Please any one suggest me some project ideas
Preetham Kumar Rai
i am a engineering student(FINAL SEMISTER) studying in visswaraiah
technological univervity in india. Can anybody suggest me some of the
revolutionary ideas/concepts that i can implement.
Regards
Preetham
preetham...@yahoo.com
Noah Roberts
> Hi...
> i am a engineering student(FINAL SEMISTER) studying in visswaraiah
> technological univervity in india. Can anybody suggest me some of the
> revolutionary ideas/concepts that i can implement.
Maybe you don't want a revolutionary idea. Improve on something if you
can't come up with your own innovation.
--
"I can't help it, the Dominating Trolls made me."
Bruce
preetham...@yahoo.com (Preetham Kumar Rai) wrote:
> i am a engineering student(FINAL SEMISTER) studying in visswaraiah
>technological univervity in india. Can anybody suggest me some of the
>revolutionary ideas/concepts that i can implement.
Yeah, I've got a revolutionary idea and I'm gonna post it on the internet
for you to implement. That's the ticket.
Here's a less than revolutionary idea. Use some idea generation techniques
to come up with your own ideas.
Noah Roberts
Dude, what an excellent idea! Too bad you just gave it away.
Roger Willcocks
"Arthur J. O'Dwyer" <a...@nospam.andrew.cmu.edu> wrote in message
news:Pine.LNX.4.58-035....@unix44.andrew.cmu.edu...
>
> On Fri, 29 Jan 2004, Preetham Kumar Rai wrote:
> >
> > Hi...
> > i am a engineering student(FINAL SEMISTER) studying in visswaraiah
> > technological univervity in india. Can anybody suggest me some of the
> > revolutionary ideas/concepts that i can implement.
>
> been kicking around in the back of my mind for some time now:
>
...
> 3. Savestruct, which I'm nominally still working on, but pretty
> much stalled. A program that can take a C or C++ 'struct' definition,
> possibly marked-up in some fashion, and produce functions to read
> and write objects and/or data structures in a portable and robust
> manner. Requires meta-knowledge of how serialization works, plus
> a decent parser and some creativity.
...
> -Arthur
Portable in the sense of source code portable or portable in the sense of
endian and alignment aware portable?
You might be interested in looking at
http://www.garret.ru/~knizhnik/cpp.html for a couple of takes at achieving
this in a source-code portable way - in particular POST (Persistent Object
STorage) which is close to what you describe, and GOODS (Generic
Object-Oriented Database System) which implements a client/server model.
From the POST readme file: "POST++ storage manager needs information about
persistent object classes to support garbage collection, relocation of
references while loading, and initialization of pointers to virtual tables.
Unfortunately C++ language provides no facilities to extract information
about class format at runtime. As far as I want to avoid use of some special
tools (preprocessors) or some "dirty trick" solutions (extracting
information about classes from debugging information), this information
should be provided to storage manager by programmer. Such class registration
can be done very easy using special macros provided by POST++."
It's freeware source code btw.
ps I have no connection with these products, I just find them interesting.
--
Roger
CBFalconer
>
> i am a engineering student(FINAL SEMISTER) studying in visswaraiah
> technological univervity in india. Can anybody suggest me some of the
> revolutionary ideas/concepts that i can implement.
A program to correct English spelling and capitalize 'I's.
--
Chuck F (cbfal...@yahoo.com) (cbfal...@worldnet.att.net)
Available for consulting/temporary embedded and systems.
<http://cbfalconer.home.att.net> USE worldnet address!
olej
Thomas Matthews
Any episode of Star Trek contains some ideas that are still
revolutionary. I still haven't figured out the transporter yet.
One of my ideas is to draw a holographic image in 3d space
and view it without any steam or smoke or any distortion.
Current systems require steam or smoke since light is not
seen unless it is reflect off of an object and back into
your eye.
--
Thomas Matthews
C++ newsgroup welcome message:
http://www.slack.net/~shiva/welcome.txt
C++ Faq: http://www.parashift.com/c++-faq-lite
C Faq: http://www.eskimo.com/~scs/c-faq/top.html
alt.comp.lang.learn.c-c++ faq:
http://www.raos.demon.uk/acllc-c++/faq.html
Other sites:
http://www.josuttis.com -- C++ STL Library book
Richard Heathfield
> Preetham Kumar Rai wrote:
>>
>> i am a engineering student(FINAL SEMISTER) studying in visswaraiah
>> technological univervity in india. Can anybody suggest me some of the
>> revolutionary ideas/concepts that i can implement.
>
> A program to correct English spelling and capitalize 'I's.
I lIked that Idea so much I wrote an ImplementatIon of It myself.
--
RIchard Heathfield : bIn...@eton.powernet.co.uk
"Usenet Is a strange place." - DennIs M RItchIe, 29 July 1999.
C FAQ: http://www.eskImo.com/~scs/C-faq/top.html
K&R answers, C books, etc: http://users.powernet.co.uk/eton
Programmer Dude
>> A program to correct English spelling and capitalize 'I's.
>
> I lIked that Idea so much I wrote an ImplementatIon of It myself.
Hmmmm.... me suspects you've haven't understood the spec. (-:
--
|_ CJSonnack <Ch...@Sonnack.com> _____________| How's my programming? |
|_ http://www.Sonnack.com/ ___________________| Call: 1-800-DEV-NULL |
|_____________________________________________|_______________________|
CBFalconer
> Preetham Kumar Rai wrote:
>
> > visswaraiah technological univervity in india. Can anybody
> > suggest me some of the revolutionary ideas/concepts that i
> > can implement.
>
> revolutionary. I still haven't figured out the transporter yet.
>
> One of my ideas is to draw a holographic image in 3d space
> and view it without any steam or smoke or any distortion.
> Current systems require steam or smoke since light is not
> seen unless it is reflect off of an object and back into
> your eye.
First implement virtual steam.
Edward G. Nilges
(1) Write a program that will read a Windows bitmap and create the
HTML for the next time some art major gives me a BMP Web page design
and I have to convert it to HTML dammit
(2) Write a program that will compare two or more source files
PROPERLY. Windiff, the Windows program that highlights differences in
a GUI is a joke, since it manages to preserve the illusion that a
program is a series of punched cards.
The program's output should be a minimal set of sensible instructions
for converting one file to another phrased in terms of syntax.
Therefore it should create the parse trees for the compared programs
and emit the changes by a tree comparision. For example, it might say
"change the damn identifier in the third statement of the procedure
named FooBar from a to b".
Edward G. Nilges
> CBFalconer wrote:
>
> > Preetham Kumar Rai wrote:
> >>
> >> i am a engineering student(FINAL SEMISTER) studying in visswaraiah
> >> technological univervity in india. Can anybody suggest me some of the
> >> revolutionary ideas/concepts that i can implement.
> >
> > A program to correct English spelling and capitalize 'I's.
>
> I lIked that Idea so much I wrote an ImplementatIon of It myself.
All you needed was my famous item parser to Isolate I's that are
surrounded by characters in the set "punctuation union blanks". But,
you'd rather be bloody right and now look at you.
Edward G. Nilges
> Preetham Kumar Rai wrote:
>
> > Hi...
> > i am a engineering student(FINAL SEMISTER) studying in visswaraiah
> > technological univervity in india. Can anybody suggest me some of the
> > revolutionary ideas/concepts that i can implement.
> > Regards
> > Preetham
> > preetham...@yahoo.com
>
> Any episode of Star Trek contains some ideas that are still
> revolutionary. I still haven't figured out the transporter yet.
>
> One of my ideas is to draw a holographic image in 3d space
> and view it without any steam or smoke or any distortion.
> Current systems require steam or smoke since light is not
> seen unless it is reflect off of an object and back into
> your eye.
A related idea would be developing the hardware and software to make
three dimensional objects from let us say non-toxic foam, cheaply.
Whereas in the United States we are surfeit with dreck (and some damn
fool is charging people money for sending their teddy bears and
business cards to the moon), perhaps this idea would have legs in
India where people aren't surfeit with dreck.
Old Enough to Know Better
"Preetham Kumar Rai" <preetham...@yahoo.com> wrote in message
news:709f593b.04012...@posting.google.com...
How about an AI that reads Usenet groups and automatically generates wise
crack replies to all posts?
Randy Howard
@yahoo.com says...
ROTFLMAO.
Mr. Nilges castigates someone for preferring to be right over using his
crap software. Well, who wouldn't?
--
Randy Howard
2reply remove FOOBAR
Randy Howard
> three dimensional objects from let us say non-toxic foam, cheaply.
It's called CNC. Look it up sometime.
Randy Howard
> PROPERLY.
Define properly. I happen to think that sdiff works quite well. Most
quality revision control packages provide comparison features, in
addition to many others.
> since it manages to preserve the illusion that a
> program is a series of punched cards.
Maybe it is just the user's delusions and antiquated beliefs that form
such an "illusion".
> The program's output should be a minimal set of sensible instructions
> for converting one file to another phrased in terms of syntax.
> Therefore it should create the parse trees for the compared programs
> and emit the changes by a tree comparision. For example, it might say
> "change the damn identifier in the third statement of the procedure
> named FooBar from a to b".
This syntax sounds more difficult than simply making edits by hand,
but why don't you publish an implementation of your intent so that
we may better understand your intent?
Elliot Marks
> You can see the icky mess I've made of this
> idea here:
>
> http://www.contrib.andrew.cmu.edu/~ajo/struct.html
I'd like to put some of my C code on my web site. How do you get
your C source on the web page in the proper format without doing
a lot of very tedious HTML coding?
EM
Richard Harter
wrote:
Do a global replace of < by <
Do a global replace of > by >
Put a <pre> tag before the code
Put a </pre> tag after the code
You can even create an edit macro to do this in one step. I wouldn't
bother, though.
The code is now ready for insertion into an HTML file.
Richard Harter, c...@tiac.net
http://home.tiac.net/~cri, http://www.varinoma.com
A man with money is always charming - pomposity is just
an eccentricity, forgivable in the rich.
Richard Heathfield
> On Sat, 31 Jan 2004 15:06:11 GMT, Elliot Marks <ema...@email.net>
> wrote:
>
>>Arthur J. O'Dwyer wrote:
>>> You can see the icky mess I've made of this
>>> idea here:
>>>
>>> http://www.contrib.andrew.cmu.edu/~ajo/struct.html
>>
>>I'd like to put some of my C code on my web site. How do you get
>>your C source on the web page in the proper format without doing
>>a lot of very tedious HTML coding?
>
> Do a global replace of < by <
> Do a global replace of > by >
> Put a <pre> tag before the code
> Put a </pre> tag after the code
BEFORE you do the replace of < by < and > by > you should do a global
replace of & with &
Also, don't forget to replace " with "
--
Richard Heathfield : bin...@eton.powernet.co.uk
"Usenet is a strange place." - Dennis M Ritchie, 29 July 1999.
C FAQ: http://www.eskimo.com/~scs/C-faq/top.html
CBFalconer
> Richard Harter wrote:
> > Elliot Marks <ema...@email.net> wrote:
> >
... snip ...
> >>
> >> I'd like to put some of my C code on my web site. How do you
> >> get your C source on the web page in the proper format without
> >> doing a lot of very tedious HTML coding?
> >
> > Do a global replace of < by <
> > Do a global replace of > by >
> > Put a <pre> tag before the code
> > Put a </pre> tag after the code
>
> BEFORE you do the replace of < by < and > by > you should
> do a global replace of & with &
>
> Also, don't forget to replace " with "
You can tell how much HTML I do by this, but are those (; and ")
necessary when within the <pre> </pre> tags?
Elliot Marks
> Richard Harter wrote:
>
>
>>On Sat, 31 Jan 2004 15:06:11 GMT, Elliot Marks <ema...@email.net>
>>wrote:
>>
>>
>>>Arthur J. O'Dwyer wrote:
>>>
>>>>You can see the icky mess I've made of this
>>>>idea here:
>>>>
>>>> http://www.contrib.andrew.cmu.edu/~ajo/struct.html
>>>
>>>I'd like to put some of my C code on my web site. How do you get
>>>your C source on the web page in the proper format without doing
>>>a lot of very tedious HTML coding?
>>
>>Do a global replace of < by <
>>Do a global replace of > by >
>>Put a <pre> tag before the code
>>Put a </pre> tag after the code
>
>
> BEFORE you do the replace of < by < and > by > you should do a global
> replace of & with &
>
> Also, don't forget to replace " with "
>
Why don't I see any of this when I look at the source view of
Arthur's code in my browser (Mozilla)?
Elliot
Elliot Marks
> Do a global replace of < by <
> Do a global replace of > by >
> Put a <pre> tag before the code
> Put a </pre> tag after the code
>
> You can even create an edit macro to do this in one step. I wouldn't
> bother, though.
>
> The code is now ready for insertion into an HTML file.
It works!! Thanks to Richard H. and also to Richard H.
