Samuel, what’s your current opinion on this issue? To avoid load issues you could restrict it to Premium users and add a maximum time and/or maximum number of items per feed.
I couldn’t agree more with your comment. The fact that this “limitation” isn’t documented anywhere is what saddens me the most, make me feel like I was deceived.
I doubt that this “feature” (that’s what Newsblur calls it) will be fixed anytime soon, as the main developer (Samuel) is busy reinventing the wheel, that is blogs.
If you find an alternative (hopefully with the training feature) let us know. There are a lot of alternatives without that feature.
The problem is that it’s an architectural restriction. There’s only 1 variable that sets this time application-wise. ATM it can’t be set by user, to premium accounts, by folder, by feed, etc.
I suppose there’s a central module that handles feed update for all feeds, and this module reads this variable. To change it, this module would need to retrieve more data for each feed, like its own config, its related account’s config, etc.
He also wanna make configuration more flexible. But all this would make that module more complex and less scalable.
Well obviously there are architectural problems, otherwise this issue would have been quickly resolved a long time ago. I just wish the developer(s) would focus on basics like this instead of fancy intelligence trainers and social features that a lot of people simply don’t care about at all.
Google Reader had 4 weeks, I have 2 weeks.
Fair enough. They are a giant company and effectively infinite resources. You have built an incredible service and I am very grateful and impressed. I know that our appreciation doesn’t come across when we are whining about often petty details. Many of us who rely on rss readers spend a great part of our day in one and minor issues can be greatly frustrating. These tools are important to us and we can get emotional about them. I’m sure you understand, I doubt you built Newsblur without using rss a great deal yourself.
But back to my whining: I still don’t like even 4 weeks, but it was enough that I didn’t realize it was happening.
Yes, everything that Jason said. And I think the last sentence was the most important: “it was enough that I didn’t realize it was happening”.
I rarely ever hit the 4 week limit in Greader. I have hit it often enough in Newsblur in the past year and a half that I actually check my blogs less regularly now because it is sort of demoralizing to know that when I want to come back and read Very Long Article later when I have the time, it will probably not be there. :-/
I’m fine with two weeks. What I’m not fine with is a lack of documentation or the ability to turn the “feature” off.
Hmmm, assuming saved stories never expire, they might fit my workflow and solve my main concern.
Once you have a sizeable chunk of saved stories you’ll find it tough to search through them though (depends on your workflow of course).
Saved story search (and tagging) is one of my top 5 post-redesign priorities. It’ll be a great feature.
I tried using saved stories for this but it didn’t work too well. It’s too manual. So, I hope one other top 5 post-redesign priority is going to be this top-1 requested thing.
So how about setting this one variable that defines the life time of the unread items to 4 or better 6 weeks, and everybody is happy? It does not really require more storage space, does it? If you increase the time, I promise to become a premium member.
On further reflection, saved stores won’t work for webcomics, which I tend to save the longest. I like to read several in a series at once, instead of one a day.
But having to go through and save each comic before it auto archives risks spoilers and is just an unnecessary step.
You can still read comics in order, just right-click the folder contains the feeds and choose “All Stories / Newest first”. You just don’t have unread counts. Many services work this way. The stories don’t go away, you just lose out on part of the benefit of unread counts.
I’m wondering what exactly is stopping you from increasing the limit to 4 weeks, at least for Premium users. Since you’re keeping old stories anyway I would think the storage cost for unread counts should be negligible?
Because I wanted 1 week and decided 2 weeks is the right balance. Beyond that, if you haven’t read it, you probably won’t. You are free to download the NewsBlur source and run your own instance with whatever period of unread you want, but then you’ll have to support the additional infrastructure to support it. The impact on my servers will be huge and that’s why it can’t go up.
You’re ignoring all the use cases everybody listed where your “if you haven’t read it, you probably won’t” statement simply doesn’t apply, e.g. travelling and feeds that update less than every 2 weeks.
Also, I’d just like to understand how storing more unread counts will hugely increase the load on your infrastructure (implying processing power, vs. simply storage as I would assume), but you’re only evading my question and telling me to run my own instance instead… if I was planning to do that I wouldn’t have paid for a Premium account.
Another suggestion I mentioned in several other comments which I’m still not sure you’ve seen and considered yet: what about a fixed number of entries per feed instead of relying on the date?
Samuel, if you have to find your own point where the unread stories start, then you’ve lost ALL of the benefits of unread counts. I feel like a broken record saying that, because I did say that, 10 months ago.
After all the messages here, the justifications and use cases and, most importantly, two previous promises to try to alleviate this problem, it’s very frustrating to see you yet again dig your heels in here and insist that you are right.
We, your customers, are telling you, directly, that your assumptions about how we read our news feeds are incorrect. Even the option to set a low total article limit on a per-feed basis – say, 50? – instead of globally raising every feed to 30 days retained would go quite far. (It would take care of the problem entirely for me; I cannot speak for everyone else.)
This topic is really interesting because it seems to be the only part of newsblur people are complaining about … and the only part of newsblur the developer doesn’t want to change … 
When i was asked to choose between paying $1 per month or $2 per month for the same Premium access, i did not hesitate and paid $2 because i had found newsblur as the best google reader replacement and wanted to give money for such a nice tool. And i like it more than Google Reader except for this 2 week limit…
I would love to have 4 weeks, i would love to have a per-feed feature to avoid charging the servers, i would love to see Samuel taking into account the fact that the load on the servers would be probably paid by the new Premium accounts or the happy Premium users that would pay next year and would not go to another service. And if i had to pay more to have more archives, i would do it.
Of course we could all get the sources and try to install our own NewsBlur but i think only 1% of the users would be able to do that and that would be a silly move not to share the news feeds.
I’m sad and i hope it could change … maybe try one day to see the results on the servers ? keep a number of posts instead of a date for small feeds ? there must be a solution to please everyone…
1 Like