| Author |
 Topic  |
|
|
Moebius
Starting Member
1 Posts |
 Posted - Nov 16 2006 : 11:57:58 AM
|
1. When working with functions from the global scope, typing :: causes a major editor delay which is quite annoying, mostly because it cannot be cancelled by typing another letter. Would it be possible to prioritize typing in this case?
2. Despite ALT+M being extremely useful I tend to create implementations of member functions in the same order as they appear in the header file. Would it be possible to add an option so "Create Implementation" would keep the implementation order the same as in header file? It could be a global option like "Keep declaration order" or "Keep alphabetic order" (or "Surprize me!", heh).
3. An option to implement all member functions from header file, that do not have a body yet would be VERY useful.
4. It would be EXTREMELY useful to finally have a tool which would show you header dependancy tree in graphical or text form. It should be able to find inclusion path from header A to header B (or state that one does not exist).
5. Integrating some minimalistic, template-based code beautifier into VAX would be great.
6. Auto-text. Very useful feature, but to my knowledge there is no way to merge my custom modifications with your updates. Or am I wrong? Couldn't you just mark my inserted/modified auto-text entries with some flag and merge/update everything except what's flagged?
I think that's enough, heh.
Anyway, great tool, but there isn't much more to say, because you have already caused a global fever. To be honest, I'm afraid of all the masochist madmen that know VAX exists but somehow don't use it yet.
|
Edited by - Moebius on Nov 16 2006 12:01:53 PM |
|
|
feline
Whole Tomato Software
United Kingdom
18749 Posts |
Posted - Nov 16 2006 : 1:47:46 PM
|
point 1, from memory this list comes from the IDE, not VA, so we don't actually have any control over this.
point 2, if you have functions from two classes implemented in the same cpp file Create Implementation will keep the new function with the correct classes functions, which is something.
You are not the first person to request more control over the ordering of methods for Create Implementation. The problem with keeping the cpp file in the same order as the .h file is what do we do when they are already out of order? Also some people will want them in different orders.
To let the user specify where in the cpp file to place the new method is:
case=2161
point 3, what about abstract or virtual functions?  To allow Create Implementation to operate on more than one function at a time is:
case=1092
point 4, you may want to have a look at this:
http://www.codeproject.com/tools/includefinder.asp
point 5, it would be about 30 seconds before we were swamped with requests for more options, more power, etc So far we are staying well away from this. One common suggestion for this is:
http://sourceforge.net/projects/gcgreatcode/
point 6, what about people like me, where my personal autotext file may have 2 lines in common with the one in the installer, but I would not guarantee it
http://winmerge.org/
What ever you do when installing both the new version from the installer and your version are backed up. It is worth pointing out that the version we ship with the installer very rarely changes, so this is not much of an issue.
Yes, stay well away from those masochist madmen, I had to work with one once, not a pleasant experience *shudder* |
zen is the art of being at one with the two'ness |
 |
|
|
sean
Whole Tomato Software
USA
2817 Posts |
|
|
feline
Whole Tomato Software
United Kingdom
18749 Posts |
Posted - Nov 16 2006 : 4:24:55 PM
|
| I may try that at some point, it looks interesting. |
zen is the art of being at one with the two'ness |
|
|
|
sl@sh
Tomato Guru
Switzerland
204 Posts |
Posted - Nov 17 2006 : 03:33:58 AM
|
quote:
Originally posted by feline
point 3, what about abstract or virtual functions?
Hmm, at least for C++ the answer should be obvious: it's incorrect to provide an implementation to an abstract function and last I looked an implementation *not* containing an implementation for a virtual but not abstract function wouldn't link. Even if it were possible to not declare an unimplemented virtual function abstract, it would be bad coding style, and I wouldn't know why VAX would want to support that 
|
|
|
|
feline
Whole Tomato Software
United Kingdom
18749 Posts |
Posted - Nov 17 2006 : 10:07:22 AM
|
Before now I have implemented an abstract virtual function in a base class. This implementation is the default form of the function, but it is up to the derived, concrete classes to work out if they can call the default form or if they have to define their own versions.
As for what VA should and should not support, remember we try to help even with code that does not actually compile, because you are in the middle of typing it in
The idea is appealing, but I am wary of trying to be to clever. We are not short of bugs or suggestions as it is |
zen is the art of being at one with the two'ness |
|
|
|
support
Whole Tomato Software
5566 Posts |
Posted - Aug 09 2012 : 01:45:10 AM
|
| case=49162 replaces case=1092 and is implemented in build 1912 |
Whole Tomato Software, Inc. |
|
|
| |
Topic |
|
|
|
A couple of suggestions