Richard Heathfield
> Richard Heathfield wrote:
>> Richard Harter wrote:
>> > Elliot Marks <ema...@email.net> wrote:
>> >
> ... snip ...
>> >>
>> >> I'd like to put some of my C code on my web site. How do you
>> >> get your C source on the web page in the proper format without
>> >> doing a lot of very tedious HTML coding?
>> >
>> > Do a global replace of < by <
>> > Do a global replace of > by >
>> > Put a <pre> tag before the code
>> > Put a </pre> tag after the code
>>
>> BEFORE you do the replace of < by < and > by > you should
>> do a global replace of & with &
>>
>> Also, don't forget to replace " with "
>
> You can tell how much HTML I do by this, but are those (; and ")
> necessary when within the <pre> </pre> tags?
AFAIK using & and " leave you a hostage to the browser (i.e. I think they're
allowed to interpret it as you intended, but are not required to). BICBW.
Bill Godfrey
> > Also, don't forget to replace " with "
>
> You can tell how much HTML I do by this, but are those (; and ")
> necessary when within the <pre> </pre> tags?
" is useful when you want a " character in a tag attribute.
<img src="..." alt="He says "Hello".">
It's not wrong to use " in a PRE element block, but unneccessary. Just
like using a instead of 'a'.
Bill,
James Dow Allen
> > ... Also, don't forget to replace " with "
>
> You can tell how much HTML I do by this, but are those (; and ")
> necessary when within the <pre> </pre> tags?
I'm certainly not an HTML expert either, but I have lots of code
samples just stuck between <pre> and </pre> at
http://freepages.genealogy.rootsweb.com/~jamesdow/Tech/index.htm
The only trouble I've noticed is in lines with both < and >
(e.g. when including a library header file). Other than that
the code displays correctly and can be copied and used as is.
(I avoid the library header include problem by not including any!)
It seems to work with 3 or 4 different browsers, but perhaps there
is a problem. Please tell me if so.
James
(From address above doesn't work.)
Richard Harter
<dont...@address.co.uk.invalid> wrote:
>Richard Harter wrote:
[snip]
>> Do a global replace of < by <
>> Do a global replace of > by >
>> Put a <pre> tag before the code
>> Put a </pre> tag after the code
>
>BEFORE you do the replace of < by < and > by > you should do a global
>replace of & with &
>
>Also, don't forget to replace " with "
Thank you for the emendations.
Richard Harter
There is a real issue here. Diff programs typically treat the
structure of the files being compared as consisting of a sequence of
atomic elements, typically lines of text. I use the term "atomic
elements" because they have no internal structure; the only visible
relationship between elements is that either they are identical or
they are not. Let me call this style of diff a flat diff.
Flat diffs actually work fairly well for code and simple text based
databases, the reason being that in code each line usually is a
syntactical unit. That is, the structure assumed by a flat diff
roughly matches the intrinsic structure.
Flat diffs usually do not work well for files that have a more complex
structure, e.g., heirarchically structured data. Examples include
differencing directories and HTML files. Similarly, applying flat
diffs to word processing documents is an unhappy experience.
The problem with using anything besides flat diffs is that the
difference algorithm needs a lot more information about the structure
and parsing of the data in the files being difference.
Edward G. Nilges
The syntax isn't that difficult. If the language is of typical
structure you can construct two parse trees using recursive descent
hand coded from the BNF of the language.
You then walk the two trees noticing what's missing and present. This
algorithm would be for me the tricky part. I have searched for, but
not found, an algorithm for "the minimal difference of two trees".
Edward G. Nilges
> In article <f5dda427.04013...@posting.google.com>, spinoza1111
> @yahoo.com says...
> > A related idea would be developing the hardware and software to make
> > three dimensional objects from let us say non-toxic foam, cheaply.
>
> It's called CNC. Look it up sometime.
You know, part of my negative reaction to your posts may be the fact
that like many tech men, you don't know how to write prose with grace
notes. But note that your prose style leads you, against what I am
sure are the better angels of your nature, to the "straightforward"
style which waxes brutal as you drink too much coffee.
Notice in a related post you speculated that building a diff based on
syntax might be difficult. Notice that I did not say "it's called
recursive descent, asshole. Look it up sometime".
That's because none of us can know everything, not even me.
What I want isn't CNC, which has been around for a while. Check out
David Noble's book Forces of Production for a left-wing history of CNC
which shows how it created "scrap at high speeds" when initially
installed, because management wanted machinists to just plug in paper
tapes prepared by dweebs in offices.
The lads in the shop protested that they had knowledge of actual
stresses and temperatures and were told under the Taft-Hartley act to
shuddup. The lads then stood back and watched as the programs failed.
The lads then negotiated the ability to modify the tapes on the job,
and CNC was a success.
No, what I was thinking of is a three-dimensional printer that would
allow me to fabricate useful plastic crap. It's related to CNC, but as
a consumer product it probably wouldn't form metal, although a
high-end product might do so, allowing, for example, auto parts shops
to replace special purpose CNC devices with a commodity Windows
system.
Of course, this would put the lads completely out of work. Management,
however, would accomplish its original goal as described by Noble,
which is labor without the laborer.
Randy Howard
@yahoo.com says...
> Randy Howard <randy....@FOOmegapathdslBAR.net> wrote in message news:<MPG.1a8514c66...@news.megapathdsl.net>...
> > In article <f5dda427.04013...@posting.google.com>, spinoza1111
> > @yahoo.com says...
> > > for converting one file to another phrased in terms of syntax.
> > > Therefore it should create the parse trees for the compared programs
> > > and emit the changes by a tree comparision. For example, it might say
> > > "change the damn identifier in the third statement of the procedure
> > > named FooBar from a to b".
> >
> > This syntax sounds more difficult than simply making edits by hand,
> > but why don't you publish an implementation of your intent so that
> > we may better understand your intent?
>
> The syntax isn't that difficult. If the language is of typical
> structure you can construct two parse trees using recursive descent
> hand coded from the BNF of the language.
You're not listening. Your admittedly incomplete description above seems
to imply that the syntax for scripting a change would be as verbose as the
part in quotations above from your original message. Namely:
> > > "change the damn identifier in the third statement of the procedure
> > > named FooBar from a to b".
I definitely think that is an overly longwinded syntax for general purpose
usage. Perhaps you meant something different, but your example wording
leaves that impression.
--
Randy Howard
2reply remove FOOBAR
"I admit that it is probably too much work ... to wade through
my prose to get to anything useful." -- Edward G. Nilges
Randy Howard
> > In article <f5dda427.04013...@posting.google.com>, spinoza1111
> > @yahoo.com says...
> > > A related idea would be developing the hardware and software to make
> > > three dimensional objects from let us say non-toxic foam, cheaply.
> >
> > It's called CNC. Look it up sometime.
>
> You know, part of my negative reaction to your posts may be the fact
> that like many tech men,
What does the phrase "tech men" mean in your world? Not a sociologist?
If so, I'm happy to be described as such. Otherwise, please be more
specific. Your prose is too abrupt to convey your meaning.
> you don't know how to write prose with grace notes.
I was not trying to. It was a one-line response to what can only be
considered a rhetorical question, given the ubiquitous presence of CNC
machinery in modern machine shops.
> But note that your prose style leads you, against what I am sure are
> the better angels of your nature, to the "straightforward" style which
> waxes brutal as you drink too much coffee.
I do not drink coffee. But please continue in your overgeneralizations,
after all, it is amusing to watch you make a fundament of yourself
repeatedly.
> Notice in a related post you speculated that building a diff based on
> syntax might be difficult. Notice that I did not say "it's called
> recursive descent, asshole. Look it up sometime".
No, you provided sample syntax that looked like it would make awk look
highly readable in comparison. BTW, I'm not the one who must continue
to add expletives to my responses when other arguments are insufficient.
What is so all-fired difficult about sed for such a purpose?
> That's because none of us can know everything, not even me.
>
> What I want isn't CNC, which has been around for a while.
CNC does EXACTLY what you want. I happen to be an investor in
a precision machining company, and I guarantee you it can achieve
your needs. More specifically, you might want to look up the
definition of "CNC".
> David Noble's book Forces of Production for a left-wing history of CNC
I am not interested in a left-wing history of CNC, but thanks for
sharing.
> No, what I was thinking of is a three-dimensional printer that would
> allow me to fabricate useful plastic crap. It's related to CNC, but as
> a consumer product it probably wouldn't form metal,
CNC equipment is not restricted to metal. It can cut anything solid
but softer than the cutting tools installed, period. Bars of soap could
be machined to .001" tolerances if anyone cared to do so. There are
also several "consumer grade" or hobbyist products that can do
such things on a smaller scale than say a Haas VF3 w/5-axis control.
You can even buy "do it yourself" books on how to make your own CNC
system for the garage workshop crowd. I believe that there are
even embedded controllers that are already programmed to read MNG
or gerber codes for plug and play integration.
Again, you haven't asked for anything new, you just took a direct,
to the point response, complained about it's lack of verbosity, then
turned out to prove it correct. What a fun ride that was.
--
Randy Howard
2reply remove FOOBAR
Willem
) The syntax isn't that difficult. If the language is of typical
) structure you can construct two parse trees using recursive descent
) hand coded from the BNF of the language.
)
) You then walk the two trees noticing what's missing and present. This
) algorithm would be for me the tricky part. I have searched for, but
) not found, an algorithm for "the minimal difference of two trees".
That would be very hard.
Probably harder than finding the minimal difference of two files.
Besides which, you have the downside that you have to have a separate
algorithm for each and every language.
And you lose all the formatting, obviously, which is a bad thing.
Tell me, what would be the benefit of munging up a piece of source code
using the grammar of the language to find the minimal difference ?
SaSW, Willem
--
Disclaimer: I am in no way responsible for any of the statements
made in the above text. For all I know I might be
drugged or something..
No I'm not paranoid. You all think I'm paranoid, don't you !
#EOT
Joe "Nuke Me Xemu" Foster
> In article <f5dda427.04020...@posting.google.com>, spinoza1111
> @yahoo.com says...
> > > > "change the damn identifier in the third statement of the procedure
> > > > named FooBar from a to b".
>
> I definitely think that is an overly longwinded syntax for general purpose
> usage. Perhaps you meant something different, but your example wording
> leaves that impression.
Perhaps he'll come up with his own COBOL. (COmmon Bogosity Oriented Language)
--
Joe Foster <mailto:jlfoster%40znet.com> Sacrament R2-45 <http://www.xenu.net/>
WARNING: I cannot be held responsible for the above They're coming to
because my cats have apparently learned to type. take me away, ha ha!
Richard Harter
wrote:
>You then walk the two trees noticing what's missing and present. This
>algorithm would be for me the tricky part. I have searched for, but
>not found, an algorithm for "the minimal difference of two trees".
Try searching on TreeDiff.
Programmer Dude
> Note the header that says "Content-Type: text/plain". This is what
> tells your browser to read and display the file as if it were plain
> old text, not HTML -- and that means you can use < and & with
> impunity.
If MSIE finds certain tags (<html>, <head>, <title> or <body>) in the
first 256 bytes, it'll render the text as HTML no matter WHAT the
Content-Type says.
--
|_ CJSonnack <Ch...@Sonnack.com> _____________| How's my programming? |
|_ http://www.Sonnack.com/ ___________________| Call: 1-800-DEV-NULL |
|_____________________________________________|_______________________|
Programmer Dude
> It seems to work with 3 or 4 different browsers, but perhaps there
> is a problem. Please tell me if so.
From the HTML 4.1 spec...
5.3.2 Character entity references
[...]
Four character entity references deserve special mention since they
are frequently used to escape special characters:
"<" represents the < sign.
">" represents the > sign.
"&" represents the & sign.
"" represents the " mark.
Authors wishing to put the "<" character in text should use "<"
(ASCII decimal 60) to avoid possible confusion with the beginning
of a tag (start tag open delimiter). Similarly, authors should use
">" (ASCII decimal 62) in text instead of ">" to avoid problems
with older user agents that incorrectly perceive this as the end
of a tag (tag close delimiter) when it appears in quoted attribute
values.
Authors should use "&" (ASCII decimal 38) instead of "&" to
avoid confusion with the beginning of a character reference (entity
reference open delimiter). Authors should also use "&" in
attribute values since character references are allowed within CDATA
[p.50] attribute values.
Some authors use the character entity reference """ to encode
instances of the double quote mark (") since that character may be
used to delimit attribute values.
Roger Willcocks
"Willem" <wil...@stack.nl> wrote in message
news:slrnc1ta3j....@toad.stack.nl...
> Edward wrote:
> ) The syntax isn't that difficult. If the language is of typical
> ) structure you can construct two parse trees using recursive descent
> ) hand coded from the BNF of the language.
> )
> ) You then walk the two trees noticing what's missing and present. This
> ) algorithm would be for me the tricky part. I have searched for, but
> ) not found, an algorithm for "the minimal difference of two trees".
>
> That would be very hard.
> Probably harder than finding the minimal difference of two files.
