discuss: Proposal


Previous by date: 22 Apr 2005 05:56:33 -0000 Proposal, Trevor Woerner
Next by date: 22 Apr 2005 05:56:33 -0000 Re: Proposal, Machtelt Garrels
Previous in thread: 22 Apr 2005 05:56:33 -0000 Proposal, Trevor Woerner
Next in thread: 22 Apr 2005 05:56:33 -0000 Re: Proposal, Machtelt Garrels

Subject: Re: Proposal
From: Rahul Sundaram ####@####.####
Date: 22 Apr 2005 05:56:33 -0000
Message-Id: <4268920D.2050703@redhat.com>

Trevor Woerner wrote:

>Hi,
>
>Back in 2003 I wrote a document as I worked to understand the kernel's 
>initcall mechanism which I subsequently published on my site. Today I 
>received an email from Machtelt Garrels suggesting that I publish this 
>document through TLDP. Seeing as TLDP has helped me numerous times in 
>the past, I would be more than happy to submit this work should this 
>project be interested.
>
>http://geek.vtnet.ca/docs.html
>http://geek.vtnet.ca/doc/initcall/index.html
>http://geek.vtnet.ca/doc/initcall/initcall.tar.bz2
>
>Understanding The Linux Kernel Initcall Mechanism: Creating Dynamic 
>Function-Pointer Call Tables
>


The abstract should be a brief explanation of the content rather than 
your experience on writing the document.  The date that you started 
writing the document is hardly relevant to the readers. The last 
revision date might be.  The kernel and version numbers of other 
software could be listed along with the main content instead of being a 
seperate section

http://geek.vtnet.ca/doc/initcall/kernel.html

Kernel  2.4.22
binutils  1.4
GCC ?


Notation and architecture could be notes. They are too short to require 
a seperate section.  The motivation section could be renamed to 
"purpose" and you could also list potential uses of this document beyond 
your personal experience if you are aware of any.   

" Notice also how the my_initN() functions were put in the general .text 
segment since I don't really care where they end up (unlike the kernel 
guys)."

http://geek.vtnet.ca/doc/initcall/review.html#ELF

The above section differentiates between various formats including EXE 
and COM ones. Is this really relevant to the document.  A simple note 
that a.out has been superceded by ELF format thats is being currently 
used might serve the purpose better.. If not, they might be better 
called as EXE and COM formats rather than  "dos.exe" and "dos.com" since 
they are shared across other operating systems like Windows?

Writing it on a third party perspective might be better than using 
statements such as these. For example this could be rewritten as " The 
function names provided here are put in the general  .text segment 
merely as an example.  Actual kernel development  would require more 
precise care about the storage of each of these data structures"

Your copyright section notes that the document is licensed under the 
"standard"  BSD license. However there are atleast two different forms 
of the BSD license used widely and yours is called as the new BSD 
license in opensource.org and as revised  BSD license by FSF. you might 
want to change your wording there or simply link to the license in 
either opensource.org or fsf.org to avoid any confusion.

You will probably need to run this through a kernel expert to get a 
technical review external to LDP.  Can you do this on this own?

The Docbook/ Linuxdoc  source format might be useful to those who want 
to check the markup. Please provide both the html and Docbook/Linuxdoc 
version of this document for every revision that you present to this  list.

regards
Rahul






Previous by date: 22 Apr 2005 05:56:33 -0000 Proposal, Trevor Woerner
Next by date: 22 Apr 2005 05:56:33 -0000 Re: Proposal, Machtelt Garrels
Previous in thread: 22 Apr 2005 05:56:33 -0000 Proposal, Trevor Woerner
Next in thread: 22 Apr 2005 05:56:33 -0000 Re: Proposal, Machtelt Garrels


  ©The Linux Documentation Project, 2014. Listserver maintained by dr Serge Victor on ibiblio.org servers. See current spam statz.