Modularised engine, Metacard, and other daft thoughts.

[Richard Gaskin]

Richmond wrote:

 > If the open source engine is to be modularised would it be possible
 > to lift it out of the Open Source version of Livecode (this naming
 > is getting clunky: there needs to be another name for this)
 > and drop it into a Metacard shell?

Any IDE that runs on the LiveCode engine -- whether LiveCode's, 
MetaCard, or anything else anyone chooses to write -- is just a 
collection of stacks.

It's always been possible for there to be any number of IDEs, and I see 
no reason why this couldn't continue regardless of any new licensing 
options.

The MC IDE was released as open source after the acquisition of the 
engine by RunRev Ltd. in 2003, under the very permissive X11 ("MIT") 
license.

Since the X11 license makes source available, the Free Software 
Foundation considers it compatible with GPL3:
<http://www.gnu.org/licenses/license-list.html#GPLCompatibleLicenses>

Given the relative ease with which one can make IDEs in LiveCode, if it 
can become open source the community will grow large enough that we may 
see a wide range of alternate IDEs available.



 > If the state of the current use-list is anything to go on, we will
 > have a Mao-like flowering; let's just hope the powers that be then
 > don't start breaking people's legs and sending them off to collective
 > farms in Tibet.

Funny you should use the Mao reference:  in the olden days, in regard to 
alternate IDEs Dr. Raney used to say, "Let a thousand flowers bloom."

RunRev has always allowed alternative IDEs, since they all require a 
licensed engine to run them.  In fact, changes to the engine in 2007 
made it much easier to do so.

The Community Edition makes this even easier, and I would imagine RunRev 
would have no problem with those that adhere to the GPL license 
requirements.

That said, while it's *relatively* easy to write IDE stacks for the 
LiveCode engine (compared to work on the engine itself), doing a 
complete job is still a non-trivial task.

As much as we've enjoyed MetaCard over the years, Ken Ray and I have had 
to put it into maintenance mode, because RunRev has been adding features 
to the engine faster than we can affordably build GUI tools to support 
them, migrating most of our efforts instead to providing tools for use 
withing the LiveCode IDE.

MC remains open and available, and anyone with sufficient time and 
interest is welcome to contribute.  But historically such contributions 
have been few, and Ken, Klaus, Jacque, and the others who've worked on 
it have had to give priority to our client projects.

If the Kickstarter campaign is successful, the resulting growth in the 
community will likely bring many eager adventurous souls on board, some 
of which may become enamored enough of the MC IDE to take it to a whole 
new level, or fork it into a radically new IDE.

All such options are available to those with time and interest, and can 
be done under either the proprietary license as we've been doing for 
years, or under the new GPL license if it's a better fit for the 
project's goals.



 > How, for instance, would we stop Mr/Ms X from taking some components
 > of the commercial variant, mixing them with some components of the
 > O-S variant and peddling the resultant chimaera?
 >
 > I wonder how, exactly, RunRev are going to police the situation, and
 > in the event of "naughty things" happening, enforce the law.

AFAIK the proposed proprietary license retains many of the restrictions 
currently in place, so distributing a fork of the proprietary version 
would seem unlikely without severe repercussions.  Indeed, the proceeds 
resulting from such a suit may be quite beneficial to RunRev. ;)

The GPL version does lend itself to forking.  This is an inherent 
feature of GPL (part of Freedom 3), and we've seen this with MariaDB 
being forked from MySQL, LibreOffice forked from OpenOffice, and others.

But those two examples illustrate why forking such complex projects 
occurs:  both MariaDB and LibraOffice were the result of broken trust 
from Oracle's management of those projects.

Forking such a substantial code base is costly, no more trivial than 
RunRev's managing of the original branch, as the cost outlays described 
on the Kickstarter page make clear.  Orgs only spend that much money as 
a last resort, when all other attempts to correct disagreements over 
management fail.

MariaDB and LibreOffice have earned the backing of many large 
organizations because they felt the original project's management could 
no longer be trusted to fulfill the projects' goals.

If RunRev's management of LiveCode were to become like that of Oracle's, 
we could expect the community to respond similarly.

But thankfully we have no reason to believe that the company will take 
such a turn.

And all the while, Oracle retains full ownership of the proprietary 
license for MySQL, just as RunRev would remain the only place 
professional devs could obtain a proprietary license for LiveCode.



Could it be distracting to see forks of the GPL version?

Indeed it would, but FOSS communities tend to frown on needless forking 
for that reason, or at least just ignore such forks.

The Ubuntu project may provide a good example of both useful and useless 
forking:

Ubuntu was formed as an extension of the Debian project because Ubuntu 
seeks to pursue a different goal, emphasizing usability for the 
individual user, while Debian's goal emphasizes stability for organizations.

Debian has a slow release cycle, and is the world's most popular 
solution for servers.  Ubuntu releases semi-annually, and is the world's 
most popular distro for desktop use.

Over the years Ubuntu has been forked dozens of times, a few rather 
popular (like Mint), and most with almost no users at all (like Hanna 
Montana Linux 
<http://desktoplinuxreviews.com/2009/07/27/hannah-montana-linux/>, and 
Ubuntu Satanic Edition 
<http://news.softpedia.com/news/Meet-Ubuntu-Satanic-Edition-Version-666-311993.shtml>).

But none have Ubuntu's 23-million-and-growing user base, and none have 
Canonical's OEM bundling agreements that make it possible to walk into 
computer stores from Italy to India and buy a computer with Ubuntu 
pre-installed, because it takes a fair bit of capital to pull that off.

Similarly, I would expect LiveCode's Community Edition to be forked, but 
I would also expect RunRev's version to remain the standard by far, and 
the rest being fringe projects, most dying off over time as most of the 
contributions continue to focus around RunRev.

--
  Richard Gaskin
  Fourth World
  LiveCode training and consulting: http://www.fourthworld.com
  Webzine for LiveCode developers: http://www.LiveCodeJournal.com
  Follow me on Twitter:  http://twitter.com/FourthWorldSys