opencd-devel

[Opencd-devel] More Thoughs

[Daniel M. <ma...@CS...>]

I think this project is a neat idea, but I can see that a lot of people on
the list have a lot of different ideas about what it should be.  I've seen
people discuss PHP and Perl, cygwin and Apache, etc.  Although I understand
people's enthusiasm for introducing others to a new way of computing, I
think that it too great a goal to accomplish in one move, and that
attempting to do so will cause this project to fail.  It is not enough to
just create a CD, we need to create a CD that people (hopefully lots of
them) will use, and will expose them to a new kind of software that does not
take away their rights.

I'll just list my ideas and some justification on why they make sense
compared to the alternatives.

This disk should target only Windows home or end users.  This means it
should only have (GUI) applications (OpenOffice, Mozilla, Gimp, etc.),
games, etc.

This means:
    1) We should not include command line unix tools, this will only confuse
and scare users, and there is really nothing about them to do with open
source anyway.
    2) We should not include any servers (Apache, etc.) or IT administration
tools.  These people already know of the existence of open source
alternatives, and there are already many CD's on the market with versions of
these tools that cost enough for companies to "take them seriously".
    3) We should not include compilers and other development tools.  Very
few people write their own software: professionals evaluate tools carefully
when making a decision, and cannot suddenly switch from VC to gcc in the
middle of a project; amateur are computer savvy enough to about know the
existence of these tools and how to find them.

For code under the GPL, its not enough (well, I don't think so, ask a
lawyer) to provide a link to someone else who hosts the code.  We should
just tar it all up and host in on our website.  We do not need to provide
the tools to untar and then build the source (but maybe just a little page
pointing people where to get the tools and how to build the code if they
want).

Ok, those were my thoughts.  I think this project will only succeed if
people begin to form a consensus as to what they are trying to accomplish.

Dan

Re: [Opencd-devel] More Thoughs

[Henrik N. O. <h....@bt...>]

> 
> Ok, those were my thoughts.  I think this project will only succeed if
> people begin to form a consensus as to what they are trying to accomplish.
> 

Jepp.  I agree.  

- 1 CD containing quality GUI stuff and nice installer.  
- Source tarballs of the relevant source (no updates) on our site and
the mirrors.  No source ISO, people can make one if they want.

Now let's get on with it.  The installer:

Did we decide on using an HTML-based system?

If so, then we can split into 2 teams.  

- One team that works on stripping down a web browser (K-meleon I
guess), so that it works in permanent Full-screen mode and launches the
normal installers.

- One team that works on a nice graphical interface in HTML with large
friendly buttons, installation help in the form of screen grabs with
explanations, etc.


Since we know that the HTML stuff will work with the installer engine
when it's ready, these two teams can work happily in parallel, and then
comine it in the end.

If people generally agree, then we should next prepare a 2-3 paragraph
working plan for each of these parts.

This work needs doing anyway, regardless of which apps we choose or
where the source goes.  We can continue debating the finer points of
that in the forum (hint) while getting on with installer development on
this list (hint, part 2 :)

Those who haven't checked out the two mock-up HTML installers on FTP
should start with that: ftp://ftp.theopencd.org

- Henrik 

-- 

Henrik Nilsen Omma             Theoretical Physics, Oxford
35 Frenchay Road               1 Keble Road
Oxford OX2 6TG                 Oxford OX1 3NP
h....@bt...          he...@th...

[Opencd-devel] More on K-Meleon and Installer Options...

[J A. F. <jaa...@ya...>]

My goodness, I go away for a weekend and I return to find a host of emails!  I
haven't gotten through everything yet, so I still may be a little behind on
some of the discussion.  If so, please bring me more up to date.

I'm not sure if anyone else has yet started work on K-Meleon or a similar light
weight web browser.  I've started looking at the source (it's C/C++) and have
gotten some response from the K-Meleon developers.  I just wanted to update
everyone on what I've found.

