the text is in exactly the same place everytime
hi .. i want to output colored output on gnome-terminal from my c program..can someone send me a good tutorial link…i m not able to find any
have you looked at curses?
where's your question?
couldnot find them
where to look
twkm, i need some tips on how to strip down everything i dont need
did you learn c, by chance?
where to look for them
twkm, im in the process of learning C
write a loop, discard until you reach the desired beginning position, stop reading when you have read "all" of it.
normally i would suggest using ''man curses ncurses''.
i was asking about colored output using ^[ escape seq
sadly i provided that to you earlier.
can u provide me again plz
i suppose, ''man console_codes''. which describes the virtual consoles, but some of it also applies to gnome-terminal.
twkm thnx
uh-huh
twkm i think theres a srt function that can do it
or part of what im looking for
i havent yet grasped the idea of pointers in a string
Mike_Oxbig, you mean you haven't grasped pointers?
troll?
yes
no, the thing is, i know how to search a string for a char
but i need to search it for a string
Then what's the prob? 
strstr()?
yes that. but …
get a c book
hold on, that finds the first occurance, and outputs the rest of the string with it right?
man strstr !
i aint on *nix
http://www.cplusplus.com/reference/clibrary/cstring/strstr.html then
but im reading web manual
no it returns a pointer to the first har that matches
thanks
mornin'
you can always loop it, starting from the end of the last string found (pointer).
err, 'char'. Morning PoppaVic
hi centosian
gimme a sec to brew this over
oh and I just dropped in so I don't have context - I was describing strstr()
centosian, there is no context to be had.
hey, there is
centosian, im trying to extract a sentance from a web page
C is probably not the ideal tool for that Mike_Oxbig
but its for a plugin
Perl ftw
your browser and a clipboard works 
zacs7++
jeees
Mike_Oxbig, if you haven't already noticed, C is not exactly "string friendly".
a plugin for what?
probably for perl !
well, I find C sorta' friendly, but he & I go waay back
C is like vi. It's friendly, but choosy who it makes friends with
im friends with vi
PoppaVic, you'd probably find FORTRAN and punch cards friendly !
I think C is a she, like the ocean
Is she hot?
hmm… C isn't that fickle
now, the compilers and platforms - those are female
Mike_Oxbig, a plugin for what?
pidgin
:o C is lesbian!
I wonder what gender C, compilers and kernels have in languages with gendered nouns
And you want to? Extract sentences from a webpage?
what is the missing number? 1 4 9 ? 25
Go find a regex lib me thinks
no, sentances.
sorry
are you serious?
penguinFunk, it's 0
zacs7, its just one sentence, its a webpage thats not changing, the text will be always in the same position
lol
nope
Mike_Oxbig, read it's offset from the start then
so just a simple find, copy will do
I'm attempting to redistribute my C program. I'm trying to use autoconf and automake to achieve this. Along with my .c file, I have two shell scripts (amazon_get and iMpd_firstrun) and one jpg file in the same directory. My configure.in is [ http://rafb.net/p/jCeiiX32.html ] and my Makefile.am is [ http://rafb.net/p/luRC3m42.html ]. Everything's fine until the make distcheck where I get an error about there being no targets to make amazon_get. Any
id
strncpy()
hehehe
!autotools
autotools is http://sources.redhat.com/autobook/
zacs7, ill dig about, see what i can do, ill come back
I can't wait Mike_Oxbig !
I read it with no worries
what text, and from what page? maybe we can provide a perl oneliner
lol
Like I said, my issue is very specific
can i paste here?
dude, my feelings for autoshit are quite specific
finishing the question might help, your post ends with 'Any id".
centosian, you could probably cram it onto 1 line of C
Mike_Oxbig, pastebin it
no use rafb from the topic
link?
:P "Any ideas why?"
it scrolled out of my reach
/topic
make[1]: *** No rule to make target `amazon_get', needed by `all-am'. Stop The exact error
Mike_Oxbig, tell me you didn't rejoin just to see the topic.
Well, I'm not very experienced, which is why I'm attempting to use autotools. Yes, the Makefile generated isn't human readable, I agree with you one that
s/one/on
http://rafb.net/p/JosZSx62.html
zacs i did
i would like it in C because i need to work more with the text
Ahh
Mike_Oxbig, Google have an API you know?
oh god
That's a very flimsy way of doing it.
last time i looked at an java api hosting my brain melted
And that webpage is due to change whenever it wants
Mike_Oxbig, you shouldn't be scared of apis
i am
im horrified
they are meant to make things easier..
i looked at pidgins API. i dont want to relive that
ever
so you don't have to do things the "flimsy" way
Mike_Oxbig, c'mon..
but this is a simple task for a quick hack
bin_SCRIPTS are supposed to be generated by sources, maybe you need EXTRA_DIST=amazon_get
do you realise what im trying to do with it? what plugin im writing?
Mike_Oxbig, simple however if you want to do it your way, read from div id=resultbox …TEXT IS HERE/div
Okay, I'll try that right away
Mike_Oxbig, do you know what an api is supposed to do?
Mike_Oxbig, if you don't like API, is programming really for you?
or what the definition of an API is?
Mike_Oxbig, what is the difference between the standardlib and any other api?
Ooh, Actually Pretty Inconceivable
APIs are good things, but …. argh, im not a programer
and im not going to become one in the future, but coding a little doesnt hurt?
EXTRA_DIST stuff won't be installed, no?
Mike_Oxbig, you realize that the C standardlib is nothing more than an api
Mike_Oxbig, I'd be disappointed if my translator broke cause the programmer decided to cut corners.
I want them to be installed in $PREFIX/bin
Mike_Oxbig, so you are using an api whenever u use C
*shrug*
Mike_Oxbig, i believe you are just wasting your time because of a word
go look it up
my C application depends on calling the scripts
no, only distributed, but bin_SCRIPTS will make it installed
At the high risk of sounding stupid, what is a for loop of the form "for(; {" this is driving me nuts, something to do with stdin? Haven't seen it in c++ B4
so I have to put it in both bin_SCRIPTS and EXTRA_DIST?
