Dan Price wrote:
>
> I would like to propose a new community called "Gardeners" based on
> some recent discussions on opensolaris-code. [1,2]
>
> The mission of the community is simple: raise the overall code quality of
> the system.
>
> The intention of the community is to provide a nexus for a group of people
> who are already active in the community and to attract new talent who
> share our values. The technical scope of the community is roughly:
>
> - Automated code checking (lint, cstyle, check_rpath, etc.)
> - Care and feeding of the build process
> - Retiring obsolete subsystems
> - Other types of code cleanup
>
> While this community has some similarity to the Linux "kernel janitors"
> [3] project, our focus will be more on structural changes than on code
> examination ("we're not here to sweep up after your broken project"). The
> community expects that it will necessarily have some overlap with the
> OS/Net community and with the Tools community.
>
> [1] http://www.opensolaris.org/jive/thread.jspa?threadID=12619&tstart=15
> [2] http://www.opensolaris.org/jive/thread.jspa?threadID=12928&tstart=0
> [3] http://janitor.kernelnewbies.org/
+1
Note that I'd like to add three points to the initial ToDo list of the
community:
- Switch the code over to make string literal |const|, e.g. compile all
sources with Sun Studio's "-xstrconst" option. This should save a
considerable amount of memory and reduces the per-user memory
consumtion.
- Start making use of C99 constructs, for example allocation of
dynamically-sized temporary memory from stack in both userland
applications and kernel code (yes, I am aware about the issue that a
kernel stack overflow is BAD (but there are always ways to work around
that))
- Track down shell scripts (sh, ksh, bash) and investigate whether ksh93
would make sense from the performance point of view. ksh93 has lots of
builtin operators (for example "~(E)pattern" is the equivalent for
"egrep pattern") which are in most cases MUCH faster (compared to
something like % (i=0 ; while [ $i -lt 1000 ] ; do echo "$i" | egrep
"^.0." ; i=$((i + 1)) ; done) # which causes 1000 |fork()|s of "egrep"
(try % ksh93 -c '(i=0 ; while [ $i -lt 1000 ] ; do [[ "$i" = ~(E)^.0. ]]
&& echo "$i" ; i=$((i + 1)) ; done)' # )) than calling the native
commands.
----
Bye,
Roland
--
__ . . __
(o.\ \/ /.o) [EMAIL PROTECTED]
\__\/\/__/ MPEG specialist, C&&JAVA&&Sun&&Unix programmer
/O /==\ O\ TEL +49 641 7950090
(;O/ \/ \O;)
_______________________________________________
opensolaris-discuss mailing list
[email protected]
