Google Reader Apocalypse Extremely Fucking Nigh

Those dicks at Google are shutting off Reader at the end of the week, and the world is in a mad tizzy trying to find a feed reader that will continue to function.

First a word on why it is stultifyingly stupid that this is even a problem in the first place:

I want to read my feeds on three devices: a desktop computer, an iPad and an iPhone; and have all three of them know what I've already read. Simple, right?

I'm reasonably happy with Reeder on all three devices, and I was reasonably happy with NetNewsWire before it.

So if you want to keep "what I've already read" in sync, all you need is to store that state in some shared area. There are probably dozens if not hundreds of different, low-impact ways to share a single .newsrc file among three devices: Dropbox, iCloud, IMAP, CalDAV... the list goes on.

Instead, these people all decided instead to depend entirely on the Google Reader web site for storing that state, and so now they're fucked.

Once Google inevitably fucked them all, Reeder, instead of doing the obvious thing and storing its state on Dropbox, has decided to cozy up with some new Google Reader clone web site, that you have to pay for, and, oh, sync only works on the iPhone version, not Mac or iPad. Wha? Is there anyone in the world who needs to sync their feeds between two different phones? I kind of doubt it.

So, Reeder's dead.

Currently everyone and their mother is scrambling to write some new web-based replacement for Google Reader, but I don't want that, even a little bit. I want a pair/trio of apps that synchronize. I have no interest in reading my feeds through a web site (no more than I would tolerate reading my email that way, like an animal).

So, of this new batch of tail-light-chasers, the best reviewed of the bunch is Feedly.

I spent an hour today using it. It's horrible. I almost gave up after ten minutes but I made myself stick it out for a full hour, since there really don't seem to be better options and I wanted to be sure.

On the plus side:

  1. It's very pretty.
  2. Importing my Google Reader feeds was easy.

But here's what's wrong with it:

  1. There is no documentation. None. You'd think a feed reader wouldn't need much in the way of documentation, but they tried to "innovate" the UI in all sorts of "creative" and non-obvious ways, so, yes, it's a problem. As just the first example you'll encounter: in a list that begins with "Today" and "All", that "All" doesn't actually mean "All", it means "All Unread". Self-documenting my ass.
  2. There is no desktop version, only a web page. There's also a Safari extension, but I can't tell what if anything it did when I installed it. Maybe it alters the web page in some way? I have no idea. See also: no documentation of any kind.

  3. There is no Next button! You go to the next article by swiping right. This might sound like a good idea to someone who has gotten high on some kind of high-octane "designer" fumes -- button-clicking, it's so passé! we swipe now! -- but as someone who actually tried to use the thing it's a disaster for two reasons:

    1. It's an RSI horror-show! Seriously, after ten minutes, both of my hands hurt from the thumb to the wrist from all the swiping.

      Using this thing will cripple me.

    2. Swiping horizontally only works about 2/3rds of the time anyway. A ridiculous proportion of the time, when I tried to swipe horizontally, it would interpret that as vertical swipe instead. And once a vertical swipe has begun, you can't reinitiate a horizontal swipe until the screen has stopped jiggling up and down in its spastic skeuomorphic dance.

      Here's me trying to go to the next article:

      • Swipe
      • Swipe.
      • Swiiiiipe.
      • Sigh.
      • Tap. Hold. Release. Wait.
      • Swiiiiiiiiiiiiiipe. There it goes, finally.

      This is the most annoying thing since the fact that you can't click on the Trash button in the iPhone Camera app until it has stopped doing it's wiggly little dance from deleting the previous photo. You have to actually sit there and wait for the animation to end before you can click it again. SKIP INTRO, MAN. Heresy.

  4. There's a list of articles, one per line, stacked vertically on the screen. After you've scanned your eyes to the bottom of the screen, how do you see more? You scroll it up, right? Ha ha ha. No. You swipe right. Madness.

  5. There's no "Unread" indicator. If I want to mark an article as unread to come back to it later, I have to pop up the "Share" menu and click on the "O" (see also: no documentation) and the only way you know whether you've clicked on it properly -- instead of missing it by 2 pixels and just dismissing the menu, which looks exactly the same -- is to notice the brief flash of text saying "marked as unread" at the top of the screen that then immediately vanishes. So I end up doing it at least twice because I never know whether it took.

  6. The UI (in the iPad version) hangs solid all the time. I think it might lock up every time it encounters a post with an embedded Vimeo video in it, but I'm not sure. What I do know is that when I encounter one of these toxic posts, this is the dance I do:

    • Realize I can no longer swipe.
    • Click the home button.
    • Either kill the app, or just launch it again.
    • It displays its splash screen for 20+ seconds, then crashes.
    • Launch it again. Wait while it re-polls all the feeds.
    • This time it works... but every article that I read last time is unread again! So as I try to page forward through them, the one that makes it crash is still lying in wait like a landmine.

    This is unforgivable. Not losing user data is the prime directive of any application. What I've already read is the only data this app generates and they lose it.

  7. There seems to be no gesture to mark an article read without selecting it -- like some kind of swipe on the article list or something -- so as far as I can tell there is no way to delete these landmine-articles without doing "Mark All Read". But, maybe there is and I just haven't discovered it -- see, again, no documentation at all.

  8. I can't pinch-zoom to zoom in on images. They scale to the width of the screen and no larger. Come on.

  9. There's no way to edit the "vertical ellipsis" menu to leave out the mysterious, incomprehensible icons of services that I do not use. (They think the Safari icon is an arrow in a circle. "Clipboard" is a piece of chain. WTF?)

  10. "Mail" always mails the whole HTML page instead of just a URL. It does not even include the URL. WTF? And it includes not one but two self-advertisements for Feedly in the HTML when it does so.

All I want is a version of Reeder that stores my .newsrc on Dropbox.

And that has a "Next" button on the right side of the screen as well as the left. Oh god please yes.

But we can't have nice things.

Previously.

Tags: , , , ,

214 Responses:

  1. How hard a requirement is a standalone desktop version for you? Newsblur, once I hunted down and turned off all of its "social" (retch) features, has treated me fairly well so far, and has an iOS client that I like substantially better than NNW.

    • jwz says:

      Critical. I have never seen a web-page-based feed reader that didn't suck.

      I suppose I could just never read feeds on a non-iOS device, but that seems... extreme.

      • You could run newsblur on the Xcode iOS device simulatoHEY OW STOP HITTING ME.

      • Hani Suleiman says:

        I use newsblur as a desktop app via fluid, along with a userscript to show unread count as a badge in the dock. Works surprisingly well.

      • Pavel Lishin says:

        Extreme, plus there currently doesn't seem to be a way to mark an item as unread on the mobile app. Which means that I never use the mobile app.

        • frossie says:

          @Pavel - you can mark a story as unread on the Newsblur mobile app - while reading any story click the 'settings' cog icon, it's the second option on the menu that pops up.

          There is a third-party desktop app for Newsblur (readkit I believe) but I haven't tried it - some time with the preferences have made the web UI fine from my POV. Note that the mobile apps essentially inherits its behaviour from settings available on the web UI, so you have to spend some time customising there (eg. setting up non-public shares and so on).

          • Pavel Lishin says:

            Not on the Android app I downloaded - the options are "Share this story", "View original", "Send to...", "Adjust text size", and "Save this story."

    • John Morton says:

      There is a user created app for the Mac called Readkit, apparently:

      http://readkitapp.com/

      No clue whether it sucks or not.

    • chet says:

      N.B. that Newsblur marks all articles read after two weeks and that the author has no immediate plans to fix this.

      • MattDarling says:

        Not the case, it's just a hard problem: https://getsatisfaction.com/newsblur/topics/do_unread_items_sunset_after_14_days#reply_12309453

        I'm annoyed by it, too, but it will be fixed eventually.