Mike_Oxbig, A set of routines that an application uses to request and carry out lower-level services performed by a computer's operating system. Also, a set of calling conventions in programming that define how a service is invoked through the application.
yes, make distcheck tries to rebuild the package from the distribution, and maybe it complains about amazon_get not being distributed
Mike_Oxbig, using apis is a little programming; writing apis is another thing
alright, I'll attempt this. Hang on for a moment
i understand why its good. i also understand that im not going to waste 100 hours finding out how an api host works, just so i can build a truly flexible application that only 1 person in the world is going to use
Mike_Oxbig, no you are not
Mike_Oxbig, if you need 100 hours to understand the usage of an API you have forefit your right to live on this planet
this is why i will write this little hack, if it works ill think. hmm, this could be done better….
an api is DESIGNED to allow easy access
do you understand?
so you should be fine understanding it in some minutes by reading a basic tutorial about it
Mike_Oxbig, I would rage it if someone took the short route and the plugin broke.
if you try to understand the apis code tho, you will probably fail without knowing the language it's written in
Thanks a bunch! I'd been working on this problem for hours at a stretch! It works perfectly now
ok, how about, i do it this way. and if it works, ill do it through API ?
How bout this? If the API supplies a method to do it, use the API, else do it the flimsy way.
well, as far as i looked, google api host doesnt provide one for translation
Mike_Oxbig, it's not going to be as simple as that I presume.
this *might* help: lynx –source 'http://translate.google.com/translate_t?langpair=fr|en&hl=en&ie=UTF8&text=television&langpair=fr%7Cen' | perl -n
-e '$_ =~ s//\n/g && print $_' | grep result_box
fetch the page, put newlines between chars, grep out the line with result_box. Voila!
takes you 90% of the way there
centosian, you are a true perl god.
apart from this has to go in a plugin in pidgin
Mike_Oxbig, who said perl couldn't be involved?
thanks, lynx and grep helped too. now wtf is pidgin
ah, http://www.pidgin.im/
dont think about pidgin, think in C
Mike_Oxbig, go and find a regex library
to parse html?
the strstr is your friend. Find the result_box text, and then find the end, then clean up.
yes
don't parse the html
or what centosian said !
what does a for loop with no var,condition,inc do, eg "for (; {" ?
pidgin provides a way to strip html, so i only need the div id=result_box dir=ltrtext/div
…
"forever"
that's the same as while(1) {
same as while(true)?
ah, tnx
obscure
npt "obscure" if you can read.
npt?
npt/not
anyway
centosian, why lynx?
wget would probably do a better job at this
no particular reason. Easiest way to fetch the page onto stdout that I could think of at the moment. What's the flag to wget to have it push the html to stdout?
curl?
centosian, i believe it's wget -O -
lwp-request works too and it's all perl
(ducking and running - that's not C!)
centosian, but then you need to disable it's verbosity which i dunno atm
:P
lynx –source , ftw
lol
nah; lynx is more code to run
nopealicious, -nv maybe?
zacs7, tried it?
or -q
zacs7, almost
q is completely verbose
-q is perfect
wget –output-file=- -q then
not that I knew, man wget ftw!
hm, doesn't seem to work
-O - then
is that no-pea-licious or nope-a-licious?
odd, –output-file=- should have worked?
zacs7, yes
iirc
-O file
–output-document=file
zacs7, no it shouldnt've
centosian, nope-alicios
bit like booty-licios ?
nope
more like "no peal lick"-ious
^^ delicous
Zhivago i love you
who doesn't !
will you marry my pet?
no.
petrock++
haha
i got my petbottle martini
extra dry
LOL
I first read it as "no peal icious"
I read it to nope and giveup
half of the names on IRC I don't even parse, I just read enough for tab-complete
its nopealicious D
http://rafb.net/p/1fk7Cv50.html
i normally call myself nihil but since it was taken..
Mike_Oxbig, shudder
Mike_Oxbig, I'm not entirely sure that's what I meant, nor what centosian meant
hmmm
im worse at C than i think
Mike_Oxbig, pch+=sprintf("%s %s",front,back);
Mike_Oxbig, indeed you are
don't forget spelling words like "I'm".
Zhivago, why im not doing it either's
you might want to review sprintf
my apostrophe key is broken
Zhivago,?
yes nopealicious !
your example is broken.
nopealicious, you do know the prototype of sprintf?
oh
well yah, forgot the str to actually write to
stupid me
Mike_Oxbig, pch+=sprintf(pch,"%s %s",front,back);
the word 'I' as in first person I ponder about because it's still legible as lower case and doesn't confuse with anything else
less of an issue with typing than handwriting, perhaps.
nopeal, hold on, im not sure what you mean
Cin, it also makes a good variable name in uppercase.
Z, I guess
I don't think he is sure either.
Mike_Oxbig, man sprintf
what does sprintf have to do with the problem of extracting that text?
Yay guessing FTW.
extracting?
http://rafb.net/p/1fk7Cv50.html
presumably you are trying to extract the text between the two tag strings?
ohs he.. ic
nope, trying to extract the text from the webpage
so, use strstr again.
Mike_Oxbig, or a finite state machine
the brackets dont matter, i can strip html host with another function
Or sscanf?
and I suggest that pch is a particularly bad variable name.
Zhivago, is it for you?
Zhivago, I couldnt agree more
Zhivago, any reason why sscanf should be avoided?
in this case?
char *start = strstr(text, front); if (start) { char *end = strstr(start, back); if (end) { … } }
because it wouldn't solve the problem?
Well, it'd half solve it, well not really
zacs7, well.. very unexpected input for example?
nopealicious, you can't talk
Mike_Oxbig, i would use a for loop and look for the end of the string with a finite state machine..
an astute diagnosis.
then i would know the amount of chars the string has and use memcpy/memmove respectively
just use strstr twice, given how half-arsed this is.
char * start = strstr(rawpage, front); char *end = strstr(start,back); // now your text is between start and end
see the example code above.
Mike_Oxbig, Sure the Google API doesn't provide something to do this?
yes, what Zhivago said. zacs7 he's scared of the API
there are plenty of apis around — expat would do nicely, also.
it doesnt provide it acctually
Mike_Oxbig, pretty sure you'd find a translator with API (even if it isn't google translator)
assuming that the page is valid xhtml. although I suppose Expat probably has support for being less strict
its just one sentace in page, how hard can it be
using expat for that is correct.. but insane
you're in C-land. String processing is a bizznitch
centosian, more insane than his flimsy method?
ok, i can use php to do this, then simple GET a string i want to translate and it will be back
depends which is more maintainable and less lines of code
simply
HAH!
