Hi list,
I've just recovered a broken image that crashed upon startup. There was some problem with a SerialPort that stayed open and tried to write/read into a physical port that didn't exist no more, or something of the sort.
So here's the magic line: $ bin/myVm shared/MyImage.image eval "SmalltalkImage backupTo: 'recover'" This makes a copy of the image into a new image called "recover.image", but as the documentation states:
So there you go!
Hope it helps someone else. :) Bernat Romagosa. |
Hi Bernat,
Interesting, I never heard of #backupTo: But if recover.image is a copy of MyImage.image, and that last one failed to start up, why then would the copy not have the same problem ? Sven On 24 Oct 2013, at 13:32, Bernat Romagosa <[hidden email]> wrote: > Hi list, > > I've just recovered a broken image that crashed upon startup. There was some problem with a SerialPort that stayed open and tried to write/read into a physical port that didn't exist no more, or something of the sort. > > So here's the magic line: > > $ bin/myVm shared/MyImage.image eval "SmalltalkImage backupTo: 'recover'" > > This makes a copy of the image into a new image called "recover.image", but as the documentation states: > > Unlike #saveAs: do not transfer the default execution to the new image. > > So there you go! > > Hope it helps someone else. :) > > -- > Bernat Romagosa. |
Because: ”Unlike #saveAs: do not transfer the default execution to the new image." So the startup and shutdown scripts won't be executed :) (Or that's what I understood) On Oct 24, 2013 1:54 PM, "Sven Van Caekenberghe" <[hidden email]> wrote:
Hi Bernat, |
Yes, I understand - I read that, but the next time you start your new recover.image they will run again, no ? Resulting in the same problem as the original image, or not ?
On 24 Oct 2013, at 14:34, Bernat Romagosa <[hidden email]> wrote: > Because: > > ”Unlike #saveAs: do not transfer the default execution to the new image." > > So the startup and shutdown scripts won't be executed :) > > (Or that's what I understood) > On Oct 24, 2013 1:54 PM, "Sven Van Caekenberghe" <[hidden email]> wrote: > Hi Bernat, > > Interesting, I never heard of #backupTo: > > But if recover.image is a copy of MyImage.image, and that last one failed to start up, why then would the copy not have the same problem ? > > Sven > > On 24 Oct 2013, at 13:32, Bernat Romagosa <[hidden email]> wrote: > > > Hi list, > > > > I've just recovered a broken image that crashed upon startup. There was some problem with a SerialPort that stayed open and tried to write/read into a physical port that didn't exist no more, or something of the sort. > > > > So here's the magic line: > > > > $ bin/myVm shared/MyImage.image eval "SmalltalkImage backupTo: 'recover'" > > > > This makes a copy of the image into a new image called "recover.image", but as the documentation states: > > > > Unlike #saveAs: do not transfer the default execution to the new image. > > > > So there you go! > > > > Hope it helps someone else. :) > > > > -- > > Bernat Romagosa. > > |
Yes, but you'll have the opportunity to fix whatever broke it before saving and opening the image again. 2013/10/24 Sven Van Caekenberghe <[hidden email]> Yes, I understand - I read that, but the next time you start your new recover.image they will run again, no ? Resulting in the same problem as the original image, or not ? Bernat Romagosa. |
Free forum by Nabble | Edit this page |