I feel like we shouldn't. More source of errors than anything I think. Also human readable names are useless when presented in search results because one can't install using a human readable name.
Shaz, I think we do need to add a doc issue. On Wed, Jul 31, 2013 at 10:30 AM, Andrew Grieve <agri...@chromium.org> wrote: > I don't feel strongly one way or the other about prefixing. > > > On Tue, Jul 30, 2013 at 6:49 PM, Shazron <shaz...@gmail.com> wrote: > >> Do we need to add a doc issue for updating this in 3.1: >> 1. >> >> http://cordova.apache.org/docs/en/3.0.0/plugin_ref_plugman.md.html#Using%20Plugman%20to%20Manage%20Plugins >> 2. >> >> http://cordova.apache.org/docs/en/3.0.0/guide_cli_index.md.html#The%20Command-line%20Interface >> >> (the docs will still work of course using the existing method) >> >> >> On Tue, Jul 30, 2013 at 3:33 PM, Anis KADRI <anis.ka...@gmail.com> wrote: >> >> > In the meantime....I renamed all plugins to package names in the >> > cordova registry. I've also updated plugman >> > >> > So anyone should be able to `cordova plugin add >> > org.apache.cordova.core.device` or `plugman install --platform >> > platform --plugin org.apache.cordova.core.device --project >> > /path/to/project` and it should just work. >> > >> > prefixing is not there yet. Do we really want it ? >> > >> > -a >> > >> > On Tue, Jul 30, 2013 at 9:58 AM, Brian LeRoux <b...@brian.io> wrote: >> > > Oh, yes, it could also be made to install from the npm registry. It >> > > does neither of those things yet. >> > > >> > > On Tue, Jul 30, 2013 at 9:43 AM, Andrew Grieve <agri...@chromium.org> >> > wrote: >> > >> A federated system just means that anyone can host a directory of >> > plugins, >> > >> yes? Anis was saying that plugman could be easily made to point at any >> > >> couchdb instance. Is that not federated? >> > >> >> > >> >> > >> On Mon, Jul 29, 2013 at 11:38 PM, Brian LeRoux <b...@brian.io> wrote: >> > >> >> > >>> No, at least I don't think so. The install/uninstall are more >> clobbers >> > >>> and plugin.xml is not a thing npm has any desire to become aware of. >> > >>> >> > >>> On Mon, Jul 29, 2013 at 11:24 PM, Andrew Grieve < >> agri...@chromium.org> >> > >>> wrote: >> > >>> > On Mon, Jul 29, 2013 at 11:19 PM, Brian LeRoux <b...@brian.io> wrote: >> > >>> > >> > >>> >> > Would this require that we use the node_modules dependency >> > structure? >> > >>> >> >> > >>> >> No. We would teach npm install about our structure. >> > >>> >> >> > >>> >> >> > >>> >> > Would we be able to enforce our <engine >> cordova-version=">3.0.0"> >> > as >> > >>> well >> > >>> >> > as our <platform name="ios" min-sdk-version="6.0" >> > >>> min-os-version="5.0"> >> > >>> >> > constraints? >> > >>> >> >> > >>> >> Yes. >> > >>> >> >> > >>> >> >> > >>> >> > Some things will be uglier to express as json I think. E.g.: >> > trying to >> > >>> >> > embed xml snippets (like for <config-file>), that contain many " >> > >>> >> characters >> > >>> >> > and newline characters. >> > >>> >> >> > >>> >> Yes. >> > >>> >> >> > >>> >> >> > >>> >> > Making things harder to search for has been pointed out as a >> > >>> >> disadvantage. >> > >>> >> > What would be the advantages? >> > >>> >> >> > >>> >> We implement a theoretically federated system. Cordova would >> > continue >> > >>> >> to use its own registry. (And thusly downstream distributions >> could >> > >>> >> use their own.) >> > >>> >> >> > >>> > >> > >>> > I thought this was already true with Anis' current setup? >> > >>> >> > >>