        • chet says:

          He has no concrete plans for WHEN it will be fixed. I've been around software enough that I feel safe assuming features that do not yet exist and are not schedule for v.next, like as not, will never show up.

          He's also busy with discovery and social and all sorts of other bullshit I'll never care about, to the extent that these things are prioritized ahead of not marking unread items read, which I find irritating.

        • Pavel Lishin says:

          Bleh. I wish he ran a parallel NewsFocus that didn't give a shit about training, but kept unread status up to date forever.

          I guess I can always fork this and work on my own copy, but I've already fed Shiloh...

        • Aaron says:

          The worse problem, and the real reason I haven't launched this yet, is that I store story ids and read stories in redis. It's extremely memory efficient, yet it still takes 18GB of RAM. This is growing at a rate of 2GB a week (on a 32GB machine), and it's the next big piece to try and scale, although I have no clue how I'm going to do it yet.

          They charge for this?

          • crowding says:

            This just confirms that (1) saving user read state and (2) centralizing feed polling are separate concerns and the former should just go on Dropbox or wherever the user wants.

  2. Liz Jones says:

    http://www.wired.com/gadgetlab/2013/06/inside_digg_reader/all/

    Damnit. I was depending on you finding a solve for this.

    • jwz says:

      Yeah, whatever. This vaporware sounds like "social blah blah blah discovery blah blah blah drive traffic to our other web property." It does not sound like "a feed reader". It sounds like a walking ad that maybe has a feed reader grafted onto it.

      • kevin says:

        Not vaporware I swear! It's gonna be good, and super fast. Let me know if you'd like an invite to check it out.

        • gryazi says:

          Portals!

        • jwz says:

          Dude, if all there is for me to download is a press release and a Wired puff-piece, then yes, it's vaporware. That's what the word means.

          But sure, I'll take an invite code, thanks!

          (If you have OSX, iPhone and iPod apps, that is. If it's just a web site, then I'm not interested.)

  3. Jim says:

    This takes me back: I wonder how many young folk know what a .newsrc (file) is?

    • jwz says:

      Yeah, someone just said to me "The people who wrote these things have no idea what a .newsrc is, grampa." I said, "Let me Google that for them."

      • One of my professional mantras, with which I have been annoying coworkers and random innocent bystanders for years: "Those who do not remember NNTP are doomed to reimplement it poorly."

        The last decade and change of the successful parasitization of the internet by "social media" proves that I may be a huge bore, but I was not wrong.

        • Ewen McNeill says:

          Amen.

          I've thought for a while that the Internet N.0 variation of .newsrc would be a REST service that allowed synchronising a list of read posts, and subscriptions. Then one could use any reader software at all, interchangeably. There's a de facto standard for subscription lists, but I'm not aware of a "read posts state" list equivalent format. (Sadly unlike NTTP there isn't as easy a shortcut of just storing message numbers which are easily summarised, so "store whole .newsrc equivalent file at once on remote file store service" seems a less than optimal API.)

          Sadly no one appears to have done so. And I've lacked the "round tuits" to implement it myself. So instead I'm still reading things as "feeds" on dreamwidth (formerely on LJ), via multiple devices, and remembering in my head if I've read things or not. It's... not optimal.

          Ewen

          PS: The same mantra is largely true for IRC. The code to implement IRC might have been... poor. But the range of platform functionality has rarely been matched again.

          • hattifattener says:

            It is sort of the problem ACAP was designed to solve. I haven't looked at it in a while, so I don't remember if there's any reason it couldn't be used for storing feed-item read-state information.

            • Ewen McNeill says:

              Thanks for the pointer.

              ACAP's design appears to have been (unfortunately) influenced by IMAP, and it appears to be hardly mentioned since 2001, neither of which are strong recommendations for reuse. The fact that it's an "always up" streaming protocol is useful here (less overhead), but for better or worse lots of networks only pass HTTP(S) these days so it's the de facto transport (hence suggesting REST).

              From other comments on this post there do appear to be a few proprietary potential protocols for storage of such feed-read state. I'm guessing some/all of them are using HTTP as the transport. It might be that reverse engineering one of those and adopting it would be more useful than inventing something new.

              Ewen

        • marijane says:

          I know a couple people who feel this way about IMAP.

          • Hey, I resemble that remark. Never mind the fact that I'm now reading my feeds by gatewaying them into maildirs on an IMAP server.

            • marijane says:

              Oh jeez, I think I even remember seeing you tweet about that, yet you were not one of the two people I had in mind when I posted that comment. I guess I should correct myself -- I know at least three people who feel that way about IMAP. And I think you know the two others I had in mind, too, Jesse.

              • Jesse says:

                laugh You know, at this point, I'm not 100% sure I _do_ know who you had in mind, but you're right that I probably know them.

            • glen says:

              I've considered going down this route several times. Do you have a writeup for doing this?

        • Matt Rose says:

          Ugh, yeah, let's use NNTP and propagate all data to servers all around the world! No thanks. I'm having a hard time figuring out how you could do NNTP worse.

          ran an INN server for far too long in the 90s

          • Ugh, yeah, let's use NNTP and propagate all data to servers all around the world! No thanks. I'm having a hard time figuring out how you could do NNTP worse.

            How about: any moderately popular blog ends up having to pay through the nose to one of a handful of (nonstandardized and hideously expensive) regional edge precacheing services (Akamai, Amazon, etc) in order to get their content anywhere near their readers. We've come so far! :)

            Don't get me wrong: the lessons of Usenet are not all positive ones. ("Spam crowds out real content," "any transport that can be made to serve porn, will" and "you can't actually trust your neighbors" are high up on the list of caveats.) But the blogosphere didn't learn those lessons either. The result: buzzfeed, huffpo, facebook and tumblr.

            "Ever get the feeling you've been cheated?" --St. Jonathan Rotten

            (Your name: slightly familiar. Were you a news.admin.* habitue back in the day?)

            • Matt Rose says:

              any moderately popular blog ends up having to pay through the nose to one of a handful of (nonstandardized and hideously expensive) regional edge precacheing services

              That is much better than the "propagate all the crap" method used by NNTP, at least you're not making some poor news server admin buy new disks to hold all of your crap (yes, this did happen to me) HTTP has a whole bunch of commands to allow optional caching to allow exactly the scenario you complain about to happen, and is IMNSHO better for it.

              if somebody has to pay for all that crap to be close, I prefer the people who actually create it.

              I was probably on news.admin.*, I was all over USENET for a while, and I was a news (along with every other type of server) admin.

        • Evil Otto says:

          My new favorite quote:

          Legacy code often differs from its suggested alternative by actually working and scaling." --Bjarne Stroustrup

  4. Peter Hollo says:

    I think the paid-for app that Reeder introduced support for was feedbin. It was just the first of many - the developer of Reeder does say he will support feedly, so you won't have to use their crap apps anymore, but the question is when. I certainly hope it'll be by this weekend!

    • Mitch says:

      When Reeder first announced Feedbin support I thought to myself "fine, I'll cough up $2/mo if it means they don't disappear" Unfortunately, although the iPhone Feedbin support was released a couple months ago, there has been radio silence for the iPad version. That's where I primarily use my RSS reader.

      When the iPhone Feedbin support was added the Reeder developer tweeted when he submitted it for approval, and 10 days later when it was available on the app store. Since there haven't been any similar updates on the iPad version I'm assuming that it has missed the deadline and will be useless after this week.

      I'd been trying to be optimistic that this would all get sorted out before the deadline and I wouldn't have to think much about it. Turns out that didn't happen.

      • Nick Jones says:

        I use Mr. Reader on the iPad. Their v2 which supposedly supports feedbin.me is in review at Apple.

