Latest version of PI frozens after executing commands

[Astrocava]

Hola!

He instalado la última versión del PI. He sido incapaz de procesar porque tras ejecutar cualquier comando el ordenador se bloquea duranta unos minutos. Abrir archivos, mover las ventanas... causan un bloqueo general. Lo he instalado en dos ordenadores y tengo el mismo problema. He vuelto a la versión anterior del PI y ya no me ocurre.


Hi!

I have installed the last version of PI. I was unable to process anything because PI frozen after executing any command. Open file, move the image windows... cause general block. I have instaled PI in two computers and both have this problems. I have unistalled this version and come back to an earlier version and the problems have disspear.

Un saludo,

Sergio

[Juan Conejero]

Hola Sergio

Esto es muy extraño; que sepamos ningún otro usuario está teniendo el mismo problema.

¿Podrías indicarme las características del equipo en el que estás observando este comportamiento? También sería interesante que probaras la última versión de PI en otro ordenador.

Gracias

[andyschlei]

Carlos,

I am having a similar problem.  I am running on Windows XP Pro, Core duo processor, and I am running the latest version (1.00.58.377).

I opened up a JPG, and the system locks whenever I move a window.  I can execute commands, undo/redo, but whenever I move a window -- image or control -- the system locks.  I can unlock it with alt-ctrl-delete, which brings up the task manager and frees up the system.

It is interesting to note that the moved window freezes after moving about 40 pixels, but the final window position depends on where you finally release the mouse button.

--Andy

[Juan Conejero]

Hi Andy,

It seems that a very reduced number of users are having the same problem. It is an extremely odd regression in build 377 for Windows. I have some ideas about what can be, but prefer not to say until I'm completely sure (they have to do with products not written by us).

For all users that are having this problem, the only workaround that I know of (for now) is uninstall build 377 and return to the previous version. On the software distribution interface there are at least two previous versions available.

For users of trial versions, we have uploaded builds 376 and 375:

http://pixinsight.com/export/PI32-win-01.00.58.376-nsi-20080324-t.exe
http://pixinsight.com/export/PI32-win-01.00.58.375-nsi-20080321-t.exe

I apologize for the inconveniences. A new version is in preparation right now, which hopefully will not have these problems.

[andyschlei]

Juan,

No problem.  I keep several version in my download file anyway.  Just a pack rat habit on my part.

Strangely, it seems to work fine on my laptop, which had some of the same symptoms on the earlier version (376).  So the underlying problem may have started there.

I'll bet that this isn't a problem on your Linux version.  

Clear skies,

--Andy

[Juan Conejero]

I'll bet that this isn't a problem on your Linux version.

And you'll win. Build 377 for Windows also works nicely on XP and Vista as VMware virtual machines under Linux <g>

So far only four users have reported this problem (all fingers crossed). I suspect this is related to "something" in Qt 4.3.4. I am working now on a new version that should be ready in a couple days. Thank you for your patience and support!

[Tommy]

Juan -

I tried versions 376 and 375 and they both locked up on my Vista laptop. Going back to the 'shared' memory thought you had, I increase the paging size to the recommend Vista amounts to between 3.5GB and 5GB. This change in the system required a reboot. Therefore I did. When my machine got back up and running, a Vista warning window pop-up saying the software package is not compatible with Windows Aero Color scheme for the desktop appearance. And it automatically set it to Vista Basic. I then tried PI again, and it worked, all three versions. Even 377 is now running fine.

Here's the strange thing. I went back to check my memory paging settings, and they were back to the default settings. So that was not the problem. AND, I was able to set the appearance back to Windows Aero scheme and PI .377 still works. So I am thinking now, that it might of been a memory leak of some sort in my RAM, and the reboot cleared it to allow PI to use what RAM it needs.

Thanks for the support and I look forward to digging into PI's capabilities.

Tommy

[Astrocava]

Hola Sergio

Esto es muy extraño; que sepamos ningún otro usuario está teniendo el mismo problema.

¿Podrías indicarme las características del equipo en el que estás observando este comportamiento? También sería interesante que probaras la última versión de PI en otro ordenador.

Gracias

Perdona Juan, he estado muy liado y no he podido mirarlo.

Lo he usado en dos ordenadores, ambos con Windows XP profesional.

