AF:
NF:0
PS:10
SRH:1
SFN:
DSR:
MID:<20080220163224.5c1f03a0@ripper.onstor.net>
CFG:
PT:0
S:andy.sharp@onstor.com
RQ:
SSV:onstor-exch02.onstor.net
NSV:
SSH:
R:<maxim.kozlovsky@onstor.com>
MAID:1
X-Sylpheed-Privacy-System:
X-Sylpheed-Sign:0
SCF:#mh/Mailbox/sent
RMID:#imap/andys@onstor.net@onstor-exch02.onstor.net/INBOX	0	BB375AF679D4A34E9CA8DFA650E2B04E0875E3CA@onstor-exch02.onstor.net
X-Sylpheed-End-Special-Headers: 1
Date: Wed, 20 Feb 2008 16:33:23 -0800
From: Andrew Sharp <andy.sharp@onstor.com>
To: "Maxim Kozlovsky" <maxim.kozlovsky@onstor.com>
Subject: Re: please review 27937
Message-ID: <20080220163323.587545bb@ripper.onstor.net>
In-Reply-To: <BB375AF679D4A34E9CA8DFA650E2B04E0875E3CA@onstor-exch02.onstor.net>
References: <20080220141254.2692db28@ripper.onstor.net>
	<BB375AF679D4A34E9CA8DFA650E2B04E0875E2B4@onstor-exch02.onstor.net>
	<20080220145031.7a6849cd@ripper.onstor.net>
	<BB375AF679D4A34E9CA8DFA650E2B04E0875E336@onstor-exch02.onstor.net>
	<20080220153459.629e37ce@ripper.onstor.net>
	<BB375AF679D4A34E9CA8DFA650E2B04E0875E3CA@onstor-exch02.onstor.net>
Organization: Onstor
X-Mailer: Sylpheed-Claws 2.6.0 (GTK+ 2.8.20; x86_64-pc-linux-gnu)
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit

On Wed, 20 Feb 2008 16:24:00 -0800 "Maxim Kozlovsky"
<maxim.kozlovsky@onstor.com> wrote:

> 
> 
> >-----Original Message-----
> >From: Andy Sharp
> >Sent: Wednesday, February 20, 2008 3:35 PM
> >To: Maxim Kozlovsky
> >Cc: Larry Scheer
> >Subject: Re: please review 27937
> >
> >On Wed, 20 Feb 2008 15:07:03 -0800 "Maxim Kozlovsky"
> ><maxim.kozlovsky@onstor.com> wrote:
> >
> >> This is not very convincing. I don't know where you got the idea
> >> that the code compiled for eee should include linux.h. If your
> >> code is compiled for linux, it will include linux.h. If it is
> >> compiled on
> eee,
> >> it should not include linux.h, because eee is not linux. If you
> >> want to include some file depending on whether it is cougar or
> >> bobcat, you can put your file in Includes/[cg|bc].
> >
> >You are not correct because you are assuming our code complies with
> >some perfect design standard, where there is no code in sm-whatever
> >that has anything to do with SSC.  That simply is not the case.
> [MK] 
> 
> It is the case though that the current rules for the include files are
> what I've just described above.

In my opinion they really weren't.

If you look at the changes for 27937 you can see that the include rules
are now actually correct; as opposed to "working" after some kludges
are thrown in like empty include files and whatnot.  Not that empty
include files are the end of the world, just that they shouldn't be
needed in a perfect world.

> Since you can fix your bug without
> changing the rules I don't see why you should change them. 

I don't know what you mean.  I can't fix the sm-seep code
w/o fixing this problem.


While I'm enjoying this debate tremendously, can I check this in?


