bannerBaka-UpdatesManga
Manga Poll
Did you dress up for Halloween this year?
Yes
No
 
mascot
Manga is the Japanese equivalent of comics
with a unique style and following. Join the revolution! Read some manga today!

RSS Feed

[Bug?] RSS Feed is incomplete

You must be registered to post!
From User
Message Body
user avatar
Member


3 years ago
Posts: 23

I've been trying to setup a custom RSS feed (for similar reasons as described in this forum thread), based on MangaUpdates' RSS feed of new releases, by using a third-party RSS feed reader plus a long regex filter to filter for the tag. However, my custom feed has always been incomplete; not all items from the Releases page were displayed in the custom feed.

Naturally, I initially suspected an error in my regex filter, but I eventually tried an unfiltered RSS feed and experienced the same issue: The RSS feed is incomplete and doesn't contain all items from the Releases page. I'm not sure why that is. Maybe some scanlation groups "backdate" their entries, and the RSS feed doesn't contain such entries? I can only speculate.

To give one example: Chapter 84 of The Greatest Estate Developer, translated by Asura on 2023-01-26, was displayed on the Releases page on that day, but it never appeared in the official RSS feed.


user avatar
Site Admin


3 years ago
Posts: 2860

I was able to simulate that day in my developer environment and the release did show up, so it's not a rendering issue. The list is limited to 100 items and cached. It's possible that enough releases were added during the cached period that it "fell off" the end of the list during the next refresh. What is your refresh interval?


user avatar
Member


3 years ago
Posts: 23

Thanks for the reply. My MangaUpdates feed via the feed reader FeedBro still doesn't see all newly released manga chapters, and I previously had a similar issue with Feedly, but I'm struggling to reproduce the issue in a consistent way.

The cache issue you point out is a likely culprit, at least a partial one. The feed reader I currently use is an offline app which refreshes every 60 minutes during the day (which should be plenty), but not at all while my PC is off at night (for 7-9h at a time, during which more than 100 new chapters might get scanlated). So I guess that's one way in which the feed in my feed reader could get incomplete. I suppose I could remedy that by subscribing to the RSS feeds of each individual manga in my lists, but constantly fetching hundreds of RSS feeds sounds like a DDOS attack on the website.


In any case, I have already noticed a separate issue that's potentially related to my original one: the order of MangaUpdates feed items in Feedbro is different from the order in the .rss feed and the releases.html page. Normally I'd suspect the feed reader to be the culprit here, but I'm subscribed to a lot of feeds and have never experienced this issue anywhere else.


So rather than my reproduction attempts, I've tried a different approach: W3C has a "feed validation service" which can check if the syntax of a RSS feed is valid and optimal. When I entered the MangaUpdates RSS feed into it, it says "This feed is valid, but interoperability with the widest range of feed readers could be improved by implementing the following recommendations." alongside three recommendations, and (behind the "Help" bracket) explanations for the reasoning behind each recommendation.

To be clear, I don't know if any of these recommendations would resolve my problem, or how difficult or costly they would be to implement, and hence whether any of them are worth implementing.


user avatar
Member


2 years ago
Posts: 23

Once again, thanks for the earlier reply. I've now tried a third feed reader (the first, Feedly, did not support ultra-long filters; the second, FeedBro, was an offline app which did not refresh feeds at all at night), and it at least seems to be more comprehensive than before. I'll observe it for some time to check whether it's actually complete now.


user avatar
Member


2 years ago
Posts: 23

I've now tried a third feed reader service and still have the problem that the RSS feed is either incomplete or not being imported in its entirety.

My current working hypothesis for why this happens is that the MangaUpdates RSS feed doesn't seem to be in chronological order, i.e. new items aren't necessarily added to the top.

The way I understand it, many feed readers check whether a feed has new items by doing the following: they cache the most recent X fetched items, then fetch or check the topmost item in the RSS feed to confirm whether it's already in the cache, and only fetch or check more items in the RSS feed if the topmost item is not already in their cache.

Unfortunately, this algorithm breaks down if a site inserts new RSS feed items anywhere else but at the top.


To give a concrete and recurring example of this problem: I never see any Weekly Shounen Jump updates in my feeds (by "Viz & MANGA Plus", update date: Sunday at ~17 o'clock CEST), presumably because they're always inserted below the topmost item in the RSS feed.


Post #803272 - Reply To (#803256) by MondSemmel
Post #803272 - Reply To (#803256) by MondSemmel
user avatar
Site Admin


2 years ago
Posts: 2860

I'll take a look.


Post #803284 - Reply To (#803272) by Manick
Post #803284 - Reply To (#803272) by Manick
user avatar
Site Admin


2 years ago
Posts: 2860

So I looked through the code and the only thing I can think of that would cause your symptoms is release approval. We order releases in the order they were added, but some releases might be approved later in the day.

Would showing more releases help? Like last 500? That would be easier to implement at this time. Ordering by approval time is also a possibility, but after we finish with the UI updates, we're planning to change the way approval works. Users will be able to see all pending releases, so there will be a special rss feed with all releases.


You must be registered to post!