tkzinc problem

[Mike Scott]

I'm having a really weird problem building a working tkzinc (a useful
but regrettably obscure, seemingly unsupported module :-( ) under linux
mint.

I've been unable to find a solution from the mint newsgroup, and
wondered if there might be people here who'd be able to help -- or
suggest a more appropriate forum. Does anyone else even use this module?
The tkzinc website isn't very helpful, being seemingly more interested
in dumpsters :-)

I'll post the symptoms if anyone says 'yes'!


Thanks.


--
Mike Scott (unet2 <at> [deletethis] scottsonline.org.uk)
Harlow Essex England

[Mike Scott]

On 22/12/12 21:17, Roger Burton West wrote:

> On 2012-12-22, Mike Scott wrote:
>> I'm having a really weird problem building a working tkzinc (a useful
>> but regrettably obscure, seemingly unsupported module :-( ) under linux
>> mint.
>> I've been unable to find a solution from the mint newsgroup, and
>> wondered if there might be people here who'd be able to help -- or
>> suggest a more appropriate forum. Does anyone else even use this module?
>> The tkzinc website isn't very helpful, being seemingly more interested
>> in dumpsters :-)
>

> I've not used it, but it looks awfully handy for a project I have in
> mind (where I was planning to, and still may, use Tk::Canvas).
>
> So I am at least potentially interested...
>
> R
>

Well, OK. I'll say at the start I'm using zinc mainly because it offers
transparency and IIRC better scaling, as well as being AFAICT rather
faster than canvas.

Anyway, the nitty gritty of my problem is that I built the thing
(tkzinc, that is; on a 64-bit mint box) some weeks ago while developing
some code to use it. It's been working just fine.

Then I tried to move my code to two other machines. On /both/ of those,
one being an older 32-bit mint-11, the other a second 64-bit mint 13, it
fails the perl test suite - anything that needs to use GLX's render
fails with an X11 BadMatch and invalid opcode RENDER. (That includes my
transparency need, and text rescaling. Ouch!) Otherwise it seems to work.

The real killer is that now, I cannot rebuild it on the original machine
- where I started and where all used to be well. (I did keep the
original library Zinc.so - just as well; if I copy this to the second
64-bit box, it's OK. No help for the 32-bit box though :-( ). I have no
idea why - I've not changed any packages in the interim, although mint
does tend to upgrade things occasionally.

The interesting thing is that my original working Zinc.so is linked
against libGLEW. The newly linked failing ones are not, otherwise ldd
reports the same list of linked libraries.

The tkzinc website is wierd, and most links seem not to work. There
doesn't seem to be any ongoing maintenance.


Anyway, I'd love to know if anyone else can build this - it's the usual
perl incantation
perl Makefile.PL
make && make test

at which point errors scream from the screen, at least now and for me.



Thanks for the interest.

[Mike Scott]

On 23/12/12 08:17, Mike Scott wrote:
....

I have a more than sneaky feeling this problem arises from an update to
the relevant mesa libraries - one seems to have been installed after I
originally built the 64-bit zinc, and before I tried installing zinc on
the 32-bit lappie.

Unfortunately, it doesn't seem too easy to revert to the older version
to check - trying to force a downgrade would cause too many dependencies
elsewhere to break.

So, I've decided to give up with zinc and use a straight Canvas. A
shame, because tkzinc was fast and flexible compared to Tk::Canvas.
OTOH, tkzinc was also broken for Windows, and portability is of some
value for this code.