody have found out in a way or another that I've been
> reviewing all the code going to the core (and not only) for quite a
> while now. And I really believe that my technical arguments were
> correct and that my comments weren't offending anyone so far.
>
Yes, I believe this is not he actual problem in this situation. But I
can't agree that nobody was offended :). Someone was - hence this
conversation.
Jesse Kuhnert
Thankfully this isn't something I have to try very hard to argue about. There are plenty of other "low" things happening here to make me angry but I'll address this one as it is easier to point out.
The changes you speak of can all be viewed in the svn repo quite easily: http://testng.googlecode.com/svn/
The new classes involved were :
http://testng.googlecode.com/svn/branches/newconfiguration/src/main/org/testng/ITestNGConfiguration.java
http://testng.googlecode.com/svn/branches/newconfiguration/src/main/org/testng/internal/config/MapConfigurationParser.java
The new method in question is named "configureAndRun" - I had wanted to replace the current "configure(Map)" method but waited to get approval.
http://testng.googlecode.com/svn/branches/newconfiguration/src/main/org/testng/TestNG.java
The change was vs. the existing configure(Map) combination along with http://testng.googlecode.com/svn/branches/newconfiguration/src/main/org/testng/TestNGCommandLineArgs.java
My ultimate intention was to replace the majority of local members you see strewn all over the TestNG.java class along with the bits and pieces of configuration in some of the methods with a more generic interface object instance that these tasks would be deferred to, so the svn repo only represents a "halted" effort.
Say what you like - but I still don't see why they weren't accepted. Clearly in this instance Cedric and Alex were wrong. And now you as well are wrong I suppose, though it's not fair to say that since you were only regurgitating sound bites you had read previously without doing any research to form a unique opinion from.
Again - wasn't really upset about the changes not being accepted. It was more the entire history of the maven integration up to this point and the point when I first saw this new Alex branch of the plugin in the maven repo that was a copy of my most recent work plus many improvements. Did you notice that there is still the horrible / unthinkable reflection code in his version as well?! Heaven forbid. .....I do actually like the way he broke up all the logic of handling different testng versions into distinct configuration class types. Very neat / clean.
Still don't think the overall combination on both sides will be better than what my eventual final revision would have looked like but who knows....
The only thing I do know is that there was a whole lot of non communication happening and zero collaboration on pretty much anything the whole time. It's overall been a huge disappointment and sort of depressing as I did sort of look up to and respect Cedric very much before all of this happened. I still do of course but it's just not quite the same kind of thing anymore - obviously. ...
I am at least satisfied that the important bits of my side have been expressed and so can move on from this thing feeling like I have some closure. I wish you all the best in your growing future. TestNG is still a kick ass unit testing framework and I have no intention of not using it anymore just because of some conflict with its creators/developers.
<snipped>
> - Jesse demands some changes to make his maven plugin
> better, but his
> patch is more bad code. Alex starts reviewing the
> existing code now
> and seeing how abysmal it is.
> - Alex and Cedric dont accept the patch for tng 5.6
<snipped>
---------------------------------------------------------------------
Posted via Jive Forums
http://forums.opensymphony.com/thread.jspa?threadID=87388&messageID=154944#154944