  5. John de Jong says:

    According to this GigaOm article (http://gigaom.com/2013/06/24/life-after-google-reader-gigaoms-guide-to-the-best-options/), Reeder will be supporting Feedly as its back-end, eventually, so you don't have to pay $20/year but also don't have to deal with Feedly.

    Eventually had better mean within the next six days.

  6. David Hoover says:

    The feedly website also has the awesome habit of not noticing that I've read things; I use the mode where it shows full articles inline and I just page down, letting it auto-mark-as-read as I pass things. But it appears there's some particular sweet spot on the screen something needs to hit for it to register, and hitting PgDn passes it all the goddamned time. So I think I've read things, come back later, and have deja-vu. Plus the last article in a category is often so short that I can't scroll down whatever magic amount it needs.

    Oh, and it's 2013 but they don't do SSL. Yay?

  7. Morrisa Sherman says:

    You're a world-famous and respected hacker emeritus. Did you send the above detailed and well-written critique to Feedly yet? I should think your list should clearly be their goals for their next update.

  8. In Feedly's iOS app you swipe up and down to mark items read/unread. Swipe a bit to act on one article, some more to act on a "page" of articles at once. I hated it at first, but I basically get the point now. The lack of documentation is another matter.

    That said, I think I'm going with NewsBlur, which has a very nice Web app and a quite decent iOS app that use real list views and have a "next" button that is appropriately sized rather than a swipe gesture. And you can pay for it, or run the open source version on your own server.

    • chet says:

      Be aware that all articles are marked as read after two weeks with NewsBlur.

      • Yeah, I guess that'd be an issue if I ever went on no-Internet vacations longer than 2 weeks. As is, if I don't read something and want to read it later I send it to Instapaper.

        • chet says:

          I mark it as a critical failure, since to me that's losing data. I use "unread" as "come back to this," and expect something I've marked unread to STAY that way.

          • crowding says:

            This bit me with Google Reader, too -- I use feeds to keep tabs on academic journals. If you track enough journals that publish stuff on your topic, it's a real firehose, and there's no workable way to actually sort through the relevant bits of the firehose other than at least glancing at the titles.

            I usually deal by waiting for a bit of downtime to sort through 1000-2000 titles to find 20 abstracts to notice and 5 articles to read in detail. Forgetting not-yet-read items just because I haven't got onto a plane flight or whatever is Not Acceptable.

  9. I've been using CommaFeed for awhile now. I wouldn't stay its the most stable-ish, but its the most feature complete in my opinion. It's open-source so you can run your own server as well if their outages cause too much of an issue.

  10. jhnc says:

    Would setting up Tiny Tiny RSS with the fever plugin and then pointing Reeder at it be out of the question?

    http://tt-rss.org
    http://tt-rss.org/forum/viewtopic.php?f=22&t=1981

    • jwz says:

      I don't know what you mean by "point Reeder at it". You can point Reeder at exactly two things: Google Reader and Feedbin. It's all hardcoded.

      • jhnc says:

        The person who wrote the fever plugin for tt-rss seems to think Reeder can be pointed at three "things":

        http://tt-rss.org/forum/viewtopic.php?f=22&t=1981

      • someguy says:

        Not according to that second link (here). They claim that by using the "Fever" account type in Reeder, you can point it at an arbitrary URL; someone wrote a plugin to let TTRSS speak that protocol and be the central backend for multiple devices.

        FWIW, I made the switch to TTRSS a few months ago after Google first decided to can Reader and couldn't be happier. I much prefer to be in control of my own data anyway where possible, and Reader's demise just gave me the incentive to move away from one more Google app.

        • Ian Young says:

          Until you look at the TTRSS source code...

          "I have to make writable directories for a fucking feed reader? WAT?"

          And cron jobs, oh god. I read my feeds every day. If I don't read all the articles in a day, I will never read them.

          And I would bet $20 that there is some kind of sql-injectable XSS exploit hiding in there somewhere.

          I just want to load xml and produce previews and link to sites, people. I don't need to re-syndicate fucking feeds or store thumbnails or whatever you're doing. Stahp.

          • icebraining says:

            You don't need a cron job. There are three ways of updating the feeds: running a daemon, setting a cron job or just updating when you have the client open.

            http://tt-rss.org/redmine/projects/tt-rss/wiki/UpdatingFeeds#Simple-background-updates-since-170

            • yabonn says:

              When I installed mine, the recommended way to do it was running some php script in a screen. Now : it works, it's dumb to point and giggle, and I don't have ideas about the best way to go to solve that problem, but the beast looks indeed strange to me.

              • icebraining says:

                It's just a process, you can use your favorite daemonizing tool to have it run in the background; you don't need to use screen. Since each distro uses a different tool (Debian uses SysV scripts, Ubuntu has upstart, Fedora uses systemd, etc), it wouldn't make much sense for the Tiny developer to include it in the source. That should be up to a package maintainer.

          • Mattyj says:

            If you think any of the current crop of feed readers work on that simple of a level, you're nuts.

            Rolled my own ttrss server some time ago and am extremely happy with it. But I'm stupid and didn't even know there were apps for Google reader. Just been using a browser on the iPad.

            Also, it's open source, which means the community, which you are a part of, is responsible for security vulnerabilities. Or deciding what level of comfort you are, ah, comfortable with.

    • I am using tt-rss and lurve it, but there's very little on the OS X side of things. Although newsbeuter syncs to it for that old-school feel.

      (Even if it have a bunch of iOS/OS X clients I can imagine a stream of invective from Jamie as once he finds a LAMP/LAPP application...)

  11. Tom Lord says:

    Look at all the nice, robust command line utils there are to slurp and merge / sort / filter RSS feeds. It should be easy to knock out a really solid feed reader of the sort you want from those tools. Which ones are everybody's favorites?

    Even if they don't exist there are some really solid and complete XSLT processors out there that will make such tools easy to write. Even in spite of the non-standard date formats RSS uses.

    It's a triumph, really, of the high standards of software engineering the dot com boom(s) engendered.

    • C says:

      Looks like rss2email or feed2imap are the go. Why re-invent the messaging wheel when we have imap clients already?

      $ apt-cache show rss2email

      Description-en: receive RSS feeds by email
      rss2email is a simple program which you can run in your crontab.
      It watches RSS (or Atom) feeds and sends you a nicely formatted
      email message for each new item.

      $ apt-cache show feed2imap

      Feed2Imap is an RSS/Atom feed aggregator. After Downloading
      feeds (over HTTP or HTTPS), it uploads them to a specified
      folder of an IMAP mail server. The user can then access the
      feeds using a mail reader like Mutt, Evolution, Mozilla
      Thunderbird or even a webmail.
      .
      The main advantage over rss2email is that feed2imap uploads
      emails itself to a specified folder. You don't have to play
      with procmail or your MUA's filtering rules to classify mails.

      • Ian says:

        rss2email is my solution, using POPfile to classify the results. After a day training it, you have an email feed of things you actually want to read, with very low false positive and false negative rates. You can also categorise things into more states, so you can have different importance / interest levels.

      • GDorn says:

        For an extra shock, have a look at the original copyright for rss2email. It's buried in the source.

  12. The mind boggles to hear that, on Feedly for iOS, you swipe left/right to navigate the list of articles, because on Feedly for Android you, yes, swipe up or down.

    Obviously the crack distribution in the Feedly office was uneven and tilted in favor of the iOS developers when these apps were conceived.

    As is, I'm quite happy with Feedly

  13. marijane says:

    A couple of weeks ago, I posted to Twitter that I was in complete denial about the shuttering of Reader. I got a half dozen recommendations for The Old Reader and nothing else. The unanimity was striking. That said, they're definitely targeting people who want the old pre-Google+ Reader web browser experience, and while they are doing a very good job of that (the keyboard shortcuts are pretty much dead on, it's been nice not having to develop new muscle memory), it doesn't sound like what you are looking for.

