[ros-dev] My attitude

[Alex Ionescu]

Thomas Weidenmueller wrote:

>Witukind wrote:
>  
>
>>I think the project would stagnate without Alex.
>>    
>>
>I doubt that'd be a problem as it didn't before he joined, so why should
>it now? People join and leave once in a while, no reason to stop. It's
>not like he's the only person with knowledge. This does *not* mean that
>I want him to leave.
>
>Arrogance certainly is an issue but I don't really care as I usually
>ignore those moments and don't comment on them, everyone has his/her
>flaws. What bothers me more is the "ownership" of code issue. These
>large "rewrites" of functions/files is annoying, especially when holding
>everything back for months and then committing everything in one single
>huge commit 
>
I now have a local SVN setup to handle my incremental patches. Other 
developers do not. Other developers still commit code in one single huge 
commit, yet you don't complain about it.

>(I'm not talking about headers fixes, because those
>certainly do require loads of changes in one commit). This of course is
>a reason then to claim (full) ownership of the code, and dare you touch
>it in such a way that it's less close to windows. For me, I do *not*
>want to clone internal things 100%, there are many reasons why this is a
>bad idea. However, since I'd write code that wouldn't be "perfect" in
>this aspect, I no longer work on the kernel as I'd interfere with this
>attitude.
>
You no longer work on the kernel because you don't have a lot of time on 
your hands; you've told me yourself. I think you're being a bit 
hypocritcal with your argument. You commited a pretty large handle table 
implementation which we talked about at length. You made it clear it was 
not Windows internal compatible. What were my comments? Did I not praise 
you for it? Did I not appreciate you for it? It's been in the tree for a 
year. Have I ever made an attempt to "rewrite it"? Did I ever said I would?

How about when I asked you to improve it with some new code. What was 
your response? It was "I don't have time with university", not "you want 
it to be perfect, like windows".

Yes, I hated your Object Capture patch because it wasn't using a public 
capture structure and we got into a big argument about it, but you might 
want to remember how much I've appreciated and loved your ExHandle patch.

> I wouldn't want to submit a patch and seeing everything gone a
>month later just because it wasn't a next to perfect clone of the
>original code, even if it worked perfectly and as expected. That's also
>  
>
>one of the reasons I abandoned win32k, i just wouldn't want to copy
>every little detail just because it "must be perfect" because MS wrote
>it. 
>
I never had anything to do with win32k. I barely know how it works, and 
you had win32k rewrites ongoing before I even joined the project. How 
can you possibly blame this on me, when you publically stated OTHER 
reasons for dropping win32k?

>This leads me to the constant IDA/ASM digging in MS binaries, I
>don't think it's good for the project to copy every little detail based
>on unnecessary reverse engineering. Testing and thinking about things a
>little longer is a good method to being able to implement many things,
>of course not all, especially completely undocumented interfaces.
>
Point taken.

> And
>just because MS did something a specific way doesn't mean they're god
>and that it's perfect. 
>
I think you and I can both find several instances where we made fun of 
the way MS wrote a certain piece of code, and found better ways to do 
it. So you can't accuse me of that.

>Also the frequent "Cutler adoration" on IRC is
>*very* annoying to me, he is *not* god.
>  
>
You aren't seriously bringing this up are you. What's next? An email to 
KJK saying "Cheesecake is not the official food of ReactOS damn it, I 
HATE CHEESECAKE!!!"?

>This are my 2 cents, I don't want anyone to take it personally,
>especially Alex. I just won't consider ReactOS a clean-room
>implementation anymore if this trend continues, which would make me
>leave due to legal issues.
>  
>
>- Thomas
>
>P.S.: I do *not* oppose reverse engineering per se, there are situations
>where it's necessary (in order to accomplish compatibility)
>  
>
Those are the only times I've used IDA.

Best regards,
Alex Ionescu