When I add chapter support, is there some podcast out there which is clearly the one I should test against first?

Are we sure we want to build our dev around Github? Microsoft just demonstrated that they will knock down a project with very low friction, at the behest of a fellow megacorp.

We _will_, at some point, be choosing free speech over corporate censorship.

Hope for the best, plan for the worst.

podbrowser.org has a startup splash; not sure it's flashy enough, will have my artist daughter give me some input.

When you push to somebody's podcast, the description gets popped up. I'm liking that!

Once an art resource is up and running, I think the podcast summary will include the art there.

The new episodes API from @dave is working great! IMHO, I think podbrowser.org is probably the best way to scan new/interesting podcasts.

Since I mostly work in web which straddles desktop and mobile, I use SVG by default. Is there an SVG version of that podcastindex icon available anywhere? I can adapt the jpg if needed. TIA!

So @dave just a note on the fact that I can only, in fact, work my way backward about 8 pages of recent podcasts. Interestingly, ID's as I work back are 471008910, 470935867, 470795844, 437335328, 304320707, 1010022, 806300, 176652, 23300, 21261

So big jumps, implying I'm missing seeing much of the recent history?

podbrowser.org is pretty near feature complete. Have bookmarks, multiple linked devices, search, podcast->episode viewing.

Can still only walk ~150 episodes deep when viewing new episodes. Not seeing categories tagged on episodes, and still can't find an episode with any sort of search. (When viewing a podcast, will see categories, though.)

All the code on my side is pretty much ready, just need @dave to switch on some of these.

TYFYS!

@vandys @csb A $5/mth server from linode or digital ocean will get you pretty far.

Ok, podbrowser.org now has a FAQ and multiple device support (i.e., sync prefs).

No foundation libraries or external modules, all hand written HTML/CSS/JS/Python. And (ahem) no tracking. 😜

It's not on my main sources page yet, but you're all welcome to peruse and laugh:

sources.vsta.org:7100/browser

I'll open source it in the normal fashion when it feels pretty complete.

Just did a round of responsive design for podbrowser.org, now it seems usable on down to fairly small screens. Will check it out on my Alcatel KaiOS phone soon!

Ok, @dave I have a filtered feed view going at podbrowser.org

When you get the q=<text> for episodes, I can generalize it to pick out episodes that match, too.

Pretty clear I need exclusion categories too. Man! Those church people crank out podcasts.

Hey @dave I have another observation. I'm bootstrapping being able to click on a feed of an episode, and nest to view the episodes of that particular feed.

But byfeedid has no concept of max, so it can return a LOT of episodes. I'm synthesizing max/before just like for recent episodes, but if you have the same sort of URL args on your server side, I wouldn't have to pull a potentially large chunk of episodes just to return the amount the client needs.

And Good Lord (literally) I need category filtering. The feed this afternoon has a whole bunch of sermons all in a row (in radio, our Sunday morning programming was called "holy rollers").

@dave Cooking up filtering UI this afternoon. It seems like if the search matches a feed, you shouldn't bother listing episodes in the feed, too. I'll need a feed UI/UX for this, something new for me to think about.

So long as the API gives me episodes which are tagged with the feed description, I'll know when it's the feed that matched, and I'll have something reasonable to look into the feed. And then further episodes won't be listed as further search results.

I'm coding up search/filter for podcastbrowser and maybe finding a few gaps in the API? First, many searches will want to spot an *episode*'s title which matches the keyword. The feed's about general history, but you want to find an episode on halberds specifically.

Also, the search needs that same before= so we can work our way backward once we've scanned the initial results, and want to go further back.

Thanks! --Andy

I'm thinking you'll eat a bunch of overhead to tee up the whole record, only to have all the marshalling wasted by the collision. I'm pretty sure it's much more efficient to do a keyed lookup, with just one value pulled (to keep overhead of assembling results low too).

Then if it's indeed new, you can do the full insert. I bet there's way more existing records than new, so it's a win to keep the common case cheap.

@dave Then you can decouple the task of physically removing the records, have a sweeper during slow times which nukes things where the lastseen timeval is older than the lastscanned timeval of the feed itself.

2/2

Hey @dave when I'm paging back into history I get no results before ID 21261, less than ten pages worth. Intentional?

'Reviews', 'Rugby', 'Running', 'Science', 'Self-Improvement', 'Sexuality', 'Soccer', 'Social', 'Society', 'Spirituality', 'Sports', 'Stand-Up', 'Stories', 'Swimming', 'TV', 'Technology', 'Tennis', 'Travel', 'True Crime', 'Video-Games', 'Visual', 'Volleyball', 'Wilderness', 'Wrestling']

(Sorry... long list, not really a fit here.)

'Language', 'Learning', 'Leisure', 'Life', 'Management', 'Manga', 'Marketing', 'Mathematics', 'Medicine', 'Mental', 'Music', 'Natural', 'Nature', 'News', 'Non-Profit', 'Nutrition', 'Parenting', 'Performing', 'Personal', 'Pets',
'Philosophy', 'Physics', 'Places', 'Politics', 'Relationships', 'Religion',

Show more
PodcastIndex Social

Intended for all stake holders of podcasting who are interested in improving the eco system