    I've been toying with trying tt-rss, but I've been hearing the UX is terrible so I might not bother.

    • Mattyj says:

      The ux is a good approximation of google reader, but it seems chatty. Not sure if its precacheing a bunch of stuff or my hosting provider is slow. I still like it a lot, though.

    • phuzz says:

      I like it, partly because every other Reader replacement is far too pretty and web2.0 for me. All I want is something that displays my feeds, 'social features' are a negative for me.
      API and mobile apps are supposed to be coming soon, but as it's someone's part time project I'll believe it when I see them.

  14. Squiddhartha says:

    For what it's worth, the open beta of the next NetNewsWire just dropped today. It doesn't have sync... yet. It's promised for the full release, but obviously that doesn't help in the immediate term. I've been making do with Netvibes. I'll give the NNW 4 beta a good poking-at, but I'm dubious about the new UI.

    • Jesper says:

      The new UI is not the problem. It basically works the same.

      The problem is the old features, in that they shook them all out. Actually, they reimplemented everything (for some value of those two words) and haven't gotten around to most of them and while I briefly sympathize on a "software developer" level, I am furious on a "person who lives in this fucking app and who thought that the people who bought it would wish to retain some of these features since no one else apparently wants to write a decent feed reader any longer" level.

      The biggest problem with new apps is that there are no allowances for patination. If you want something that you can make work a little more comfortably every day until it's an extension of your hands five years down the road, fuck you. You'll have an email client that looks like Twitter's iPad UI at the time we designed it and you'll like it. Get Info is yesterday's design pattern; Jony Ive would not have appreciated being able to have opinions about how stuff works in the apps he uses. Everything I can't conceive of using myself is clutter and bloat. And if some of these kids that you're trying to sell to are RSSing on a fifth-grade level, fuck them, they're nerds and they'll have unusually cruel feature requests and it's not worth catering to them.

  15. Ben Donley says:

    You're paying someone to aggregate feeds so your phone doesn't have to poll a hundred websites in various states of disrepair. It's not like paying for a .newsrc - it's like paying for NNTP. Since RSS feeds rarely follow the specs, your RSS-NNTP has to normalize everything and your RSS-.newsrc can only work with one aggregator.

    They're all terrible. Amongst all these POS aggregators, Feedly is one of the few that won't even export OPML, so if one day someone builds the right tools, it would be a total pain to switch again.

  16. Hi Jamie, This is Edwin from feedly. Sorry for the frustrations. Let me try to address the lack of documentation.

    1. Yes. There is no documentation. On mobile, we use an Android convention: if you long press on an icon, you should get a sign with information about what the icon does.

    2. There is indeed no desktop version. The safari extension is a work in progress tool to make it easier to discover and add RSS feeds.

    3. The most common workflow among our users is to swipe vertically to go from card to card, tap to drill down into an article if it is interesting, read and use the swipe up to close or the arrow to go back to the summary level and swipe more vertically.

    It seems that your workflow is inline one article and then swipe right to go to the next inlined article. I can see how this could be a lot of swipes if you have to read lost of articles. I will talk to the team about adding a next button to go from one inlined article to the next.

    4. I guess you mean swipe up because this is how you swipe away one card to see another card with more articles.

    5. Here is a hidden gesture (I know sorry, no documentation): when you are looking at an article summary, a short right to left swipe will mark an article as read, a long right to left swipe will mark the entire card as read. The opposite direction will keep unread.

    6. If you can email me the steps to reproduce the hanging (feed and navigation), I will open a bug for us to look into the issue and fix it.

    7. See 5. at the summary level, horizontal swipe will mark as read and keep unread.

    8. We are working on this.

    9. Press and hold on the icons of the ... menu should show a sign.

    10. Please add that to feedly.uservoice.com I am curious to see how many people would like to have an option to only email the title. Note: the link to the article is embedded in the email. And yes, there are two links to feedly: we are a small bootstrapped startup and do what ever we can to try to grow organically and try to keep the service free.

    11. Feedly is offering an API to allow other readers to sync their state across devices. 50 developers are working on apps. 9 have already developed apps. Reeder is working on it for his next update. If you think through the problem, it is a little more tricky than simply sync'ing state. On phones, it is unlikely that you can start the phone and start polling the N feeds you subscribe to see if there are updates, specially if you need to join that information with the user read state across multiple devices.

    We love candide feedback and we love to iterate. Expect to see some of the points your raised being addressed in the July update.

    If you have any other questions, I am edwin@feedly.com

    • ix says:

      In addition to 6 and to echo something jwz said, it is especially annoying that Feedly's data sync is less than bulletproof. I've had this happen where I got logged out of feedly, logged back in, and suddenly my feed history of the last week was completely gone (I guess it didn't sync with Google Reader for a while, then did a fresh sync).

      I know these things can be hard, but losing the user's data is simply unforgivable (not to mention hugely annoying). Every time it happens your users trust you a little less. I wouldn't even be looking for alternatives if this kind of stuff didn't happen, even though I'm not a big fan of the UI.

    • Michael Fischer says:

      On mobile, we use an Android convention: if you long press on an icon, you should get a sign with information about what the icon does.

      I don't think this was a wise decision. If you want an app to feel native, it's best to stick with the standard conventions. This gesture is not well-known among iOS users (I'm unaware of any other iOS app that uses them), and therefore would be non-intuitive to us. We wouldn't even think to to try it.

      If you really need the equivalent of a "tooltip" in your iOS app, either your icon needs a redesign, or your app should start with a tutorial the first time it's launched, as many iOS apps do.

    • jwz says:

      Hey! Thanks for replying...

      3: Yes, I read my entire feed. That's why I subscribe to these things: to read them, not skim summaries.

      4: No, I meant what I said, swipe right. I'm looking at the "All" list. There is a vertical list of 8 article summaries on it. To see the 9th, I have to swipe right to reveal a new page, instead of swiping up to scroll. It's a horizontally-stacked series of paginations, not a long scrollable document like... well... everything else everywhere always.

      6: I don't currently have an example of a public RSS URL that causes the hang, but I've extracted one of the culprits from a private feed and will mail you.

      • Hey! Thanks for the clarification...

        Regarding 4. If you go to the advanced settings (link at the bottom of left nav panel), is your "Look and feel > Transition" setting set to the default "Stack" or "Swipe"? Stack (which should have been the default) should make the cards with 8 articles on them transition vertically. "Swipe" is horizontal.

        Thanks for the email regarding 6, one of our engineers is looking into why this could lock the phone.

    • J. Peterson says:

      Thank you for responding to JWZ's post.

      Please fix:

      On my Samsung tablet (Galaxy Tab 2 10.1) Feedly has a bug where after reading the third post on a feed, I see the same posts again. If I long press on a post, -then- I see the forth post. It's a refresh bug. Ugh.

      With the Android app, when I click on a link in a post please send me to the web browser. I do not want to surf the web in Feedly. Feedly is not a web browser. And let the Android "back" button take me to where I was in Feedly.

      That Android app "scroll up too far and the article blips off the screen?" No. Wrong. Do not overload scrolling with closing.

      Android App, compact title view. Half the time tapping on an article's title just marks it saved. No, I wanted to read the post. Please, if I want to save it, I'll open it first. Don't overload those functions.

      On the web page, "Index" is the view I want. Front and center, not way down at the bottom of the list.

      Your designers are not smarter than Google's. Just copy Google Reader and everybody will be happy. Google won't care, honest.

      • GDorn says:

        Your designers are not smarter than Google's.

        You might want to qualify this with a date. Google UX circa 2005 was functional. Google UX today is a clusterfuck, a joke nobody is laughing at anymore.

