Has anyone else noticed that when deplying using Lagoon, on the status line
of the deplyment dialog, as the classes flash by that are being removed, there are a couple of DNU #onClassRemoved messages? (e.g. ClassHierarchyModel doesNotUnderstand: #onClassRemoved). My .exe seems to be created just fine but I was wondering it that is normal? Tim |
"Tim M" <[hidden email]> wrote in message
news:[hidden email]... > Has anyone else noticed that when deplying using Lagoon, on the status > line of the deplyment dialog, as the classes flash by that are being > removed, there are a couple of DNU #onClassRemoved messages? (e.g. > ClassHierarchyModel doesNotUnderstand: #onClassRemoved). > > My .exe seems to be created just fine but I was wondering it that is > normal? > Sometimes you will get this because it takes a while for the event registrations for dead ClassHierarchyModels (or any other object for that matter) to get cleaned up. These rely on weak collections getting updated so it might cause some of them to hang around after the classes themselves have actually been removed. Generally speaking, however, you should not see this if you deploy from a clean image in which you have not been using development tools during the session. I would tend to take this as an indication to check the deployment log to make sure that the development system is being removed cleanly. A good way to view the deployment log is using the Executable Browser (aka Environment Browser). As an aside, the #browseUnimplemented command of this (which got lost from the menus when updating all browser to include a common set of commands, and which you must therefore issue by opening an inspector on the browser using a Visual Object Finder so that you can send it #browseUnimplemented) is very useful for identifying over zealous stripping, although quite often you can identify such issues before deployment using the same command in the package browser, although you have to be able to identify the false positives which takes a little experience. Regards OA |
> Generally speaking, however, you should not see this if you deploy
> from a clean image in which you have not been using development tools That explains it - it hasn't been a clean image (I'm still working on automating that bit ;-) > As an aside, the #browseUnimplemented > command of this (which got lost from the menus when updating all > browser to include a common set of commands, and which you must > therefore issue by opening an inspector on the browser using a Visual > Object Finder so that you can send it #browseUnimplemented) is very > useful for identifying over zealous stripping I forgot all about that menu option - it took a bit of fiddling with the ObjectFinder to see what you mean. It seems its available in the Package browser and the Executable browser - you have to click on the right pane and then look at the parent presenters sub presenters - and find a PackageSelector or an EnvironemntBrowser shell. As I was packaging a Swazoo app and notice a few strange odditities in the latest Swazoo port. Tim |
"Tim M" <[hidden email]> wrote in message
news:[hidden email]... >> Generally speaking, however, you should not see this if you deploy >> from a clean image in which you have not been using development tools > > That explains it - it hasn't been a clean image (I'm still working on > automating that bit ;-) > >> As an aside, the #browseUnimplemented >> command of this (which got lost from the menus when updating all >> browser to include a common set of commands, and which you must >> therefore issue by opening an inspector on the browser using a Visual >> Object Finder so that you can send it #browseUnimplemented) is very >> useful for identifying over zealous stripping > > I forgot all about that menu option - it took a bit of fiddling with the > ObjectFinder to see what you mean. It seems its available in the Package > browser and the Executable browser - you have to click on the right pane > and then look at the parent presenters sub presenters - and find a > PackageSelector or an EnvironemntBrowser shell. > It shouldn't be that hard to locate. It's implemented directly on the shell, so just using the object finder and clicking over the caption of the browser will bring up an inspector you can use to send it. Probably easiest, though, to just edit the Executable Browser menu bar to put it back. Just add a command an appropriately named that sends #browseUnimplemented. It should already be on the package browser menus, unless that is something we've added back since 6.02. Regards OA |
Hi Support at Object Arts,
> It shouldn't be that hard to locate. It's implemented directly on the > shell, so just using the object finder and clicking over the caption > of the browser will bring up an inspector you can use to send it. As I pretty much use an Idea-Space in D6, I'm not sure how to use the Object Finder on the caption, as the caption is for the entire Idea space? Hence my convulated way around - and yes you are right, I'll add it to the menu bar by editting the view, its much simpler ;-) |
"Tim M" <[hidden email]> wrote in message
news:[hidden email]... > Hi Support at Object Arts, > >> It shouldn't be that hard to locate. It's implemented directly on the >> shell, so just using the object finder and clicking over the caption >> of the browser will bring up an inspector you can use to send it. > > > As I pretty much use an Idea-Space in D6, I'm not sure how to use the > Object Finder on the caption, as the caption is for the entire Idea space? > Hence my convulated way around - and yes you are right, I'll add it to the > menu bar by editting the view, its much simpler ;-) > > For future reference, just click over the tab. Regards OA |
Free forum by Nabble | Edit this page |