Whenever my blog remains quiet for more than a few days, you can bet the reason is that I'm nose-down in development, working on something new. Which begs the question: what have I been working on since I last posted? I'll answer that with a screenshot:
Now, I imagine those of you who are testing the FeedDemon 1.6 beta are more than a little surprised to see that FeedDemon 2.0 is in the works, so I'll explain what's going on.
A few months ago I posted about the possibility of changing FeedDemon to subscription-based software, and the reaction to the idea was...less than positive. As a result, we changed direction and decided not to make FeedDemon subscription-based. Instead, FeedDemon would require a NewsGator subscription in order to use features like synchronization, but you'd still be able to use FeedDemon - indefinitely - even after your subscription expires.
That sounds like a simple change, but given that NewsGator's site (including our ordering system) is based around the idea of software by subscription, a significant overhaul of our site was required - and that process is still underway. That left us with a choice: either release FeedDemon 1.6 after the site redesign, or ship it now as subscription software. Given that we'd promised not to make FeedDemon subscription-based, the only decision we could make was to wait until after the redesign.
Rather than let FeedDemon stagnate, though, we decided to plow ahead with v2.0, and release it as a free upgrade from the current release. So while the downside is that you'll have to wait longer for the new release, the upside is that you'll get some pretty nice new features (including the oft-requested feed treeview) for free.
I'll post more about specific features in FeedDemon 2.0 later on, but for now, I wanted to give a "heads-up" to what's in store :)
PS: If you happen to be testing FeedDemon 1.6 right now, please don't think your efforts are in vain. Even though FeedDemon 2.0 is dressed in a new UI, underneath it's using the same "engine" that drives FeedDemon 1.6, so issues uncovered in 1.6 will still apply to 2.0 (a beta of which should be available in mid-January).
Wicked! - this will make FeedDemon PERFECT! ...
Are you going to do any work on the speed of rendering ? I went back and played with Thunderbird RSS and found the downloading / rendering of feeds soooo much faster than FeedDemon (1.5 and 1.6a). No idea why but feels sluggish. Have top end PC and a 8mb connection.
Posted by: Woody | Tuesday, December 13, 2005 at 05:23 PM
Funny you should ask that, Woody. I spent much of today recoding the newspaper rendering, and it's now roughly 2x as fast as before.
Posted by: Nick Bradbury | Tuesday, December 13, 2005 at 05:32 PM
Now if you could only hide the feeds with no new content, then you will truly have a perfect app.
I've always wanted to share some ideas with you, but never got a response :(
Here is a direct copy and paste from my blog, saving you a click about what I think OPML should be able to do. Now don't concern yourself about the OPML bit. Think of what I said in this post, as a way to redefine a user interface, and how we sort our feeds. i still say death to hierarchies and bring on tagging goodness. anyway, here is the post:
While both SSE and OPML are both young specifications, I want to propose a new form of thinking. I like SSE there is no doubt about that, and I really like Dave Winer’s work with OPML, but it is too static.
Take for example the first MP3 players that ever came out. You sorted the MP3’s into folders that you manually had to change through, and once put on the MP3 player this was the only way to navigate.
Then as people got more and more music, and space became larger and larger, this way of sorting your music became a bit too cumbersome. Metadata, known as ID3 tags came to save the day. Apple’s iPod is a clear example of metadata navigation done correctly. You can sort by genre, year, artist name, album name, it’s quite nice.
Now relate those MP3 files to data. As OPML is in it’s current state, it is quite a chore to put everything into specific folders, hell even them some people just dump everything into one folder. OPML needs to evolve right now while it is still young so that people don’t have to deal with it later when changing the specification can be a difficult transition.
How should OPML evolve? Tagging. Not quite like MP3’s which have fields for all their metadata, just a line with words separated by commas, each word being a tag. In practice, all you need is to add to the OPML specification is a line of XML, so that every item has a tag(s). Block tagging should also be in place, so that instead of folders like you have now, you can have your first 10 items be given the tag “Technology” the second chunk of 10 items in your OPML file be tagged “Personal”.
Why should you listen to me? Because OPML is nice and cute for bloggers to share their blogroll when they subscribe to less than 100 feeds or so, and the author properly maintains his OPML file in some special order. What about people like Robert Scoble who has more than 500 feeds? More importantly, he keeps them all in one folder! How in the world am I supposed to know what’s what? I would have to subscribe to his entire OPML and look at all the feeds to get an idea of what I like.
Tagging in the above situation fixes everything. Imagine an OPML viewer, that isn’t static at all, but instead very dynamic. When you load up an OPML file, it lists all the tags in alphabetical order, yea you heard that right, ALL the tags, however many that may be. Then a user clicks on anyone of these tags, and now all the data tagged with this piece of information is displayed AND all the other tags that are related to it.
Take this example; I am subscribed to 10 feeds in my OPML file. 4 of them are tagged Blogs, of these 4, 1 of these 4 blogs also have the tag Microsoft. Now I also have another 6 items tagged News, and of these 6 items, 3 of them also have the tag Microsoft. So let me make this easier to visualize:
OPML File
Item 1: Tags: Blog
Item 2: Tags: Blog
Item 3: Tags: Blog
Item 4: Tags: Blog, Microsoft
Item 5: News
Item 6: News
Item 7: News
Item 8: News, Microsoft
Item 9: News, Microsoft
Item 10: News, Microsoft
Now when I load up this OPML file into an OPML viewer, there is no predefined folder structure. I merely see all the tags present: Blog, News, and Microsoft. Now when I click on the Microsoft tag I should see 4 items that have this tag, along with two other tags: News, and Blog. I can now click News and Blog much like a sub heading to further break down my list of feeds. So I went from 4 items with the Tag Microsoft, I click the tag Blog which is a subheading and now I am only looking at 1 item.
