Re: LEFT Error (Or PICNIC Error...) REFORMATTED

[Larry Bolhuis <lbolhuis@xxxxxxxxxx>]

> Sorry to sound like I'm pounding on you, but frankly to me it sounds
> like you might be translating your lack of familiarity of PC-based tools
> to deficiencies in WDSC.  WDSC acts just like every other PC-based IDE
> out there.  The things you're complaining about are the same in Visual
> Studio and JBuilder and Visual Age for Java and so on.
>  
You are correct. Since most of my coding is CL it's mostly done in SEU. 
In DOS I usually use Edlin... OK Not really, usually Notepad. In linux I 
like joe or pico. For the level of coding that I do the big power 
functions of the IDEs don't usually help me much. Sure I like having the 
exploere type bar on the left and the multiple members open and cut and 
paste and clickable errors on the bottom after a failed compile, all 
good stuff, but I don't push the capability of these things hard at all.

>  
>
> >    And I use % as in %substr and %len but when the super duper
> > integrated RSE editor doesn't know that there is no LEFT either I
> >    
> tried
>  
>
> > it.
> >    
>
> I don't understand this.  There is a LEFT function in REXX, which RSE
> supports, so when you searched LEFT in the Help, it brought you to the
> help text for REXX's LEFT function.  The problem is that there is no
> LEFT function in RPG, and you assumed there was one.  Although I guess a
> note somewhere on each page identifying the chapter you were in might
> help.  Although with IBM documentation, you're still not guaranteed that
> it will help <smile>.
>  
The point was if I'm editing a source member that is of type RPGLE the 
Help system SHOULD ( now I know it doesn't) should limit it's context to 
RPGLE and possibly the eiditor itself.    I figured by now RPG should 
have a LEFT function. Jon and Susan and Clause and all their Canadian 
firends and even that Irish guy keep telling me the language is getting 
better so I keyed it in figuring the IDE would coplain if LEFT wasn't a 
function.  Biggest problem was I didn't put in the % sign so RPG took it 
as an array and then complained about the delimiter. Not understanding 
the delimiter complaint I head off to the help to find ou there IS a 
LEFT function (for REXX as it turns out) and the syntax is to use a 
comma as I have done. And off I go spinning down the wrong path.

> Just for a pointer: if you decide to mess around with EGL, be prepared
> for a shock.  The EGL documentation constantly refers to z/OS.  Thanks a
> lot, guys! :)
>  
Thanks for the warning. I think I'll avoid it.

>  
>
> > > Why would you assume that something you happened to look up in the
> > >      
> help
>  
>
> > > would apply to RPG?
> > >
> > >      
> >    Because I was in an RPG source member when I pressed F1.  It'a all
> > about the context.  Check out every politician who ever held an office
> > and they will tell you their opponent tooke some statement out of
> > context which (usually) twisted the meaning around 180 degrees.  Go
> > google RPG. Now google iSeries RPG. Totally and completely different
> > context with totally and completely different resutls.
> >    
>
> Sorry, I must be slow.  I don't understand what this means, either.
>  
Same as above. I'm in RPG source members. I want help about RPG. When I 
open the manual for my Dodge RAM, because that book is in the context 
(glovebox) of my Dodge, it talkes about 4WD and towing and maximum 
loads.  When I open the manual for my Prius, because that book is in the 
context of my Toyota, it gives stern warnings not to mess with the 300v 
battery and talks about the magic motor that starts all by itself. It 
doesn't give me any towing our hauling capability at all, the Prius has 
no towing function.  At my house there is no combined manual for all the 
cars in my garage (thank goodness) but my Craftsman toolchest has tools 
that work on all three (hammer, vice grip, duct tape, WD40!)

> > > If you hit % (the prefix of all built-in functions) and hit Ctrl-space,
> > > you will get a list of all built-in functions.  It's actually pretty
> > > useful, and the Ctrl-space key combination is standard if you've used a
> > > PC-based IDE before.  WDSC is, for better or worse, a PC-based IDE.  I
> > > suggest you take the Eclipse tutorials.  They'll give you a lot of good
> > > insight into how Eclipse (and thus WDSC) works.
> > >      
Likely if I get into significantly more RPG the tutorial sounds like a 
good idea. Unfortunatly I've been trained by someone that you find these 
things with right and left clicking of the mouse. Memorizing even more 
ctrl and alt and shift key combinations isn't hight on my priority list 
at the moment!

> Hit "Help" in Word.  Or Excel.  Or any other Windows program.  You get
> the help for the entire tool.
>
>  
Please PLEASE do not let Micosloth set the bar for help!  It used to be 
you could trip over that bar but I think it's below ground now.

> Ctrl-space is the Windows standard content-assist key for just about
> every Windows editor out there.  And I'm going to avoid the temptation
> to tell you to use Ctrl-Alt-Del <grin>.
>  
Fortunatley I use few windows based editors. I learned Ctl-Alt-Del in 
1982. Took us a least a day to figure out how to reboot that machine 
without using the power switch. The problem with all these bizarre key 
combos is that you can't label them, ya gotta learn 'em.

>  
>
> >    Just different variable names. In the original LN was already
> > evaluated to %LEN(%TRIMR(@IFC)) to simplify debugging had it ever
> > compiled.
> >    
>
> My point was just that I don't understand how IPAddress and @IFC are
> related.  Why can you take the length of the trimmed value of @IFC and
> use it to extract data from IPAddress.  Those two variables must somehow
> have an implicit relationship.  Since I can't assume anything like that,
> the code doesn't make complete sense to me.
>  
Yes, thaey ARE related in that @IFC is the IP Interface that I'm 
interested in while IPAddress is the current Interface being retrieved 
from a user space supplied by an API. Problem is the IPAddress is null 
padded by the API to 15 bytes while @IFC is space padded. So I want to 
compare only the number of non-space characters in @IFC to the same 
number of characters of IPAddress.  Technically this compare could 
result in a false positive (i.e. @IFC of '172.16.1.1     ' would match 
IPAddress of '172.16.1.10nnnn' (where 'n' = null))  I still need to 
resolve that but at least I get some positive matches now!!

> Joe
>
>