10 Thou shalt foreswear, renounce, and abjure the vile heresy which claimeth that “All the world’s a VAX”, and have no commerce with the benighted heathens who cling to this barbarous belief, that the days of thy program may be long even though the days of thy current machine be short.
This particular heresy bids fair to be replaced by “All the world’s a Sun” or “All the world’s a 386” (this latter being a particularly revolting invention of Satan), but the words apply to all such without limitation. Beware, in particular, of the subtle and terrible “All the world’s a 32-bit machine”, which is almost true today but shall cease to be so before thy resume grows too much longer.

Ten Commandments of C Programming

This some times reminds me of how I feel about assumtions. Uusally met with me thinking.

“I’m not running Windows, !@#$ you”

“I’m not running Linux, !@#$ you”

“I’m not using MS-DOS, !@#$ off!”

“I’m not working in Visual BASIC or Visual C++, really un-!@#$ you”

“What if I’m not on a 32-Bit system, oh go !@#$ your self”

lol.

Reading

Some good links

http://www-128.ibm.com/developerworks/power/library/pa-ctypes1/index.html

http://www-128.ibm.com/developerworks/power/library/pa-ctypes2/index.html

http://www-128.ibm.com/developerworks/power/library/pa-ctypes3/index.html

http://www-128.ibm.com/developerworks/power/library/pa-ctypes4/index.html

Maybe I’m werid but it is interesting stuff ^_^

1 Thing I’d love to see. Is a simple yet powerful language like Perl/Python/Ruby thats got a C/C++ style compiler like GCC. Like being able to run the C pre-processor on the files before compilation and being able to link the object files together into an executable.

But also, a simple interpreter that implements the standard library and can run single file ‘scripts’ like a Bourne Shell. Yet with the options of being able to ‘link’ to pre-compiled libraries of code to extend itself in order to run the script. That way, people could use the interpreter for use interactive (like irb / python) yet run scripts and not have to have a development environment (compiler and friends) and any precompiled libraries could be distributed with the script.

Like

import math_module
import sockets_module

Compile the app and link it to a math and socket library. Or i f we’re using the interpreter.

foo -l lib1, lib2

or even in the script

#!/usr/bin/foo -L /usr/lib/foo/

Would be kinda cool.

The power, flexibility, and speed of a compiler yet the simplicity of an interpretor. But with keeping the ability to also use various libraries of code with the scripts. One could probably do some preprocessor work to include the library code in the script before using it too and save disk space.

Wouldn’t it be awesome if one could work on a Project in Python instead of C++ and keep a compiled executable. Yet when ya need a little script for some ‘loose ends’ to just go ahead and write a script to do it in the same language as the project ? Hehehehe.

I think D has the ability of doing a script, like #!/path/compiler -run at the top of the file. No clue if its a dual-purpose compiler/interpretor or if it just compiles an executable in temp space and runs it or what.

rf.c semi-final

Well to make a long story short. I found out that calling setlocale() made my program (rf) crash when ever I used to -b switch. But strangely on OpenBSD it ran perfectly fine.

Sure enough after 2 days of tracking I found a few bugs and the little blighter I’ve just squashed was just the one I expected to find. It was just a matter of how. Just a bad pointer, my guess is after the setlocale() call. It was pointing some where FreeBSD didn’t like and without the call it wasn’t enough to get ballocked at. Who cares how thats possible… as long as its fixed hehe. Well as best as _I_ can tell any way.

The projects grown and I’ve learned quite since the start. Both about working in C, using Makefiles, dealing with GCC, GDB in more depth, and groff e.t.c. For sear size we are at:

#wc [ lines | words | bytes | filename ] 
16 121 778 COPYING
25 68 735 Makefile.optimize
255 1023 6517 rf.c
11 114 3636 rf.o
87 333 1819 rf.1
9 68 404 tags
403 1727 13889 total

I have 3 Makefiles, a generic makefile for building the application on i686. Makefile.debug that does likewise but includes debugging symbols. And Makefile.optimize that has a few extra flags with mostly optimization options in it (hence the name). Each also can do a ‘make install’ for me. All in all gcc gets called from Makefile.optimize with:

-Wall -Wpointer-arith -Wcast-qual -Wcast-align -Wconversion -Waggregate-return -Wstrict-prototypes -Wmissing-prototypes -Wmissing-declarations -Wredundant-decls -Winline -Wnested-externs -std=c99 -march=i686 -fforce-mem -fforce-addr -finline-functions -fstrength-reduce -floop-optimize

I made some other changes too. I rewrote parts of main, its still not pretty but a wrapper on fopen() I think makes it easier to dick with. I finally added a -v option to see the file name. I was trying to figure out how to make a variadic macro and stumbled onto the __FILE__ and __LINE__ additions to C hehehe. I also decided to have it clean up after its malloc’ing. AFAIK a programs memory is free()’d after it exits so I didn’t bother. Now I did for completeness sake. clean_up() is called after handling the output done in read_bottom (which actually uses read_all() to printout the file). So that if for some odd reason theres really low memory or some thing and it takes *awhile* to handle malloc’ing and seeking semi-end-to-end through a linked-list. I figured it’d be a good ‘just in case’ even if its not likely.

