Re: Re: Bug: Current CVS in Playlist.pm

On Wed, Oct 24, 2001 at 07:48:42PM +0100, Richard Smith wrote:
> Mark's case was totally worse case scenario. Not average case. All Win2k,
> WinME and recent Office users have the Windows installer tech. And they have
> it if they've installed anything else that used it. Only Windows 95 users
> MIGHT need DCOM (but then, so do loads of other programs, so they probably
> have it already) and if they do - it's on the AP site for download.

Assuming that these things already exist can be dangerous.  We assume
nobody has any of the Perl modules, yet we're willing to make other
assumptions about the software on people's machines.  At best, that's
inconsistent.

> 
> The sliMP3 code wasn't Windows compatible (but was easily fixed) around the
> time Mark's friend bought his player. And I refuse to believe it wasn't
> possible to make it work on NT4 after the code bug was sorted.
> 

It wasn't possible, your belief notwithstanding.

> And yes, all that's required is a double click on server.pl, no DOS
> required. And the unzip could be sorted by making a .zip that was
> self-extracting. Easy enough.

Not according to the directions.  And not as long as the windows
version ignores or otherwise misparses the information in the config
file (last time I checked, two weeks ago, it still ignored the path
for the MP3 dir in the config file.  And if you'll remember, I
submitted bugs after those fixes were implemented.  Like, f'rinstance,
using single backslashes in the code.  Good for Windows paths; bad for
Perl interpretation.)

> 
> In fact, for the majority of users, downloading AP, running the
> self-extractor for the server, and double clicking server.pl is all that
> should be needed. Using some other setup program (like the Nullsoft one) you
> could even make server.pl go into the start menu. Use perl2exe or whatever
> it is, and no AP is required.
> 

That's great, but until that's _true_, my points stand; right now,
today, the server is opaque, unintuitive, and difficult to work with
for non-*nix users.  It's not an easy install, it's not simple to run,
the instructions are unclear, the config file parsing is (or was
recently) severely broken, and it doesn't run out-of-the-box on at
least one NT4SP6 system I've encountered.


> So now we're down to one file that can be installed using a variety of
> methods for even the simplest Windows user. Wasn't hard, was it?
> 

Dunno.  Hasn't happened yet.  Ask me again when it does.  I don't
understand why you're being defensive about what amounts to
suggestions to improve useability.  Have you helped non-power-user
Windows people through the install process?  Could you share some of
their feedback and concerns with us?  It would be nice to have a
counterpoint to the feedback and concerns I'm bringing to the group
 from other users.

> And like Sean says, the networking part is the real stumbling block right
> now, NOT the server software.

Really?  So these "average" windows users who understand Perl and can
mysteriously get the server code up and running bug-free have a mental
block against understanding dotted-quads?  Your "average" windows
users have a decidedly narrow amount of computer knowledge,
particularly since use of the device assumes the presence of at least
some hub-like device, or awareness of crossover cables.

Look, it's simple, really.  Take the 1.6 stable tarball, pick 10
"average" Windows users who among them span Win95-XP (not the power
users, coders, or gamers; the Moms and Pops of the world), and ask
them to set up the SliMP3 device and software.

Take notes.

Then fix the problems.

You may find I'm not far off in my assessment.



I'm happy with the product; it does what I want, and I can make it do
other things if I desire.  I'm merely trying to improve it so _other_
people can enjoy it.  That's an altruism that I can easily do without,
in terms of both time and, now, stress.  My experience with "average"
users informs my opinions.  Perhaps your definition of "average" and
mine differ.  But I can say with some certainty that there are more of
my type than there are of yours.

If Sean wants to improve the saleability and market for his product,
that's his business, quite literally.  If you want to debate whether
or not there are actually people having these problems and expressing
these desires, be my guest.  I'm sharing my actual experiences and
those of others I know with the server code.  Whether you choose to
believe that those incidents occurred is up to you.  But to ignore
them and assume that any "average" person will certainly never
encounter these problems or express these desires is just foolish.
Yes, the early adopters understand intuitively how to set up and use
the software.  Many of them are subscribed to this list, and a good
fraction of them are making code contributions.  Unless I'm
misunderstanding something, this group of early adopters is not the
final target market for the device.  






-- 
Mark C. Langston
[email protected]
Systems & Network Admin
http://www.bitshift.org

To unsubscribe from this group, send an email to:
[email protected]

 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/