PHP!? Dun dun dun.
well, why not call google and ask them to code an API
*snap snap*
I already have
strstr is so sufficient for this task
Cin i knooooooooooooow
Mike_Oxbig, then dooooooooooooooooooo something
instead of whining aorun dhere
Mike_Oxbig, I don't think 'know' has that many O's
i am, im trying to write something while talking here
strstr is sufficient for the trivial problem, but it won't scale or compose
why aren't you using a half-arsed scripting language like python?
Zhivago, that was my point before.
Mike_Oxbig, try writing something while NOT talking here
good point
just write something to parse a hard-coded example string
he want to write a plugin for some im client
brb
centosian, I can't wait to use his plugin for pidgin then google changes their translation page
hmm, if we were to eliminate the deluded, insane and mentally incompetent the world would be a much lonlier place.
lonelier, even.
then don't wait
Yeah, would you miss me?
see, much quieter
Yes, it's nice.
SHHH!
Sorry!
damn i need ciggies
nopealicious, make your own out of toilet paper.
and tea leaves.
:x
all people with a religion of faith - so most of the worlds population….
and, of course, people who can't use the apostrophe.
not to be confused with diluted.
Encourage everyone to conserve apostrophes. Use them only where theyre needed. Example - contractions work perfectly well without 'em
Ill thought, is it not?
its a standing policy of mine to promote language to evolve to a more efficient configuration.
savetheWorld, That's a poor excuse not to use them?
damn my spider phobia! I tiny black spider was on my arm and I automatically jerked my arm like hell
it is a less efficient configuration that you use.
its a standing policy of mine to promote language to evolve to a more efficient configuration.
Cin, send him to /dev/null !
I read that complete idiots (as in, very low intelligence) can still have the capacity for langauge, going on and on about stories in intricate detail, as fictional as they may be
if you meant to say "I am illiterate," then that would be both more efficient and correct.
you should get into lojban and (um) esperanto
Uh oh, they're onto me!
we'll have to disagree about that. Of course since I actually have a linguistics and analysis background Im inclined to trust my notions about it. 
well, he can read and write, but not properly
hmm, I'd ask for your money back.
properly as defined by grammarians or linguists? the opinions of grammarians are, by definition, always obsolete.
Note that he used an apostrophe in "we'll".
otherwise, I suggest that you take some basic perceptual psychology and measure reading speeds.
you'll note that there's a significant speed cost involved in parsing your hideously mangled text.
Quartus, that means his apostrophe key only has infinity-1 pushes remaining.
Implementing an actual Speech recognition product introduces one to the realities of the way people actually use language, rather than the imaginary way grammarians think people "should" use it.
Zhivago, especially if you use a screen reader.
savetheWorld word up
Are you using a screen reader?
what do screen readers have to do with it?
Zhivago, language is a tool for communication; savetheWorld is obviously to communicate his thoughts (at least to me) which makes his language be fine
go and do what I suggested above.
obviously able*
please rewrite that sentence so that it makes some kind of sense.
I dont see a suggestion.
http://rafb.net/p/8JCxPQ11.html
nopealicious, C is also a tool, doesn't mean you can write it against the standard and the way it was intended "just because everybody else does"
Zhivago, language is meant for communication not pedantry
as long as a message is transfered successfully over a voice channel the language is fine
and i can udnerstand savetheWorld
you'll note that there's a significant speed cost involved in parsing your hideously mangled text.
Mike_Oxbig, what is '72y49872369'?
Are you using a screen reader?
just some mangled text
what would that be relevant to?
you seem to be unable or unwilling to realize tiny variations in language don't matter but for some ppl its fine if they get the message
you express pedantry through communication, and communication is done via language
research shows otherwise.
Microsoft Sam for the win!
Cin, it doesn't matter as long as the remote end GETS the message correctly
er, okay
Zhivago, shows?
Cin, don't u think?
actually research shows that nopealicious is correct. So say the PHd's that designed and built Dragons System's Naturally Speaking.
you appear to have defective reasoning.
Well, I'm understanding about 50% of what is being said, and I naively speak English. Night all.
Cin, Zhivago it sah neeb dais taht teh hunam midn is cabaple of asweome thinsg
it just sounded like you contradicted your point, so I have nothing to add to that
and it is your problem if you're unwilling to use your brain
no?
how does that support your argument against spelling properly?
you can have an opinion. That doesn't make it right. *plonk*
My mum is not here, can someone tuck me in over IRC?
nopealicious, it sounds to me as though you've never been in a position where it was important — even vital — to send and receive communication in as unambiguous and precise a manner as possible.
ah, you can't answer rationally.
Quartus, THIS IS NOT SUCH A PLACE
its not like anyone is going to do something important with the information gathered here
I need a tuck in
It's cold.
hmm, you were lying about linguistics, weren't you?
has revealed themself a prescriptionist grammarian. The worst kind (Hi Miss Grundy!)
Zhivago!*@* added to ignore list.
Miss Grundy!?
Zhivago, are you as "stupid" as a computer? unable to recognize patterns you haven't seen yet?
savetheWorld, I can relay what he says for you
favour returned
Zhivago, what does 'Zhivago' mean in Korean/Chinese?
I'm able to recognise new patterns — I see no reason that I should tollerate laziness in others.
nopealicious, I disagree; but even assuming if you feel the discussion here would not benefit of additional clarity, the skills for so doing are of cross-spectrum benefit; also they are trivial.
Zhivago, that is your problem
nothing in Korean.
And In Chinese?
I don't know if there is any meaning associated with it, but it's unlikely.
Quartus, what does it matter in the big scale if someone uses a ' less or misses some charcter somewhr
okay, just wondering. Night all.
or were you unable to read the last sentence without a problem
it matters because it imposes a cost upon everyone else.
nite zacs7 thanks for help
i mean it's ridiculous, your brain does the work anyways..
and it matters if you want to be a programmer.
Mike_Oxbig, lol what help?
it's not like you needed to concentrate for reading
no, your brain does a lot less work if you write properly.
ha
Zhivago, yes if you talk to a "stupid" computer
read some studies on reading speed and comprehension in the presence of errors.
ok Zhivago i will take you as a "stupid" computer then
nopealicious, accuracy and clarity in communication is absolutely important in every meaningful human pursuit.
Zhivago, I have, I can read upside down
no u read something about languageinterpretation in the human brain
what does upside down have to do with it?
It's slower?
please rewrite that in english.
Quartus, what is a meaningfull human pursuit?
what do you call it when someone tries to prove a point by making reference to their experience and other external sources that supposedly back up their point for them?
if u cn rd ths u r gud
yes, but that's an issue of orientation rather than orthography.
one that doesn't find it's with the explosion of the sun in 4 billion years?
centosian, Oh my goodness, I'm god!