I really would like to rewrite read_top() b/c I don’t know of any comparable function to fgetln() on windows and a Macro that inlines a rough equivalent to my beloved err() function too. Which likewise is available from the BSD and GNU C Libraries and both are marked as ‘first appeared in 4.4BSD’ in the man pages. And AFAICT not available from Microsofts implementation.

One thing I need to do is learn more about groff and the man page macros and stuff. So far its mostly been a matter of reading my systems man pages and trial/error to get a fairly decent manpage.

I try to keep things small, my displays usually 80×25 or 80×35 if I’m working on Windows (GVIM) and 1 thing I really hate. Is a huge function thats so freaking big, you don’t even remember the NAME of it by the time you hit the closing bra}e. So lol, I do my best not to write functions that long. At 60 some lines with comments I think read_bottom should be broken up into 2 smaller functions, the extra function call is negiable overhead I’m sure. I just think it would take longer to read. Really if you snip off the variable declarations and list startup (which could be function’ized) it almost fits on 1 screen. be_verbose man as well be a macro or inline function too. Although a compiler could take care of that choice it self I’m sure.

To be honest, I just like functions that are small, do whats asked, and avoid being swiss-army knifes.

/*-
* Copyright (C) 2007
* Terry M. P*****. All rights reserved.
*
* permission to use, copy, modify, and distribute this software for any
* purpose with or without fee is hereby granted, provided that the above
* copyright notice and this permission notice appear in all copies.
*
* THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
* WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
* ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
* WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
* ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
* OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
*/


// vim: set noexpandtab ts=8 sw=4 ai :
// vi: set ai nu ts=8 sw=4 :

#include <err.h>
#include <stdio.h>
#include <stdlib.h>
#include <limits.h>
#include <locale.h>
#include <unistd.h>

struct lnpos {
long nl;
struct lnpos *next;
} lnpos = { .next = 0, .nl = 0 };

static FILE *open_file( char * );
static void read_all( FILE *, int );
static void read_top( FILE *, int );
static void read_bottom( FILE *, int );
static void clean_up( struct lnpos *);
static void be_verbose( void );
static void usage( void );


static const char *this_progname;

/*
* rf - read file to standard out v1.27
*/

int
main( int argc, char *argv[] ) {

char *erptr;
int t_flag = 0, b_flag = 0, v_flag = 0;
int ch, lncnt;
FILE *fp;

(void)setlocale( LC_ALL, "" );

this_progname = argv[0];

while ( (ch = getopt(argc, argv, "b:t:v")) != -1 ) {
switch (ch) {
case 'b':
b_flag++;
lncnt = strtol(optarg, &erptr, 10);
if ( *erptr || lncnt <= 0 )
errx( 1, "Improper line count -- %s", optarg );
break;
case 't':
t_flag++;
lncnt = strtol(optarg, &erptr, 10);
if ( *erptr || lncnt <= 0 )
errx( 1, "Improper line count -- %s", optarg );
break;
case 'v':
v_flag++;
break;
case '?':
default:
usage();
/* NOTREACHED */
}
}

/* This is ugly but cleaner then w/o open_file(). */
if ( argv[1] == NULL ) {
usage();
} else if ( (t_flag < 1) && (b_flag < 1) && (v_flag < 1) ) {
fp = open_file( argv[1] );
read_all( fp, lncnt );
} else if ( (v_flag != 0) && (t_flag < 1) && (b_flag < 1) ) {
fp = open_file( argv[2] );
read_all( fp, lncnt );
} else if ( t_flag > 0 ) {
fp = open_file( argv[3] );
read_top( fp, lncnt );
} else if ( b_flag > 0 ) {
fp = open_file( argv[3] );
read_bottom( fp, lncnt );
} else {
usage();
/* NOTREACHED */
}

if ( v_flag != 0 )
be_verbose();

fclose( fp );
return 0;

}


/* Simple fopen wrapper to keep the if...else if...else blockage from
* getting even uglier. Since doing other wise would defeat the purpose of it.
* open_file() halts the program if fopen failed.
*/
static FILE
*open_file( char *arg ) {

FILE *fto;

fto = fopen( arg, "r" );
if ( fto == NULL )
errx( 1, "File can not be opened or"
" does not exist -- %sn", arg );
return fto;
}


/*
* print out an open file to standard output
*/

static void
read_all( FILE *fp, int lncnt ) {

while ( (lncnt = fgetc( fp )) != EOF ) {
printf( "%c", lncnt );
}
}

/* Read n lines from the top of the file.
* note that it was very inspired by the head(1) implementation of BSD
*/
static void
read_top( FILE *fp, int lncnt ) {

char *cp;
size_t error, rlen;
while ( lncnt && (cp = fgetln( fp, &rlen )) != NULL ) {
error = fwrite( cp, sizeof(char), rlen, stdout );
if ( error != rlen )
err( 1, "stdout" );
lncnt--;
}
}


