Reinhard Doelz (doelz) wrote:
: [Discussion is whether a command-line driven program package (example:
: Genetics Computer Group, GCG Inc, Madison) is evolving correctly into
: an X-based Graphical User Interface or whether Mosaic can be used].
: [the earlier thread on this can be reviewed using the following URL:
: <a href=gopher://gopher.bio.net/77/.wais-sources/biosci?mosaic+and+przemko+not+merck> X-GCG </a>]
:micha at amber.biophys.uni-duesseldorf.de wrote:
: : Przemko (przemko at reks.uia.ac.be) wrote:
: : : Or (just dreamin') a Mosaic access...
: [...]
: One very important argument why MOSAIC _cannot_ be used is that the client
: by definition does not implement 'conditional input' - i.e. the evaluation
: of parameters depending on other input.
This really doesn't matter! It the user enters a bad combination, the
software better send a error. Especailly since GCG is already set-up
for batch mode, hence the command line options.
: As the FORMS evaluation takes place as 'batch' you need to specify all and
: everything before you even start and, thus, have very little chance to
: be realistically implemented in a one-shot page.
It would mainly be implemented as a huge form with all the defaults
values. Then people can change them.
The CGI interface would need to make a URL to the output files or grab
them and make a html.
Timeout for larger runs might be a problem.
: Sequence searching, in particular, is not suited for synchronous requests.
That depends on the machine you run them on.
Neal
