Coll auto save in patch fails

    artelse
    Oct 14 2012 | 10:53 am
    I try to save coll data in the patch by setting the embed flag in the inspector. When I close the patch, coll data is not saved and max doesn't ask to save the patch before closing. I tried sending a [flag 1 0] message with a closebang, but this doesn't work either. How is the flag different from the embed? When I send the [flag 1 0] message, the embed flag gets set..
    Any insights?
    • KrisW
      Oct 14 2012 | 12:39 pm
      im not sure , if u send [embed 1] message it will embed [coll]s data it into your patcher . but u need to save your patcher if u want to include it inside
    • artelse
      Oct 15 2012 | 2:35 pm
      Yes, it has saved a data set once and it never wants to update the changed data. Still puzzled by [flag 1 0] vs [embed 1].
    • Chris Muir
      Oct 15 2012 | 5:14 pm
      Any chance that you have a saved file on disk that the coll is reading? Try adding a 1 after the name of the coll so that it will not try to read from disk.
    • artelse
      Oct 15 2012 | 7:02 pm
      That's already what I do: [coll name 1] Documentation isn't very extensive on these options have to say.
    • KrisW
      Oct 15 2012 | 7:27 pm
      it might be not your problem but ,i had some issues when ive tried to store links that has been bad formated ... ive used "" instead of "/" ... in my case [coll] did not readed ,stored or keept anything . just to let you know that it also could be the case if you are using [coll] for such data .
    • artelse
      Oct 15 2012 | 8:59 pm
      Tried quite a few things, even stuff like name space conflicts and the like, but can't get changed coll data to save in the patch. I now reverted to saving to and loading from file with a closebang and loadbang construct. At least it works now, but might be a bug.
    • KrisW
      Oct 16 2012 | 9:56 am
      can you see the data in my coll ? could do the same and post ?
      you should try to prove it on a fresh patch PS: it might also be that you did not NAMED your coll . im guessing
    • artelse
      Oct 16 2012 | 2:20 pm
      Here's my coll with data. Not sure this is useful as creating a new coll in a new patch, copies from memory the data and it's not a problem with data in memory, but saving it in the patch.
    • KrisW
      Oct 16 2012 | 3:04 pm
      strange ,ive saved your code as patch and reopened it - data is here . can you confirm it with my attachment ?
    • artelse
      Oct 16 2012 | 8:21 pm
      That works fine. Perhaps it has something to do with bpatchers?
    • Chris Muir
      Oct 23 2012 | 8:20 pm
      So what do bpatchers have to do with it? If you're saying that the coll is in a bpatcher, and it doesn't get saved when the top level patch gets saved, that is the way things work. You are not saving the bpatcher when you save the top level patcher, so the contents of the coll doesn't get saved.
      What I do in this case, is have a "Vault" subpatcher at the top level (not an abstraction) that holds all my colls, tables, etc. I use arguments to abstractions/bpatchers to get them to use the colls in the "Vault."