/* Read n lines from the bottom of the file
*/
static void
read_bottom( FILE *fp, int lncnt ) {

int hmany = lncnt;
long nlnum = 0;
long where;

struct lnpos *root = 0;
struct lnpos *cur = 0;

root = malloc( sizeof(struct lnpos) );
if ( root == NULL )
err( 1, "can't init the list" );
root->next = 0;
cur = root;

cur->next = malloc( sizeof(struct lnpos) );
if ( cur->next == NULL )
err( 1, "can't add first node" );
cur = cur->next;

/* read the file, count every 'n' and store them in a new member of
* our linked list.
*/
while ( (lncnt = fgetc( fp )) != EOF ) {
if ( lncnt == 'n' ) {
nlnum++;
cur->nl = ftell( fp );
cur->next = malloc( sizeof(struct lnpos) );
if ( cur->next == NULL )
err( 1, "can't store line feeds" );
cur = cur->next;

}
}

/* Let's mark the end of the list and move to it */
cur->next = malloc( sizeof(struct lnpos) );
if ( cur->next == NULL )
err( 1, "can't terminate the list" );
cur = cur->next;
cur->next = NULL;

/* rewind our linked-list and seek to b_flag segments. So readall() starts from
* the correct fseek offset to print till EOF. - This keeps down on
* unnecessary code here !
*/
cur = root;
while ( hmany < nlnum ) {
cur = cur->next;
nlnum--;

}

where = fseek( fp, cur->nl, SEEK_SET );
if ( where != 0 )
err( 1, "could not seek through the filen" );

read_all( fp, lncnt );
clean_up( root );
}

/* Simple destructor - walk the list and free() the memory before the
* program exits.
*/
static void
clean_up( struct lnpos *rp ) {

struct lnpos *t = rp;
struct lnpos *atpos = rp;

while ( atpos != NULL ) {
free( t );
atpos = atpos->next;
t = atpos;
}
}

static void
be_verbose( void ) {

printf("==> %s <==n", __FILE__);
}

static void
usage( void ) {

(void)fprintf( stderr, "usage: %s [-t count | -b count] "
"[-v] [file ...]n",
this_progname );

exit( EXIT_FAILURE );
}

I’ve learned so much in the past 2 months, its awesome. Really this is the largest project I’ve ever worked on in my ~/Programming directory. 255 Lines isn’t much but when your a bumbling baboon trying to learn in the wilderness. It kinda helps to learn by reading and writing code I guess.

I also want to start assembling a little personal library of functions and macros that I can draw on to do odds and ends without rewriting them. After all with my little ‘toys’ I might have use of other stuff from time to time hehe.

Time for BEEEEEEEEEEEEEEEEEEEEEEEEEEDDDDDDDDDDDDDDD!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

You know… if I was a smarter man I might pursue learning more about Particle Physics.

Its fascinating…

TBH I don’t hate mathematics per say. Its just the way school tends to put me to work on’em.

plan:

A table that gives A-Za-Z and 0-9 a nummerical value per char. So going to foo in table gives us bar.

Find our position in the table by looping through the table. But instead of A-9 or some thing. Do it sorted via commonality.

Increment our position in the table by an arbitrary rotation value. If we’re at the end of the table, start from the beginning again.

Remember that !!!

also, gotta figure some thing out /w that PBI. Dunno a good way of doing it any better then now for the NIC list.

Backup took about 2 hours to take, plus a lot of set up time because of networking issues… It was to much bugger working with the install disk as a subsitute R&RO disk.

So booted and dropped to single user, mounted stuff RO after having fsck’d it with the install disk. Dumped it ‘cross the network after sorting out a minor issue with PF ! Alls good but next time I think I need more compression and less speed on the gzip.

Only problem is I need to get the archive to a partition I can trust, that would be a FAT32 one but the max file size is 4GB and the backups a bit fatter then that. Tried to copy it to an ext3fs so I could test doing a restore to a Linux partition. While ext3 I think can handle files 16GB++ to a few TB. I could only create a 2GB file on it so bonkers.

If I can bzip or 7z it down to the right size I might be able to burn it to DVD-R. I think my best bet would be to *try* and copy the the dump from the ufs2 slice to my ntfs one using an ifs driver for it on XP.

Caesar

Ok so I’m a crazy stooge !

I spent the night trying to evade my allergies by implementing a simple rot13 program based on some super-short implemnetations of the algorithm from USENET.

Shes got interactive and file based rot13, a -r option set up to use a different rotation then 13… but I havn’t implemented those yet hehehe.

Had a lot of fun working on it and learned a few things while working on the manual page that even let me clean up rf.1 !

Gotta look at: http://synergy2.sourceforge.net/

Taking a dump(8) to – (stdin) and piping it to gzip and using |ssh over the network to use dd to create a backup might not be a bad idea.

dump -0f – / | gzip -2 | ssh -c blowfish Terry@IP dd of=vectra_date.gz

hmm.

later

Got to remember to back up my home directory.

Think it’d be some thing like

cd ~
su
mv Music /somewhere
pax -wf /Backups/_usr-home-Terry.pax ./*
cd /Backups
gzip -9 _user-home-Terry.pax

I’ll worry about it later