Your question takes me back to my belief that you aren't engaged in any, and thus don't appreciate the point.
Cin, an invalid proof
on topic people, please
Zhivago, iwillmakesuretotypelikethatasoftenaspossible justforyou
haahhaahhahhaahaha
willhelpyouunderstand "lazy" ppl
An apt description.
Potentially helpful, especially when the group includes 8 year olds as well as extremely experienced (not just old) people. It helps the listener (not to prove a point) but to focus on what might be more important points.
my question?
repetition creates neurons. what wires together fires together and vica vers
oh come on, dont be so mean
sorry Was AFK, watching my daughter's magic tricks..
Oh My I was banned? heh ! 
Cin, nopealicious' question
Sad how threatened the Miss grundy's of the world become when their "proper" notions are shown to be flawed.
"appeal to authority".
ah, that might be it
please learn how to use the apostrophe properly.
i'm affraid to ask for help again…
bring a sense of humor and a thick skin, Mike_Oxbig
apostrophes denote ownership or omission, not plurality. who/what is Miss grundy?
centosian, I just dont want to be told "get a C book" because I left it in england when i moved
"get another C book"
hm, on the other hand Mike_Oxbig , we can't teach you C as well as a book.
damn, didn't see that coming
lol
a C book is fairly well essential for the basics.
we're far happier to answer intellient questions.
is there an online C book ?
intelligent
Zhivago, yea i understand
I*
this looks OK : http://www2.its.strath.ac.uk/courses/c/
Explore http://www.iso-9899.info/
hi !
thanks centosian and tiny
this looks ok too: http://www.physics.drexel.edu/courses/Comp_Phys/General/C_basics/
please, your misspellings are making my brain work harder than necessary.
please learn how to use the singular form appropriately.
oh yea, I read that one before
Hi guys. I have a super basic question - I'm not a C coder but I just need to mod one program quickly. I have a string that's set with strdup and cleared with free() .. I need to know if it's currently set or not set. I'm trying if ( mystring != NULL ) but that doesnt seem to work right - is
that the right syntax?
that last url's first example is incorrect.
variables are always 'set'.
Zhivago:that is a shame
TomJ_ i think if (!mystring) will work
initialize it to NULL if you want to see if it was changed to something else.
what content of the string denotes its being set or not set? (every programmer can do this a different way.. )
well the content is variable.. I just need to know if it contains something. I know it gets cleared with free(), so I guess I'm asking "has it been cleared with free() or is there something in it?" .. so I could just set it to NULL myself then check it's not that? that makes sense
It's up to you to maintain the state. free doesn't/can't set the variable/pointer to NULL.
free() doesn't set a pointer to NULL
I am trying to work with arrays of structs, it's segfaulting, can anyone please tell me whats worng? typedef struct { char *str; } HS, int main() { HS **hs; hs = malloc(sizeof(HS *) * 2); hs[0]-str = malloc(sizeof(char) * 16); }
ok, I'll just manually set it to NULL after the free() call, then I can test for NULL elsewhere
that would work
thanks very much guys
hmm, the problem with C tutorials on the web is that they tend to be full of lies.
your program will have to set up and follow some internal set of rules to tell you|itself when a string has been cleared. perhaps someplace in the code previous programmers have done this, or they may not. You will have to determine if they did or not.
i think you want 'HS *hs'; and 'sizeof(HS) * 2'
assuming you want hs to be a pointer to a malloced, single-dimension array of HSs
yes, what pizza_ said. This compiles and runs: http://rafb.net/p/RRv5d388.html
you have not allocated any structs.
what did you allocate?
I will try, but with char, I've always used char **arr; then arr = malloc(sizeof(char *) * rows); and arr[0] = malloc(whater); It's a little bit diferent with struct, right?
you did not allocate any structs — what did you allocate?
yes, because your struct already contains a char *; you need to malloc room for the actual struct, not a pointer to it
I see.
"I see" doesn't answer the question.
I didn't
i'm looking for guides to using automake/autoconf/other build tools, or just make in general; anybody got any recommendations?
google.com
and lots and lots of patience
pizza_, Zhivago, centosian, thank you.
and hide any sharp objects or firearms
basically I just want my project to not be frustrating to future maintainers.
yw
(Including me.
start small and go from there. Some people pointed me to the gnu 'hello' program when I asked a similar question
the autotools book might help too, and I think there's an IRC channel on that topic but I forget what it's called. Anyone?
Hello
hello
OK, thanks
real horrorshow like
"Yo whatup"
side splitting
do you know what number of pages in statistics of the project in sourceforge.net means?
No.
ok so i can do an extremly dirty solution, just vecause i cant think of anything else
Is there some sort of description of XOPEN2K anywhere? I tried google, but all it gives me are a bunch of header files
which part of C involves XOPEN2K ?
at least pthreads
hello
r_r
i need to do a scan of a matrix in zig zag order. can you point out any algorithms in c?
brb
i need some help again
http://rafb.net/p/3J2eB833.html
oh no, i forgot something
http://rafb.net/p/lT8MI767.html this one
overuse of * ? what's the problem anyway?
juanmabc, im trying to extract a srting from a string
you start with a haystack, search for a needle, and copy - from the needle - to what you want to now call "a string"
simple.
but i dont know the needle
then yer gefuckt
i only know that the needle is div id=result_box dir=ltr*/div
where * could be anything
dude, you are seriously messed-up.
?
um, I told you how to do this already …
Zhivago, and im using you solution
everything in between the strings is what i want, so i get their lenght, and copy
ok, so what's the problem?
err, it doesnt work
pastebin it and someone will probably notice your logic error
what behaviour surprises you?
no acctually
or tell us what you expect, and what you get, and maybe someone can guess the problem
but i was hoping for something at least
s/guess/deduce/
ok, but first i need to printf the result
please try to be coherent.
with the right flag i mean. ok nevermind, ill post it in a second
what flag?
http://rafb.net/p/hZAV7f60.html
Hello
I am working in a small C project, and would like to use m4 as a preprocessor
is there a recommended way of working with it?
pre-process with it before compiling.
make may make this easier
Should the extensions of the files be
..c.m4
?
..h.m4
etc.
Yeah, if you are going to use m4, you already need to encapsulate the cpp interp and C
I'd use a single letter for convenient — perhaps .k
Is there some sort of description of XOPEN2K anywhere? I tried google, but all it gives me are a bunch of header files
alternately, use .c and pre-process to .k or something horrible like that
Maybe this is already an m4 question… but
if I'd like to have a standard set of macros
is there a way of _not_ having to include(`macros.m4') in every file?
