Thanks Nathan,  I checked in a text editor for old channel/layer names, and
there is one very basic gizmo in use (unless you can count gizmos copied to
groups).  Again, there are no clones.  Anyway we seem to have moved on but
maybe I can forward the script to the Foundry.

Howard, I did start the copy/paste thing but not exhaustively.

John, the duplicated comp trees are very modest and I think I only have
about 25 reads so I don' t think it's the "to many open files" issue.  I
tend to only hit that when I've had hundreds of Reads.



On 21 August 2011 14:07, Nathan Rusch <[email protected]> wrote:

>   The causes I’ve seen for the blank error message behavior when the
> script file is otherwise intact are bad clones and a few Gizmo-related bugs.
> Check any Gizmos in the script file for references to non-existent or
> non-default layer/channel names (the most common occurrence I’ve seen is a
> Gizmo with a linked ChannelMask knob that has a value of {rgba.red
> rgba.green rgba.blue rgba.alpha} or something similar that shouldn’t be
> saved in the script). Typically deleting the values and letting the knob
> revert to its default state will let the script load normally. Also,
> checking for Gizmo name collisions is probably worth a shot as well.
>
> Just some more ideas... Good luck.
>
> -Nathan
>
>
>  *From:* Howard Jones <[email protected]>
> *Sent:* Sunday, August 21, 2011 1:49 PM
> *To:* Nuke user discussion <[email protected]>
> *Subject:* Re: [Nuke-users] Re: Disappearing nodes on script re-open
>
>  try copy and paste in sections then til you find the node(s) thats
> causing it?
>
> Howard
>
>   ------------------------------
> *From:* Michael Garrett <[email protected]>
> *To:* Nuke user discussion <[email protected]>
> *Sent:* Sunday, 21 August 2011, 20:36
> *Subject:* Re: [Nuke-users] Re: Disappearing nodes on script re-open
>
> Thanks Ken, that is all very useful information.  The script file is intact
> because it just so happens the most recent version was left open for a local
> gui render the night before but if we copy/paste from that into a new script
> then we get the missing nodes.
>
>
> On 21 August 2011 12:18, Ken Littleton <[email protected]> wrote:
>
> The leading cause of truncated scripts is saving the script file to a
> overloaded network server.  Either Nuke quits before the script is saved or
> more likely, Nuke is killed by the user thinking it has hung.  This can
> happen locally as well if the machine is out of memory.
>
> I would recommend changing the autosave path to a local path.  This will
> speed up autosaves and act as a local repository in case the server screws
> you.  Another safety net is to have a copy of the script saved to the output
> folder when submitted to a render farm.
>
> I haven't seen the clone issue in quite a while.  Take a look at the script
> with a text editor, if it's truncated, it was a save issue.  If it's
> complete to the Viewer node, then the script has an issue.
>
>
> On 8/21/2011 11:33 AM, Michael Garrett wrote:
>
>  A bit more info:  in this case, there are four shots in the one script
> with four very similar node trees for each shot, and the last tree is
> completely gone except for the backdrop nodes.  Of the first three, they all
> have deleted nodes at different points in the tree even though the tree is
> very similar.
>
> We have managed to salvage stuff to the point where we are mostly OK but
> obviously we would like to know what happened.
>
> On 21 August 2011 11:27, Michael Garrett <[email protected]> wrote:
>
> Excuse my poorly worded email, I'm still waking up here.  There was only
> one kind of "blank error" message that is occasionally appearing on script
> open.  but some of the nodes are gone regardless.
>
>
> On 21 August 2011 11:19, Michael Garrett <[email protected]> wrote:
>
> I'm getting the random blank error message when re-opening a node graph
> from last night and some nodes have disappeared.  I'm also occasionally
> getting the "blank error" message on script open.
>
> There are no gizmo changes overnight or cloned nodes that can typically
> cause this kind of issue. Historically, merging the script or deleting the
> Viewers in a text editor could help  but there are other solutions I can't
> remember.
>
> Is the best way to fix this issue just to comb through the script in a text
> editor and copy/paste sections in?  Any magic bullets?  We're in 6.2v4
>
> Thanks,
> Michael
>
>
>
>
>
>
> _______________________________________________
> Nuke-users mailing [email protected], 
> http://forums.thefoundry.co.uk/http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>
>
>
> _______________________________________________
> Nuke-users mailing list
> [email protected], http://forums.thefoundry.co.uk/
> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>
>
>
> _______________________________________________
> Nuke-users mailing list
> [email protected], http://forums.thefoundry.co.uk/
> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>
>  ------------------------------
> _______________________________________________
> Nuke-users mailing list
> [email protected], http://forums.thefoundry.co.uk/
> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>
> _______________________________________________
> Nuke-users mailing list
> [email protected], http://forums.thefoundry.co.uk/
> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users
>
_______________________________________________
Nuke-users mailing list
[email protected], http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-users

Reply via email to