  17. Barf McStarf says:

    I was in the same position. Here's what I did: switched to using ReadKit (App Store, under $10), which is like a shitty version of Reeder but can read Instapaper too. Feeds are retrieved from NewsBlur, which is free if you read less than 64 feeds and maybe a couple of bucks a month for unlimited. This should handle all your dirty, nasty feeds.

    • Barf McStarf says:

      And NewsBlur's app on the iPhone is OK. It works especially well for fast-fingering through all the crap that Jon Gruber posts on a daily basis.

  18. Guillermo says:

    Reading feeds on a webpage like animals? We are in the age of webapps, having a standalone feedreader IS how "animals" read their feeds. Feedly is great. BTW, there's this very prominent button called "save for latter", just alongside the social sharing buttons, which SAVES FOR LATTER. I don't have a clue about the swiping issues you describe, i find it very convenient on my phone and I don't find any swiping in cloud.feedly.com...
    I've only found two inconveniences:
    1. the one you described about not having a next button
    2. It continuously asking for login on my phone.
    Other than that, it does things pretty much the same as Google Reader used to do them, sometimes even better.

  19. AnonMyAss says:

    What I want to know is why Emacs hasn't written a new feed reader itself when it realized, before Google made the announcement that Google's reader was doomed.

  20. brett says:

    newsblur.com

    I've been quite happy with it.

  21. Patrick Useldinger says:

    I've happily moved to InoReader. The web version is just like Google Reader. The mobile version (.../m) is lame which improves my iPhone's battery life a lot.

    • J. Peterson says:

      Thanks for the tip! The web version works great on my tablet too. There's no "share" function, but I can get that by just opening the page in Chrome.

  22. Jason says:

    Well if you literally want to share a .newsrc file, you could read using Gwene: http://gwene.org/

  23. Jeena says:

    I use TinyTinyRSS and had to write my own very simple Feedreader for it http://jabs.nu/feedthemonkey

  24. jwz says:

    One of the things that is crazymaking about all these people trying to re-invent feed-reading as a web site is the blather that goes: "It's a feed reader but it's also... and it's... and social! ... and discovery! ... and sharing!"

    I just don't get it. Discovery: that's what the act of reading feeds is. Sharing: I have a blog, you're reading it. Social? I don't even know what that means in this context if it differs from "Sharing". My blog has comments. Maybe it's that.

    • Jesper says:

      Comments are way too social to actually be "social". "Social" means a million people each going into Grand Central Station with a megaphone and going "HAY GUYS". "Stupid" is another adequate six letter word for this.

    • Mattyj says:

      Registers antisocialreader.com. Has delusions of grandeur.

    • Kevin Lyda says:

      To be fair, "discovery" is about finding new feeds. If you just read the same feeds you will never be exposed to new ideas and you will be stuck in the echo chamber, blah, blah, blah.

      I suspect you're kind of stuck. Desktop apps aren't a huge priority anymore. Most "apps" will be a website and mobile apps. Your best bet would be an active opensource one and writing a desktop app that uses the api. However the most active opensource one seems to be tt-rss and I suspect a giant ball of php is something you should avoid for your health.

      • Nick Lamb says:

        "If you just read the same feeds you will never be exposed to new ideas"

        Unless you already read a feed that exposes you to new ideas? I don't own any sort of Apple devices but I'm pretty sure Jamie would have mentioned by now if those don't have hyperlinks.

        For example I have this blog in my feed. Suppose I somehow didn't know about Oglaf. Well no worries because this blog links to Oglaf and now I know about it. See how that works?

        I don't want to receive "recommendations" based on some crappy AI algorithm's attempt to classify feeds ("Hey, you like Antipope, try this SF publisher feed that's just a long series of adverts?"), let alone based on the fact that somebody whose brother I work with likes a feed, and I especially don't want to receive "sponsored recommendations" which is the monetization step for this type of "social feed reader".

  25. volk007 says:

    +1 for NewsBlur, switched immediately, using ever since.

    • homer says:

      i stopped using it once i realized arrow keys were bound to next/previous articles instead of scrolling down/up.

      • Pavel Lishin says:

        Not defending that idiot default decision, but it is something you can change.

        The worse news is that you have to adjust the percentage by which it scrolls, instead of using the browser default like a sane person would.

      • Kevin says:

        This is now a preference that can be changed.

  26. aerique says:

    I went with Blogtrottr. Covers syncing if you've got IMAP and works on all devices that can read e-mail. With procmail one can make the filtering as crazy as you want (I just dump all of it into an RSS folder).

  27. Ben says:

    Here's a partial solution: Use TheOldReader. Although it is web-based, as of yesterday the (free) iphone/ipad app FeedlerRSS supports it, and no swiping or other gesture controls required. It even has a 'next item' button. What will they think of next?

    So, you're still stuck with a web interface for the desktop until they get an existing RSS reader app to support their API, but at least the website also accepts keyboard controls and doesn't try to do anything too clever.

  28. Wood says:

    Check out https://www.inoreader.com/ - tried the usual suspects like Feedly, OldReader etc and did not like them. I have not found anything bad yet about inoreader and about as close to the Google reader as I could get.

    No iOS app yet :-( but I hear they are coming.

    NetNewsWire might have have a new version soon so keep an eye on that as well.

    • jwz says:

      That appears to be a web site, not an application. Was I less than clear about my disdain for that approach?

      • Wood says:

        Well basically your screwed! There is nothing! I am sure someone will build something or as they say if you have that much of a personal need its time to get your hands dirty :-)

  29. https://github.com/swanson/stringer - anti social personally hosted RSS reader(albiet on heroku, unless you're smart enough to self-host it) with ReaderApp support. Doesn't solve JWZ's immediate problem (until reeder's updated on the ipad/desktop), but might fix someone else's problem.

    Using feedly, until I find a desktop that that decided does not suck. I do not have high hopes for this.

  30. For iOS, I've found newsify pretty good. They can use feedly as a backend. http://newsify.co/

    • jwz says:

      Trying it now. It seems ok, but there doesn't seem to be any way to turn off or speed up the dumb animation when going from one page to another; and the "Next" button is tiny, and at the bottom of the screen on the right.

      What is it with these people who don't realize that "Next" is the single most important command in the UI, and persisting in making it hard to execute??

      • NotTheBuddha says:

        About a year ago I mentioned that your your DNA picture albums were the only place I'd seen that had single-click to advance.

        It still is.

      • nooj says:

        it's braindamage brought on by apple. in every aspect of the ui, non-swipe navigation is always pushed into the corners or edges.

  31. Jeroen says:

    The most sensible take on the reader apocalypse is this one: http://the.taoofmac.com/space/blog/2013/06/22/2322#the-long-farewell-to-rss

    He describes how he uses rss2email to store everything in IMAP and then use an email client to read the news. This allows for synchronization, and you get many many clients that support all kinds of alerting, browsing and favoriting stuff.

  32. MSD says:

    What about newsbeuter, yes its a CLI application but, you your site makes me think you're not afraid of the CLI.

    http://www.newsbeuter.org/

    SSH to the box hosting newsbeuter from your 'device' - read you feeds. Done.

  33. kellyu says:

    I'm giving Newsblur a long term trial - the iPad version hangs a bit sometimes, but that might be related to it being an iPad1; it has been a bit dodgy for everything recently.

    Once I switched off a few things it worked better, but I really need to clean out the 'Saved Stories' folder, as I think it comes from the recently starred collection, and it rubs me the wrong way to see 176 unread items in Saved.

  34. Alberto says:

    putting your .liferea dir on a dropbox and symlinking to $HOME works, apparently (just tried now), but then you sync only between linux clients. i keep on dropbox an opml list of my feeds, and sort of sync manually using feed readers that understand opml. you're right, it's such a trivial feature, it's a pity it's not there.

