Webmail Evaluation - Spring 2000 - Meetings
April 19, 2000 - Agenda
Minutes
Evaluation period
Based on testing progress to date, we need to extend the
evaluation period. Please let Peggy know if you can't work into the
beginning/middle of June.
- we have a lot more testing to do
- we may need to tweak the installs
- we need to write the final recommendation
SilkyMail
- no responses from vendor, we're going to eliminate from list
for this year
IMP 3.0
- available in beta
- has missing features
- suggested: if IMP is selected (v2), that we put IMP 3.0 up
later? basic agreement
- anyone who is interested can query re: further enhancements on
TWIG
Server installs
- Don will be finishing IMP on Solaris and Linux by the end of
this week
- John still working on install scripts/test form for TWIG
- Shaheen still working on IMP/IIS/NT installs, still problem
with cc/bcc
- WorldPilot is almost ready on Apache
- Worldpilot is expecting to find a mail folder, so (for any
platform) create the default folder by starting ELM (log in via
ssh client); then try WorldPilot again. Andy is working on patches
for this problem
- behavior that brings you back to the login screen with no
error message is likely either no mailbox (see above) or password
error (log into spinner to confirm password, and launch ELM while
you're at it).
- Shaheen will do a form to capture error messages that goes to
server installs
Raw data
available so we can double check if necessary (password protected,
ask someone who was at the 4/19 meeting how to log in)
User Testing
Note that we need to resolve discrepancies in testing results.
This means that the test teams need to find some means of getting
together to identify their discrepancies and determine whether they
are valid or not (and then comment them).
Possible reasons for discrepancies (Peggy and I encountered all of
these when we reconciled our IMP/Tru64 results today)
- eval data not transferred to matrix correctly
- tester provided incorrect result
- different behavior on different client OS's or different
browsers/versions
- different interpretation of criteria
- testers used different steps to test the criteria
|
