Re: [PD] 100k lines of code (was libpd separating gui from core)

2014-03-22 Thread Billy Stiltner
IOHannes, re: 'dir for vanilla' well a quick look for the TCL script to expose the innards of pd's file browser dialog turned up empty. browsing my old code I came up with this. http://www.geocities.ws/billy_stiltner/code/cgidir.tar.gz directory services through http. with that idea pd could use

Re: [PD] 100k lines of code (was libpd separating gui from core)

2014-03-22 Thread Billy Stiltner
found them in wheredoesthisgo 0.1 looks like there is no reference to the variables used in http://www.tcl.tk/man/tcl8.5/TkCmd/getOpenFile.htm maybe it is in the tk source. anyways why not just make an abstraction that calls tcl exec for linux send the path sent on an inlet to ls for windows

Re: [PD] 100k lines of code (was libpd separating gui from core)

2014-03-19 Thread IOhannes m zmoelnig
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 On 2014-03-18 21:37, Billy Stiltner wrote: what's wrong with making the file select dialog an atom? allready works in all the oses. just fan it's innards out some outputs do you have a public decryption key for your message? i'm totally lost.

Re: [PD] 100k lines of code (was libpd separating gui from core)

2014-03-18 Thread Billy Stiltner
what's wrong with making the file select dialog an atom? allready works in all the oses. just fan it's innards out some outputs On Mon, Mar 10, 2014 at 3:03 PM, Jonathan Wilkes jancs...@yahoo.com wrote: On 03/10/2014 12:56 PM, IOhannes m zmölnig wrote: On 03/10/2014 05:38 PM, Jonathan

Re: [PD] 100k lines of code (was libpd separating gui from core)

2014-03-10 Thread IOhannes m zmoelnig
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 On 2014-03-09 14:29, me.grimm wrote: Hi Miller, I know you probably have more pressing problems but it would be nice to get something like [getdir] in vanilla before you hit those 100k lines of code OR 50 years are up :) that's probably badly

Re: [PD] 100k lines of code (was libpd separating gui from core)

2014-03-10 Thread Dan Wilcox
Heaven forbid we add the 50 odd lines of totally manageable code in getdir.c that have largely been working since it was released in 2005. Nobody needs to get the current canvas dir, except for the recurring request for exactly this every 6 months or so on pd-list! With that argument, the

Re: [PD] 100k lines of code (was libpd separating gui from core)

2014-03-10 Thread i go bananas
I'd also like to see [getdir] made vanilla. There is literally no way to do that without the external. you could get rid of [value] to put [getdir] in. I could count on one hand the amount of times i've actually seen that used in patches. On Mon, Mar 10, 2014 at 11:46 PM, Dan Wilcox

Re: [PD] 100k lines of code (was libpd separating gui from core)

2014-03-10 Thread Dan Wilcox
It doesn't have to be a direct include, could be done with any sort of mechanism. I do agree that [getdir] is one of the few essential externals I still need after I ported my patch lib to vanilla ... that and [stripdir] / [splitfilename] but the latter will be possible with the new [list

Re: [PD] 100k lines of code (was libpd separating gui from core)

2014-03-10 Thread Jonathan Wilkes
In Pd-l2ork: [dir( | [canvasinfo] And for dir of Pd binary: [dir( | [pdinfo] And don't take IOhannes' bait.  He's implying that Pd Vanilla is a community project-- that if  _we_ coded up a currency converter, or some much more pressing functionality, Miller would accept that code.  That is

Re: [PD] 100k lines of code (was libpd separating gui from core)

2014-03-10 Thread IOhannes m zmölnig
On 03/10/2014 05:38 PM, Jonathan Wilkes wrote: Additionally, IOhannes also knows that Miller wants the [initbang] functionality in the form of a backwards-compatible [loadbang] which takes arguments. [...] thanks for the insights. i didn't know that i knew *that*. i would therefore be

Re: [PD] 100k lines of code (was libpd separating gui from core)

2014-03-10 Thread Jonathan Wilkes
On 03/10/2014 12:56 PM, IOhannes m zmölnig wrote: On 03/10/2014 05:38 PM, Jonathan Wilkes wrote: Additionally, IOhannes also knows that Miller wants the [initbang] functionality in the form of a backwards-compatible [loadbang] which takes arguments. [...] thanks for the insights. i didn't