Re: AfxGetApp() returns NULL

The problem with the higher-level approach is that it has a clumsy UI; in the abstract, it
would be good, but the last time I tried it in VS2003 (with 80K lines of unknown source) I
found the actual usage clumsy. Since I needed to find many kinds of things, I kept having
to scroll things around, redo searches, getting myself back into yet a different context
to find the next thing (having gone off and looked for some other stuff in the meantime),
and so on. So I reverted to my older editor and things went much faster. The problem I
have with the VS IDE group is that they fail to do user studies; they see each aspect of
the IDE as living in total isolation from a workflow concept. Taken independently, the
ideas may be good, but without user studies of real users, and a concept of workflow
optimization, even the high-level methods are clumsy to use. And the worst part of it is,
if I DO need to do editing, I'd be forced into using the IDE editor, which can't even get
something as simple as indentation right! So to do real editing, I have to go to an
editor that actually knows how to indent code properly (the code in a {} list is indented
one position from the {, not directly under it, for example!)
joe

On Mon, 17 Dec 2007 06:23:38 GMT, "David Ching" <dc@xxxxxxxxxxxxxxxxxxxxxx> wrote:

"Joseph M. Newcomer" <newcomer@xxxxxxxxxxxx> wrote in message
news:os2cm31j0g276skq7gur1kbd0oo5ohaifd@xxxxxxxxxx
As to Find In Files, it is my choice, and my choice alone, to decide if
the contents of a
search are no longer useful; the fact that it chooses to delete the
contents of the
previous find, or give me some clumsy method that requires I keep choosing
different
windows, seems pointless. I typically will do a "delete lines that match"
if the find was
a little too generous.

The chances I will have a method with the same name as a variable is
really vanishingly
small.

Also, the scale of my projects usually obviates the need for incredibly
sophisticated
tools; there's not a lot of sophistication required to deal with source
sets in the range
of 100K-200K lines, but if I had a megaline of code, it would be a
different issue.
Finding definitions is generally a couple keystrokes. The number of times
I need to find
all uses is so infreguent that a two-step process imposes no serious
effort; most times
when I need to find all uses, a simple search string in the one-and-only
module that uses
the class is sufficient. Again, working on small scales like 200K source
lines means that
using clumsy tools doesn't buy much, and find in files is certainly badly
done.

One serious failure, which Microsoft is merely the latest perpetrator of,
is the notion
that output windows are read-only. There is no justifiable reason that I
cannot edit
EVERY window (in particular, delete lines based on pattern matches); I end
up having to
copy the lines to a real editor so I can process them. This should not be
necessary. I
find the VS editor gets in my way more than it helps me. I have very high
expectations of
a text editor, and it isn't even close. I have high expectations of file
find, and
replace-in-files, and VS isn't even close to the tools I use continuously.
I see it as a
compiler/debugger world, but it doesn't make it as a source editing
environment.

My source bases are typically in your definition of small. I used to
confine myself to the tools you do, but no longer. For example, even when
finding all uses of a private data member such that it is in only one .h and
one .cpp file can be done simply by doing a Find/Find Next, it is very
helpful to get a list of all uses simply by right clicking on the data
member and selecting Find References. Your mind does not have to go into
the mechanics of finding all references by decomposing the problem into
Find/Find Next, loading the next file, repeating the Find, etc. Granted,
this is no big deal and comes naturally after all these years, but I'm
really getting into the .NET world where they cut through this low-level
nonsense (miniscule though it may be) on all fronts, so I readily recognize
it for what it is --- nonsense. I can deal with it fine, but at least
recognize that it's a crude process merely approximating the more
higher/abstract one.

-- David

Joseph M. Newcomer [MVP]
email: newcomer@xxxxxxxxxxxx
Web: http://www.flounder.com
MVP Tips: http://www.flounder.com/mvp_tips.htm
.

Relevant Pages

  • Re: AfxGetApp() returns NULL
    ... search are no longer useful; the fact that it chooses to delete the ... or give me some clumsy method that requires I keep choosing ... that output windows are read-only. ... find the VS editor gets in my way more than it helps me. ...
    (microsoft.public.vc.mfc)
  • Re: Netfetch images
    ... You're missing a "as I am used to something different" here, ... comparison clearly shows Windows to be clumsy and unhelpful." ...
    (comp.sys.acorn.apps)
  • Re: Netfetch images
    ... XP usually to be clumsy and unhelpful ... You're missing a "as I am used to something different" here, ... "as I have used Windows and at least one other OS for a long time, ...
    (comp.sys.acorn.apps)
  • Re: Two user profiles, but only one signature?
    ... Rich Pasco wrote: ... I considered two user logins at the Windows level, ... >> clumsy, but it might work for you. ...
    (microsoft.public.outlook)
Loading