Portatil HP nx6110 Intel Pentium M 1,73GHz 504 MB Ram
Fijo: Core 2 Duo  6400 2.13GHz 2GB de Ram

[Borja]

Hola.

A mí también me pasa.
Ya me pasó hace días pero dejó de pasarme y no se había reproducido ... hasta ahora.
En mi caso (WIndows XP SP2 con IE7) lo que he podido comprobar que me bloquea el PI es la siguiente secuencia (no os riáis  :wink: )

- Abro el PI (v1.00.58.0376) y cargo un archivo (un TIF o CR2 de la Canon, es indiferente)
- A continuación, abro el Notepad o la calculadora de Windows (mediante Inicio-->Ejecutar)
- PI se bloquea inmediatamente  :shock:

A ver si lo podéis reproducir.
Si necesitáis alguna otra información o que compruebe algo me lo decís.

Un saludo
    Francisco

[David Serrano]

[English below]

- A continuación, abro el Notepad o la calculadora de Windows (mediante Inicio-->Ejecutar)

Hmm, eso es muy extraño...

Yo tengo cuelgues similares incluso sin ejecutar pixi. Hasta lo he desinstalado ya del güindons. Cabe decir que es vmware, y por eso no he dicho nada hasta ahora (me parece que es cosa de vmware).

Consigo "avanzar" a base de pulsar Ctrl-Alt-Del varias veces. Cuando finalmente aparece la ventana correspondiente (suele pasar a la segunda vez), la cierro y encuentro que todos los clics hechos previamente surten efecto de manera inmediata, como si el güindons tuviera una cola de eventos (que no sé si tiene) y se vaciara de repente.

Si esto no le pasa a nadie más, sabré que mi problema es otra cosa distinta.

En fins...

--

Hmm, that's very strange... (reproducing the problem by loading an image in PixInsight and then running notepad or calc from Start --> Run).

I have similar hangs even without pixi. More, it's even uninstalled from that windows machine. Worth mentioning that it's vmware, that's why I didn't tell anything until now (I blame vmware).

I get a little forward by doing Ctrl-Alt-Del a few times. When the window appears (which it usually happens at the second try), I close it just to see that all clicks and keystrokes I previously did are immediately interpreted, just as if windows had an event queue (I don't know if there's such thing) and it got emptied at that moment.

If this happens to nobody else, I'll know that this problem is different.

Anyway...

[Juan Conejero]

Hola

David, yo también corro PI en Windows XP SP2 sobre VMware Workstation, y todo funciona a la perfección. Estos problemas son realmente extraños; además, parece que se están produciendo desde hace dos o tres versiones, no antes. Puede tener algo que ver con Qt 4.3.4, pero lo realmente extraño es que tanto a Vicent con Vista Premium como a mí con XP SP2 (tanto sobre máquina real como en VMware) nos funciona perfectamente y jamás da ningún problema de este tipo.

Estoy terminando la nueva versión 1.1; mientras escribo esto se está compilando la aplicación principal Debería ser mucho más estable que 1.0, sobre todo porque toda la interfaz gráfica es código nuevo y no usa "cosas raras" heredadas de versiones viejas de Qt. Esperemos que todo esto desaparezca...

En fins...
Anyway...

Eso es lo que yo llamo "libertine translation"  :lol:

==================

Hello,

David, I run PI also on Windows XP SP2 under VMware Workstation, and all works perfectly. These problems are really odd; furthermore, they seem to happen since two or three versions, not before. It could have something to do with Qt 4.3.4, but what is really strange is that both Vicent and I, he with Vista Premium and me with XP SP2 (both on a real machine and under VMWare), are unable to reproduce these issues; we never have had any problem of this kind.

I'm finishing new version 1.1; as I'm writing this the core application is compiling It should be much more stable than 1.0, especially because the whole graphical interface is new code and doesn't use "rare things" inherited from old Qt versions. Let's hope that all of these issues will be gone...

[David Serrano]

Eso es lo que yo llamo "libertine translation"  :lol:

Pues es lo único que miré en el diccionario xD.

EN FINS xD.

[Borja]

Hola.

En la nueva versión no ocurre este problema.

Saludos

[Juan Conejero]

Bien!

Gracias por avisar Borja.

[Astrocava]

Comprobado: en la nueva versión no ocurre