[ros-kernel] RE: [ros-cvs] CVS Update: reactos

[Hartmut Birr]

Hi,

if it nice that you have an answer of Eric's questions. It seems, you have
no answer of my question. Please explain, from where comes your knowledge
about the definition of some types and functions which you have contributed.
This types and functions was never published on the web or on an other
public area. If your knowledge comes either from the IFSDDK or from the W2K
source which M$ has lost some months ago, it will violate the M$ copyright.

- Hartmut     


> -----Original Message-----
> From: ros-kernel-bounces at reactos.com 
> [mailto:ros-kernel-bounces at reactos.com] On Behalf Of Alex Ionescu
> Sent: Thursday, June 24, 2004 11:16 PM
> To: 'ReactOS Kernel List'
> Subject: RE: [ros-kernel] RE: [ros-cvs] CVS Update: reactos
> 
> 
> Eric,
> 
> I am aware that some of them were placed in the wrong files. 
> It is a work in progress, as the other devs have been made 
> aware of. There are other people working on the stubs whose 
> job is to alphabetize them and put them in the correct files. 
> I've done my best to place them where they should be, but 
> I've probably made mistakes, which is why other devs have 
> volunteered to clean them up. So it's not a "waste of time", 
> they volunteered to do so. I asked if a branch would be 
> better, but everyone at the time said merging into HEAD would 
> be ok, as long as it's not a complete mess (and I don't think 
> 5 or 6 stubs out of 300 placed in the wrong files is a mess).
> 
> I made sure not to copy paste anything, because I've asked 
> Steven Edwards and he said to re-write everything. I use MSVC 
> and obviously use the MS indentation because that's what I 
> have in front of me. But I can name one difference, in the 
> way they add spaces and I don't (After the function name). 
> Notice that I *did* copy some of them from Internet sites 
> (stubs which weren't in the MS headers), since I don't think 
> that violates anything (and that's like 5% of the stubs). If 
> it does, I'll gladly remove and re-write them.
> 
> As for prototypes being put in internal header files, the 
> reason is probably because 1) I made a mistake for those 
> prototypes (which the people who volunteered to clean up will 
> fix, or I myself)
> 	2) Some of them I thought belonged there because 
> similar ones are placed there.
> 
> Now Eric, I'd like to add that I didn't wake up one morning 
> and decide to commit 500 stubs because I felt like it. There 
> was a lot of feedback on the #ros IRC channel, and I've asked 
> over 12 developers, including Vizzini if adding stubs was ok. 
> I've talked to some people to help me organize them (place 
> them in the right files) and also on how they should be done 
> (UNIMPLEMENTED/DEPRECATED). I am very conscious about the 
> dangers of violating Microsoft copyright, and I find it 
> strange that as a person who usually doesn't appreciate and 
> has taken a stance against such actions would actually do them.
> 
> I've been working for over a month on these 50% of stubs, and 
> have taken diligent care in respecting all the rules and 
> constant feedback from others. I appreciate your comments but 
> I feel personally attacked for some non-issues. The stubs are 
> *BETA* and they will be properly arranged by me and others. 
> It won't kill anyone that Ipi is accidentally in DPC for some time.
> 
> Finally, stubs are actually useful, even though I've heard 
> you hate them. For example, Steven is trying to get an FS 
> driver for NT working, and needed the Cc Stubs. Many more 
> drivers and apps will load with stubs in place, rather then 
> just not crashing.
> 
> Best regards,
> Alex Ionescu
> 
> 
> -----Original Message-----
> From: ros-kernel-bounces at reactos.com 
> [mailto:ros-kernel-bounces at reactos.com]
> On Behalf Of Eric Kohl
> Sent: June 24, 2004 6:30 AM
> To: ReactOS Kernel List
> Subject: Re: [ros-kernel] RE: [ros-cvs] CVS Update: reactos
> 
> 
> "Hartmut Birr" <Hartmut.Birr at gmx.de> wrote:
> > Hi,
> >
> > you added new prototypes or stubs to the ros source. I've not found 
> > any public informations about many of this functions. In some 
> > declarations,
> the
> > types of the parameters are out-commanded and replaced with 
> VOID. Over
> this
> > types is also no public information available. This is 
> suspicious for 
> > me. Please explain from where your informations comes.
> >
> > - Hartmut
> 
> I'd also like to add that a lot of functions were obviously 
> placed in the wrong files. For example, KiIpiServiceRoutine() 
> does not belong into ntoskrnl/ke/dpc.c because it is no DPC 
> function but is a multiprocessor support routine (IPI = 
> Interprocessor Interrupt, see Intel MP specification).
> 
> A lot of stubs and prototypes look like they were cut & 
> pasted from MS headers because they use exactly the same 
> indentation. We cannot accept such contributions as they 
> violate Microsofts copyright!
> 
> Finally I'd like to mention that several prototypes were 
> place in internal header files although they belong into the 
> ddk headers.
> 
> 
> Please stop adding stubs because you are wasting your time an 
> the time of those who have to clean-up the mess.
> 
> 
> Regards,
> Eric Kohl
>