  35. Daniel says:

    I would have thought you'd read this (why the "just putting .newsrc in DropBox" approach doesn't work), especially given some of your epic rants on file data structures in Netscape Communicator.

    http://inessential.com/2011/10/25/why_just_store_the_app_data_on_dropbo

    • jwz says:

      I hadn't read that, but I had thought of those problems, and I believe that that page is an odd combination of "giving up too easily" and "not accepting that less than 100% multi-year fidelity is actually totally ok."

      The only real argument against "just store a .newsrc" is bandwidth: that it's more efficient for a remote server to poll the feeds of a hundred thousand people than letting each phone do it themselves, but I'm not convinced that makes much difference to the user's wait-times, in This Modern World.

      If there was a centralized, reliable RSS-polling-and-redistributing service like Google Reader used to be, that sure would be handy -- but that's how we got into this mess in the first place, and it's an object lesson in relying on someone else's infrastructure. They can take their toys and go home at any time.

      • Daniel says:

        Not losing user data is the prime directive of any application.

        not accepting that less than 100% multi-year fidelity is actually totally ok.

        Which of these statements is correct? Because if your syncing mechanism is not 100% no-BS-bulletproof, the combination of multiple clients updating and the 'exciting' reliability of mobile data pretty much guarantees a data consistency clusterfuck.

        See also: the debacle that is Apple's Core Data syncing

        • jwz says:

          All problems can be reduced to simplistic, single-sentence aphorisms.

          Usenet taught us that beyond a certain time threshold, you can just throw everything away and if it's important, it will come back. That's the fundamental and essential difference between personal email, and drinking from the firehose.

          Opinions on what that threshold is will vary, of course, but I posit that nobody will bitch if your system doesn't support reading an article from a feed; having the article disappear from the feed for a year and then somehow come back; and having it still be marked as read. Which is the kind of 100%-fidelity scenario that author seems to be assuming is a requirement for declaration of victory.

  36. ukasz says:

    Have you tried bazqux.com? You do have to pay, but the "feel" is simmilar to reader imho.

    • Josh Jordan says:

      Good call! Bazqux (first 30 days free, $9/year thereafter) implements the Google Reader API, so it works with any of the old Reader mobile apps with a trivial change: the app developer only has to change the hostname from google.com to bazqux.com. I use it with Mr. Reader on my iPad.

  37. bleh says:

    Gwene. Gnus.
    Now we just need a good newsreader for iOS/Android/whatever. With .newsrc-on-Dropbox sync support.

  38. phuzz says:

    In other 'death of RSS' news, Twitter have killed off their 1.0 API, so you can no longer get an RSS feed of twitter posts.
    I've started trying http://www.twitter-rss.com but those feeds keep popping up with adverts and 503 errors.
    The only other workaround I've found is to create your own dummy twitter account and use the new API (which requires a login before it will give you any information).

    • extra88 says:

      I tried that Twitter RSS host for about five minutes before experiencing those problems. In that time I also realized I was an idiot because I was already in the process of setting up Tiny Tiny RSS as my Google Reader replacement so why not host my own Twitter parser? I set up Twitter-RSS-Parser this morning, seems to work. The problem with this approach is the Twitter API rate limiting.

      I've been using NetNewsWire 3 synced with Google Reader on my desktop and Google Reader's mobile web interface on my iPod touch. For now I'm giving up on a desktop app (I'm not that wedded to it) for Tiny Tiny RSS's native web interface and trying out mobile web clients that use its JSON API. TTR's developer recommends ttrss-mobile and it works but its typical jQuery Mobile UX is too slow and lacks features I want. I'm going to try g2ttrss-mobile which is patterned after Google Reader's mobile design.

      • phuzz says:

        I've been trying to avoid having to set something up that I have to run myself, but right now it's looking inevitable.
        So, what's cheaper and easier? 'Proper' hosted server, or AWS instance or something else?
        Maybe I should just ask my boss if I can stick something in our server room, I do have a Raspberry Pi sat around doing nothing...

        • extra88 says:

          You can find guidance on setting it up on Amazon's free tier, a Raspberry Pi, or as a Heroku app. Heroku might actually be the easiest, even though it's a weird platform compared to what the TTR developer expects, because I saw an article that appeared to detail the steps pretty well and provides scripts to simplify it. I don't know what it would take on Heroku to add extras like a separate web client interface or something like the Twitter parser (btw, I definitely prefer g2ttrss-mobile). I also didn't look to see what the options are for something like cron job so feeds don't have to be fetched at the moment of login.

          The Pi would probably take more work because you'd have to first set it up with a LAMP stack.

    • jwz says:

      Yes, this is why I had to change the default text-source in xscreensaver from the Twitter firehose to the Wikipedia firehose.

      The new Twitter API even has a "sample, debug stream" that returns a small selection of recent twits -- not the full stream, not at full speed, but just some -- but even that requires you to log in first.

      Jerks.

      Even people who work there think this was a dick move.

      Oh well. If you don't want me to use your service, I'll go ahead and just not use your service. May your alienation of long-time users slide you toward irrelevancy.

      • Tim says:

        I've been wondering why they're being jerks. What's in it for them? There must be some perceived value to them.

        Also offensive are the dumb lies offered up in justification. Nobody uses RSS anymore, it's so yesterday. That is why we are moving to more difficult and rate-limited APIs which require login. Users will surely find this more convenient.

        • extra88 says:

          How do you monetize anonymous, unlimited access to the data you control? With a walled garden you can show ads to account holders and you can get other companies to "partner" (i.e. pay) for access to better APIs.

          "Who needs RSS? Sites should just tweet what they want you to know and people should follow them." I'm seeing sites, even non-commerical ones, following this trajectory, seeing RSS as the domain of tech nerds and replaced by Twitter follows, Facebook page likes, etc.

          • Jeff Warnica says:

            They monetize the data by also including paid advertisements. As relatively valuable targeted ads to well-profiled eyeballs are, there is value in relatively random ads, as well. Such things can still be GeoIP'd and what not.

            I'm not sure I disagree with the strategy of skipping RSS postings for Twitter or Facebook updates. I mean, I'm sure the math supports it.

    • Michael says:

      Fortunately Twitter has been putting the "tweet" in a DIV with a predictable class which hasn't changed since I've been scraping (at least 6 months).

      With BeautifulSoup, tweets = soup.findAll("p", {"class": re.compile(r"tweet-text")})

      Glue to some RSS generating library, serve with CherryPy or your other Python web solution of choice.

      A long way from a real solution, but it let me get my Twitter feeds back in to my RSS reader (incidentally I use Yarssr and quite like it) with a minimal time investment, and it lets me feel pleasantly like I am giving them the middle finger while I do it.

      Fuckers.

    • phuzz says:

      Update for interested persons:
      Now twitter-rss.com seems to have disappeared, I ended up following the instructions here, which contain a script that can run out of your google drive (ok, I don't really understand that bit), and will provide an RSS feed of pretty much anything you can get out of the twitter API. The downside is that you have to create your own twitter account to get an API key.

      Still, the method works fine, and required very little interaction to get it running.

      I'm now trying to work out if Miniflux is going to be any better than theoldreader.

  39. paul says:

    I've ended up using ownCloud - its RSS plugin does everything I need (folders, a "Mark all as read" button, star things to save for later). Was using Feedly here too, but TBH the Reader thing has me wary of just hopping between things hosted elsewhere that could decide to shut up shop whenever they feel like it.

  40. Hendrik Mans says:

    Complains about a free, ad-supported service everyone was relying on shutting down and then bashes a sustainable, independent effort for charging 2 fucking US dollars a month.

    Seriously, if even smart people are this short-sighted, I give up.