The biggest trick to getting K-Meleon to work for us will be getting it to run
off of the CD-ROM.  To do this requires digging around in some of the Mozilla
code.  The other features we require are fairly easily implemented in K-Meleon.
 I'm hoping to have a feel for what needs changed at least by the end of the
week.

However, I do have some issues to raise:

Are we targetting only the Win32 platform?  If so, K-Meleon as our installer
engine might really be the way to go.  If not, then we should probably look at
some browser based on a cross-platform language like java, python, perl, etc. 
This way we would be able to easily run our stripped down web broswer on
multiple platforms.

If we do want to go with a cross-platform browser, then we face other issues. 
I've been looking for a while, and I haven't found many good java/python/perl
browsers.  The most complete one I could find was a java-based broswer called
XBrowser, but I don't think it has a full screen mode (like K-Meleon).  Grail
is a python based browser, but it's lacking in some features as well (Doesn't
support all of HTML 3.2).  If anyone knows of any other viable broswers, please
let me know.

So I suppose my biggest question is, are we only targetting the Win32 platform?
 The answer should significantly influence our installer application.

If anyone else is already looking at K-Meleon, or a similar browser, please let
me know.  I'm away on 'vacation' this week, but I'll see what I can do.

jaaron

__________________________________________________
Do You Yahoo!?
Yahoo! Health - your guide to health and wellness
http://health.yahoo.com

Re: [Opencd-devel] More on K-Meleon and Installer Options...

[Henrik N. O. <h....@bt...>]

On Tue, 2002-05-07 at 21:48, J Aaron Farr wrote: 
> 
> However, I do have some issues to raise:
> 
> Are we targetting only the Win32 platform?  If so, K-Meleon as our installer
> engine might really be the way to go.  If not, then we should probably look at
> some browser based on a cross-platform language like java, python, perl, etc. 
> This way we would be able to easily run our stripped down web broswer on
> multiple platforms.
> 
I suggest we go ahead with win32 for these reasons: 

1)  90% of desktops use Windows 
2)  Mac users are in less need of 'saving' because OS X is aleady a
decent system based on the open source UNIX-like Darwin (as far as I
understand).  So, Mac users already have some introduction to OSS, and I
would say they are less likely to switch to Linux. 
3)  If we make the installer HTML- based then it can for the time being
run in a standard MAC browser (although the user may need to click 'yes'
in a few places.  Later we can work on a custom Mac browser. 
4)  Although we have agreed (it seems) to refrain from MS-bashing in
this project, I think we'll all agree that it's the MS stranglehold we
want to reduce. 


But most importantly, I think it's important to get one CD under our
belt first, if so only on one platform, and then we can think about
diversifying for the next release.  So, my suggestion is: let's keep it
simple and get one platform done.  The choice of software will also be
different for other platforms, adding to the confusion. 

- Henrik 


-- 

Henrik Nilsen Omma             Theoretical Physics, Oxford
35 Frenchay Road               1 Keble Road
Oxford OX2 6TG                 Oxford OX1 3NP
h....@bt...          he...@th...

Re: [Opencd-devel] More Thoughs

[Toby I. <to...@go...>]

On Mon, 6 May 2002 13:37:55 -0400
"Daniel Marques" <ma...@CS...> wrote:

DM> For code under the GPL, its not enough (well, I don't think so, ask a
DM> lawyer) to provide a link to someone else who hosts the code.

This is true. Legally we have to supply the source ourselves if asked for it. However, if we give links to someone else who hosts the code, it is unlikely that anyone *will* ask for it. If they do, I'm sure we can make arrangements.

-- 
          Toby A Inkster, Esq. ~ http://www.goddamn.co.uk/tobyink/
              mailto:tobyink<at>goddamn.co.uk ~ gpg:0x5274FE5A
  jabber:tobyink<at>amessage.de ~ icq:6622880 ~ aim:inka80 ~ yahoo:tobyink

In a hotel in Athens: Visitors are expected to complain at the office
between the hours of 9 and 11 A.M. daily.