Hmm, my english skills leave me here;-) A kind of ontology is what I was aiming too. Not a date but data - in the sense of "information".
Did you get me wrong here or do I get you wrong...

Things like "audioInterfaceType" (jack, alsa...) and "audioPluginInterfaceType" (lv2, ladspa...) as well as applicationCategory (DAW, Effekt, Notation...) and all we find important would be extractable and searchable. (And Google and other search engines might make things more searchable as well...)

Holger

13. Januar 2017 14:40 Uhr, "Chris McKenzie" <kristopolous@yahoo.com> schrieb:

Also this is all arguably just props up anb insufficient solution.

For me at least, the only reason I care about the date is that it's the best *indirect* indicator of whether something say, works with jack or lv2 or what have you.

An ontology that would be far more useful than the indirect reporting of a date would be to (and some of this is done on the site)

list the types of inputs (midi, jack, etc) and outputs of a program

Whether they are meant for low latency "live" performances or not

List what kind of plugin standards the thing works with

Include, if possible, similar programs that may be more difficult or less difficult to learn, for people of various dedication and skill.

The "ontology" part of this is important because ideally there could be a "web app" where you could string your stack together of different applications.

These could also be used for demonstration purposes to help other people understand the various ways things can fit together.

Because in practice, at least I have spent a lot of time in trying to learn what some software is good at, what it does for me, and how I could possibly connect it with everything else.

These things are relatively simple to communicate but currently as far as I know, isn't happening.

 

On Jan 13, 2017 5:19 AM, Chris McKenzie <kristopolous@yahoo.com> wrote:

 

Interesting idea. There's potentially a far quicker execution of it that doesn't need a buy in from everyone.

You take all the repo links for the projects and then have a script that pulls them say twice a month.  Then with a little git or svn magic you can find out if a newer release has been tagged and when it came out.

Also you can piggy back on apt repos for other things

apt-cache show Xxx

Generally has a pretty good description if the wiki is lacking along with an up to date website.

So I'm pretty sure some basic parts can be readily scripted within a few hours.  There's various models I can think of so the autopopulation is useful and there when you need it but doesn't necessarily clobber human created content.