I agree with Z, I think you need to look at using make and rules for this mess you intend
Aha, ok
maybe he needs a filename like .c4 ;-
Thanks for your help
.h from .c4, and .c from .c4 - nothing like a good explosion
you could probably use some horrible shellscript
I like you guys
people in other channels tend to criticize ugly hacks
heh, your at the worng time mate
it is an ugly hack, but it's your blood, sweat and pain, man
please learn how to spell "you're".
Thanks again
sure
guys, what would be your avice on which language to use to develop a web services on a linux platform?
Zhivago, Done
what should this web service do?
it should execute cli commands on the linux based hosting services machine
why?
it's for a control panel
not on my machine you ain't
what about CGI?
basic staffs
perhaps a half-arsed scripting language such as python would be a good choice?
because I need to do it within a web page and a desktop application
oh.. All you mean is using a browser for an interface.
so I thinked about a web servie, since I need it to be web based and also desktop based
Zhivago . PoppaVic
I'd suggest a half-arsed scripting language such as python, once again.
there is an example… Cups, yeah - Cups does this.
that's completely arsed Zhivago
not half
I have to do it as a compiled program, because code should not be accesible
hmm, were you dropped on your head as a child?
shouldn't be accessible to whom, and why?
doesn't matter if they can see the code if the source is r/o.
to anyone who could access the program
is this some kind of 'security measure'?
yes Zhivago
then you are an idiot.
really?
why do you think that trying to hide the source will help there?
could you assure that?
cipher your code using AES-2!
if you like — where is a screwdriver?
ssdd
lsd?
where?
well.. a program whichs code is not accesible is far more secure than other which exposes the source code to anyone who have acess to the machine
Toerkeium, bullshit.
no — it just sounds that way to an idiot.
Toerkeium, how does that logic work ?
do you think that the source code defines the program?
well.. if you can see how a program works, there are more chances that you can hack it
yes but things work both ways
is that why *nix boxes are so secure?
in a real secure app having the source code is irrelevant
or are regarded to be secure (when correctly set up)
you don't need the source to see how a program works …
and having a compiled program, I can make sure no one else can use it without my permission
nonsense.
omg
internally yes
heh
knowing the how a secure system works does not compromise security, which is why all secure ciphers are open
btw, I love sudo.. Wonderful tool
however, if your app is not secure, then yes, showing people the source may speed its compromise, but that doesn't mean it won't be compromised anyhow
if there is a security hole, you have more chances to get hacked if your code is exposed, and even for more unexperienced persons
externally nonsense as well.
ah, you're only worried about incompetent people …
just write really crappy code — that will confuse them.
instead of ROFL, instruct me
so that's why Windows doesn't ever get hacked right?
we really don't exist to "tutor" you. Yer welcome to fall-down-go-boom forever.
what /do/ you exist for
the only way to keep your app from being exploited is to make it secure
and if your code is open, more people can review it and alert you to problems or submit fixes themselves. as The current world populations of computers demonstrates, having closed source seems to be almost no barriers to having one's security compromised.
Laughing of others ignorance, is very stupid
yer welcome to tutor him - in a nice /msg win
and brainless
Virtually all the zombie botnets are comprised wholly of Windows desktop machines
that's not what I asked, though
Toerkeium, lets start with some basics.. A is too B as B is too _____ ?
yes, I am aware of that.
hmm "almost wholly" (i better fudge a little.. )
i'd have to agree with Zhivago — if you're concerned about security, pick a language that doesn't make secure programming as hard as C
if there is a security hole and someone really wants to get in, they'll find it. the trick is to not have the security hole.
I lean towards python but there's a lot of choices
Umm, I don't believe Z said that
heck, if you're intent on compiling to native code, use ocaml, heh
so basically, you won't do what I need in C or C++, correct?
yeah i wouldn't
you would do it in a more high level programming language?
I've done it in C. So what?
pizza, i thought the trick was risk mitigation after the fact?
nearly the only thing I'd do (am doing) in C is implement higher-level programming languages.
the trick is to put up a security consultancy firm… err
connect with bloody SSH ! problem solved
definitely an important part of the picture that some people forget
hmm, why aren't you just compiling to C then?
I have far less expectations of what a programming language can do for me
vlii, outsource your security responsibilities to a security firm with contractualy SLAs and payout penalties for violations?
sounds cool
ah, bonded securitors
simplicity, desire to reduce runtime dependencies
so, understood your point.. but what if doing it with a compiled program is mandatory, what would be your choice?
zhiv, yeah, usually at that point it devolves into the hiring company having to prove a security violation to the security company to collect a payout
tell the programmers you'll kill a kitten for each flaw found.
when i get ready to compile to native code it'll probably be via LLVM.
so, what's the C for?
zap0, i already kill kittens because i… cough… nevermind
a simple bytecode vm
erugh
and the runtime
sounds like more of the same "fighting the insurance company" - which doesn't help dick when entire systems of confidential info is compromised.
ocaml! or if it has to be something non-obscure, java.
poppavic, indeed.
simplicity and auditability are more important than performance to me at this stage
valgrind is an amazing piece of work.
damn, let me ask you, why not C or C++ ?
yeah, so, why are you using a horrible byte-code system
as opposed to what
just write a compiler in ocaml or lisp or something civilized.
which compiles to ocaml or lisp or whatever.
hell, just watching the news about folks fighting identity-theft will infuriate you, let alone when you are fighting the telco/cc-co, etc
then bootstrap off that by writing a compile hosting in itself to whatever.
Eh. Too hard to optimize, i'd like to avoid the dependency, too little OS integration, too little control over memory layout
also somebody else is implementing this language in CL already
so I don't feel the need
well, just use a CL compiler to generate it to native code
or to C, for that matter.
why?
the main problem is that C offers too many opportunities for basic security problems like buffer overflows
well, you have a bunch of decent compilers already available
hmm?
so you might as well use them to bootstrap it.
oh, ok.. understand why …
"hmm?" is not a sensible question.
I'm not necessarily a sensible person.
Anyhow, it's a marketing issue as much as anything
if someone is doing a CL implementaton, unless they have some severe problems, they'll be compiling it to CL.
Yes. That is what it does.
given that, you can then use the CL implementation's compiler to compile it to native code or C or whatever.