> Besides which, you have the downside that you have to have a separate
> algorithm for each and every language.
> And you lose all the formatting, obviously, which is a bad thing.
>
> Tell me, what would be the benefit of munging up a piece of source code
> using the grammar of the language to find the minimal difference ?
>
>
> SaSW, Willem
A very similar problem came up a couple of months ago in the TUHS (The Unix
Heritage Society) mailing lists where there was a discussion of how to find
common code between System III Unix and System V. You can probably guess
why.
The problem is that variable names and bracketing conventions may have
changed, even though the underlying algorithm remains the same. For instance
(a rather contrived example):
int function(struct foo arg, int operation)
{
if (operation == 1)
arg.bar = 6;
return 0;
}
should be flagged as being similar to (and worth checking out further):
int boogle(struct baz var, int opcode)
{
if (opcode == 1) {
var.bar = 6;
}
return 0;
}
Pretty clearly this is best tackled by creating and comparing parse trees.
See http://minnie.tuhsorg/Programs/Ctcompare/ for one (token-based, not
tree-based) solution. "The purpose of this set of programs is to allow you
to compare two sets of C code trees on a token basis, rather than on a line
by line basis. The programs help to identify similarities between snippets
of code in both trees."
--
Roger
Roger Willcocks
...
>
> First implement virtual steam.
>
Vapourware...
--
Roger
James D. Veale
knowledge of a language. The "atomic elements" can be words where words
are separated not only by blanks but by common programming language
delimiters.
This is what members of the "Complite File Comparison Family" do.
This works fairly well out of the box. But customizations in the form
of keyword and phrase processing, DNA expansion, and more are available
to expand the useful domain to various application specific text files
such as log files, assembler listings, link maps etc.
Time-limited demonstration versions are available on my web page at:
http://world.std.com/~jdveale
Jim Veale
c...@tiac.net (Richard Harter) writes:
...
bob7094
"Edward G. Nilges" <spino...@yahoo.com> wrote in message
news:f5dda427.04020...@posting.google.com...
Try http://www.stratasys.com or do a google search along the lines of "three
dimensional rapid prototype"
This kind of technology is available, but it doesn't look like it will find
its way into many hobby modeler's basements.
-- Bob Mills
CBFalconer
> "Arthur J. O'Dwyer" wrote:
>
> > Note the header that says "Content-Type: text/plain". This is
> > what tells your browser to read and display the file as if it
> > were plain old text, not HTML -- and that means you can use <
> > and & with impunity.
>
> If MSIE finds certain tags (<html>, <head>, <title> or <body>)
> in the first 256 bytes, it'll render the text as HTML no matter
> WHAT the Content-Type says.
<sarcasm> Are you suggesting that people should cater to the
existence of buggy software that ignores standards? </sarcasm>
Ben Pfaff
> Programmer Dude wrote:
>> If MSIE finds certain tags (<html>, <head>, <title> or <body>)
>> in the first 256 bytes, it'll render the text as HTML no matter
>> WHAT the Content-Type says.
>
> <sarcasm> Are you suggesting that people should cater to the
> existence of buggy software that ignores standards? </sarcasm>
I don't think <sarcasm> is an HTML tag.
--
Ben Pfaff <b...@cs.stanford.edu>
Winter 2004 CS140 teaching assistant
CS140 webpage: http://www.stanford.edu/class/cs140
Richard Harter
<jdv...@TheWorld.com> wrote:
> There is an intermediate approach, that doesn't require a detailed
> knowledge of a language. The "atomic elements" can be words where words
> are separated not only by blanks but by common programming language
> delimiters.
>
> This is what members of the "Complite File Comparison Family" do.
> This works fairly well out of the box. But customizations in the form
> of keyword and phrase processing, DNA expansion, and more are available
> to expand the useful domain to various application specific text files
> such as log files, assembler listings, link maps etc.
>
> Time-limited demonstration versions are available on my web page at:
>
> http://world.std.com/~jdveale
>
> Jim Veale
AFAICT what you're doing is still a flat diff but with an expanded
set of set of delimiters. This is a trivial (ignoring display issues)
modification for ordinary diff routines. The interesting feature is
that you are building diffs upward, i.e., there is a heirarchy in
which a change at a lower level implies a change of the containing
level.
Programmer Dude
> No, what I was thinking of is a three-dimensional printer that would
> allow me to fabricate useful plastic crap.
Go Ogle for "stereolithography" and learn something. It's been
around for quite a while. Our CAD/CAM guys have been using it
for nearly a decade.
Programmer Dude
>> If MSIE finds certain tags (<html>, <head>, <title> or <body>)
>> in the first 256 bytes, it'll render the text as HTML no matter
>> WHAT the Content-Type says.
>
> <sarcasm> Are you suggesting that people should cater to the
> existence of buggy software that ignores standards? </sarcasm>
<seriously> No, but if one is going to be a web author, I do think
it's smart to understand the, um, idiosyncrasies of a very common
browswer! </seriously>
&grin;
MSCHAEF.COM
CBFalconer <cbfal...@worldnet.att.net> wrote:
...
><sarcasm> Are you suggesting that people should cater to the
>existence of buggy software that ignores standards? </sarcasm>
I think that the most important standard in web browsers, unfortunately,
is "whatever works in MSIE", and that everybody else has to accomodate
that. Such is the power of a 90% market share...
This guy, who works on Safari, spoke to this issue in the 1/18-1/20
timeframe (about halfway down the page on my machine)
http://weblogs.mozillazine.org/hyatt/
-Mike
--
http://www.mschaef.com
Willem
) Randy Howard <randy....@FOOmegapathdslBAR.net> wrote in message news:<MPG.1a8514fa9...@news.megapathdsl.net>...
)>
)> It's called CNC. Look it up sometime.
)
) ...
)
) syntax might be difficult. Notice that I did not say "it's called
) recursive descent, asshole. Look it up sometime".
^^^^^^^
See the difference ?
MSCHAEF.COM
Edward G. Nilges <spino...@yahoo.com> wrote:
...
>(2) Write a program that will compare two or more source files
>a GUI is a joke,
I wholeheartedly disagree. Windiff is one of the most useful development
tools that I regularly.
>since it manages to preserve the illusion that a
>program is a series of punched cards.
Which is largely a correct "illusion" unless you're continually
reformatting your program.
>The program's output should be a minimal set of sensible instructions
>for converting one file to another phrased in terms of syntax.
>Therefore it should create the parse trees for the compared programs
>and emit the changes by a tree comparision.
Georgia Tech did something like this to detect cheaters in
programming assignments. It's called cheatfinder, and works by
doing a more intelligent compare of two programs' structure. It
flags programs that match more than a pre-determined level.
It's not quite a diff, but it's the same general idea.
> For example, it might say
>"change the damn identifier in the third statement of the procedure
>named FooBar from a to b".
If I had a diff program that produced output like that, it'd be borderline
useless. Windiff (and command line diff) are so useful primarily because
they (show the code) that's changed, rather than producing an
elaborate description of the change. I'd rather interpret the results of a
diff output record myself.
-Mike
--
http://www.mschaef.com
Edward G. Nilges
> In article <f5dda427.04020...@posting.google.com>, spinoza1111
> @yahoo.com says...
> > Randy Howard <randy....@FOOmegapathdslBAR.net> wrote in message news:<MPG.1a8514fa9...@news.megapathdsl.net>...
> > > In article <f5dda427.04013...@posting.google.com>, spinoza1111
> > > @yahoo.com says...
> > > > A related idea would be developing the hardware and software to make
> > > > three dimensional objects from let us say non-toxic foam, cheaply.
> > >
> > > It's called CNC. Look it up sometime.
> >
> > You know, part of my negative reaction to your posts may be the fact
> > that like many tech men,
>
> What does the phrase "tech men" mean in your world? Not a sociologist?
> If so, I'm happy to be described as such. Otherwise, please be more
> specific. Your prose is too abrupt to convey your meaning.
What's a "sociologist"? My degree was in philosophy and math. I would
never "do" sociology as practiced in an American university for the
very good reason that American sociology has been terrorized by
corporate power into silence.
Is a "sociologist" some female who speaks to our isolation and
despair? Then OK, I'm a sociologist except for the part about being
female.
>
> > you don't know how to write prose with grace notes.
>
> I was not trying to. It was a one-line response to what can only be
> considered a rhetorical question, given the ubiquitous presence of CNC
> machinery in modern machine shops.
>
OK, perhaps you are brusque and then find yourself getting flamed.
> > But note that your prose style leads you, against what I am sure are
> > the better angels of your nature, to the "straightforward" style which
> > waxes brutal as you drink too much coffee.
>
> I do not drink coffee. But please continue in your overgeneralizations,
> after all, it is amusing to watch you make a fundament of yourself
> repeatedly.
>
> > Notice in a related post you speculated that building a diff based on
> > syntax might be difficult. Notice that I did not say "it's called
> > recursive descent, asshole. Look it up sometime".
>
> No, you provided sample syntax that looked like it would make awk look
> highly readable in comparison. BTW, I'm not the one who must continue
> to add expletives to my responses when other arguments are insufficient.
>
> What is so all-fired difficult about sed for such a purpose?
sed? awk? Are these tools from the Dark Ages now what we must use?
Even computists are now realizing that unix servers are no longer
cost-effective, and that Linux servers are stolen code.
>
> > That's because none of us can know everything, not even me.
> >
> > What I want isn't CNC, which has been around for a while.
>
> CNC does EXACTLY what you want. I happen to be an investor in
> a precision machining company, and I guarantee you it can achieve
> your needs. More specifically, you might want to look up the
> definition of "CNC".
>
I see, an investor. As such, I don't think you have the sort of
insight that a machinist must have to actually work with the
technology. In fact and in some ways, you may be part of the problem.
You see, thanks to the misapplication of CNC, good machinists were
forced out of work when the investors were deluded into thinking that
CNC programs could be written solely at the front office.
It's my view that "investors" have too much power and too much control
of technology.
Even from your self-interest as an investor, please read Forces of
Production by David Noble. That's because he shows how you will lose
money if the company in which you invest thinks that advanced CNC will
allow it to deskill machinists.
> > David Noble's book Forces of Production for a left-wing history of CNC
>
> I am not interested in a left-wing history of CNC, but thanks for
> sharing.
What's interesting about this comment is its closed-mindedness.
Randy, for twenty years, investors and venture types have acted as if
they know technology and the result is that the only growth area,
today, is...computer games.
The result is that the premiere system for manufacturing automation in
the enterprise is a GERMAN written product, written in fact by
programmers who take four week vacations. SAP is of higher quality
because it was written in a society in which actual programmers, in
addition to investors, have a say in the direction of work.
David Noble was the director of the Smithsonian under Carter and later
a professor of technical history. He did his homework, bigtime, in
Forces of Production and what's especially interesting is that there
are NO "right wing" histories of CNC technology.
That's because its installation in American factories in the 1950s was
a complete joke.
Small companies were forced to install overcomplex systems that could
bend metal, not for consumer needs but to Air Force specifications
because the defense industry and anti-labor was driving the
technology.
For his contribution, Noble appears to have been forced out of the
Smithsonian (which is now itself a joke; a recent exhibit of "American
culture" celebrated the Dukes of Hazzard).
>
> > No, what I was thinking of is a three-dimensional printer that would
> > allow me to fabricate useful plastic crap. It's related to CNC, but as
> > a consumer product it probably wouldn't form metal,
>
> CNC equipment is not restricted to metal. It can cut anything solid
> but softer than the cutting tools installed, period. Bars of soap could
> be machined to .001" tolerances if anyone cared to do so. There are
> also several "consumer grade" or hobbyist products that can do
> such things on a smaller scale than say a Haas VF3 w/5-axis control.
> You can even buy "do it yourself" books on how to make your own CNC
> system for the garage workshop crowd. I believe that there are
> even embedded controllers that are already programmed to read MNG
> or gerber codes for plug and play integration.
OK, CNC can cut soap. Your point being?
What you don't say, because you don't care, is that American business
used CNC rhetorically in the 1950s, to make the false claim that in
the future, nobody would actually have to work in a factory to make
things, and, we'd all live like the Jetsons.
The reality is that any machine, properly used and in a human fashion,
needs tending, and the moral reality is that its designers should plan
for this...even when the "investors" don't care.
Of course, manufacturing was never "automated" in America, or Britain.
Instead, a generation of decent working machinists was made sick at
heart by the way in which their skills were belittled and they stood
by while scrap (and the Ford Pinto) was created at high speed.
Then, as we know, manufacturing disappeared, but it went to China as
is coding, today.
The result is that many Americans are living on home equity and credit
cards because, literally, they don't produce anything any more. The
result is that we're perceived as fat and stupid.
"Investors" can't invest if the machinists are no longer trained, but
thrown into jail for drug use instead.
Paul E. Black
> Randy Howard <randy....@FOOmegapathdslBAR.net> wrote in message news:<MPG.1a8514fa9...@news.megapathdsl.net>...
