Re: Do C++ and Java professionals use UML??
Gene Wirchenko wrote:
John B. Matthews wrote:
[snip]
I would argue that a single click from any included class or interface
qualifies as "easily found." Naturally one may wish for more. When Java
5 - a particularly eventful revision- was new to me, I kept a copy of the
API under local version control for the purpose of adding/updating
useful links. YMMV
One also has to know that it exists. I do not remember seeing
the stuff at page top that I see now. I was primarily concerned with
individual classes, so I must have missed it.
This is also a problem with hardcopy docs.
One big disadvantage of on-line documentation is that sometimes,
bits of it are hidden or not in obvious places. Hard-copy
This is also a problem with hardcopy docs.
documentation has the advantage of nothing being hidden. If you go
I guess I'm less fortunate than you. The hardcopy docs I've used over
the years often had references in other volumes, links to material in
far-removed locations within a doc set, incomplete definitions
through the whole book, you get all of the content.
So why don't you have the docs in hardcopy, then?
Then you'd have no excuse to lack the information, according to
your logic.
All you have to do is go through the whole book, right?
Also, you assume only one book's worth of documentation. That
is optimistic in many contexts.
I've used hardcopy programmer's docs and online, and while you
might find it easier to use the hardcopy multi-volume sets with
footnotes pointing you to volumes not in hand, I personally find
online docs both easier to use and easier to search.
--
Lew
If I'm ever stranded on a desert island with only one book,
I want that book to be an unabridged dictionary because it
contains every other book ever written. All you have to do
is read the words out of the dictionary in the right order.