2004-01-06 08:59 AM
2008-04-27 03:38 PM
2008-04-29 05:06 AM
TomWaltz wrote:Archives are files stored to backup media not a file format. The PLA often doesn't even work as an archive by itself since it can't incorporate macros called by parameter strings.
I thought it was a bad idea then and I think it's a bad idea now.
Archives are just that: archives. You're not supposed to edit them.
PLP files make it readily obvious that a file is shared. No one can try to open with any other way than signing in.There are certainly simple ways to indicate that a file is shared without requiring a different file format.
It makes no more sense than making MOD, GSM, and BPN files all the same as well.This is a silly argument, but thanks for the strawman. MODs are (effectively) assemblies - at least as good as we've got for now. GSM are (generally) components. Their current form and function could not be consolidated in any sensible way (unless perhaps you have some suggestions that haven't occurred to me). Perhaps in some future grand unification all the files become some form of assembly, superassembly, or subassembly each forming components of those above them in the continuum from a screw to a city, but I don't see any near term use for this except perhaps as material for a PhD thesis.
Library management requires that: MANAGEMENT.Old parts generally do work in newer versions. (I often grab stuff from my old libraries that still do the same useful stuff they did ten years ago.) They just lack the new features. In any case what I propose does not in any way prevent updating library parts to newer versions. In fact it facilitates it with better MANAGEMENT tools.
Old parts may not work on new versions. Continuing to load them forever is just asking for problems. At some point, you will have to revise objects for current usage.
This entire idea sounds like one giant workaround. Instead of fixing any problems it just makes do with the ones we have.I don't see you offering any better proposals. I believe that this does eliminate many of the problems we already have while also opening opportunities for improvements and features which would be difficult to implement otherwise.
2008-04-29 09:18 AM
2008-04-29 05:36 PM
2008-04-29 07:35 PM
Jere wrote:It's true there would be duplication of the parts stored within the projects and their counterparts in the reference/source libraries. This is why my proposal would require good library management tools. I envision something akin to file sync software (Chronosync is a good example on the Mac, I forget the names of the ones I've used with Windows). This would provide a list of all the parts in use in the project and their counterparts in the reference libraries, show which (if any) have changed, and provide the means to update one the other or neither.
If every project had localized libraries, what happens when ArchiCAD's main library is updated? Would there not be many duplicates of library files stored in different project locations?
2008-06-18 06:06 PM
2009-02-24 04:10 AM
2009-02-24 07:56 AM
I have suggested to the powers that be that with all the talk of intelligent objects that maybe we now have the resources to create a single cumulative object library, with parts that are intelligent to 'respond' with the appropriate nested code to whatever archicad version is loading them...Bruce,
2009-02-24 01:06 PM
2009-02-24 08:34 PM
Rob wrote:
I agree but this would require much more powerful GDL code base which I think is unnecessary. The overall goal should actually be moving from coding as such to more graphic tools.
Rob wrote:Agree it is not really feasible to include a converter but you should be able to save back to earlier versions if you want to and run the risk of it not working. I have mistakenly edited an object in AC12 several times when running AC11 in parallel - there is no recourse other than your backups. This is completely ridiculous when I know for certain all the code is actually AC10 compliant.
Secondly it would be a huge overhead to program an internal interpreter that would allow for 'degrading' a code to earlier versions. In fact I think it would not be possible at all in some cases (SEO, movable hotspots etc)