> No, what I was thinking of is a three-dimensional printer that would
> allow me to fabricate useful plastic crap.
I'm trying to think of a reason to have Toy Builders
http://www.toybuilders.com/
make me something. They have half a dozen different machines
and technologies, all for hire. Maybe for my birthday ...
-paul-
--
Paul E. Black (p.b...@acm.org)
Programmer Dude
> I'm trying to think of a reason to have Toy Builders
> http://www.toybuilders.com/
> make me something. They have half a dozen different machines
> and technologies, all for hire. Maybe for my birthday ...
Oooooh, nice!!
When I worked for Division Engineering and had access to the SLA
(stereolithography) department, I wanted a Starship Enterprise,
but never did find any decent CAD files for it (this was prior to
"The Web"...bet I could Go Ogle for them now!).
OT: What I'd like now is the diagram Paramount made showing the
Enterprise (D) overlaid on a map of the Paramount lot. They did
that to give writers a sense of the scale of the thing...
Randy Howard
> > If so, I'm happy to be described as such. Otherwise, please be more
> > specific. Your prose is too abrupt to convey your meaning.
>
> What's a "sociologist"? My degree was in philosophy and math. I would
> never "do" sociology as practiced in an American university for the
> very good reason that American sociology has been terrorized by
> corporate power into silence.
You did not answer my question. Try again.
> > > you don't know how to write prose with grace notes.
> >
> > I was not trying to. It was a one-line response to what can only be
> > considered a rhetorical question, given the ubiquitous presence of CNC
> > machinery in modern machine shops.
> >
> OK, perhaps you are brusque and then find yourself getting flamed.
I don't feel flamed unless the author actually makes a valid point.
I've yet to have that feeling from you.
> > What is so all-fired difficult about sed for such a purpose?
>
> sed? awk? Are these tools from the Dark Ages now what we must use?
They are tools that solve a particular class of problems in ways
that are well understood by those capable of comprehending them.
If you need to reinvent the wheel, or more likely don't even
realize that the metaphorical wheel already exists, I can see
why you might be surprised to find out that others do not have
the same limited scope of imagination.
> Even computists are now realizing that unix servers are no longer
> cost-effective, and that Linux servers are stolen code.
Troll. Repeating the same lie over and over and over and over .....
Boring.
> > > What I want isn't CNC, which has been around for a while.
> >
> > CNC does EXACTLY what you want. I happen to be an investor in
> > a precision machining company, and I guarantee you it can achieve
> > your needs. More specifically, you might want to look up the
> > definition of "CNC".
> >
> I see, an investor. As such, I don't think you have the sort of
> insight that a machinist must have to actually work with the
> technology.
No, but they do. They assure me that your need to do some 3d
shaping as originally described in this thread is trivially
easy to accomplish on even an entry level piece of CNC equipment.
Going off on tangents about marketing stories from 50 years ago
does not change that. This whole diatribe of yours is a result
of you not liking hearing that the first time. Too bad. If you
disagree, prove me wrong.
> You see, thanks to the misapplication of CNC, good machinists were
> forced out of work when the investors were deluded into thinking that
> CNC programs could be written solely at the front office.
Wrong. CNC equipment cannot be programmed correctly without an
understanding that the "obvious, direct" method to a programmer doesn't
work with physical materials of varying properties. All you wind up
doing is damaging equipment and wasting stock material without an
in depth understanding of proper machining methods. That may have
been a problem at some point in the past, but high quality machine
shops don't suffer them today.
> It's my view that "investors" have too much power and too much control
> of technology.
Here you go again, off on another rant. I never set foot in the building,
and I have no control of what parts they make. About the only "control"
I exercise is the ability to call them up and say, "hey, can you make one
of these....", which happens rarely anyway. Do you ever grind your axes
privately? Just curious.
> Even from your self-interest as an investor,
Self-interest? How do you figure? A machine shop that specializes in
fixtures for a very narrow technical field could hardly hope to gain
financial profit by me sending you to look for CNC work. I imagine
you couldn't afford even a single prototype of their work, much less
make a volume commitment for same. You're reaching, even more than
usual.
> That's because he shows how you will lose
> money if the company in which you invest thinks that advanced CNC will
> allow it to deskill machinists.
I invested in the company solely BECAUSE it is staffed with skilled
machinists and knows what is doing, and is highly sought after in
the field for that very reason. Stop making crap up just to keep
your keyboard dust-free.
> > > David Noble's book Forces of Production for a left-wing history of CNC
> >
> > I am not interested in a left-wing history of CNC, but thanks for
> > sharing.
>
> What's interesting about this comment is its closed-mindedness.
Why would *anyone* intentionally seek out a politically biased history
on a machining technology? It's ludicrous.
> Randy, for twenty years, investors and venture types have acted as if
> they know technology
I claim to know very little about CNC machining myself, but I do know
enough about it (as does almost any sentient creature in a modern
country) to realize it can solve the original problem you posed. If
you are too ignorant to realize that, I can't make you, so I won't
bother further.
> there are NO "right wing" histories of CNC technology.
There doesn't seem to be a need for biased histories of CNC, in either
direction. The fact there is a biased one just means that a surfeit
exists.
> That's because its installation in American factories in the 1950s was
> a complete joke.
Next time I need information on the 50s, you'll be the first person
I ask. In the mean time, when we are discussing current issues, perhaps
we can stay within at least a few decades of them in our discussions?
> For his contribution, Noble appears to have been forced out of the
> Smithsonian
If he writes politically biased histories, that is hardly surprising.
> OK, CNC can cut soap. Your point being?
Since you are too dense to see it, even when spelled out, why bother
explaining, AGAIN, that it is more than capable to solve the problem,
which AS YOU DESCRIBED IT, was precisely a CNC system. Whether or
not you put that name on it, the function you described was, by
definition, a CNC device. WAKE UP.
> What you don't say, because you don't care, is that American business
> used CNC rhetorically in the 1950s
True, or untrue, it has nothing to do with the problem you described.
There would therefore be no logical reason for me to say anything
about it one way or the other here. Same goes for you btw.
> The reality is that any machine, properly used and in a human fashion,
> needs tending
Probably so, but as usual, completely besides the point.
> The result is that we're perceived as fat and stupid.
I see you are speaking for yourself again. Good for you, come right
out and share these things, you are brave to do so in a public forum.
> "Investors" can't invest if the machinists are no longer trained, but
> thrown into jail for drug use instead.
ROTFLMAO. Are you implying that I am using drugs because I invested
in a company that uses CNC equipment? Your logic processing capability
is almost as well-tuned as the athletic abilities of PeeWee Herman,
but not quite.
James D. Veale
>... This is a trivial (ignoring display issues)
>modification for ordinary diff routines.
Turns out it's not quite as simple as it might seem, especially
when you take into account the esthetics of what humans like to see.
Consider the following files:
----- FILE ONE ----- ----- FILE TWO -----
if (someFunction) if (a gt 0
thendo and b gt 0)
something; thendo
end something_else;
end
if (a gt 0 --------------------
and b gt 0)
thendo
something_else;
end
--------------------
The 'words' in file one consist of: "if" "(" "someFunction" ")" ...
The 'words' in file two consist of: "if" "(" "a" "gt" "0" ...
A straightforward application of a diff algorithm can match up
the "if" and the "(" from file one and then
skip to the "a" "gt" "0" from file two.
This is perfectly valid, but it's not what people would like to see.
You'd rather have it match up the entire "if (a gt 0)" line not just
a part of it.
The "Prefer Lines" option in both the heuristic 'Comp' utilities
and the rigorous 'Tig' utilities help to avoid this problem.
Also the "Contiguous Match Factor" in the 'Tig' utilities helps to avoid
similarly annoying artifacts.
Jim Veale
Nils Petter Vaskinn
> Richard Heathfield <dont...@address.co.uk.invalid> wrote in message
> news:<bvee3k$jem$3...@titan.btinternet.com>...
>> CBFalconer wrote:
>> > A program to correct English spelling and capitalize 'I's.
>>
>> I lIked that Idea so much I wrote an ImplementatIon of It myself.
>
> All you needed was my famous item parser to Isolate I's that are
> surrounded by characters in the set "punctuation union blanks". But,
> you'd rather be bloody right and now look at you.
perl -pi -e 's/(\W)i(\W)/$1I$2/g'
That will do the job. If that can get famous anything can!
--
NPV
"the large print giveth, and the small print taketh away"
Tom Waits - Step right up
Richard Harter
[snip]
Agreed. It turns out that this is a problem that I (and I suppose
anyone who has implemented a reasonable diff algorithm) am familiar
with. The essence of the matter is that when there are duplicates
(lines or tokens as the case may be) there has to be a resolution
method for which atom pairs with which.
For code it suffices to match all of the unique items first and then
match duplicates within the gaps between unique items. That is a
heuristic but it is a very effective one generally matching natural
expectations.
When the atoms are tokens rather than lines there are many more
duplicates so it is probably the case that special heuristics are
needed.
John W. Krahn
>
> On Fri, 30 Jan 2004 18:15:41 -0800, Edward G. Nilges wrote:
>
> > Richard Heathfield <dont...@address.co.uk.invalid> wrote in message
> > news:<bvee3k$jem$3...@titan.btinternet.com>...
> >> CBFalconer wrote:
> >> > A program to correct English spelling and capitalize 'I's.
> >>
> >> I lIked that Idea so much I wrote an ImplementatIon of It myself.
> >
> > All you needed was my famous item parser to Isolate I's that are
> > surrounded by characters in the set "punctuation union blanks". But,
> > you'd rather be bloody right and now look at you.
>
> perl -pi -e 's/(\W)i(\W)/$1I$2/g'
>
> That will do the job. If that can get famous anything can!
That won't work if the 'i' is at the beginning or end of the line.
perl -pi -e's/(^|[[:space:][:punct:]])i([[:space:][:punct:]]|$)/$1I$2/g'
John
--
use Perl;
program
fulfillment
Edward G. Nilges
> On Fri, 30 Jan 2004 18:15:41 -0800, Edward G. Nilges wrote:
>
> > Richard Heathfield <dont...@address.co.uk.invalid> wrote in message
> > news:<bvee3k$jem$3...@titan.btinternet.com>...
> >> CBFalconer wrote:
> >> > A program to correct English spelling and capitalize 'I's.
> >>
> >> I lIked that Idea so much I wrote an ImplementatIon of It myself.
> >
> > All you needed was my famous item parser to Isolate I's that are
> > surrounded by characters in the set "punctuation union blanks". But,
> > you'd rather be bloody right and now look at you.
>
> perl -pi -e 's/(\W)i(\W)/$1I$2/g'
>
> That will do the job. If that can get famous anything can!
I'll have to trust you on this one. Last time I used perl was several
years ago.
But why does the code look like an explosion in a gnome factory?
Seriously, regular expressions have to be tested and documented
carefully in order to be maintainable in real code. Also, there is no
consistent standard for how they are executed. They are executed
consistently in perl, but the regular expression can't be re-used in
another processor.
Edward G. Nilges
> Edward wrote:
> ) The syntax isn't that difficult. If the language is of typical
> ) structure you can construct two parse trees using recursive descent
> ) hand coded from the BNF of the language.
> )
> ) You then walk the two trees noticing what's missing and present. This
> ) algorithm would be for me the tricky part. I have searched for, but
> ) not found, an algorithm for "the minimal difference of two trees".
>
> That would be very hard.
> Probably harder than finding the minimal difference of two files.
> Besides which, you have the downside that you have to have a separate
> algorithm for each and every language.
No, you don't.
You have a separate PARSER for every language...which produces the
same TREE.
Producing the PARSER is a known problem whose solution is yacc, bison,
or manual recursive descent.
But once you have an algorithm for "the minimal difference of two
trees", you're done.
Consider the tree a(b(cd)e) where the root is a, it has two children b
and e, and b has in turn two children c and d.
Compare a(be). The minimal difference would be cd. The "instruction"
to get from tree 1 to tree 2 would be: "remove all the children of
node b". The "instruction" to get from tree 2 to tree 1 would be "add
the children cd to node b".
OK, consider the two famous code fragments
/* Nilges' famous blunder */
for (intIndex1 = 1; intIndex1 < strlen(strGoop); intIndex++) malarkey;
/* A simple fix to Nilges' famous blunder made by a techie */
int intLength = strlen(strGoop);
for (intIndex1 = 1; intIndex1 < intLength; intIndex++) malarkey;
If we ignore the comment, fragment A has the structure
forStmt(forKeyword, forHeader(forInitializer(assignment(lValue,
assignOp, expression)),
forTest(expression(...variable...expression(functionCall...),
semicolon...
B has the structure assignment...forStmt(forKeyword,
forHeader(forInitializer(assignment(lValue, assignOp, expression)),
forTest(expression(...variable...expression(functionCall...),
semicolon...
What we want is a sequence of instructions to the maintenance
programmer:
(1) Add "intLength = strlen(strGoop)" as the first statement
(2) In the for statement, change "strLen(strGoop)" to "intLength".
