making stand-alone Tcl app for QNX4

[Will Parsons]

I would like to convert a Tcl application (no Tk) into a stand-alone
executable for QNX4. In the past, I've created stand-alone executables for
Windows using both freewrap and tclkit, but have used binary packages for
both of these. I don't think binary packages for either is available for
QNX4, so I am looking for a solution I can compile on the target platform.
Looking at the info on its website, it looks to me if building freewrap for
QNX may not be too straightforward, and tclkit does not seem to have
source packages. Can anyone give any advice on what is the path most likely
to lead to success on this?

- Will

[Jeff Hobbs]

The full tclkit set of sources is open source. It is just Tcl
with a few extra modules (vfs, rechan, pwb and metakit). metakit
is C++ though ... don't know if that won't stump you on QNX.

--
Jeff Hobbs, The Tcl Guy
http://www.ActiveState.com/, a division of Sophos

[Schelte Bron]

On 01/11/06 02:16, Will Parsons wrote:
> tclkit does not seem to have source packages.

Instructions on how to build tclkit from scratch can be found at
http://www.equi4.com/218

Schelte.
--
set Reply-To [string map {nospam schelte} $header(From)]

[Will Parsons]

I don't have a C++ compiler on QNX, so that eliminates that possibility...

- Will

[Cameron Laird]

In article <slrnds9va...@L1422169.w-intra.net>,
Will Parsons <william....@us.westinghouse.com> wrote:
.
.

.
>I don't have a C++ compiler on QNX, so that eliminates that possibility...
>
>- Will

Nah. While I might agree that it's not worth pursuing,
it's surely more feasible than might first appear.
Jean-Claude uses a relatively restricted field of C++,
AND there are still cfronts (think of a C++ -> C
translator) around. I'll bet there's a satisfying way
to do this.

In any case, good luck, Will. I find better Tcl
capabilities under QNX exciting.

[Michael Schlenker]

Cameron Laird schrieb:

And maybe Tclkit-lite would be an option for Will, which is not using
C++ AFAIK.

Michael

[Cameron Laird]

In article <b6ff93...@lairds.us>, I confused matters by writing that:

Aiee! My response was misleading. I should have said this:
Tclkit does NOT have an absolute dependence on Metakit. For
many purposes, Tclkitlite, which does *not* depend on C++,
suffices. <URL: http://www.equi4.com/tclkitlite.html > pro-
vides details. Tclkitlite, by the way, should interest many
people beyond those focused on QNX; as the page just mentioned
highlights, it brings the base size of a (Linux x86, for
example) Tcl application down below 450K. That's impressive.

So, Will: do not let C++ get in your way.

Thanks to my private correspondent for the opportunity to
correct my mistake.

[Will Parsons]

Cameron Laird wrote:
> In article <b6ff93...@lairds.us>, I confused matters by writing that:
>>In article <slrnds9va...@L1422169.w-intra.net>,
>>Will Parsons <william....@us.westinghouse.com> wrote:
>> .
>> .

>>>I don't have a C++ compiler on QNX, so that eliminates that possibility...

<snip>

> Aiee! My response was misleading. I should have said this:
> Tclkit does NOT have an absolute dependence on Metakit. For
> many purposes, Tclkitlite, which does *not* depend on C++,
> suffices. <URL: http://www.equi4.com/tclkitlite.html > pro-
> vides details. Tclkitlite, by the way, should interest many
> people beyond those focused on QNX; as the page just mentioned
> highlights, it brings the base size of a (Linux x86, for
> example) Tcl application down below 450K. That's impressive.
>
> So, Will: do not let C++ get in your way.
>

Thanks - I'll try to see if I can get Tclkitlite to work.

- Will

[Dennis LaBelle]

Will Parsons wrote:

I've looked at building tclkit from the sources and I think that it is more
complicated than building freeWrap. That's one of the reasons I still
maintain the freeWrap application.

I'd be interested in knowing whether you still have the same opinion on
which is easier to build after you try the tclkit approach.

Alternatively, you could get me access to a QNX4 system (via the internet??)
and I could build freeWrap for you. (Darn, I just can't stop promoting this
stuff!)

Dennis LaBelle (The freeWrap Guy)

[Will Parsons]

Thanks for the offer, but my QNX development system is in a corporate
environment behind a firewall, so I cannot provide you access to it.

My first cut at building tclkit(lite) for QNX has run into a snag because
of a problem using exec (as described in another post "problems with exec
under QNX"). If I can't get tclkitlite to work, I'll probably give
freeWrap a shot. I'll report back if I have success.

(Oh, and thanks for freeWrap!)

- Will

[Cameron Laird]

In article <slrndsbatk...@isis.thalatta>,
Will Parsons <elleno...@yahoo.com> wrote:
.
.

.
>> Alternatively, you could get me access to a QNX4 system (via the internet??)
>> and I could build freeWrap for you. (Darn, I just can't stop promoting this
>> stuff!)
>>
>Thanks for the offer, but my QNX development system is in a corporate
>environment behind a firewall, so I cannot provide you access to it.

.
.
.
Heh-heh:
1. I'd really like to see Dennis get a chance at a QNX host.
2. If your organization is against such access, I shan't
argue.
3. If it's merely a technical question, though, don't let a
little firewall get in your way. QNX supports VNC, for
example, and there are well-documented ways to tunnel VNC
through all sorts of firewalls.

[Dennis LaBelle]

Cameron Laird wrote:

Yes, all I need is access to a plain QNX host (it doesn't need to have any
company secrets on it).

I imagine that this is sort of out-of-the box thinking for most companies.
Making a host like this available to open source authors (of the company's
choosing) who are willing to "lend a hand" would probably be very
beneficial.