[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

SEUL: stuff



responsees to assorted things, and new stuff...
a bit disorganized, excuse the writing; i know  haven't slept enough.

leaders: currently juhana and omega are the web-people, as mentioned,
and i am taking the role as software dev leader.

design: i am currently working on a draft of a design doc, which will
encompass some statements about our goals, and lay out a rough development
plan and project analysis.  this should make holes and dependencies more
obvious than just freeform discussion, as well as forming the first rough
draft of an actual design and work plan.
i hope to have it done by 4am...

interface vs. base imp: should we do the interface or the installer core
first?  you'll see my exact views in gory detail soon enough, but if you care
about my opinion in general...  it's not really clear issue under discussion.
since the interface can be referring to the final system ui or the installer
ui.  it should be picked apart a bit further and organized differently,
and i will do this...
i can't say there's a hard and fast rule for whether to do ui or core
first.  i have always done both simultaneously, first sketching out a
full, modular design (including figuring out what actually fits in a desired
time frame), and then delegating out modules.  a good abstraction for the ui
modules usually come out of this, so further design refinement can then
be done by a specialized ui group.  by this time, the structure of the ui
that inffluencse input order and dependencies (and vice versa) has been largely
determined.  other aspects of depencies between modules also get determined
at this level.  it is actually perfectly valid to argue for _either_
ui-first or code first design for a single smaller module, depending
on what the desired outcome is.
in a larger project you get a more mixed bag of dependenies, and they
have to be ironed out individually.  some parts will seem to do ui
first and others not - though each is actually dictated by a specific
case analysis, not one single design rule.
the most important rule i think is the old "90% design, 10% code".
a good high-level design makes lower levels follow [more] easily.

testing!
i know a few non-techies in my area who would love to test-drive seul
and tell us what sucks (or doesn't), and i am planning on giving them
that chance.  others who find appropriate tester-people should let
them do so as well.  this should be BETA testing, or at least late-alpha
testing, however.
i've tried giving programs to non-programmers to test even at
the early beta stage, and the feedback has been useless, because
they report all the obvious bugs that are already listed in the
liner notes.  these people are important for testing, but we must
be careful about what we hand them, and structure the form we
want for responses... but this will all come in due time.
these people will be most useful for actual test-driving once seul
(or pieces thereof - we can let them try individual utilities out
as well), is relatively stable.  this will be a while yet. :)
but when that time does come, there should be an appropriate place
for them to send reports to.  it may also be good for us to try
betas on people we know, get verbal feedback, and then write up
the report ourselves, translating their reactions into more concrete
design/spec-change suggestions, which are what we like to deal with.

along these lines, i've also been talking to these same types of
people recently, as well as technical people (read "sysadmins who wish
there were a better linux distribution so they could use it")
survey results are forthcoming...  (along with everything else)


last thing:
  if you think you will be doing coding on seul, let me know!
i know only some of the people on this list are programmers, and
some are pretty good ones.  i would like to know what we have to
work with.  if you have expertise in a particular area, please
mention it.
  this is not a commitment of any sort.  i just need to start
getting to know who's out there.  so if there's any vague chance you'll
be coding, drop me a line.

and a reminder that there are still more leaders needed.  look at
the list posted and let us know if you can help.
i'll also need some people to help out with management of the
development group, but that will come later.  let me know if
you are interested in leading a some chunck of the software project
though.  i will be eternally grateful to you.

i'll now go off and hide for a few hours while i finish this
design draft.  get back to you later...

-luka
----------------------------------------------------------------------------
Simple End User Linux Mailing list
To be removed from this mailing list send a message to majordomo@txcc.net
with the line
unsubscribe seul-project
in the body of the letter.
----------------------------------------------------------------------------