A straightforward tree match would produce the skeleton, which can be
nicely decorated with qualifiers identifying syntax classes as above!
> And you lose all the formatting, obviously, which is a bad thing.
Yes, bad. The answer is to have the lexical analyzer, which would seem
to be a necessary prepass to the parsing (and produced automatically
with lex or by hand), RETAIN position information. In fact, I do so
religiously in my stuff because I like the way I can pinpoint errors.
In the quickBasicEngine parser for the Apress book, the lexical
analyzer produces a collection of .Net objects, qbToken, which not
only expose token values and types but also start index and
length...giving you end index "for free". I find that the object
oriented approach in fact makes it economical to have a rich token
structure which retains formatting.
>
> Tell me, what would be the benefit of munging up a piece of source code
> using the grammar of the language to find the minimal difference ?
>
See above. The parser writer has a clue about the real needs of the
programming proletariat and codes in solidarity with the just struggle
of world peoples (sorry, couldn't resist mein Herr). Therefore, he
captures not only token type and length but also position. Therefore
he is able to rebuild the code with formatting intact.
The only problem would be when the change is a reformatting such that
the tokens are the same but have different starting positions. But, if
white space is a token, then the reformatting becomes the instruction
to add, delete or change white space.
However, your point is well taken. A former colleague did some
graduate work on how to represent formatting as syntax and when the
change modifies format, the result could be ugly.
>
> SaSW, Willem
Edward G. Nilges
Bad day at the gnome factory, huh? I rest my case. Regular expressions
are too unreadable to represent a long term solution. They are a HACK.
>
>
>
> John
Edward G. Nilges
> Edward wrote:
> ) Randy Howard <randy....@FOOmegapathdslBAR.net> wrote in message news:<MPG.1a8514fa9...@news.megapathdsl.net>...
> )>
> )> It's called CNC. Look it up sometime.
> )
> ) ...
> )
> ) syntax might be difficult. Notice that I did not say "it's called
> ) recursive descent, asshole. Look it up sometime".
> ^^^^^^^
>
> See the difference ?
OK, I misread your intent based on some nasty things ya said earlier.
Look, an overly brusque tone can imply the automatic insertion of
"asshole" and that is one reason I am more verbose than the average
poster. Sure, I get mad but for the most part I avoid the brusqueness
that triggers flame wars.
You called my German phony when I say that no use of language can be
phony in and of itself (read Kant: read Gadamer). This caused
unnecessary bad feeling what with a police agent, Richard Heathfield,
loose about the shop doing his best for pay to troll, disrupt, and
create dissension.
>
>
> SaSW, Willem
Edward G. Nilges
> In article <f5dda427.04013...@posting.google.com>,
> Edward G. Nilges <spino...@yahoo.com> wrote:
> ...
> >(2) Write a program that will compare two or more source files
> >PROPERLY. Windiff, the Windows program that highlights differences in
> >a GUI is a joke,
>
> I wholeheartedly disagree. Windiff is one of the most useful development
> tools that I regularly.
>
> >since it manages to preserve the illusion that a
> >program is a series of punched cards.
>
> Which is largely a correct "illusion" unless you're continually
> reformatting your program.
>
> >The program's output should be a minimal set of sensible instructions
> >for converting one file to another phrased in terms of syntax.
> >Therefore it should create the parse trees for the compared programs
> >and emit the changes by a tree comparision.
>
> Georgia Tech did something like this to detect cheaters in
> programming assignments. It's called cheatfinder, and works by
> doing a more intelligent compare of two programs' structure. It
> flags programs that match more than a pre-determined level.
I find the idea of such a program reprehensible, since it might be
used in a false accusation, and if the instructor does his job
(emphasizing good style), his students' programming style might well
converge leading to false positives.
In fact, the whole thing sounds like something from Georgia, in which
the Bush-inspired assumption is that (1) students will have no joy in
learning and (2) the whole purpose of education is constant testing
with the constant expectation of cheating.
Anhedonic teachers, who struggled through graduate school hating the
work, are the ones who expect cheating and don't even "see" the
student who doesn't cheat because he wants not to get rich but to
LEARN.
Many programming shops in fact demand that the programmers have a
uniform style, such that you cannot tell who's the author. In school,
the students might well be taught to conform to this expectation and
wouldn't this give cheatfinder false positives?
>
> It's not quite a diff, but it's the same general idea.
>
> > For example, it might say
> >"change the damn identifier in the third statement of the procedure
> >named FooBar from a to b".
>
> If I had a diff program that produced output like that, it'd be borderline
> useless. Windiff (and command line diff) are so useful primarily because
> they (show the code) that's changed, rather than producing an
> elaborate description of the change. I'd rather interpret the results of a
> diff output record myself.
>
I agree that it's often faster just to use windiff. The problem is in
code (for example, in 250 stored procedures I recently had to convert)
where the change is so systematic that windiff just gives up, and
shows a mass insert and delete.
The tree approach could be insensitive to consistent identifier
changes and might be useful in communicating changes to offshore
programmers by email.
> -Mike
Richard Heathfield
> /* Nilges' famous blunder */
> for (intIndex1 = 1; intIndex1 < strlen(strGoop); intIndex++) malarkey;
>
> /* A simple fix to Nilges' famous blunder made by a techie */
> int intLength = strlen(strGoop);
> for (intIndex1 = 1; intIndex1 < intLength; intIndex++) malarkey;
The "fix" is broken, since it uses the wrong type for the intLength object.
This is better:
size_t intLength = strlen(strGoop);
for (intIndex1 = 1; intIndex1 < intLength; intIndex++) malarkey;
It's still not very good, because the names are crap, but it's better than
it was.
<snip>
> What we want is a sequence of instructions to the maintenance
> programmer:
>
> (1) Add "intLength = strlen(strGoop)" as the first statement
> (2) In the for statement, change "strLen(strGoop)" to "intLength".
(3) scrap the lot, and write a solution that actually works.
--
Richard Heathfield : bin...@eton.powernet.co.uk
"Usenet is a strange place." - Dennis M Ritchie, 29 July 1999.
C FAQ: http://www.eskimo.com/~scs/C-faq/top.html
K&R answers, C books, etc: http://users.powernet.co.uk/eton
Richard Heathfield
> You called my German phony when I say that no use of language can be
> phony in and of itself (read Kant: read Gadamer). This caused
> unnecessary bad feeling what with a police agent, Richard Heathfield,
> loose about the shop doing his best for pay to troll, disrupt, and
> create dissension.
Mr Nilges's claim that I am a police agent is mistaken.
Mr Nilges's claim that I am paid to troll is mistaken.
Mr Nilges's claim that I seek to troll is mistaken.
Mr Nilges's claim that I am paid to disrupt is mistaken.
Mr Nilges's claim that I seek to disrupt is mistaken.
Mr Nilges's claim that I am paid to create dissension is mistaken.
Mr Nilges's claim that I seek to create dissension is mistaken.
In fact, he is 100% wrong.
Programmer Dude
> You have a separate PARSER for every language...which produces the
> same TREE.
Just a thought here....
In some of the more event-driven languages, the program structure
is more a field of small bushes than a tree. Given that these
small bushes may be functionally identical, but not ordered the
same nor named the same, a tree-based comparitor is going to suffer.
If, for example, one form has...
Private Sub Command1_Click()
...at the top of the source, while a revision has...
Private Sub b_OK_Click()
...at the middle (and assume the variables have changed and the
source has been re-ordered while retaining function), it's going
to be very tough for the comparitor to realize they are, in fact,
the same function.
None of which makes the idea a bad one. It's just not going to
be the End All Be All of program diff.
BTW:
I'm quite pleased with Compare It!, a shareware program I've found.
It works nicely with another product (also from Grig Software),
Synch It! (directory tree comparitor), and now features color syntax
highlighting of source code. It's not perfect, but very, very nice.
http://www.grigsoft.com/wincmp.htm
http://www.grigsoft.com/winsin.htm
Programmer Dude
> Regular expressions are too unreadable to represent a long term
> solution. They are a HACK.
Well, perhaps a bit. Most coders who do implement them in production
code tend to document them pretty well. They are certainly terse
and arcane, but surely no worse so that advanced mathematics, eh?
I'll say this. I use them regularly--daily even--and in large enough
numbers that I wouldn't *want* a verbose syntax.
I'll say this, too. REs are just one of those things I'd expect any
programmer to learn, just as I'd expect a mathematician to know all
the ins and outs of upper level calculus.
After all, they come from the not-a-telescope part of the art. (-:
Willem
) Look, an overly brusque tone can imply the automatic insertion of
) "asshole" and that is one reason I am more verbose than the average
) poster. Sure, I get mad but for the most part I avoid the brusqueness
) that triggers flame wars.
There is no such thing as 'tone' in a purely text message.
) You called my German phony when I say that no use of language can be
) phony in and of itself (read Kant: read Gadamer).
Well then, I'd have to say you didn't actually *use* the language, you just
stole the odd tidbit from it in an otherwise english sentence, for whatever
reason. And furthermore, these tidbits were riddled with errors.
Noah Roberts
>
> size_t intLength = strlen(strGoop);
> for (intIndex1 = 1; intIndex1 < intLength; intIndex++) malarkey;
>
> It's still not very good, because the names are crap, but it's better than
> it was.
Wouldn't you also then need intIndex to be size_t? size_t is usually
unsigned int, correct? Meaning that if you happened to have a string
who's length > 2^(sizeof(int)-1)-1 your loop would never end.
--
"I can't help it, the Dominating Trolls made me."
Richard Heathfield
> Richard Heathfield wrote:
>
>>
>> size_t intLength = strlen(strGoop);
>> for (intIndex1 = 1; intIndex1 < intLength; intIndex++) malarkey;
>>
>> It's still not very good, because the names are crap, but it's better
>> than it was.
>
> Wouldn't you also then need intIndex to be size_t?
Yes, of course.
> size_t is usually unsigned int, correct?
Er, it /must/ be an unsigned integral type of at least 15 (in C90) or 16 (in
C99) bits. Nowadays, unsigned long is common, and I reckon unsigned long
long can't be far round the corner.
> Meaning that if you happened to have a string
> who's length > 2^(sizeof(int)-1)-1 your loop would never end.
My loop? MY loop? My dear chap, I wouldn't be seen /dead/ using that loop.
It's a Nilges loop. You can tell from the rather pathetic attempt at
Hungarian Notation (and, by implication, the incorrect types).
James D. Veale
Because it compares on a word-by-word basis, it can deal with changes
on every line. In addition the keyword processing can be used
to equivalence changes in variable names as well.
If you'd like to email me that code, I'd be happy to try it.
Or you could download the demo from the website and try it yourself.
http://world.std.com/~jdveale
Jim Veale
spino...@yahoo.com (Edward G. Nilges) writes:
>.....
Randy Howard
> "asshole"
What kind of person would automatically insert such a word into a
reply based upon an attempt, which is obviously flawed, at "reading
between the lines" in a text message?
> and that is one reason I am more verbose than the average poster.
Understatement of the epoch.
> Sure, I get mad but for the most part I avoid the brusqueness
> that triggers flame wars.
You *are* the trigger for flame wars. Which is of course your
motive.
> You called my German phony when I say that no use of language can be
> phony in and of itself
OK, you are a phony programmer, that makes phony claims about CS topics,
and posts phony German in an attempt to disguise your phony attempts to
make yourself into a phony intellectual. Capiche?
Randy Howard
> > programming assignments. It's called cheatfinder, and works by
> > doing a more intelligent compare of two programs' structure. It
> > flags programs that match more than a pre-determined level.
>
> I find the idea of such a program reprehensible, since it might be
> used in a false accusation, and if the instructor does his job
> (emphasizing good style), his students' programming style might well
> converge leading to false positives.
And it would be quite easy for an instructor to interview the students
identified by this during office hours and immediately establish
which, if any, of them knew the material and which did not. It's a
filter, not a jury. Also, "hello world" assignments from a first
semester course probably all do look alike, but the more advanced
assignments later in a degree program are unlikely to pass well
designed tests for such commonality. Even if they do, the interview
process would be sufficient to demonstrate the requisite knowledge
of the assignment. It would certainly be more than sufficient to
demonstrate for example that Mr. Nilges cannot read and understand
C code, despite claiming such knowledge on his resume.
> In fact, the whole thing sounds like something from Georgia
He has no progressed to ad hominem attacks against entire states.
Illinois allows him to live there, without exile, which is far
worse of a statement about Illinois than anything one might
dream up to say about Georgia.
> the Bush-inspired assumption is that (1) students will have no joy in
> learning and (2) the whole purpose of education is constant testing
> with the constant expectation of cheating.
Given the abilities of Mr. Nilges demonstrated previously in this
newsgroup, it is completely understandable that he should have a
strong negative reaction towards testing for competence.
> Many programming shops in fact demand that the programmers have a
> uniform style, such that you cannot tell who's the author.
Irrelevant. They don't have to worry about cheating, because such
business situations do not usually have more than one programmer
perform the same work.
