Virtual Registry and File system?

  • Thread starter Thread starter wally
  • Start date Start date
W

wally

There is a brilliant application that allows you to wrap your EXE and all
associated DLLs, OCXs, etc. into a single executable and run the executable
on Windows OSs with no install and nor registration of the files. It will
even include only the portions of the .Net framework that are required for
your application.

The app is called Thinstall. the problem is the cost. $4,000 license fee
PER APPLICATION PLUS A PER PC LICENSE FEE for the "virtual OS" on each
client PC that runs your application. This is clearly ridiculous. I bought
the program when it was only $795 for unlimited use and I thought that was a
bit steep for most developers.

I and want to build a cheaper version of this application. I would
appreciate any links to source code or theory on virtual registries and
virtual file systems (as these seem core to the success of the application).
Maybe we can make this an open source project to eliminate the need for
installations altogether and ease the frustrations of sys admins and
software shops everywhere.

Thanks for your help!
 
wally said:
There is a brilliant application that allows you to wrap your EXE and all
associated DLLs, OCXs, etc. into a single executable and run the
executable on Windows OSs with no install and nor registration of the
files.
Click to expand...

Why on earth would you want to pack an executable into one file together
with dynamically linked libraries instead of just linking statically?
 
[Warning: I may wander OT at times.]

wally said:
There is a brilliant application that allows you to wrap your EXE and all
associated DLLs, OCXs, etc. into a single executable and run the executable
on Windows OSs with no install and nor registration of the files. It will
even include only the portions of the .Net framework that are required for
your application.
Click to expand...

I hesitate to call this idea good, much less brilliant. But I'll move on...
The app is called Thinstall. the problem is the cost. $4,000 license fee
PER APPLICATION PLUS A PER PC LICENSE FEE for the "virtual OS" on each
client PC that runs your application. This is clearly ridiculous. I bought
the program when it was only $795 for unlimited use and I thought that was a
bit steep for most developers.
Click to expand...

It may not be so ridiculous. Niche-market software can be incredibly
expensive. Otherwise, the developers wouldn't recover their costs. I once had
the experience of reviewing a particular piece of smart motion detection
software for digital surveillance systems. It was $25,000 plus $5,000 per
simultaneous camera feed. It was considered a moderately-priced smart motion
detection system at the time.

In more common software, the industry-standard AutoCAD computer-aided design
software can easily run over $4,000 for a full version, because that's what
the market will pay.
I and want to build a cheaper version of this application. I would
appreciate any links to source code or theory on virtual registries and
virtual file systems (as these seem core to the success of the application).
Maybe we can make this an open source project to eliminate the need for
installations altogether and ease the frustrations of sys admins and
software shops everywhere.
Click to expand...

Careful. Check for applicable patents first. A piece of software at the
$4,000+ level probably has at least one and maybe several patents protecting
it. Even if you go open source, you can't violate existing patents in doing
so. Consult a good patent attorney to find out more.

Sean
 
Sean M. DonCarlos said:
[Warning: I may wander OT at times.]

wally said:
There is a brilliant application that allows you to wrap your EXE and all
associated DLLs, OCXs, etc. into a single executable and run the
executable
on Windows OSs with no install and nor registration of the files. It
will
even include only the portions of the .Net framework that are required
for
your application.
Click to expand...

I hesitate to call this idea good, much less brilliant. But I'll move
on...
The app is called Thinstall. the problem is the cost. $4,000 license
fee
PER APPLICATION PLUS A PER PC LICENSE FEE for the "virtual OS" on each
client PC that runs your application. This is clearly ridiculous. I
bought
the program when it was only $795 for unlimited use and I thought that
was a
bit steep for most developers.
Click to expand...

It may not be so ridiculous. Niche-market software can be incredibly
expensive. Otherwise, the developers wouldn't recover their costs. I once
had
the experience of reviewing a particular piece of smart motion detection
software for digital surveillance systems. It was $25,000 plus $5,000 per
simultaneous camera feed. It was considered a moderately-priced smart
motion
detection system at the time.

In more common software, the industry-standard AutoCAD computer-aided
design
software can easily run over $4,000 for a full version, because that's
what
the market will pay.
Click to expand...

IMHO, the market will pay those prices only as long as there is no
alternative. When alternatives are brought to market, the original widgets
generally drop substantially in price. That's what we need here.....an
alternative.

Xenocode is trying to do a .Net version of Thinstall, but it is not nearly
as robust and cannot currently wrap 3rd party controls or components with
your .Net code - so, IMHO, it is very limited.

