On Sep 4, 2013, at 12:10 PM, Norbert Hartl <norbert@hartl.name> wrote:
Am 04.09.2013 um 11:43 schrieb Esteban Lorenzano <estebanlm@gmail.com>:
On Sep 4, 2013, at 10:54 AM, Sven Van Caekenberghe <sven@stfx.eu> wrote:
On 04 Sep 2013, at 10:47, Norbert Hartl <norbert@hartl.name> wrote:
In all my other projects using zinc and RFB I didn't see the problem. But then I don't save images in production.
Images are very cool and very useful, but saving images in production is a no go, it is asking for trouble, IMHO.
yes, but preparing an image for production often includes opening RFB and save it in that state :)
It is a good thing to avoid that. Why? Because it is not automatically reproducible. If it is it might be better to put it in a start script. I let my images build by CI which I copy then on the target machine (in a script of course :) ). Then there are start scripts that e.g. start the RFB server. And those things always works until I save the image and change this way the prerequisites of the process. Or did you mean something different?
in practice is the same. you have a "preparation script" that opens a RFB and saves the image. If you say that doing it that way is not good, then I disagree: pharo is image based and if we do not keep the power of having a "ready to start" image (in production too), then we are losing something important. ... and we are transforming a necessity (not doing something because is buggy) into a virtue... and that's just plain wrong :) but I think we are now talking of different things :)
Norbert