I don't want to compile to C, though, I want to use a VM
given that, it should be easy enough to write a compiler in that language to whatever you want to target.
why do you want a VM?
well.. thanks to all for the comments and suggestions
simplicity, mainly. also, I plan to do some amount of scheduling of multiple threads/processes within the runtime
in the same general spirit as erlang
and I can't say I really know how to do that in the context of generated C
plus! not having a runtime dependency on gcc would be nice.
your idea is to do lightweight threads?
No doubt a compiler to something native will be a reasonable addition to this project, but I'd want a VM as well.
yeah
pth works nicely
i know im nagging, but can you help me with something
just ask your question
http://rafb.net/p/K0DQkP32.html
hmm
err… length = end - start, no?
think about this: what is the type of char[i]?
what type is 'start'?
i is a number
int
start is a char (or array of chars) right?
both wrong
what type is it declared as?
well, int i = 0; ?
right
understanding what type your variables are will help you, i swear.
you liar! 
)
i know what types are my variables… i declared them that way
char * start = strstr(example, front); /* what type is 'start'? */
start is a char (or array of chars) right?
yes…
what type is 'start'?
you'd do it with whatever threading system you normally use in C — pthreads or whatnot.
a char
wrong
Yeah, that's the thing I'd want to avoid.
if your argument is that you want to support a lot of coroutines, then a VM may make slightly more sense.
pizza_, an array?
char c = 'A'; — char
i get a segmentation fault on a specific line where i try to print 3 variables, i try it with gdb, and try displaying each variable one by one, and they all get printed correctly. and another fact in linux only it segfaults in freebsd it works fine, what could that be?
wrong
a string dammit
wrong
most folks don't see coroutine vs thread
the coroutines could explicitly pass the control to the scheduler
note the '*', what does * mean?
aaaaaaaaaaaa pointer
correct
'start' is a char pointer
what type is 'start[0]'?
usually, that's true anyway - or they toss their "timeslice" to a specific alt-coroutine
right, that's what i'd have to do
an array? or character array?
I mean at language level, not at implementation level
eh? what do you mean
even printf("\n\n"); gives a segfault…
why could that be…
your computer's broken
throw it away and buy a new one
wrong and wrong
that doesn't happen in my computer, but i don't think that is the case
I really hate pthreads for mixing the idea of process and thread. It's screwed folks up bad.
the coroutines in the language you are implementing, may explicitly have a "yield"
no, the language doesn't have coroutines
is there code around the printf?
also, is it possible that something closed stdout?
ahh, ok, sorry then
symmetric coroutines need an explicit yield.
hmm
!testcase plutonas
basically aiming for something like erlang's processes.
A testcase is a minimal compilable example exhibiting your symptoms. Please paste a testcase at http://rafb.net/paste to help us assist you.
well, all coroutines need an explicit yield.
yes it is…
ohhh
whath could close stdout, that is possible
or some damned way to pass control around, anyway
pizza_, start[0] is an array. why wouldnt it be? (but i didnt declare it anywhere)
well, I think making a test case (see what candide said) would be a good place to start
nope, i can write testcase printfs but i don't i could reproduce that
does the problem happen consistently?
i comment out the printf where the problem happens
and then it happens on the next printf
etc
did you write the program that's doing this?
if 'start' is a char * then how can 'start[0]' be an array?
yes, but here on my computer and fbsd it doesn't
i don't see 'array' mentioned in 'start' so why would the [0]th element of 'start' be an array?
on the uni's computer that runs linux it happens, and only with big input…
but the data i want to print is correct, with gdb's display i can see it…
how large is the program?
pizza_ ah i know what you mean, i ment start as a whole was an array with start[0] being the element
'start' is a char pointer, not an array. 'start[0]' is a char.
so what would stop start[0] from coppying into "it"
!man strcpy
string.h - strcpy(char * restrict dst, const char * restrict src); - strcpy () functions copy the string src to dst (including the terminating `' character - http://www.iso-9899.info/man?strcpy
what is the type of the second parameter of strcpy()?
char
wrong, try again
yer head-banging cinderblock
does cinderblock float?
yes i don't think this will work long-term
start[0] is a char
nah; learning what pointers mean is a hard bridge to cross
yep.
i didn't ask what you passed it, i asked what the type of strcpy()'s second parameter was. it's listed above by candide
is it about the i?
no.
what is the type of the second parameter of strcpy()?
hello !
ah you mean what it sould accept
lo
pizza_, a pointer to a char
correct
it = strcpy(it, start[i]); /* what type is 'start[i]'? */
but start is a char *, start[i] is a char
correct
so what is the problem?
so i need to change cast ?
wait wait wait
no, you need to supply the correct type
no wrong word
casting the wrong thing doesn't fix anything
but if i cast it from chat to char * ?
it might make the warning go away but it won't make the program work
then you will have a very tiny pointer.
no, you need to supply strcpy() with a char *
but start[n] will always be a char
correct, so obviously that will not work
can i do char * start[]; ?
atring is not a char, it's not a char* - yer head is on sideways, mike
you can fix it by removing two characters and adding 1
you can do anything you like, but that will break your existing program
at least to give it the right type.
wait, im more confused now
char * my_char[]="bla bla bla bla ";
and maybe even correct semantics.
the big deal is that char and char* are different types.
please don't give bogus example code.
instead, i propose you buy a book.
that's wrong too
You cannot do anything you like
I want to eat almond roca
I can't.
you can eat glass if you like; C is flexible.
fortunately you can stop talking.
that's so mean
lol
sorry had a crash, on another linux university machine, it now seems to work correctly
with a newer kernel… isn't that strange?
whats the difference between start and start[n]?
weird
start[n] is *(start + n)
anyway, i'll put a note to the professor to run it on that specific machine
the difference between start and start[n] expressed in what units?
just to be clear, the compile turns *(start + n) into something like *(start + n * sizeof(char)), right?
as usual, wrong question
It's not the wrong question
Units are important.
units?
units.
strcpy wants a char *, i have start[0] which is a char
what type is start?
you've also got start, which is a char*
a pointer to char*
argh
no
a char*
a pointer !
how did you declare it?
a pointer to a char
'a pointer' isn't a C type
+ to a char
buy a book; goto book;
how did you declare it?
PoppaVic, goto spaghetti
char * start
ok, then the type is char *
dude, this dude has been at the same stupidity for hours.
use goto when it is the least horrible option — just like anything else.
amen.
