Re: phreeqci ?

["David L Parkhurst" <dlpark@xxxxxxxx>]

> The idea of a helpful interface to phreeqe is fantastic.  But because of
my difficulty using it, I'm compelled to make some comments.

> After several years away from phreeqe, I was surprised by how much effort
it is taking me to re-learn the key word concept and overwhelmed by the
effort that it takes >  me to learn how to use the windows front-end.

I assume you are using version 2, not version 1
(https://wwwbrr.cr.usgs.gov/projects/GWC_coupled/phreeqc).

> I failed to simply re-create the input file for example 1 using the
on-line help, interface intuition, and geochemistry experience.  A simple
step-by-step tutorial would be really helpful to show just how files are
created.  For example, I cannot determine how the words "as HCO3" appear
after alkalinity in "ex1" short of typing them in myself.  But that cannot
be right since that would require my absolute and precise knowledge of the
syntax - kind of defeats the utility and beauty of the windows front-end.
In other words, my problem is the interface, not my geochemistry.  I fear
that it will take me days and days to learn and experiment with phreeqe
before I am comfortable with applying a real problem.

It is really difficult to determine how people respond to cues in
interactive programs. We have tried to make things as easy as possible,
while maintaining the generality of PHREEQC. If we hard code the options
for the gram formulae, then addition of a new element will not be handled
consistently. A column labeled "as/gfw" is available to change the
conversion factor for an element concentration from mass to moles and the
"Description of Input" box gives more details, with some specific
information about alkalinity.



(Embedded image moved to file: pic24590.gif)

> Also, while I appreciate interest in protecting the original code
integrity and formats, I feel that forcing a user to deal directly with
"key words" and expressions like "as HCO3" bogs down the user with syntax,
jargon, and unnecessary typing.  Windows programming allows for a far
superior way of isolating the user from such details.  Simply "checking" a
checkbox labeled "as HCO3" as opposed to checking the one marked "as CaCO3"
could invisibly place the necessary code language into the input file.  A
good user/geochemist should use his/her skill to know what boxes to check
and not what syntax to use.  Also, the user can always look at the input
file, but if you think about it, that shouldn't be necessary and "messing"
with that file directly can potentially lead to errors that are hard to
find except by very experienced users.

Our experience has been that people use a combination of the screen input
and editing. It has definitely required more effort in development to allow
the direct editing of files, so maybe that was a mistake on our part. I
certainly get frustrated if I have to go through several screens to correct
a single typo. You can enter and edit all of the data (we're still working
on a few keywords, but will hopefully finish before too long) with the
screen input which will preclude any syntax errors, so the choice is
yours. I hope you appreciate all of the Windows programming that has been
included, some of which is quite difficult to get to work seamlessly
throughout the program.

> Consider a modflow analogy:  modflow used to be tedious mostly because it
required such direct user attention to syntax and columns and columns of
cell data values.  Then, excellent windows software appeared in the 90's
which virtually eliminated such headaches and modeling became a virtual
laboratory for experimental hydrogeology and not a data entry and syntax
nightmare.

Modflow with all of its packages is very much "keyword" driven. I think it
is a very similar approach to organizing a wide variety of process
information.

> Again, I am struck by the excellent front-end for phreeqe, but I think it
needs to go a few steps further to be a real user-friendly tool.  Is there
any interest or support in my position on this topic?

"Excellent" is perhaps not the right word considering your comments. I am
committed to producing the best interface that we can put together. In the
future, I am considering the use of  wizards for common tasks. However,  I
want to finish the remaining keywords of the interface,  so currently, I do
not support your position on this topic. This is the second interface that
we have developed. Version 1 was much clunkier but perhaps clearer in some
parts of the design. It was so tedious, that I did not use it. I think
version 2 is a major improvement, to the point that I actually use some of
the screens as opposed to simply typing the keywords directly. I think it
makes sense to finish at the current level and decide what people have the
most trouble with. Your input is welcome and will be considered in further
improving the interface.

David

David Parkhurst (dlpark@xxxxxxxx)
U.S. Geological Survey
Box 25046, MS 413
Denver Federal Center
Denver, CO 80225

Attachment: pic24590.gif
Description: GIF image