And, while the original effort to make such software may have indeed been
substantial for the original writer, this may have been due to his/her/their
lack of programming knowledge, OS knowledge or other reasons. So, the
effort required by the original software designer is not necc the same that
will be required by the second widget designer. And, the ideas and
approaches of the original designer will not be exactly those of the second
writer.
Careful. Check for applicable patents first. A piece of software at the
$4,000+ level probably has at least one and maybe several patents
protecting
it. Even if you go open source, you can't violate existing patents in
doing
so. Consult a good patent attorney to find out more.

Sean
Click to expand...

Patents that cover software are hard to prove. If I come up with a new way
to do something independent of the source code of their apps, it may still
be covered by patents, but patents are easily nullified. To nullify a
patent, your new widget must show new capabilities that would not be readily
apparent if one were to look at the old widget. Your widget must show some
type of innovation.

This innovation may require some thought....but thinking is free and
Thinstall is not.

And, if need be, I can have the code designed in Europe - where software
patents have, for the most part, been tossed aside.

Besides, they'll let me know if I'm doing something wrong. I feel most sure
of it.
 
Patents that cover software are hard to prove. If I come up with a new
way to do something independent of the source code of their apps, it may
still be covered by patents, but patents are easily nullified. To nullify
a patent, your new widget must show new capabilities that would not be
readily apparent if one were to look at the old widget. Your widget must
show some type of innovation.
Click to expand...

And, as far as innovation goes, Thinstall is not innovative in and of
itself. It is linking. Something compilers have done for years. And, it
adds a virtual registry and virtual file system....also things done long
before thinstall. And, just how innovative is the idea of compressing or
even encrypting an executable?

Are not all of these things things that a competent programmer could
conceive of combining in one form or another? There is no special
innovation in Thinstall, only the gathering of different existing
technologies into a single application.

I don't think it would stand a chance in a patent lawsuit. IMHO, the
patents it claims are simply paper tigers.
 
OT? Programming virtual registries and file systems using C++ is OT in
comp.lang.c++?
Click to expand...

Yes, because comp.lang.c++ is about language questions and issues, not about
windows programming questions.

--

Kind regards,
Bruno van Dooren
(e-mail address removed)
Remove only "_nos_pam"
 
IMHO, the market will pay those prices only as long as there is no
alternative. When alternatives are brought to market, the original
widgets generally drop substantially in price. That's what we need
here.....an alternative.
Click to expand...

unless the software is wildly popular (like office) any alternatives are
usually inferior to the real product.
the reason is that most open source developers would rather work on exciting
new features than
spending days and weeks ironing out all the little bugs.
Xenocode is trying to do a .Net version of Thinstall, but it is not nearly
as robust and cannot currently wrap 3rd party controls or components with
your .Net code - so, IMHO, it is very limited.

And, while the original effort to make such software may have indeed been
substantial for the original writer, this may have been due to
his/her/their lack of programming knowledge, OS knowledge or other
reasons. So, the effort required by the original software designer is not
necc the same that will be required by the second widget designer. And,
the ideas and approaches of the original designer will not be exactly
those of the second writer.
Click to expand...

Some things are just very difficult, or can take a lot of time to get right.
commercially developing a file system driver for example costs several
hunder thousands of dollars.
ditto for things like VMWare.
the same goes for the software in all kinds of spectrum analysers and other
devices.
this has nothing to do with limitation of the programmer, but with the
enormity of the task.

maybe you can look at the sources for WINE on linux?
it seems to cover at least some of the things you are interested in.
for virutal file systems you might want to check out an open source virtual
machine (bochs?)

the sys admins I know gladly pay thousands of dollars for high quality apps,
instead of using cheaper buggy apps.
for some systems, downtime is so expensive that 10000$ is peanuts.
Patents that cover software are hard to prove. If I come up with a new
way to do something independent of the source code of their apps, it may
still be covered by patents, but patents are easily nullified. To nullify
a patent, your new widget must show new capabilities that would not be
readily apparent if one were to look at the old widget. Your widget must
show some type of innovation.
Click to expand...

even if you think the patent does not apply, a large company can simply
spend
you to death before you even reach the court.
This innovation may require some thought....but thinking is free and
Thinstall is not.

And, if need be, I can have the code designed in Europe - where software
patents have, for the most part, been tossed aside.

Besides, they'll let me know if I'm doing something wrong. I feel most
sure of it.
Click to expand...

No. they'll wait until you have somthing going on and then they claim
license fees.

really, I don't want to discourage you, but this will be a very difficult
task technically,
and possibly you have to do some legal work as well.

good luck.

--

Kind regards,
Bruno van Dooren
(e-mail address removed)
Remove only "_nos_pam"
 
Back
Top