PoppaVic, it was just a joke, nothing against goto
C should have a goforth keyword
seeing goto should make yer hackles rise. Seeing how they use it should either sooth or make you want to kill the author.
hehehe
Hi! I am trying to resolve a hostname, but the compiler gives me dereferencing pointer to incomplete type
goforth && spread(love);
target_address.sin_addr.s_addr = *((struct in_addr*) target_hostent_ptr-h_addr);
something like that
goto is good for "break"ing multiple loops at once
yeah
Perl has named loops for that, so you can "last OUTER"
yup
yes lets recap the goto usage now
and coming to a common fall-thru point in a func, too
one way or another, goto is useful and elegant when used appropriately. all these people that say gotos are evil are just spouting bs their prof told them.
well, the prof is well aware of the WORST - and it's hard to teach "better", "best" - so.. A flat denial is suitable in a lot of courses
correct
One can use many features well, but it doesn't mean folks should just jump into featureitis
maybe if folks where graded on use of goto, applicability and such, we could get past it. Not too likely in huge classes
I'd like to see a good alternative to goddamned setjmp/longjmp, myself. For some reason, that muck just grates me hard.
could somebody tell me what an "incomplete type" is supposed to be?
typedef struct foo_ Foo; // foo_ is incomplete
….
well, its just that i didnt define the stu
the struct
struct foo_ { … }; //completes foo_ and Foo
hi, anybody knows a good (and free) wavelet library in C?
fat-lobyte: is struct in_addr defined in a header that is include by your source file?
pizza_, damn you were right, that was ist…
pizza_, forgot to include "netdb.h"
fat-lobyte: that's the only thing type you dereferenced in that statement
s/thing //
yes, well ok. But still difficult to understand "incomplete struct" in my opinion
it's simple
struct foo_; // I am incomplete, until resolved Only use a ptr to me
it's a "forward"
PoppaVic, I think not. I would have expected something like "type or identifyer unknown"
no size or members or order is know, but you specify a symbol that you can use as a ptr until resolved.
knowN
so you are allowed to "declare" a struct before defining its appearance?
fat-lobyte: but you are allowed to use a pointer to an incomplete type
aha… interesting
fat-lobyte: but you cannot access the data underneath because the compiler doesn't know what is there
not "a struct" - just a referencial symbol to a later resolved size.
pizza_, ok, i think I understood now. Just didnt now that this works
but what is the sense of this?
im sure there is one
if you can't puzzle it out, then don't ever do it
as you see i didnt indend to
just forget to include a header
headers are lovely things.
compiler error messages could try and be a little more helpful
Agreed
if it told you it couldn't locate the definition of struct in_addr specifically you could figure out the problem in no time
ah well
could be worse, too - look at doze and message#
hah, true
speaking of that, i couldn't find any built-in err#-description functions for winsock error messages
somebody program with cobl ?
dude, I sympathize.. I'd never code for doze again.
wow, pizza_ i just put &start[i]
and i totaly understand why
you really do want to get punted out the door with crusehed nuts, don't you?
i prefer start+i, which is equivalent
but i'm glad you sorted it out, just telling you wouldn't have really helped
pizza_ i wasnt asking you to tell me ;0
*
What can you do?
i don't speak english,
pizza_ but now i have another problem, i think its segfaulting or something
i dont understand your message
hey, i found a good tutorial on pointers
you need to buy a C book. you need to understand types and pointers. i swear, it will make your life easier.
!k&r Mike_Oxbig
k&r is The C Programming Language, 2nd edition, by Kernighan and Ritchie, http://cm.bell-labs.com/cm/cs/cbook/ - be sure to see the errata as well, at http://cm.bell-labs.com/cm/cs/cbook/2ediffs.html
yea, ill get it from amazon
thanks
you're welcome, excellent decision
Any idea why this segfauls?: http://rafb.net/p/1uOei427.html
where does it fault?
PoppaVic, in which step?
oh, nm - because it's stupid
PoppaVic, stupid?
a ptr is not a string, strings are not ptrs or chars, and C doesn't automagically allocate for people
a ptr is not a string, strings are not ptrs or chars, and C doesn't automagically allocate for people
I thought sprintf would make a pointer point to a given string in memory.
Well, the start of it.
don't "think" like that - rtfm
:P
Thanks.
Unless otherwise documented, you just deal with ptrs-to-things, or things - and a str is a convention, not a "thing"
sprintf is the same as printf, except instead of writing to stdout it writes to a location in memory
Yeap.
pizza_, I understood that.
But in another sense
folks are welcome to add all the complexity they like ;-
Heh.
I think the worst time I had with strings was back in-the-day of rebuild cp/m to add a driver.. And learning that the "kernel" used $ as a terminator - while I was trying to learn C.
Well, now the result does show… plus an abort trap.
rafb
http://rafb.net/p/orA3Vd52.html
wrong keystroke, sorry
same shit, different place
T-T
strcat still isn't creating new space for you
and, every cstring is at LEAST ''
PoppaVic, ah, I forgotten that it actually modifies its operands.
yeppers
with C, everyone needs to remember "C is stupid and literal and you burn in hell for bad expectations"
"abc"
It's just a silly interface to an assembler
Heh.
a string-compare?
PoppaVic has invented the C language.
strcmp(left,right);
PoppaVic, no, no.
PoppaVic, something to group characters into a string.
alright, time to have a rest and watch fantomas
A simpler function/syntax thing.
I know "a""b" works.
ok, stop speaking of string like that.. You want a buffer or a dynamic string/buffer
I need a pointer to an array of characters whose values I set.
and I dunno where you got "abc"
That ends in ''.
adelfino, isnt this what i was asking pizza_ ?
Mike_Oxbig, don't know; didn't read your question.
so what? use a nice, comfortable-sized buffer. Sprintf into it, and strdup it to create a dynamic return that is disposable.
my friend is writing an irc bot using a seperate pthread for each connection. I told him not to, buut…
The thing could be solved with [0] = 'a' … [1] = 'b' [2] = ''
nope
PoppaVic, why not?
you'd STILL need a buffer to create-into.
Yeah, test[3].
stop being cheap
xlq:just help make sure he learns about what actually cause his problems, when they come around
But I only need two characters.
today
And user does not have access to it.
tomorrow it's 40 or 65 or 1024, etc
ashr, I am already
he says he wants to respond to messages at the same time as each other, because it'd be 'leet'
I said it wouldn't do anything at the same time
anybody have any experience etching their own surface mount PCBs at home?
if it's that simple, you'd not need to "create" it, you'd just code it all once and do strcmp
and it creates more problems than it's worth
if i listen on a raw socket, what packets do i receive? i dont seem to get tcp packets that get to my host…
edlikestoboogie, I've etched some pcbs before, but only one (very simple, 1 transistor) smt board
PoppaVic, I don't get why do you mention strcmp; I don't need a comparation.
