Vectelopment system is not an instant thing.
It have been changed several times.
This history may help you to understand some
inconsistency and bugs.

THE USABILITY POOL
------------------
It was started by the pool, 
asking what is the most expected VL features.

VASM2
-----
VASM2 was started, to address some issue.
The most important thing was to experiment with
better navigation (consistent back, ok, cancel,
menu, etc).
After tried several options, dialog+XDialog
remains the best tool-kit.
The new framework for better navigation
uses a recursive wizard system.

  
VSETUP
------
The VSETUP begun to port the successfull wizard
system from VASM. I started from the isolinux of VL 4.0.
Open the initrd.img and start to explore the fascinating
installation process.

Wow, I was so impressed, and eager to hack it.
Fisrt, port the new wizard system.
Alas, the dialog program on isolinux is so primitive !!!
It does not understand --ok-label, etc.
I tried to bring a newer dialog, it could not work,
since incompatible gcc library.
So, I borrowed isolinux from Uncle Slackware.
Dang, it does not have parted.

So, stay with VL 4.0 isolinux. Accept ESC key for BACK.
Add a new utility called probepart to accompany the probedisk.

Several things was improved according to the POOL:
- Make friendlier partition resizer, and more informative cfdisk
- to do list before begin. Community want side information,
  but dialog cannot do that. When trying to make side information,
  I made the flying CREDIT instead.
- capable to use / and /home. Easy to select partition.
  But DUMMY was added since I accidentally formatted my own HOME !!!
  I lost 3 months data there :(
- Then come the biggest trick, packages selection.
  What is the best BULK or PACKAGES ?
  Combination is the best. So, lets start thinking about 
  bulk, packages, and ISO.
  Hey wait a minute. We have VL standard and SOHO.
  How about integraring them too ?
  

THE-PROPOSAL
------------
So the propsal about integrated build directory was launched.
Partly, because I was so dissapointed with SOHO development.

SOHO was started November 2003. I thought it was a piece of cake.
What is so difficult about adding KDE + Open Office on top of
well running VL standard ?
Instead, SOHO was trying to catch the newest thing, 
and abandon the best foundation of VL standard.
The result was only a release candidate. Too bad.

To avoid such a tragedy again, an integrated build directory 
should be used.
VL standard, soho and all flavours from the same version
MUST use the same packages ! Period.


VECTELOPER & VECHIEF
--------------------
What can I say.
I just made it, 
and produce the vl-gx and then vl-dynamite a long the way.
In the end, it is like I have an Asian food store.
What flavours do you like ?
It will be ready in 1 or 2 weeks.


PACKAGES AND DEPENDENCY
-----------------------
The main respons about the proposal was dependency.
A response noticed me about two approaches:
- layered
- network

LAYERED APPROACH
----------------
To adopt this approach, the packages directory
was changed from FLAT structure (like Slackware)
to TREE structure.
Thus bigger picture of dependency can be shown by the
position of the packages in the tree.
If you still found FLAT structure somewhere, thats a legacy.


NETWORK APPROACH
----------------
Slackware has swaret to do this.
I read the swaret source code, and realized the power of ldd,
but also its complicated dependency information using
MANIFEST, etc.

Given that Vector Linux uses bulk install for its CORE system,
full on-the-fly upgrade is imposible.
Thus, we can make a simpler system by limiting the 
system to work for the packages of the same MAJOR version.
The ldd is used to solve dependency during package compilation,
then the dependency information is added to each package
(not centralized like swaret or apt-get).
The installer (vecpkg) will use the information during the download.


VECPKG2.
--------
Thus begin the making of vecpkg2.
There are two aims.
Capable to use the tree structure,
and able to do the dependency things.
The first goal has been achieved.
The second goal is not yet. 
(If still nothing after this, then nobody has done it).


.........