[PD] Re: Rradical [originator] with no [prepend] solution....
leechun at leechun.freeserve.co.uk
Thu Jan 6 01:08:40 CET 2005
My apologies for this late reply. I am still a bit preoccupied with
configuring Gentoo at the moment. I have been "shaving yaks".
>> While reinstalling my system, I have encountered this problem with
>> [originator] in the memento of rradical abstractions.
>> Because there is a [prepend RRADICAL.dummy] in the originator.pd and
>> although I have a [prepend] but its not the right [prepend] wanted by the
>> [originator]. As far as I know, there are at least 2 [prepend] our there.
>> I did a quick work around using [pack],[symbol] and [route], which seems to
>> work so far and it eliminates the [prepend] dependency. Please see the
>> attached patch.
>This only works for (re)storing single floats, it does not work for
>storing symbols and it cannot work for storing lists of unknown
>length, which is the main reason I need to use prepends there.
Argh, I did not think that far. I knew its not that simple......
>maybe it helps you out of your [prepend]-nameclash-confusion.
>for me, it behaves like i would expect it from a [prepend]-obj (it's not
>nice code, but i think it works)
>it needs some zexy-objs: glue, a2l, lister.(i hope everyone uses zexy)
Thanks for the patch, but I think having a single "iem_prepend.pd_linx"
built as suggested would be a better way.
>> i made an prepend_alternative-abstraction, because i don't want to
>> use a whole library only for one object (for a project with
>> different pd-users i'd like to keep the number of involved libs as
>> low as possible).
>If you just build the externals from CVS using the "build" system
>installed there, you get the "right" prepend as a single external. No
>need for libraries at all then, just the standard CVS externals.
>Iemlib-prepend is build as iem_prepend.pd_linx(dll...) there.
Sure, I will look into doing that.
>In my experience: If Pd crashes, when someone is trying to use
>Memento/RRADical, then that someone has some old ggee prepend
>installed. If Memento doesn't work as expected, then he is using
>iemlib as a locally installed library instead of the CVS version of
>iemlib so that iem_prepend "overwrites" the CVS prepend.
Umm, I got all the sources individually from each project web sites this
time. Maybe I will try using the cvs next time round. Do you use CVS for
More information about the Pd-list