Re: Detail first? Or Big Picture?

For me, it varies a lot on the project.

Since I've mostly been working long-term for companies, as opposed to contracting, and I've pretty much exclusively worked in a single-developer environment, I've rarely had a need to "clean up" my code or the comments in it.

When I release something to the public, though, I'll usually go through and look to see that I've done things as cleanly as I'm able, and commented any "inelegant" parts where I've either deemed speed to be a priority over good coding practices, or I've deliberately hacked my way around a problem/bug/whatever.

In terms of development, that also varies. If I'm working interactively with someone, like my boss, and need to know that my design is going to work for them, I'll usually do the GUI first and make sure it works roughly correctly, but with no actual functionality involved beyond what's necessary to demonstrate. If I'm puzzling out how to do something, I'll usually set up a test procedure and work out the basic concept to make sure that it can do what I want it to, then I'll work it up into a proper class/form/whatever. More often than not, though, especially if the coding is fairly straight-forward, I'll do top-down class design first, test that functionality, then integrate it into the GUI/higher class (or skeleton thereof), etc.

Ummm...does this highlight the fact that I have ADD, or is this normal? :)


Rob

MM wrote:
What do YOU do when your classic VB project is nearing completion? Do
you spend a few days "tidying up"? Removing those redundant comments
that could be confusing rather than helpful if left in? What is your
development style? To address the details first and build your product
from them; or look at the big picture and decide how the application
is going to be used, knowing that the details will sort themselves out
later?

I am firmly in the latter camp. I always build some scaffolding, hang
some windows and doors off them (controls), stand back and contemplate
the overall look at feel. Maybe I'll remove a window and add a
chimney. Or change the angle of the roof pitch, or the colour of the
bricks. All these bricks and mortar are a metaphor for my "school" of
design - Rapid. Application. Development. Can't be rapid enough in my
book. I know of others who spend weeks designin' away, with ne'er a
sign of any application taking shape. But they are happy. They keep
their RAD picture in their heads. Fair enough.

Now I am approaching the end of my project, well, its initial
prototype at least, and I am in the process of removing those
redundant comments, removing obsoleted procedures or functions called
OldMakeLongInt or SearchString_2, which a sudden insight when I got
down to the details made me think of a better way to do something.

I am completely comfortable with my approach, though I know others
won't be. Finally, I'll probably add line numbers and error handlers
everywhere, though as I have long since had an add-in to automate this
task, it won't be the chore it once was.

MM
.

Relevant Pages

  • Re: Detail first? Or Big Picture?
    ... The grand Design approach, while rigorous, is often why Grand Corporate and Government IT projects fail miserably, overrun, and cost billions more than projected. ... Removing those redundant comments ... All these bricks and mortar are a metaphor for my "school" of ... Can't be rapid enough in my ...
    (microsoft.public.vb.general.discussion)
  • Re: Six word SF
    ... spreads out a lot further than with the lid up. ... I'll bet that varies so much with the design of that particular lid-pan ...
    (uk.rec.sheds)
  • Re: Loading the design from Compact Flash...
    ... through RS232 port and I have in fact populated the local BRAMs with ... Did you try removing the "-target mdm" as well? ... way to make sure that my design is actually loaded and working (the ... SysAce controller unless you intend to write to the compact flash - it ...
    (comp.arch.fpga)
  • RE: calculating active directory replication traffic on w2k3
    ... information with a grain of salt, as it varies greatly based on design, ... environment, topology, number of users/sites, etc. ...
    (microsoft.public.windows.server.active_directory)
  • Re: interpretting F & t-vals from regression
    ... If a level within your design turned ... out to be useless in the light of observed data, ... understand what you mean by "removing" it. ... Which duration levels I see a slope and which levels I do not is ...
    (sci.stat.consult)