Getting this error since the latest update of repository installer importer, long press on empty space – add item -…
Getting this error since the latest update of repository installer importer, long press on empty space – add item – app – this message gets displayed TrianguloY
D5 Creation
< ![CDATA[
The bug is caused because something is launching the script with “” as data.
Is it only shown when doing long click, add, app?
Is the bug shown when you open the app drawer?
I guess you have the ‘on Resumed’ event of the drawer, or one item, to that script (intentionally? )
I’ll add a check, so this gets ignored. But check the resumed events
]]>
< ![CDATA[
Yes it is shown when entering the app drawer now also, the only script I have in the app drawer is the dynamic calendar icons set to display on Google calendar and today widget.
]]>
< ![CDATA[
And paused events too.
]]>
< ![CDATA[
Yes, resumed was set to run a script, but somehow the script was erased from the script editor, this is a restored template so maybe it didn’t survive the backup and restore..
]]>
< ![CDATA[
Strange, this means the backup messed the scripts id (the repository script can’t change their own id, and it didn’t add itself to other events)
Pierre Hébert maybe you should check this
]]>
< ![CDATA[
TrianguloY waiting for an answer of Adam on his other post. If it is what I fear, RootTools library may be unreliable.
May not be lightnings fault
]]>
< ![CDATA[
I saw them after.
Sorry I can’t help with the root importer way 🙁
]]>
< ![CDATA[
Backup don’t change scripts identifiers, but template can, if merged. Was it the case ? If so this mean that somewhere a script id is translated so that it doesn’t conflict with another existing script, but not everywhere as it ought to be.
]]>
< ![CDATA[
Yes Pierre Hébert when I restore, I restore template first then the backup is this the correct procedure? I read one of your replies to a previous post a while back asking if just restoring a backup was sufficient to get a complete restore and you advised to restore the template also?
]]>
< ![CDATA[
Hum, I don’t remember having suggested to restore a template then a backup, but this could work although restoring the backup will probably be useless.
The issue with backup is that they leave widgets untouched: LLX don’t remove them and don’t create them. If widgets are removed between 2 backups they are lost.
The issue with templates is that it tries to recreate widgets, but Android doesn’t support this in all cases, so it may fail.
At the moment this is pretty confusing. I plan to fully rework this.
If you intend to backup/restore your setup to keep it safe, but don’t change your widgets then use the backup.
If your goal is to try various setups with various widgets, then prefer templates instead.
]]>
< ![CDATA[
But I am not sure, did you use the restore template feature with the merge option or without ? And if it was with the merge option, do you think the bug appeared precisely at this time ? Lukas mentioned RootTools, I didn’t follow the whole conversation so I may have missed something.
]]>
< ![CDATA[
No, I restore without merge, I only check “Replace. Fully replace the setup” . I may have misunderstood your response to the question about restoring maybe you just advised to restore the template rather than the backup? As regard to widgets restoring, for myself, when I restore a template all my widgets are restored perfectly, but the app drawer does not restore to how I’d left it previously, so when I then restore the backup afterwards the app drawer settings are then restored for example, my edited “My Drawer” but the widgets then need applying again. I posted a couple more screens which is why the conversation is fragmented, sorry about that, and my extra long comment.. 😄
]]>
< ![CDATA[
Pierre Hébert no, no (I should have delete my comment)
The importation process is ok.
The bug was with the repository importer app and the root way to import the script. The launcher is ok, no bug
]]>
< ![CDATA[
ok, I have trouble to gather all informations this morning…
]]>
< ![CDATA[
My fault guys, I should have fully checked the issue and what had actually happened before reporting, sorry about that..
]]>
< ![CDATA[
Pierre Hébert in short: the repository importer uses a library called RootTools to detect the already used IDs in LL by looking up all files in the scripts folder. As the Repository importer is now the script with ID 0, the library seemed to fail to find these files, and the first got overwritten.
]]>