> I agree that it's often faster just to use windiff. The problem is in
> code (for example, in 250 stored procedures I recently had to convert)
> where the change is so systematic that windiff just gives up, and
> shows a mass insert and delete.
A pragmatic person might suspect that Mr. Nilges simply is incapable
of using the tool properly, like so many languages he has trouble
with.
--
Randy Howard
2reply remove FOOBAR
"I admit that it is probably too much work ... to wade through
my prose to get to anything useful." -- Edward G. Nilges
CBFalconer
> Richard Heathfield wrote:
>
> > size_t intLength = strlen(strGoop);
> > for (intIndex1 = 1; intIndex1 < intLength; intIndex++) malarkey;
> >
> > It's still not very good, because the names are crap, but it's
> > better than it was.
>
> Wouldn't you also then need intIndex to be size_t? size_t is usually
> unsigned int, correct? Meaning that if you happened to have a string
> who's length > 2^(sizeof(int)-1)-1 your loop would never end.
Er, I think that snippet generates an infinite loop for any
strGoop longer than 0 chars.
--
Chuck F (cbfal...@yahoo.com) (cbfal...@worldnet.att.net)
Available for consulting/temporary embedded and systems.
<http://cbfalconer.home.att.net> USE worldnet address!
Richard Heathfield
> Noah Roberts wrote:
>> Richard Heathfield wrote:
>>
>> > size_t intLength = strlen(strGoop);
>> > for (intIndex1 = 1; intIndex1 < intLength; intIndex++) malarkey;
>> >
>> > It's still not very good, because the names are crap, but it's
>> > better than it was.
>>
>> Wouldn't you also then need intIndex to be size_t? size_t is usually
>> unsigned int, correct? Meaning that if you happened to have a string
>> who's length > 2^(sizeof(int)-1)-1 your loop would never end.
>
> Er, I think that snippet generates an infinite loop for any
> strGoop longer than 0 chars.
Oh, so it does! Well spotted!
Just when you think Mr Nilges's C can't get any worse, it suddenly does.
John W. Krahn
>
> "John W. Krahn" <kra...@acm.org> wrote in message news:<40234DD8...@acm.org>...
> >
> >
> > perl -pi -e's/(^|[[:space:][:punct:]])i([[:space:][:punct:]]|$)/$1I$2/g'
>
> Bad day at the gnome factory, huh?
No, I don't use Gnome, I use Icewm.
> I rest my case.
Which case is that?
> Regular expressions are too unreadable
Well, since you can't read C or BASIC I can understand that.
> to represent a long term solution. They are a HACK.
John - Just another regular expression hacker,
Randy Howard
says...
> Just when you think Mr Nilges's C can't get any worse, it suddenly does.
I think it's intentional. Nobody could get it *that* wrong, *that*
consistently by random chance.
Edward G. Nilges
> CBFalconer wrote:
>
> > Noah Roberts wrote:
> >> Richard Heathfield wrote:
> >>
> >> > size_t intLength = strlen(strGoop);
> >> > for (intIndex1 = 1; intIndex1 < intLength; intIndex++) malarkey;
> >> >
> >> > It's still not very good, because the names are crap, but it's
> >> > better than it was.
> >>
> >> Wouldn't you also then need intIndex to be size_t? size_t is usually
> >> unsigned int, correct? Meaning that if you happened to have a string
> >> who's length > 2^(sizeof(int)-1)-1 your loop would never end.
> >
> > Er, I think that snippet generates an infinite loop for any
> > strGoop longer than 0 chars.
>
> Oh, so it does! Well spotted!
>
> Just when you think Mr Nilges's C can't get any worse, it suddenly does.
OK, this is an example of why I believe the troll at the bottom of
this pond to be a police agent, Richard Heathfield.
FYI, the discussion is with Willem and it is meant to show how a
syntax-driven product can be, perhaps, more effective than windiff.
Therefore, I posted two code snippets as examples of code.
In fact, the very fact that they contain errors is realistic due to
the 80% failure rate of enterprise systems, and the fact that nasty
people, who seek out other's errors while making their own, create a
climate of intimidation which results in more...errors.
However, the code snippets were meant only as examples, and the point
(which due to a lack of computing and general culture the police agent
missed) was that both have a syntax representation which I sketched
using parentheses to represent trees adequately in ascii text.
But because it appears that Mr. Heathfield is a paid employee of an
undisclosed third party, he has trolled a productive discussion.
His behavior has caused Chris Sonnack to make a fool out of himself
because the latter has sent foolish libel to my publisher. This
especially upsets me because unlike Randy Howard, Richard Heathfield,
and Richard Harter, Chris is a working programmer and as such is my
intended audience.
However, Messrs. Howard, Heathfield and Harter have displayed, almost
as in a laboratory, how police agents, noncoding managers, and
noncoding investors divide working programmers so as to retain control
of the product.
Following Mr. Heathfield's lead, Randy Howard has posted repeated
abuse based on misaprehended tertiary sources for a month. Following
Mr. Heathfield's lead, Mr. Harter has pompously posted out of date
references on the philosophy of mathematics.
I have no way under usenet bylaws of preventing nonprogrammers
including editors of C books from abusing this newsgroup by absenting
themselves. However, I can point out that my egregious misreading of
Chris' code was caused by stress, and by the fact that I pulled an
all-nighter...debugging a 10K compiler for subset quick basic written
in VB.Net.
This compiler has been tech reviewed by several authorities, and they
will be only amused by libelous email from Chris, and because Chris is
a working programmer, this makes me very sad indeed. They will at best
only agree that I am rusty on C. I'm rusty on C because it is an out
of date language.
PEOPLE NEED TO USE THIS NEWSGROUP IN THE ORIGINAL SPIRIT OF USENET.
THIS IS ONE OF FREEDOM, AND COMPASSION.
AN ERROR IS AN OPPORTUNITY TO LEARN.
RICH RETIREES AND POLICE AGENTS SHOULD DEPART.
Randy Howard
> as in a laboratory, how police agents, noncoding managers, and
> noncoding investors divide working programmers so as to retain control
> of the product.
Presumably, I am the "investor" above. I'm not quite sure why you
are so anti-investment, without it, there would not be much business,
or many jobs. But, since you are usually unemployed, such economic
trifles may not matter (or be comprehensible) to you. Regardless, I am
not "noncoding". After seeing your attempts at coding, I dare say you
are the closest thing to a "noncoding" sociologist I have ever
encountered, bar none. Even so, I've never met a programmer as poorly
grounded in the basics as you seem to be.
I'll probably write more code "for fun" this year than you will write
for the rest of your life. More importantly, mine will actually work.
After doing so, I'll also be able to recognize the C code of others,
even if I haven't slept in a while beforehand.
> Following Mr. Heathfield's lead, Randy Howard has posted repeated
> abuse based on misaprehended tertiary sources for a month.
Ahh, back to "tertiary sources", you must have found that misplaced
sticky note after all. I suppose that you consider a direct examination
of your own posts to this newsgroup as tertiary? Perhaps you are simply
not capable of understanding it, because I have told you repeatedly that
I used your own posts, first hand, to arrive at my conclusions.
Why is that so hard for you to understand? Perhaps if you wrote it down
on a sheet of paper, say 25 times in a row, you would be able to remember.
Tie a string around your finger, or whatever works for you.
Furthermore, it's pretty hypocritical to call me abusive, after the crap
you've heaped on people for years here, not to mention the unbelievable
capacity that you have for public self-abuse. As someone else said
recently, it is you that have painted the big bullseye target on your
chest for all to see.
> I have no way under usenet bylaws of preventing nonprogrammers
The only nonprogrammer in evidence here has the initials EGN. (I
apologize to anyone else unfortunate enough to share those initials
with the Spewer of Nilgewater, that is indeed unfortunate)
> However, I can point out that my egregious misreading of
> Chris' code was caused by stress, and by the fact that I pulled an
> all-nighter...
You had two years to figure it out. Also, most of us have pulled more
than a few "all-nighters" without being unable to recognize source code
when we see it the next morning.
> debugging a 10K compiler for subset quick basic written in VB.Net.
Debugging your code would tend to drive one off the deep end. Upon
further reflection, I believe you. In fact, I think the damage may
have been permanent. (Aside: Does it really take 10K LOC to write
such drivel?)
> This compiler has been tech reviewed by several authorities
Are they willing to be named publicly?
I'm willing to bet right now that the code and text published in
"your" book will be riddled with errors. Your poor co-author should
have his name removed before it is too late to avoid being associated
with you.
> They will at best only agree that I am rusty on C.
You can't claim to be rusty on a topic you know nothing about. You
claiming that you are "rusty on C" is like me claiming that I am rusty
on open-heart surgery or the reactor procedures for a Trident submarine.
> I'm rusty on C because it is an out of date language.
You are an out of date sociologist pretending to be a programmer, and
therein lies the whole problem.
> PEOPLE NEED TO USE THIS NEWSGROUP IN THE ORIGINAL SPIRIT OF USENET.
We do, in every thread in which you do not participate.
> THIS IS ONE OF FREEDOM, AND COMPASSION.
Compassion does not include tolerance for intentional trolling,
repeated lies and almost criminal incompetence. You are free to
make a fool out of yourself, as you are obviously aware, but you
are not ENTITLED to a compassionate reaction to your lies and
false accusations.
*****
When will you have the HONOR to apologize to Chris for falsely
accusing him in a poorly veiled attempt to cover up your own
incompetence?
*****
If you want compassion, hang out in alt.flame for a while, I'm sure
they'd just love you in there.
> AN ERROR IS AN OPPORTUNITY TO LEARN.
That is very true. It is stunning that you should be aware of it, as
you have never exhibited such behavior here.
> RICH RETIREES AND POLICE AGENTS SHOULD DEPART.
This is so tiring. Anyone that isn't as pathetic as you must leave.
Well, if that was carried to its logical conclusion, you would have
all of usenet to yourself.
If you weren't incompetent, after as much time as you have been in
this industry, you would be retired too. If you're jealous, that's
understandable, but you have no one to blame but yourself.
BTW, I continue to program for my personal enjoyment, as well as
occasional projects for people I have worked with/for in the past,
so don't try and pretend like you hold the "currency high ground"
here, or over anyone else for that matter.
I don't have to dig up some distant reference to the 1970s to
describe projects I have worked on, or where the code is in current,
active use as you do Mr. Nilges.
Richard Heathfield
> Richard Heathfield <dont...@address.co.uk.invalid> wrote in message
> news:<c025u3$31g$2...@titan.btinternet.com>...
>>
>> Just when you think Mr Nilges's C can't get any worse, it suddenly does.
>
> OK, this is an example of why I believe the troll at the bottom of
> this pond to be a police agent, Richard Heathfield.
Mr Nilges's belief is mistaken on three counts (one of which is that this is
not a pond).
> FYI, the discussion is with Willem
Private discussions belong in email. Usenet is for public discussions.
> and it is meant to show how a
> syntax-driven product can be, perhaps, more effective than windiff.
> Therefore, I posted two code snippets as examples of code.
>
> In fact, the very fact that they contain errors is realistic
It certainly provides good evidence that you wrote them.
> due to
> the 80% failure rate of enterprise systems, and the fact that nasty
> people, who seek out other's errors while making their own, create a
> climate of intimidation which results in more...errors.
Would this be the climate of intimidation in which these nasty people
threaten lawsuits against people who notice, and point out, the errors that
these nasty people make? Well, those of us who are not easily intimidated
don't worry too much about silly little people making silly little threats
over silly little things. And those nasty people who insist on writing 80%
of their code incorrectly (perhaps as some kind of political statement) are
just being silly. Let's not confuse them with real programmers, who have
rather better than a 20% success rate.
> However, the code snippets were meant only as examples, and the point
> (which due to a lack of computing and general culture the police agent
> missed) was that both have a syntax representation which I sketched
> using parentheses to represent trees adequately in ascii text.
>
> But because it appears that Mr. Heathfield is a paid employee of an
> undisclosed third party, he has trolled a productive discussion.
The premise, the step of "reasoning", and the conclusion are all incorrect.
> His behavior has caused Chris Sonnack to make a fool out of himself
> because the latter has sent foolish libel to my publisher.
Mr Sonnack is not a fool, and would not foolishly expose himself to a libel
case.
> This
> especially upsets me because unlike Randy Howard, Richard Heathfield,
> and Richard Harter, Chris is a working programmer and as such is my
> intended audience.
>
> However, Messrs. Howard, Heathfield and Harter have displayed, almost
> as in a laboratory, how police agents, noncoding managers, and
> noncoding investors divide working programmers so as to retain control
> of the product.
I find those two paragraphs fascinating, albeit rather off-topic. On what
information are they based? Are working programmers forbidden from being
investors? Are working programmers forbidden from being responsible for
other working programmers? Are working programmers even forbidden from
being police agents? (Of course, I am not a police agent, but Mr Nilges's
deduction appears to be based on the assumption that I am.)
