RE: help with c++ if test

  • From: "Sina Bahram" <sbahram@xxxxxxxxx>
  • To: <programmingblind@xxxxxxxxxxxxx>
  • Date: Wed, 9 Feb 2011 14:48:19 -0500

 
Kind of, but yes.
 
It's a simple size versus speed trade off.
 
Take care,
Sina
From: programmingblind-bounce@xxxxxxxxxxxxx 
[mailto:programmingblind-bounce@xxxxxxxxxxxxx] On Behalf Of Littlefield, Tyler
Sent: Wednesday, February 09, 2011 2:45 PM
To: programmingblind@xxxxxxxxxxxxx
Subject: Re: help with c++ if test
 
Wow, that's sweet. so essentially you just have the pointers to the functions 
or points to jump to, spaced out. so you would just
xlat to the number you want to compare against, then jump to that. for things 
like ranges, you'd just repeat the pointer, makes
sense.
On 2/9/2011 12:38 PM, Sina Bahram wrote: 
In assembler you mean? 
  
Wel, it's pretty simple. Take some of your conditionals, and then simply put 
them into your data segment by spacing out your data an
equivalent spacing based on your conditional, so 32 steps away if you are 
checking for 32, and so on and so forth, then use xlat and
a single jump to basically translate, jump, translate jump, and you've got 
yourself a super efficient, very fast piece of code,
which has some extra stuff in the datasegment, but who cares. 
  
Incidentally, this is the first trick I'd use if writing a piece of software I 
wouldn't want subject to static analysis. Dynamic
would still handle it, especially with concepts like taint analysis, but static 
stands no chance. 
  
Take care, 
Sina 
  
  
  
From: programmingblind-bounce@xxxxxxxxxxxxx 
[mailto:programmingblind-bounce@xxxxxxxxxxxxx] On Behalf Of Littlefield, Tyler
Sent: Wednesday, February 09, 2011 2:35 PM
To: programmingblind@xxxxxxxxxxxxx
Subject: Re: help with c++ if test 
  
xlat looks cool. You were mentioning that that's how you would optomize code, 
which I didn't understand after I read up on it. It's
awesome for indexing into tables, but I'm not really sure how that helps in 
optomization.
On 2/9/2011 12:31 PM, Sina Bahram wrote: 
Yes, good luck doing that in c, lol, haha. 
  
Actually, what you really should do is use function pointers dereferenced out 
of arrays, using your input as a way of selecting
where to jump to next. 
  
And, oh my god, that actually is what they do . still no goto, *smile*. 
  
*sigh*, I miss the xlat instruction. That was my favorite op code. 
  
Take care, 
Sina 
  
From: programmingblind-bounce@xxxxxxxxxxxxx 
[mailto:programmingblind-bounce@xxxxxxxxxxxxx] On Behalf Of Ken Perry
Sent: Wednesday, February 09, 2011 2:18 PM
To: programmingblind@xxxxxxxxxxxxx
Subject: RE: help with c++ if test 
  
Yes and instead of goto in kernel programming you really should use jumps. 
  
ken 
  
From: programmingblind-bounce@xxxxxxxxxxxxx 
[mailto:programmingblind-bounce@xxxxxxxxxxxxx] On Behalf Of Sina Bahram
Sent: Wednesday, February 09, 2011 2:11 PM
To: programmingblind@xxxxxxxxxxxxx
Subject: RE: help with c++ if test 
  
Yes, regardless, goto is bad. 
  
But absolutely you can break out of the inner loop, but then you'll be in the 
top level loop, in which you can do a conditional. 
  
This is still more efficient, since it is definitionally true that you would do 
a check in the inner loop before your goto, and if
you are doing a check in the inner loop, then that means you are doing a check 
M*n times, where m and n are the dimensions of the
outer and inner loops, respectively. 
  
If you remove that, and simply do the check in the outer loop, then you are 
only doing a check m times, which is an entire order of
complexity better. 
  
Goto is bad. This is the one rule where because I say so and other inane 
excuses from teachers is acceptable. Goto is just bad.
There is no reason, none, what-so-ever, that you need one. 
  
Now, if we get into kernel programming where assembler is being linked in, etc, 
etc, then we can talk . but in something as straight
forward and earth shatteringly simple as the logic for an adventure game? Goto 
is absolutely in no way necessary under any possibly
conceivable permutations. 
  
Take care, 
Sina 
  
  
  
From: programmingblind-bounce@xxxxxxxxxxxxx 
[mailto:programmingblind-bounce@xxxxxxxxxxxxx] On Behalf Of Littlefield, Tyler
Sent: Wednesday, February 09, 2011 2:05 PM
To: programmingblind@xxxxxxxxxxxxx
Subject: Re: help with c++ if test 
  
Sina:
Break is awesome when you have one single loop, but you can't break out of two 
loops, can you? It was just the first thing that came
to mind because I've seen it used (and used it) like that.

On 2/9/2011 11:50 AM, Sina Bahram wrote: 
tyler, this is a horrible example of goto, my friend. 
  
That's what break is for. 
  
Take care, 
Sina 
  
From: programmingblind-bounce@xxxxxxxxxxxxx 
[mailto:programmingblind-bounce@xxxxxxxxxxxxx] On Behalf Of Littlefield, Tyler
Sent: Wednesday, February 09, 2011 1:17 PM
To: programmingblind@xxxxxxxxxxxxx
Subject: Re: help with c++ if test 
  
I really really highly recommend you avoid goto. This isn't basic, and they're 
not very useful except for in some odd cases, far and
few between. Such as jumping out of two nested loops like so:
int i, j;
for (i = 0; i < 100; i++)
{
for (j = 0; j < 100; j++)
{
if (i+j == 100)
goto botttom;
}
}
bottom:
//do something here
if you want to compare strings, do something like this:
if (input == "north")
{
GoNorth();
}
else if (input == "south")
...
else ...



On 2/9/2011 10:52 AM, Kristoffer Gustafsson wrote: 
Hi. 
I'm writing if tests in c++ with the goto command. 
I want to do text games and I'm using these commands for that. 
But it doesn't work, so I must do something wrong. 
Can you give me an example how this is done with a string please? 
/Kristoffer 
  
--       
       
Thanks,      
Ty      
  
--     
     
Thanks,    
Ty    



--   
   
Thanks,  
Ty  




-- 
 
Thanks,
Ty

Other related posts:

  1. Home
  2. programmingblind
  3. Archive
  4. 02-2011