PoppaVic, the "" thing was "to".
at home and commercial proto. It's not for the faint of heart and it's all EPA-screams and toxins
ahh
PoppaVic, that's why 'a' & 'b' % '' to "abc".
Sorry "ab".
And % is & :
:P
{ 'a', 'b', '' } ain't "abc"
Heh.
My keyboard is buggy (?)
So, there's no syntax to do that, right?
As in "a""b" is "ab".
who is building the [sub]string? And where?
PoppaVic, haven't done that yet; it's trivial; but wanted to see if there was bult-in solution.
have u guys ever tried using ur laser printer to print the pcb design and then fuse them onto a copper clad board with a clothing iron?? then etching.. and surface mouting?? my question is.. even if i make my board with hundreds of conductive leads, will my board come out with little to no
problems if i do the procedure correctly?
basicly what i'm trying to say is.. can i make homemade SMD circuit boards with moderate complexity at home?
ed, i used to do that
doe sit work mike?
3mm
and can i make moderately complex designs with it?
but this isnt the place
"you"
smt mounts are a bitch at home
do you know the place to ask such things?? another irc room?
nope
lol
how small you want the tracks?
in fact, better pm me
they don't need to be that small
it's not trivial. Someone is building something to use.. The preprocessor? compiler? executable?
ok
its gonna annoy the crap out of Z
PoppaVic, hmm, I'm not sure what your question is :/
asking for less than 15 mill traces will piss off the workd
you need to know who is building what, when and where is a bonus.
why would a multi-threaded IRC bot be bad?
cpp is one how; C is another how - and often where. Compiletime, runtime, etc
you need to cogitate it a little… C uses the programmer as a preprocessor
gbb, so may synchronization issues, because messages need to be passed between threads (between irc connections)
PoppaVic, I'll finish writing this and I'll show you have I want to do.
well, hopefully you will egt a lightbulb and it makes sense.
get
so?
"multi-threaded" always makes my guts churn.
as long as you use appropriate locking/synchronization mechanisms, you're fine
heh, why's that?
most of my work is multi-threaded
because of the way folks percieve it all. It's not what most folks believe and few docs can tell the diff.
multiprocessing made it worse
you mean multiple processors, right?
yeah
processors, processes, heavy, light, coprocess.. events, etc
What's up?
I need something like sleep(), but shorter
I suspect, gbb, we need a nice "language" for the collective concepts
usleep?
thanks, ashr
usleep or select
though i believe usleep is deprecated
use nanosleep instead
non-blocking sockets vs. select()?
nanosleep/usleep
non-blocking sockets work well *with* select…not vs
nb means "deal with an event", and select means "there is an event"
thanks all
ok
right
I sort of messed up
np, I have major issues thinking of "event programming"
heh
actually that's what I'm doing ATM
I suspect I need to abstract my head more
it's been a topic on & off for at least a few months.. Most "event-driven" stuff really demands an fd.. Not sure what the answer is
try to program a bunch of lemmings, and then let them do lemming stuff (using event drivern programming).
retro-man: well, a flood is a flood - what unit-size?
retro-man: that sounds pointless
retro-man: event-based programming for a lemmings game, I mean
I don not think simulating stuff this way is pointless.
since when you have you been opposed to doing pointless things
what is a "lemmings game"?
you raise an interesting point
try to program a bunch of lemmings, and then let them do lemming stuff (using event drivern programming). You have to bring them somewhere safe
retro-man: the idea of "simulate" is just part of it.
you are missing out if you don't know what lemmings is!
Sorry, wriong mess.
well, I know of and about lemmings - games I am low on.
The game with the little creatures walking straight, and wont stop untill they hit someting, and then turn. they even walk down a cliff.
You have to bring them save somewhere, … how, with gadgets and stuff.
pingus is a lemming based game I guess.
retro-man: you *guess*?
retro-man: well, sorta' - they are basically hamsters/gerbils - until they swarm and run over cliffs.
I am not sure….
retro-man: I'm positive
ok… :s ???
we could also debate "the fuzzy-wuzzy falacy", but why?
Hmmmm, lets ask Prof. Dijstra what to do.
*Prof Dijkstra
retro-man: every bit of coprocessing, threading, processing is about "this is a UNIT that does something that could affect the whole"
when, how and given-what is the iffy
A thread is like a program counter, right?
no
I mean a pointer pointing to some instruction.
retro-man: currently, there are at least 2 forms of thread (and I believe it is three)
hmmm, sounds interesting
retro-man: usually there's also a stack
a 'thread' is more like an entire Vcpu with interfacing possible between.
yeah, there are lots of things, but mainly, a PC (program counter).
retro-man: if you're just talking about a pthreads sort of thing
yeah
all the threads have to have on there own, they have to have on their own
retro-man: it gets worse… They may even have their own hw/cpu and need to cross-communicate
I heard of that, isn't that vector processing?
or they can be in the same space and need to share
share memory? share tics?
retro-man: I have no idea what new terms folks twist-up and misuse or coin. But, suffice to say the entire issue needs to get it's collective shit together.
what do I need to do with a variable, so the change of one side of fork() makes a difference to the other side?
you need to use a pipe or some other form of IPC for that.
ashr, sounds complicated, but thanks
vfork()?
yep.
xlq, that won't do, they need to work in parallel
or clone() or something?
a fork makes a 100% copy of the old available to the new. After that they go their separate ways.
which makes it harder to create fast processors, due to long main memory access times.
retro-man: well, making chips is only part of the issue; mem/busses have always been slower.
shared memory could work for you, though i'm not sure if you want to add that kind of complexity
I remember people talking about memory being as fast as CPU in the eighties.
retro-man: for a few years - maybe a decade - folks fealt that daughter-cards were the answer
daughter-cards?
I will wikipedi that
retro-man: complete cpu/ram - but SOMEONE, somewhere is still playing crossing-guard
gbb, i even did shared memory before, but of course by now it's all forgotten and i need to learn it anew )
I've never heard of any solution, in over 20 years, where the whole mess didn't require a Master that acted as at least an interface to the daughters/threads/processes/etc
http://rafb.net/p/wWmElC65.html