> Following Mr. Heathfield's lead, Randy Howard has posted repeated
> abuse based on misaprehended tertiary sources for a month. Following
> Mr. Heathfield's lead, Mr. Harter has pompously posted out of date
> references on the philosophy of mathematics.
Mr Howard's criticism of Mr Nilges appears to be based on Mr Howard's
reading of articles posted by Mr Nilges. These are not "tertiary sources"
but "primary sources". I am no mathematician (and, as Pascal wrote to
Fermat, "that is a great defect"), so I make no comment on Mr Harter's
contributions to this newsgroup, one way or the other.
> I have no way under usenet bylaws of preventing nonprogrammers
> including editors of C books from abusing this newsgroup by absenting
> themselves.
Usenet doesn't have any byelaws. It has only conventions, which Mr Nilges
seems to go out of his way /not/ to observe. I'm not sure who he means by
"nonprogrammers including editors of C books" but he clearly can't mean me,
since I am not a non-programmer.
> However, I can point out that my egregious misreading of
> Chris' code was caused by stress,
Mr Nilges may find his life less stressful if he uses his time more
effectively.
> and by the fact that I pulled an
> all-nighter...debugging a 10K compiler for subset quick basic written
> in VB.Net.
If Mr Nilges didn't spend so much time writing long, off-topic essays for
posting to this newsgroup, he might have more time for debugging. And I
think it's fair to say that he obviously /needs/ more time for debugging.
After all, according to him, 80% of his code doesn't work.
> This compiler has been tech reviewed by several authorities, and they
> will be only amused by libelous email from Chris, and because Chris is
> a working programmer, this makes me very sad indeed. They will at best
> only agree that I am rusty on C. I'm rusty on C because it is an out
> of date language.
Mr Nilges is of course incorrect. His rustiness on C has nothing to do with
whether or not the language is out of date.
>
> PEOPLE NEED TO USE THIS NEWSGROUP IN THE ORIGINAL SPIRIT OF USENET.
In the original spirit of Usenet, people tend to use mixed case. Upper case,
in the original spirit of Usenet, is considered to be rather uncomfortably
like "shouting".
> THIS IS ONE OF FREEDOM, AND COMPASSION.
With the freedom to post what one likes comes the responsibility to post in
appropriate newsgroups. Regular contributors to this newsgroup, when
they're not distracted by debunking the idiocies perpetrated by a certain
Mr Nilges, regularly show compassion to those who need it, by answering
their programming questions.
>
> AN ERROR IS AN OPPORTUNITY TO LEARN.
How true. What a shame that Mr Nilges does /not/ learn from his errors. For
example, he /still/ can't get the return type of strlen right, a good
nineteen months after this was first pointed out to him.
> RICH RETIREES AND POLICE AGENTS SHOULD DEPART.
This newsgroup is about programming. *Anyone* who wishes to discuss
programming - including rich retirees and police agents, should there be
any such who wish to discuss programming - is free to discuss programming
*here*.
Those who wish to discuss other things (e.g. economics, sociology, Fascism,
Nazism, and other such stuff) would do better to find newsgroups where such
things are topical.
pete
>
> Edward G. Nilges wrote:
>
> > Richard Heathfield <dont...@address.co.uk.invalid> wrote in message
> > news:<c025u3$31g$2...@titan.btinternet.com>...
> >>
> >> Just when you think Mr Nilges's C can't get any worse,
> >> it suddenly does.
> >
> > OK, this is an example of why I believe the troll at the bottom of
> > this pond to be a police agent, Richard Heathfield.
> > Therefore, I posted two code snippets as examples of code.
> >
> > In fact, the very fact that they contain errors is realistic
>
> It certainly provides good evidence that you wrote them.
>
> > due to the 80% failure rate of enterprise systems,
I think it's amazing that a programmer's error rate
can be tied to the failure rate of enterprise systems
and not be considered a characteristic of the programmer.
--
pete
CBFalconer
> Edward G. Nilges wrote:
>
>
> > Following Mr. Heathfield's lead, Randy Howard has posted repeated
> > abuse based on misaprehended tertiary sources for a month. Following
> > Mr. Heathfield's lead, Mr. Harter has pompously posted out of date
> > references on the philosophy of mathematics.
>
> Mr Howard's criticism of Mr Nilges appears to be based on Mr Howard's
> reading of articles posted by Mr Nilges. These are not "tertiary sources"
> but "primary sources". I am no mathematician (and, as Pascal wrote to
> Fermat, "that is a great defect"), so I make no comment on Mr Harter's
> contributions to this newsgroup, one way or the other.
I believe that the mathematics in question are concerned with the
4 color theorem, and were published roughly 20 years ago. Since
these are apparently out of date, what should we do about things
over 2000 years ago, such as the Pythagorean theorem, or Euclids
algorithm? I might be able to find more examples of out of date
mathematics, which many are still erroneously using.
As you can see from the quote levels, this posting is partially
based on a tertiary source.
Nils Petter Vaskinn
> Nils Petter Vaskinn <n...@spam.for.me.invalid> wrote in message
> news:<pan.2004.02.04....@spam.for.me.invalid>...
>> perl -pi -e 's/(\W)i(\W)/$1I$2/g'
>>
>> That will do the job. If that can get famous anything can!
>
> I'll have to trust you on this one. Last time I used perl was several
> years ago.
Actually it doesn't deal with start/end of line issues, so it depends on
the specification. Are "start-of-file" newline and "end-of-file" to be
considered blanks?
> But why does the code look like an explosion in a gnome factory?
Because regular expressions are designed to be terse. I agree they can
look horrible, and I could spend as long time reading/writing one line of
regexp as I would reading/writing several lines of ordinary code, but
usually not as much as would be required for code that essentially
implements the same thing as the regexp.
All programming languages involve some degree of tradeoff between
human-readable and machine-readable. Usually the easier you make it for
the human the harder you make it for the machine, and you often lose
flexibility too.
> Seriously, regular expressions have to be tested and documented
> carefully in order to be maintainable in real code.
Yes, but documenting and testing a regular expression is no harder than
documenting and testing several lines of code that does the exact same
job.
Also perl allows whitespace and comments in regular expressions (if you
use a switch) So for more complex expressions you can split it up and have
inline comments.
> Also, there is no
> consistent standard for how they are executed. They are executed
> consistently in perl, but the regular expression can't be re-used in
> another processor.
http://www.pcre.org/
Perl compatible regular expression, in a library. Distributed with a BSD
style lisence with advertising clause. that allows it to be embedded in
GPL or LGPL programs with (L)GPL having presedence (making it GPL
compatible)
Any language that can call a library (which means most) can use it.
And even if there are numerous incompatible regexp implementations out
there: SO WHAT! Many languages (most?) have the concept of a loop but
there are several different syntaxes in use. But it isn't a problem. And
when hopping languages you quickly pick up the different syntax because
the concept is the same. Same thing with regexps, same concept differences
in syntax, all you have to do is bother to look at the documentation.
--
NPV
"the large print giveth, and the small print taketh away"
Tom Waits - Step right up
Edward G. Nilges
> Richard Heathfield wrote:
> > Edward G. Nilges wrote:
> >
> ... snip ...
> >
> > > Following Mr. Heathfield's lead, Randy Howard has posted repeated
> > > abuse based on misaprehended tertiary sources for a month. Following
> > > Mr. Heathfield's lead, Mr. Harter has pompously posted out of date
> > > references on the philosophy of mathematics.
> >
> > Mr Howard's criticism of Mr Nilges appears to be based on Mr Howard's
> > reading of articles posted by Mr Nilges. These are not "tertiary sources"
> > but "primary sources". I am no mathematician (and, as Pascal wrote to
> > Fermat, "that is a great defect"), so I make no comment on Mr Harter's
> > contributions to this newsgroup, one way or the other.
The problem is that Randy has no basis for his absurd charges, because
the only material I published with code was in May of 2002, and these
were minor, "toy" programs meant as illustrative of a subsequent case,
which I decided not to make when I encountered your trolling, Richard.
Randy's opinion is based on the disbelief that it is possible for one
person to include programming in a wider range of interests, and this
is because so many programmers are so poorly prepared for the
intellectual challenge of programming, they are forced 24/7 to abandon
any other intellectual interests, and they are moronized by their
work, becoming idiot *savants*.
But in a passive-aggressive and self-defeating way, they are
consistently attracted to low-level platforms which make the mastery
of their trade that much harder.
This in turn is based on a sexist fear of inadequacy found in the
phrase, hurled as an insult, "script kiddie".
As I pointed out in Dec 2000 in my review of Steve McConnell's book on
certification, the infantilization and hyper-sexualization of American
culture, on display in a racist fashion during last week's Super Bowl,
MEANS that in fact American men, including male programmers, are in
fact deeply anxious about their place in society. In programming this
causes the psychic economy of deliberately choosing inadequate
platforms so that the programmer can redisplay as in a dream his
mastery and thus, as in a dream, quell his fears of his inadequacy.
Let me be blunt. C is used as a fashion statement.
You will have the opportunity in May of this year to buy "Design Your
Own .Net Languages and Compilers" and to see 20K LOC by yours truly,
and I will have to ask you to buy a copy. I welcome you then to post
your critiques of the code.
Randy probably hasn't read the code of the only examples, "truth
table" and "item parser" which were written in a language, C, which I
no longer regularly use. But he has no right (1) to make global
statements as he's done and (2) infer from a small set the right to
make those statements.
Ordinary courtesy would bid him shut his yap if I'd posted the "Quick
Basic subset compiler's code" and for some reason he concluded from
this code I'm a bozo.
Ordinary courtesy and the avoidance of interpersonal Fascism would
cause him to want, each time he expresses the opinion that I don't
know what I am doing, to carefully justify this opinion not only with
a detailed critique but also an alternate solution: this would be
truly "on topic" in a way his Rants are not. Of course, both of you
clowns specialize in merely repeating the insulting claim without not
once backing it up.
You claim you want people to "learn". Pray, what have they learned
from TWO YEARS of your trolling?
Richard, you are, in all probability, a paid police agent who
emotionally manipulates fools into doing actual work, and Randy is, in
all probability, a retiree who's "worked on" "operating systems and
device drivers" in an undisclosed fashion which doesn't imply he's
capable of anything but low-level maintenance. I haven't any respect
for either of you.
Ordinary common sense would bid Mr Howard shut his yap when the only
code posted was, apart from extempore whiteboard code, the two
examples in question. He has made himself a fool.
Richard Heathfield
<snip>
> Let me be blunt. C is used as a fashion statement.
People can use anything as anything if they try hard enough, but /I/ use C
not as a fashion statement, but as a programming language.
> You will have the opportunity in May of this year to buy "Design Your
> Own .Net Languages and Compilers" and to see 20K LOC by yours truly,
> and I will have to ask you to buy a copy. I welcome you then to post
> your critiques of the code.
Mr Nilges's code is already known to be flawed. There seems little point in
paying him money simply to have this knowledge reinforced.
> Randy probably hasn't read the code of the only examples, "truth
> table" and "item parser" which were written in a language, C, which I
> no longer regularly use. But he has no right (1) to make global
> statements as he's done and (2) infer from a small set the right to
> make those statements.
Mr Howard has, in fact, publicly stated that he has read the item parser
code. His statements, of course, are not "global"; he understands
topicality, and posts his articles only where they are relevant.
> Ordinary courtesy would bid him shut his yap if I'd posted the "Quick
> Basic subset compiler's code" and for some reason he concluded from
> this code I'm a bozo.
Those who know Mr Nilges's writings of old will smile grimly when he speaks
of courtesy.
<snip>
> You claim you want people to "learn". Pray, what have they learned
> from TWO YEARS of your trolling?
Mr Nilges seems to be asking himself a rhetorical question.
> Richard, you are, in all probability, a paid police agent
Mr Nilges here betrays his utter misunderstanding of the words "in all
probability". In fact, I can reveal that the probability of my being a paid
(or indeed unpaid) police agent is 0.
> who
> emotionally manipulates fools into doing actual work, and Randy is, in
> all probability, a retiree who's "worked on" "operating systems and
> device drivers" in an undisclosed fashion which doesn't imply he's
> capable of anything but low-level maintenance. I haven't any respect
> for either of you.
Mr Nilges's respect for me - or lack thereof - is utterly irrelevant to this
newsgroup, which (in case he's forgotten) is about programming.
Randy Howard
> > > reading of articles posted by Mr Nilges. These are not "tertiary sources"
> > > but "primary sources". I am no mathematician (and, as Pascal wrote to
> > > Fermat, "that is a great defect"), so I make no comment on Mr Harter's
> > > contributions to this newsgroup, one way or the other.
>
> The problem is that Randy has no basis for his absurd charges,
Of course I do. You have demonstrated, for at least a solid month or
more, that you know absolutely nothing about programming. Not just
in your source code, but in your lack of knowledge of programming languages,
O(n) notation, ADT's, the difference between sound algorithms and micro-
optimization, etc., etc., etc. The list is darn near endless. Just in
case you really are too blind to see it, I'm not the only one, by a long
shot, to come to the exact same conclusions.