  41. jim says:

    who the fuck uses a feedreader in 2013?

    • phuzz says:

      You see all those comments above yours? Yeah, all those people use RSS.
      What's that? You don't use one? Wow, what a special fucking snowflake you are.

    • tape says:

      everyone I've ever heard of EXCEPT you, apparently.

      of course I've never heard of you, so.

  42. Jason says:

    Been warming up to https://mnmlrdr.com (Minimal Reader) lately. I recommend it.

  43. http://www.noowit.com/ past a good google reader replacement that turns any feed into a digital mag you can also curate and publish your very own mag which can edit itself according to who’s reading it based on their interests.

  44. Andy Bennett says:

    Hi,

    Have you tried newspipe? You give it your OPML file, specify an eMail address and it sends you a mail for each item. It includes headers such that you can filter them easily it also includes a link to the original article. It can be configured on a per-feed basis to send the actual article or the RSS summary. SEEN state and syncing is managed by your eMail client so works the way that you might be used to.

  45. Miranda M. says:

    I have no idea what you just wrote, but it was hilarious. that is all.

  46. Brian says:

    I have taken to using newsbeuter on my home server, and connecting to it through SSH wherever I am to read the news. I use w3m to open any pages I'm interested in reading more about. If it is something I need a GUI based browser to interact with, I just open it in chrome on the local machine. I find this to be an amazing solution, and there are SSH clients for all major desktop OS's. I also believe there are SSH clients for iOS and Android. I think I have found my solution. The best part is it lives on a box I have full control over, and is an open source project, so if the guy maintaining it gives up on it, I or someone else can always adopt it and keep it up. If you are comfortable at the command line this is by far the best option IMHO.

    I am adding your blog to my feed right now.

  47. lairdb says:

    Mr. Reader on the iPad, hands down, backed by any of several aggregation services (bazqux.com is the cleanest UI.)

    Doesn't satisfy the desktop app desire, but Mr. Reader is definitely the cleanest, purpose-focused, feature-configurable of the iPad readers.

    • kanye says:

      I'm not seeing an option to select bazqux.com - its hardcoded to google reader. Or have you installed a dev version?

    • EllenH says:

      Mr. Reader version 2 is out today. I've never used it before (was a Reeder addict, on iPad) and am quite impressed. I'm using it with Feedbin, because I prefer paying and being a customer, not a product... I've explored most of the other possibilities and they all have things I really don't like. All except Mr. Reader, for now anyway. My problem seems solved! Yay!

    • Jay Allen says:

      I could not agree more! It's gotten to the point that I actually prefer reading RSS feeds on my iPad because of Mr. Reader, which is a huge endorsement for me.

      In fact, I've basically whittled down my options to only services that work directly with Mr. Reader which, thanks to the developers embrace of interoperability, are a LOT of them.

  48. Ryan says:

    Oh. My. God. A whole $2 a month? Too rich for my blood.

  49. Thomas W says:

    RSS feeds are the new newsgroups.

    You're right. We can't have nice things. :P

  50. john smith 69 says:

    Call me old school, but I don't understand this requirement to have all your news presented in a consistent format. Why not just open your web browser and go to the news web sites you like?

    And what's up with "remember which stuff I've already read"? Let me explain how I do it: I read the news articles until I start seeing stuff I remember seeing before. Why do you need software to do this?

    • Because not everything is news? Because visiting 100 web sites every day when less than half of them have posted content is a terrible use of time, energy and bandwidth? RSS readers enable following many, many low-frequency sites with very low mental overhead. My comics folder alone has 30 feeds in it. Update frequencies include daily, weekdays, three-a-week, weekly, and completely unpredictable. If I kept a folder of 30 bookmarks and memorized the update schedules for all of them, I wouldn't have time left over to feed trolls. When new content is posted, the reader shows me. If there's nothing new, I haven't expended any time or energy looking for things that don't exist yet.

      Call me old school, but I also prefer to read posts from oldest to newest. While feed readers let me sort articles that way, I'm not aware of any web sites that are willing to rearrange themselves similarly.

    • chad says:

      Scott Ventura above gives a good reason, but the reverse of his good reason is also a good reason: it's not just about having a computer do the boring, repetitive work of caring about when several/many/dozens of websites update, but also about being able to look at things when I have time, and having the computer do the boring, repetitive work of keeping track of which things I've looked at.

      In my case, I have about 150 feeds in 7 categories. In a normal day, I read ~20 of them (with 0-3 articles each). On a slow day, I might read and/or skim 30-60. On a busy but not overwhelming day, either 24 (the web comics), or 1-3 prose text feeds that seem interesting and quick. There is no way I would do that by hand.

  51. ENKI-][ says:

    I hear good things about 'newsbeuter', which is 100% curses -- no GUI so far as I can tell. I don't use it, because I don't really use RSS feeds, but I know someone who subscribes to a couple thousand feeds and keeps the damn thing open for months at a time, and has never had any major problems with it.

    On the flipside, this guy claims that he has never had any major problems with Gentoo (after I tell him about the last three times it nuked my disk), so maybe you should look for a second opinion from someone who doesn't emit supernaturally strong anti-bogon fields.

  52. Naum says:

    I (mostly)? like the Feedly web app. Like GR, can be driven by keyboard and while I understand the grievances listed, I don't use RSS like that -- I subscribe to 3K+ feeds and view items as a river that I wade in, sometimes in more depth, depending upon time. But I like that I can tap 'GG' and quickly get to an index of item links for a particular site. The only gripe I have is that the Feedly overlords deactivated the Search function because of the huge uptick in new subscribers and resultant performance demand -- but it is on the todo list to restore.

    The Feedly iOS / Android apps, however, are not to my suiting -- for the reasons enumerated above but also the font size is not adjustable (important for 45+ year old eyes) for summary and title text / and doesn't obey global Android font setting either.

  53. Chris says:

    +1 for Curata Reader. It's minimal, unfussy, and all about reading.

  54. Steve Webb says:

    Jamie -

    I've been a fan since your spinny netscape icon thing. You're +1 in my book.

    My advice: Fuck the cloud. Fuck Dropbox. Fuck them all. I switched to tt-rss. I host it myself on a tiny VM in my basement. Nobody will take it away from me. The Android app for it is awful so I've reverted to only reading in a browser but at least the UI is GReader-like. Has hotkeys. Easy to modify.

    • crowding says:

      I suspect that jwz does not want to sysadmin his morning newspaper.

        • crowding says:

          I'm a bit younger than jwz but I tried building feed scrapers with XSLT and whatnot around 2003 as well. The lesson learned was to to treat the lack of a feed as a sign that the website in question does not actually want my attention.

          • reeses says:

            Well, that's just insane. But I meant the portalizer. I had elisp for site scraping because every site was a special bloody case and gnus was there anyway.

  55. jonathan says:

    I've started using Newsify on the iPhone. I don't know if there is an iPad app or not. It uses the feedly back end, but the interface is actually understandable, unlike feedly's.

    I understand why all these apps are going to feedly. It's the path of least resistance. Feedly was the first to reimplement the google reader API. And the whole advantage to using google wasn't just the shared state, it was that you didn't have to really dick around with rss to create a reader. It's just json objects and a single request, instead of 200 simultaneous http requests all spitting back different XML schemes that call themselves rss or atom or whatever.

    That said, it's not terribly hard to do that.

  56. crowding says:

    No one's mentioned Feeddler. That's what I was using on iOS, and I was casting about for alternatives because there was no word from the dev for a while, tried feedly and hated it.

    But Feeddler free version just got updated with support for Old Reader, BazQux and FeedHQ backends, with the paid version on the way, so I'm going back.

