[WiX-users] Windows Installer XML toolset roadmap - proposal

[Rob M. <ro...@us...>]

By now you've seen Fred and Nick's email about the WiX toolset website.  I
asked them to pick up the website project because it has been languishing in
my queue for over six months.  The reason that is relevant now is because I
originally wanted to unveil the "Windows Installer XML Toolset Roadmap"
(think bright shiny marquee with rotating lights and stuff) on that website
about six months ago.  However, it has become increasingly clear to me that
I can not do as much as I want and need to get done so I'm falling back to
simpler plans and redistributing responsibilities to those who want it.  So
on with the simpler presentation of the proposed Roadmap for the WiX
toolset.


Defining the "core" of the WiX toolset

At the beginning of the year, I would have classified the "core" WiX toolset
to be candle.exe, light.exe, lit.exe, winterop.dll (then wixcab.dll),
wix.dll, and wix.xsd.  However, through amazing strides by Robert Flaming to
greatly improve the quality of the decompiler, I was "forced" to add
dark.exe to the "core" WiX toolset.  For the rest of this proposal the core
WiX toolset can be defined as cande.exe, dark.exe, light.exe, lit.exe,
winterop.dll, wix.dll, and wix.xsd.  All other tools in the WiX toolset
(such as tallow, and votive) are what I call "support" tools.


WiX Toolset Roadmap

The following is the priority order of things I think that we should focus
on right now.

1.  Stabilize the core WiX v2.0 toolset.  This has my primary goal for the
past 2 years (since WiX v2 was first created).  A stabilized core toolset
means there will be no more schema changes, only bugs without workaround
will be fixed, and the WiX v2.0 toolset will be marked "Production Quality"
on the main SourceForge site.  Think August.

2.  Improve the WiX Documentation.  Ensuring the toolset doesn't crash or
come up with the wrong answers is a vital step towards improving adoption
but I believe documentation (and that includes examples/templates) will be
the real difference whether people adopt the WiX toolset outside of this
community of ours.  Think December.

3.  Create the WiX Website.  When the tools are stable and the documentation
is improving, we'll need to have a place that showcases the WiX toolset and
directs people to the various facets of the toolset (core toolset download,
support tools, documentation, current status of project, other interesting
facts, etc).  Fred and Nick have not provided a schedule, yet (mostly
because I did not ask).

4.  Continue investment in support tools.  We need to continue to invest in
the tools that are built on the core WiX toolset.  Justin will continue to
lead the charge on Votive.  One day someone will pick up the lead for tallow
and hopefully improve it (while taking into account the Component Rules).
There is another project in the works that I think is very exciting and will
be debuting soon (a month or three).  Think 2006.

4.  Improve and grow the WiX CustomAction Library.  At the same time the
support tools move forward, I want to move forward with the WiX CustomAction
Library.  It feels like if we fixed all of the issues in the current
CustomActions and added about 10 - 20 more we'd cover maybe 80% of people's
CustomAction needs.  This is a very exciting area to me and where I want to
focus in 2006.

5.  Start WiX v2.1 toolset.  Finally, with WiX v2.0 toolset locked down and
stable, we will need a place to add the new features and take the bug fixes
that we weren't willing to destabilize the v2.0 for.  Derek has already
started the process for moving to v2.1 and is holding my feet to the fire
for each bug I suggest should be fixed in v2.0.  Think post-v2.0 stable.


Errata

My original dream (sometime in October of 2004) was that we would stabilize
the core toolset by January of 2005.  Obviously that was just silly.  Then I
adjusted that dream to April of this year knowing that date was extremely
aggressive.  We obviously missed that one as well.  My latest date, August
2005, is based on what I'm seeing in the bug trends and discussion on the
mailing list.  Far less communication seems to be about what doesn't work
(although there is that) and more about how to make the installs work.  The
latter is why I think documentation is so important.

Documentation is a no-brainer.  We will never have enough but we certainly
can do better than what we have now.  This will become my focus after the
core toolset is reaches production quality.  One of my primary goals is to
try to get to a place where the community can help produce more permanent
documentation than just the mail archive.  Which leads to the third point,
the WiX Website.

Fred and Nick are leading this charge.  I didn't put any restrictions on
them so I really can't say what they are going to deliver or when.  All I
can say is that I like the feedback I've seen on the mailing lists so far.
I'm sure they would appreciate your support (especially since Nick said he
isn't real big on web development... <grin/>).

The two #4 bullet points are not a typo.  I do see those as equal
priorities.  With a solid core toolset and documentation in place it will
become the tools that help people get their job done faster that are needed.
Votive has a great start but it needs a lot more features.  Justin has many
plans and would love others to join him.  Tallow could use a lot of polish
and maybe add a few more things it handles (out of proc SelfReg capture
anyone?).  Also, there is a "top secret" tool that Reid is leading a small
group of devs internally right now.  They are trying to get the "first
draft" working quickly and will release it here for everyone else to use and
contribute to shortly (I'm hearing two - three months).  I am very, very
excited about the potential of this project and can't wait until Reid gets
it working.  Also through 2006 growing the WiX CustomAction Library to
handle many of the standard things developers need done make a lot of sense
to me.  

Finally, as we go into 2006 there will be updates and changes to the core
toolset we'll want to make that would destabilize v2.0.  As mentioned above
Derek is already moving on this so that when we want to fix bugs but think
they might destabilize v2.0 we can put them in v2.1.  Thus v2.0 will be for
those people who want to use a stable version of the WiX toolset and v2.1
will be for those who want all the new bells-and-whistles and are willing to
deal with breaking schema changes and crashes.
 

Comments, suggestions, feedback?

Feel free to send feedback to wix-users.  I may not have time to respond to
everything but will definitely read everyone's comments.  I will then try to
incorporate them into another draft of the roadmap and send it around again.
Please, keep in mind, this is the high-level thinking.


virtually,

	Rob Mensching
	http://blogs.msdn.com/robmen
	http://wix.sourceforge.net