> and these were minor, "toy" programs meant as illustrative of a subsequent case,
> which I decided not to make when I encountered your trolling, Richard.
Funny, I sort of doubt you are capable of writing anything BUT toy programs.
> Randy's opinion is based on the disbelief that it is possible for one
> person to include programming in a wider range of interests
Wrong. It's based upon your demonstrated incompetence, on an almost daily
basis since before Christmas.
> and this is because so many programmers are so poorly prepared for the
> intellectual challenge of programming
You being the most obvious example of that theory.
[persiflage snipped]
> As I pointed out in Dec 2000 in my review of Steve McConnell's book on
> certification
You forgot to repeat "popular". You're slipping Mr. Senile.
[more gibberish elided]
> Let me be blunt. C is used as a fashion statement.
Really? You must hang out with some very poorly dressed individuals
to come to that conclusion. I thought it was a programming language.
Hmmm, but you don't even recognize it when it is shown to you, so
why should you not be confused about its purpose as well.
> You will have the opportunity
*cough*
> in May of this year to buy "Design Your Own .Net Languages and
> Compilers" and to see 20K LOC by yours truly
Where 5K would have likely been sufficient?
> and I will have to ask you to buy a copy. I welcome you then to post
> your critiques of the code.
I won't buy a copy. I might check it out at the local bookstore,
and see firsthand how longwinded you can be in book form. But I
won't donate one lousy dollar to propping up your continued
fraudulent activities as a pretend programmer.
> Randy probably hasn't read the code of the only examples, "truth
> table" and "item parser" which were written in a language, C, which I
> no longer regularly use.
You really don't know how to listen, or read do you? There is a
reason you have two eyes, two ears, and only one mouth.
> But he has no right (1) to make global statements as he's done and
Sure I do.
> (2) infer from a small set the right to make those statements.
I can say what I please, particularly when my statements are backed up
so completely by your own self-destructive posts here. I am not the
first (or I'll wager the last) to notice this.
> Ordinary courtesy would bid him shut his yap if I'd posted the "Quick
> Basic subset compiler's code" and for some reason he concluded from
> this code I'm a bozo.
I'll make a deal with you. As soon as you stop posting gibberish to
this newsgroup, I'll stop talking about you and/or your pathetic
code. Promise. But, as soon as you start lying again, or misleading
people on purpose, or just plain trolling, or any of the usual
suspects, I'll be on your case again. BTW, if your book is as full
of it as it is likely to be, that'll be cause for discussion as well.
> Ordinary courtesy and the avoidance of interpersonal Fascism would
You have no grounds to talk about courtesy. EVER. You started the
death spiral in that area over 2 years ago. Your repeated attempts
to skirt by the Godwin's Law invocation while associated people with
war criminals is also less than courteous, you raving lunatic.
> cause him to want, each time he expresses the opinion that I don't
> know what I am doing, to carefully justify this opinion not only with
> a detailed critique but also an alternate solution: this would be
> truly "on topic" in a way his Rants are not. Of course, both of you
> clowns specialize in merely repeating the insulting claim without not
> once backing it up.
It's been backed up dozens of times. You're denying it doesn't make
them go away. The alternate solution is simple. Stop pretending to
be competent in this field, and start giving speeches to the
ladies auxiliary club about the philosphy of earthworms and the
socio-economic crises arising from their employment in gardens.
Everybody wins.
> You claim you want people to "learn". Pray, what have they learned
> from TWO YEARS of your trolling?
You have an amazing ability for self-caricature. Do you realize
that?
> Richard, you are, in all probability, a paid police agent who
> emotionally manipulates fools into doing actual work
Basis? Come on, let's here it. I know you wouldn't post such a
thing without backing it up. That would be hypocritical.
> Randy is, in all probability, a retiree who's "worked on" "operating
> systems and device drivers" in an undisclosed fashion which doesn't
> imply he's capable of anything but low-level maintenance.
You are free to believe what you want. I'm more than happy that
the balance is decidedly in my favor on that score, and you aren't
capable of understanding even the simplest of code, so I see no
point in trying to explain it to you.
> I haven't any respect for either of you.
Well, we finally found something we have in common. However, since
my opinion of a philosopher is off-topic here, I guess that's enough
about that.
> Ordinary common sense would bid Mr Howard shut his yap when the only
> code posted was, apart from extempore whiteboard code, the two
> examples in question. He has made himself a fool.
I'll let the record stand for itself. You haven't fooled anyone but
yourself.
Programmer Dude
EXAMPLE #1
Willem wrote:
> There is no such thing as 'tone' in a purely text message.
I respectfully must disagree with that, because--at least in my
experience--sometimes pure text messages carry a lot of tone.
To name just two examples, they can be friendly or unfriendly....
regards,
your internet pal,
-Chris
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
EXAMPLE #2
Willem wrote:
> There is no such thing as 'tone' in a purely text message.
[snort] You lame brain noobie!! Any idiot knows better!!
Next time, engage yer brain, ya doofus!
FOAD!
Willem
) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
) EXAMPLE #1
)
) Willem wrote:
)
)> There is no such thing as 'tone' in a purely text message.
)
) I respectfully must disagree with that, because--at least in my
) experience--sometimes pure text messages carry a lot of tone.
) To name just two examples, they can be friendly or unfriendly....
)
) regards,
) your internet pal,
) -Chris
)
) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
) EXAMPLE #2
)
) Willem wrote:
)
)> There is no such thing as 'tone' in a purely text message.
)
) [snort] You lame brain noobie!! Any idiot knows better!!
) Next time, engage yer brain, ya doofus!
)
) FOAD!
Nice try, but both examples had all the meaning right there in the words
themselves. The 'tone' of these messages, however, is not at all clear:
For example, in a real-world conversation, the first reply could have been
said in a warm, friendly tone, conveying that you are sincere. However,
you could also have used a cold, sarcastic tone, indicating that you
actually have little respect for me and are making fun.
The second reply could have been said in a shouting, harsh voice,
accompanied by a growling face, in which case I would infer that you are
mad at me. It could also have been said in a joking tone, with a smile on
your face, conveying the message that you're joking.
Programmer Dude
> [...] Chris is a working programmer and as such is my intended
> audience.
This audience member left the building years ago. IF NOTHING ELSE,
you should try to recognize that your message misses its mark.
> However, I can point out that my egregious misreading of Chris'
> code was caused by stress,...
Then apologizing for your--now proven to be false--insults should
be a simple matter.
> AN ERROR IS AN OPPORTUNITY TO LEARN.
Indeed. What have you learned here?
Thomas Matthews
> Hi...
> i am a engineering student(FINAL SEMISTER) studying in visswaraiah
> technological univervity in india. Can anybody suggest me some of the
> revolutionary ideas/concepts that i can implement.
> Regards
> Preetham
> preetham...@yahoo.com
Another idea is to create an application that can
create C or C++ souce code from an executable.
There are plenty of requests for this kind of tool
from the language newsgroups.
--
Thomas Matthews
C++ newsgroup welcome message:
http://www.slack.net/~shiva/welcome.txt
C++ Faq: http://www.parashift.com/c++-faq-lite
C Faq: http://www.eskimo.com/~scs/c-faq/top.html
alt.comp.lang.learn.c-c++ faq:
http://www.raos.demon.uk/acllc-c++/faq.html
Other sites:
http://www.josuttis.com -- C++ STL Library book
Kenneth Chiu
I think this depends on how you define "tone". Written
language can clearly connote as well as denote. But just as
clearly, face-to-face communication includes channels not
available when writing.
As an analogy, watching a play is a different experience
than reading a book. I can't say one is better than the
other, or that theater is more expressive than literature,
but I at least think they are different.
Programmer Dude
W> There is no such thing as 'tone' in a purely text message.
>> EXAMPLE #1
>> [...]
>> EXAMPLE #2
>> [...]
>
> Nice try, but both examples had all the meaning right there in
> the words themselves. The 'tone' of these messages, however,
> is not at all clear:
But already you admit they do appear to contain 'tone'. Whether
you are able to interpret it *correctly* is another matter. You
have already identified the obvious, first-level modes, as well
as possible underlying modes.
> For example, in a real-world conversation, the first reply could
> have been said in a warm, friendly tone, conveying that you are
> sincere. However, you could also have used a cold, sarcastic
> tone, indicating that you actually have little respect for me and
> are making fun.
Only if you supply an external (rather paranoid) context. In the
absense of such, the text carries--at least--a superficial mood.
Whether this is a deceptive cover for a deeper level meaning is
not only not the issue--it demonstrates just how subtle the 'tone'
of text/plain can be.
Do you realize what you are arguing? That when you read Shakespeare
or Stephen King (or whatever fiction you prefer) you get no sense of
tone from the text?
Do you really mean that?? If you do, you have my genuine heartfelt
condolences! What bleak, bleak world!!
Willem
) Only if you supply an external (rather paranoid) context. In the
) absense of such, the text carries--at least--a superficial mood.
When you take a text, without any tone, and look at its meaning, you get
some kind of mood. Of course, there is a tone that accompanies that mood.
However, this tone is completely derived from the written meaning.
) Do you realize what you are arguing? That when you read Shakespeare
) or Stephen King (or whatever fiction you prefer) you get no sense of
) tone from the text?
Of course, but that sense of tone is always concurrent with the actual
content of the text. It has to be, because that's the only channel of
information you have.
) Do you really mean that?? If you do, you have my genuine heartfelt
) condolences! What bleak, bleak world!!
You can guess the tone from the content of the message. Usually, this
guess is correct. However, the message itself does not have 'tone'.
But still, if somebody claims to infer something from the 'tone' of a
written message, he is making a fool of himself, because this 'tone' is
something he himself adds to the message, derived or not from the actual
text or the message. Which means that either this 'tone' is also present
in the actual text, or it is a fabrication.
Programmer Dude
W> There is no such thing as 'tone' in a purely text message.
> I think this depends on how you define "tone".
Okay. I define it as "emotional subtext". How 'bout you? (-:
> Written language can clearly connote as well as denote. But
> just as clearly, face-to-face communication includes channels
> not available when writing.
Absolutely! VAST amounts of emotional subtext are carried in
the audible and visual communications channels!
> As an analogy, watching a play is a different experience
> than reading a book. I can't say one is better than the
> other, or that theater is more expressive than literature,
> but I at least think they are different.
I agree completely. (I once wrote a long essay comparing some of
the significant differences between plays, "tele-plays" and films
(one key diff: they become progressively more immersive for a
variety of reasons). Books are *almost* not on that continuum,
because they are so different.)
But the fact *here* is that, oh, yes, Virginia, there IS tone
in pure text.
Programmer Dude
W> There is no such thing as 'tone' in a purely text message.
> When you take a text, without any tone,...
The only kind of text "without any tone" tends to be non-fiction
(yet there is certainly very toney non-fiction!), such as tech
manuals. MOST fiction has a "point of view" and a "mode". One
aspect of genre classification of fiction is based on the mode.
Compare "slapstick" and other "frenetic" comedy to comedies with
their basis in irony or sly wit.
Anyway, didn't mean to interrupt, please continue...
> ...and look at its meaning, you get some kind of mood.
So, you're saying that even text with no tone engenders mood in
reader. And what about when the author deliberately *tries* to
put a subtext into the writing?
> Of course, there is a tone that accompanies that mood. However,
> this tone is completely derived from the written meaning.
That meaning can be subtle, with multiple *levels* of meaning,
and can tie into required external knowledge of the reader. For
example, British or Chinese humor can be much funnier if you know
about the respective cultures.
>> [When] you read Shakespeare or Stephen King (or whatever fiction
>> you prefer) you get no sense of tone from the text?
>
> Of course, but that sense of tone is always concurrent with the
> actual content of the text. It has to be, because that's the only
> channel of information you have.
No, as above, the read brings some context to the reading. But I
agree, the bulk comes from the only possible source: the writing.
Thus, IF you get tone from writing (and it seems you do), then
you must agree--by your own logic--the tone is in the writing.
> You can guess the tone from the content of the message. Usually,
> this guess is correct. However, the message itself does not have
> 'tone'.
If the writer has played fair, the guess is most likely correct.
The only reason it wouldn't be is deliberate deception on the
writer's part, or ignorance of context on your part, or failure
to understand the subtext (also on your part), or (for that matter)
failure of the writer to successfully embed the subtext.
AS A WRITER, I take extreme umbrage to the idea that the text has
no tone. As a simple exercise, read some Lovecraft. Then read
some PG Wodehouse. Then come tell me text has no tone.
> But still, if somebody claims to infer something from the 'tone'
> of a written message, he is making a fool of himself,...
Not if he parses the tone out correctly. IOW, not if he (or she)
sees what's really there.
> ...because this 'tone' is something he himself adds to the
> message, derived or not from the actual text or the message.
Nope. I know it's there, because I put it there! (-: