Is this using the iHackernews API or did you roll your own? (I'm guessing the latter, since you reference jsoup.org?)
The iHN API has been very unstable for me lately, but all other HN readers for Android seem to use it (or their UI is just so bad it borders on unusable).
Do you think the CPU cost of parsing the HN soup yourself is significant compared to pre-chewed JSON from an API (say, iHackerNews)?
You're correct, I rolled my own parser. The iHackernews API is nice, but I found it pretty unreliable in my testing. I've also noticed most other Android apps report API errors frequently. I'm planning on open-sourcing the library I wrote for this, but I need to clean it up first. Honestly, it's not the best code I've written.
I don't really have a good feel for the CPU load of the parsing (I was mostly concerned with how much time it would take). In my testing it seemed fine. Probably more concerning is the memory usage, as I see the GC kick in whenever I refresh the feed.
Comments
Is this using the iHackernews API or did you roll your own? (I'm guessing the latter, since you reference jsoup.org?)
The iHN API has been very unstable for me lately, but all other HN readers for Android seem to use it (or their UI is just so bad it borders on unusable).
Do you think the CPU cost of parsing the HN soup yourself is significant compared to pre-chewed JSON from an API (say, iHackerNews)?
You're correct, I rolled my own parser. The iHackernews API is nice, but I found it pretty unreliable in my testing. I've also noticed most other Android apps report API errors frequently. I'm planning on open-sourcing the library I wrote for this, but I need to clean it up first. Honestly, it's not the best code I've written.
I don't really have a good feel for the CPU load of the parsing (I was mostly concerned with how much time it would take). In my testing it seemed fine. Probably more concerning is the memory usage, as I see the GC kick in whenever I refresh the feed.