    * It doesn't have a box layout. You get a list of feeds in the left panel and a list of articles in the right. Joy!
    * The UI has you performing actions by touching visible targets with your fingers! What will they think of next?
    * One of the visible targets is a fucking next button! You can configure what side of the screen it's on, even!

    • Jon Konrath says:

      I used Feeddler up until the apocalypse, and thought it was okay, but looked for alternatives because it's got a single hobbyist developer, and almost no documentation aside from a personal blog, so I was almost certain it would never get updated. It's good to hear it may support Feedly as a backend, but I'll believe it when I see it.

      I've been using Feedly since the switch, and I don't love it, but I don't hate it. I do 80% of my reading in Safari with the plugin, and it's bearable, but only because it supports keyboard shortcuts. The swiping in the iPad app is annoying, but at least they didn't use device shaking for controls. It's occasionally clunky on my first-gen iPad, and I get frequent crashes when I open a web page from within it, but then I'm getting more and more app crashes in the old iPad, and the only fix seems to be upgrading to a newer one.

      My biggest fear with Feedly is that they'll get bought by google and then shut down a day later.

  57. dude says:

    who is this 'jwz' person, why does he still use rss feeds, and why am i reading fluorescent green text on a black background?...

    • phuzz says:

      You'll find JWZ's wiki entry as the third item if you search for 'jwz'.
      He uses RSS feeds because they're a useful way of getting information out of lots of web pages at once.
      As for the third one, don't worry, that's just the drugs kicking in...

  58. Michael Lenaghan says:

    I ended up going with Feed Wrangler:

    http://feedwrangler.net/

    They have a web UI, an iPhone app, and an iPad app--and they handle syncing through their own back-end. (A desktop app is "in the works.") It's paid-only; for me that's currently a plus. I don't want people trying to make money by selling what they know about me.

    Here's a pretty balanced review:

    http://www.macstories.net/reviews/feed-wrangler-a-new-rss-reader-with-smart-streams-filters-read-later-integration/

    I agree with the conclusion: it's definitely 1.0 in terms of feature set, but it's a very good--and very stable--1.0. (Having said that, like Feedly it has no documentation. Shock: not everything about the product is immediately obvious.)

    • Jay Allen says:

      Dammit. Reading over Feed Wrangler's feature and that review made me think I'd found my ultimate solution. But I try to sign up and simply get redirected to the main page. I try to login and get Invalid username or password. Retry ad infinitum. Am I missing something?

      • Michael Lenaghan says:

        After signing up you do get redirected back to the main page, which then becomes your "dashboard". Obviously I have no clue what's going wrong in your case. All I can tell you is that I'm currently logged in and it's currently working. Not much help, I know...

        Over the last few weeks I've tried Digg Reader, Feed Wrangler, Feedbin, Feedly, Go Reader, and NewsBlur. Feed Wrangler is still my preferred choice given my particular workflow.

  59. Rich Forster says:

    The Feedly web and app interfaces may be shocking, but I think the value lies in the backend, Normandy. Lots of the Google Reader based apps are moving over to use Normandy, so in theory you can keep using the apps you like and they just sync to Feedly rather than Google. I'd much rather stick with Byline than use that awful Feedly app.

  60. Pablo Massa says:

    First world problems.

  61. Chris says:

    The Old Reader is the best replacement I have found, tried Feedly and few others. The Old Reader is nice and simple.

    • jwz says:

      And I quote:

      Currently everyone and their mother is scrambling to write some new web-based replacement for Google Reader, but I don't want that, even a little bit. I want a pair/trio of apps that synchronize. I have no interest in reading my feeds through a web site (no more than I would tolerate reading my email that way, like an animal).

  62. edoceo says:

    Thanks to all who commented. It validates some of my assumptions about needs of RSS reader consumers, showed me some new things too. Thanks to JWZ for starting the conversation

  63. Ken says:

    I'm in Jamie's boat about this situation. I will not use a web-based RSS reader. I do probably 95% of my RSS reading on the Mac, where I have used NNW for years. I use Byline on iOS, mostly because I can limit the number of feeds that show up there. Byline has been updated to sync with feedly. That's fine with me since I have no big opinion on any of the new services; I just need a backend to sync with. Since NNW3 and Byline have been happily syncing with Google Reader all this time, the feed list and state I have in feedly are horribly out of date, and feedly doesn't seem to have any mechanism to either import my OPML or reset the sync from Google Reader. And since, if I want to keep using Byline, no Mac RSS reader supports feedly yet, nor will any likely support it before July 1, I'm in a total state of confusion about what to do. I still want to use NNW, but am really pissed at Black Pixel for refusing to say ANYTHING about their syncing plans, even if they'll have anything at all in place before July 1. I consider that really disrespectful to their customers.

    • jwz says:

      I certainly eagerly await the day where I can host my own back-end for my feeds and then point Newsify-or-whatever at my server. Because it's none of Newsify's or Feedly's or Google's goddamned business what feeds I subscribe to.

      Hell, Newsify doesn't even talk to Feedly over SSL. What year is this?

      • Ken says:

        I'd love to just host my own server for my feeds, but I just don't have the facilities for that.

      • reeses says:

        The thought experiment starts with "I will add a zone file to my own DNS to spoof the service" and ends with hitting my head as hard as I can on my desk.

  64. Rubix says:

    Why are you people so against using a website to read mails or feeds?
    Is the webrowser such a bad platform? It comes with portability across all OSes and devices, and features instantaneous deployment of software updates. Those are good for both developers and users.
    It seems to be possible to create UIs in JS as good as with client-side technologies.
    It is about the loss of control of the user on its data? That's about the only possibly serious reason I could find.

    • jwz says:

      It seems to be possible to create UIs in JS as good as with client-side technologies.

      So you say. Maybe this is the case in theory, but in my experience, practice differs dramatically. Web-based UI is always shit compared to "real" applications. I'm sure web-based UI will catch up eventually, but it's absolutely not there yet.

      Also they don't work offline. One of the fantastic use-cases of RSS reader apps is that you can pre-load everything and then read your feeds on a plane or train with no network.

      • Ken says:

        Agreed. Oddly, probably my favorite update for iOS 6 was that the Mail app finally stopped throwing up tons of dialog boxes when I opened it in the subway. I LOVE apps that behave offline. Another reason I love Byline.

      • Rubix says:

        Caching for offline reading is very useful, but often the full content isn't contained in the feed.
        Apps like Pocket or Readability handle offline caching well enough, at least for text, so I'd rather have the RSS reader send them the urls. I don't know if any tool exist for this, I currently do it manually.

        I hope that one day we will have web apps that can work offline. And I sure hope that their UI won't suck as much as they do now. Gmail seems good enough for online mail-reading, though.

        • jwz says:

          Caching for offline reading is very useful, but often the full content isn't contained in the feed.

          It is for all the feeds that I read, because if they don't include full bodies, I unsubscribe. I interpret that as them saying, "please don't follow this web site" and I comply.

          • Rubix says:

            Yeah, but at the end of the day, if it's a content of quality, you still want to read it. Using RSS only for the notification of new content is still better than hitting F5 manually everyday.
            Maybe I should just find some tools that generate RSS by parsing a webpage.

            • Ben says:

              page2rss.com is the one I use. It's crappy, but it makes me feeds for a lot of 1.0 pages.

  65. Alex says:

    The Sage RSS extension for Firefox uses the bookmarks API to store your feeds. It would be a decent project to get it to store .newsrc information somewhere that Mozilla Sync picks up and synchronises. That would provide synchronisation across anything that will run FF (which you don't).

  66. Frank Wales says:

    Meanwhile, since Google Reader was the only RSS client in the world, unrelated RSS feeds at Google have also been turned off.

  67. Helena Martinez says:

    I have been using feedly for few days. Now I have switched to http://silverreader.com which is much faster than any of readers and also I got news much faster on sites that publish often.