> >> By the way, why do you need to port sm-seep to linux? We already
> >> have everything ported, cm_read_seep_bc() etc.
> >
> >Feel free to take a look at the changelist:
> >
> >Change 27934 by andys@ripper on 2008/02/19 11:56:49 *pending*
> >
> >        Fix for TED00022316 - system show version gives wrong boot
> > dev
> >
> >        Port to Linux: make values of devices displayed by commands
> >        correct for Linux or OpenBSD.
> >
> >        reviewed by
> >
> >Affected files ...
> >
> >... //depot/dev/nfx-tree/code/sm-seep/env-api.h#2 edit
> >... //depot/dev/nfx-tree/code/sm-seep/env.c#6 edit
> >... //depot/dev/nfx-tree/code/sm-seep/linux.h#1 add
> >... //depot/dev/nfx-tree/code/sm-seep/openbsd.h#1 add
> >... //depot/dev/nfx-tree/code/sm-seep/seep-api.c#7 edit
> >
> >
> >
> >> >-----Original Message-----
> >> >From: Andy Sharp
> >> >Sent: Wednesday, February 20, 2008 2:51 PM
> >> >To: Maxim Kozlovsky
> >> >Cc: Larry Scheer
> >> >Subject: Re: please review 27937
> >> >
> >> >On Wed, 20 Feb 2008 14:18:14 -0800 "Maxim Kozlovsky"
> >> ><maxim.kozlovsky@onstor.com> wrote:
> >> >
> >> >> What does "It" in "it builds" mean? Did you do a clean build? I
> >> >> am sure I did this for the reason and I don't appreciate your
> >> >> calling this "abuse". The world of Oss is not limited to linux
> >> >> and
> openbsd,
> >> >> you know. EEE is also an OS in some sense.
> >> >
> >> >Actually, I do hear what you're saying.  I started down the path
> >> >of creating a new variable, SSC_OS and setting OS_INCL based on
> >> >that, but I encountered some problem that involved intracacies
> >> >with our include files around nfxnis and rpc/xdr and stuff.  I
> >> >couldn't get that
> >> resolved
> >> >after trying 3 or four different things, and then went this way
> >> instead.
> >> >While I can see a possible need for a 3-way variable, it seems we
> >> >don't really need one here, at least not now, as all the eee.h
> >> >files were empty anyway.
> >> >
> >> >I didn't file a defect, but the problem is that if you say
> >> >
> >> >#include OS_INCL
> >> >
> >> >in a source file, that should be either linux.h or openbsd.h, not
> >> eee.h.
> >> >Because OS_INCL was based on the makefile variable OS, in some
> >> >directories it was 'eee', causing those source files to want to
> >> >include eee.h. But they either actually needed either linux.h or
> >> >openbsd.h, or neither. There were at least 3 places where we had a
> >> >to include an empty eee.h file to make the code compile.  One of
> >> >those was ssc-genlib, but I cleaned that up a few weeks ago.  That
> >> >fix was not a general one, however, and was actually just a
> >> >consequence of changing the way certain things worked in all that
> >> >code.  OS was also being used to set default include directories,
> >> >and those were correctly 'eee' in a small number of cases.
> >> >
> >> >The problem came along when I needed to port sm-seep to Linux, and
> >> >including OS_INCL wouldn't work because that code is compiled with
> >> >OS=eee.  Since I'd had the problem before, I decided it was time
> >> >to implement a general fix.
> >> >
> >> >So it's not a black/white case, it's more like a grey area, but
> >> >one that had to be resolved.
> >> >
> >> >I didn't mean to ruffle any feathers, just trying to keep things
> >> >light. By abusing everyone.  Because that's so humourous.  Abusing
> >> >variables is bad, abusing people is OK as long as it's not me.
> >> >
> >> >
> >> >> >-----Original Message-----
> >> >> >From: Andy Sharp
> >> >> >Sent: Wednesday, February 20, 2008 2:13 PM
> >> >> >To: Larry Scheer; Maxim Kozlovsky
> >> >> >Subject: please review 27937
> >> >> >
> >> >> >I'm sending this to both of you because it involves makefiles
> >> >> >and/or code you may have written or know about.  Yes, I am a
> >> >> >bit nervous about it.  It builds on both cougar and
> >> >> >bobcat/openbsd, but...
> >> >> >
> >> >> >Change 27937 by andys@ripper on 2008/02/19 14:20:57 *pending*
> >> >> >
> >> >> >        Ai-eee.
> >> >> >        Fix abuse of OS make variable.  This variable is
> >> >> > limited to the two SSC operating systems, either openbsd or
> linux.
> >> >> >        eee is not one of those.  Really this variable should
> have
> >> >> >        been called SSC_OS or something, but that's probably
> >> >> >        already used for something completely different.
> >> >> >
> >> >> >        reviewed by
> >> >> >
> >> >> >Affected files ...
> >> >> >
> >> >> >... //depot/dev/nfx-tree/Includes/eee/ipm-if.h#1 delete
> >> >> >... //depot/dev/nfx-tree/Tools/defs.mk#11 edit
> >> >> >... //depot/dev/nfx-tree/code/sm-elog/eee.h#1 delete
> >> >> >... //depot/dev/nfx-tree/code/sm-ipmd-lib/eee.h#1 delete
