[PD] finding differences between almost identical patches

João Pais jmmmpais at gmail.com
Sun Feb 20 15:19:13 CET 2022

from my work as a music engraver I can suggest another idea, but which 
only works on limited contexts: export each canvas as a graphic file, 
and overlay the different versions convert (pdfstudio does this). But as 
anyone can notice, there are too many conditions necessary for this to 
work (find and export all canvases, patches can't be moved, ...).

Or: comment your patches.

> On Sun, 2022-02-20 at 13:03 +0100, rolfm at dds.nl wrote:
>> hi list,
>> is there a smart way to find the difference between 2 versions of a
>> patch?
> Actually, the Pd file format is not really well "diff"able.  Even if
> you edit only one object within canvas with many objects, the implicit
> order of the objects changes, which leads to all subsequent 'connect'
> statements to change. A really tiny edit can (and most often does) lead
> to a quite large diff. Also, this makes it hard to concurrently work on
> patches and merge changes. It will almost inevitably lead to merge
> conflicts.
> Still, I think a tool for tracking differences in text files like 'git
> diff' or Kdiff3 is your best bet. If I am looking at a diff for the
> purpose of figuring what has changed, I consider mainly the 'obj'
> statements and try to identify new or deleted objects while ignoring
> the 'connect' statements. Also, the 'canvas' statements seem to create
> a lot of noise when stuff has been moved around.  Often, a change
> happened within a certain subpatch and looking at the diff gives you a
> clue about which subpatch is affected. To see and experience the actual
> difference, I open both version with Pd and compare visually and
> experimentally.
> Roman
> _______________________________________________
> Pd-list at lists.iem.at mailing list
> UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list

More information about the Pd-list mailing list