These seems a little bit excessive for an OPML file with 10 items, but now imagine an OPML file like Robert Scobles with 700+ items. I can easily load it up, click on Technology tag, this list of 700 now gets cut down, now I click on the tag HDTV and this even smaller list is made now even more refined! With two clicks, I sorted through 700 feeds to show me all the Technology feeds Scoble is subscribed to, and within those technology feeds I made only the Technology feeds with the HDTV tag viewable as well.
What I am trying to say is, I love Channel 9 a lot, and I watch all the videos. My first Channel 9 video was when Scoble interviewed Amar Gandhi from the RSS team, and from then on, I was hooked. The thing you guys do with Virtual Folders is amazing, using Metadata to sort information; I figured this could be applied to the OPML specification as well.
I’ve been trying to contact Dave Winer but I suppose he is a busy man since I haven’t heard back from him. As the father of RSS, he has changed my life on the internet dramatically, now I just wanted to get my ideas out to him and to the rest of the world. I might be wrong, I might be right, I’m just telling you from a user perspective how I would like to manage a list, which will only get larger, of items that are of importance to me.
Keep it up RSS Team!
You can contact me for further feedback at [email protected] I hope to help you guys as much as I can
Posted by: Stefan Constantinescu | Tuesday, December 13, 2005 at 05:32 PM
Ah! I was wondering when the next FeedDemon RC would come out. Looking forward to 2.0. Hope the betas come soon.
Posted by: Manuzhai | Tuesday, December 13, 2005 at 06:13 PM
As always your good for a surprise Nick!
BTW, how's TopStyle doing? ... ;)
Posted by: Stefano F. Rausc | Tuesday, December 13, 2005 at 06:39 PM
hahaha sweet :D
I only realised the other day it's been nearly a month since you last posted regularly, and indeed immediately figured you were busy coding. Didn't expect more work on FD though! It looks fantastic!
I do agree that releasing 1.6 as subscription-only is a bad idea. It would burn so many of your existing customers who don't want to use NG sync - I've seen a lot of queries about that on the forums.
Next question - what is NG going to do about the 2 year subscription we were offered? If you make it from the release of 2.0, with the amount of polish you give through beta testing, by the time 2 years is up it will have been as much as 4.5 years since I purchased FD upon release... that's insane value for money.
Posted by: Andrew Herron | Tuesday, December 13, 2005 at 07:55 PM
is it possible now in v2.0, to update all the groups?
Posted by: Anonymous | Wednesday, December 14, 2005 at 03:38 AM
That's a strange (or weird?) but interesting way to render overflowing nodes in a tree view. I guess I've to see it live with some horizontal scroll action to make a decision whether I like it or not :]
Posted by: Björn Graf | Wednesday, December 14, 2005 at 04:35 AM
So just to confirm: I bought FeedDemon back in February of this year, before it was bought by NewsGator... when FD 2.0 comes out will I be able to upgrade for free and use NewsGator's serices? Can I use them now?
Thanks for the clarification. I stopped using FeedDemon for a while to try out an online RSS reader, and then I slapped myself and woke up and said, "Why am I not using FeedDemon? I paid $25 (got a $5 discount from can't remember where) bucks for it and most importantly it's the best RSS reader out there."
Best,
Luis
Posted by: Luis F. Escalante | Wednesday, December 14, 2005 at 05:25 AM
Nick: I don't like the idea of multiple folders.
I don't want just another thing to keep organized in folder.
I just want to throw all my feed and read them when they have unread items.
Can you make, pleaaaase, a way to only see feeds with unread items?? thanks!
Posted by: edddy | Wednesday, December 14, 2005 at 07:40 AM
I love the idea of multiple folders! In essence I think all RSS aggregators need to become OPML browsers too. Look at Taskable.com for example - it makes RSS reading absolutely click and point simple for computer newbies because they can navigate to their choice of feeds by drilling down through the OPML hierarchy.
BTW, I bought FeedDemon and as a power user will be sticking with it but it still needs to incorporate OPML browsing IMHO, even for power users.
Posted by: James Corbett | Wednesday, December 14, 2005 at 10:12 AM
To reiterate Manuzhai's comment, how's TopStyle 4.0 coming along? Please don't spend all your time on this Johnny-come-lately RSS stuff! ;-)
Posted by: Alan Palmer | Wednesday, December 14, 2005 at 11:47 AM
Nick, are you still using Delphi? If not, then what?
Posted by: Mike | Wednesday, December 14, 2005 at 12:19 PM
Nick, you guys continue to impress me. Always do right by your client (which you guys are obviously doing!) and you will reap the rewards. Nice work!
Chris
Posted by: Chris | Wednesday, December 14, 2005 at 12:31 PM
So pleased to see the treeview - can't wait!!
Posted by: Martin Dolphin | Wednesday, December 14, 2005 at 01:24 PM
I can't think of a better present from old Saint Nick ;)
Posted by: Morgan Pugh | Wednesday, December 14, 2005 at 03:47 PM
Those of you still holding out for TS 4.0 are kidding yourselves. There will be no 4.0 as we know (want) it - not in the near future. Nick is completely in love with the RSS scene - even debating/developing standards - this is much more rewarding (to him) than the tired old Top Style code I would imagine. Can't say I blame him either, although we are left wondering about the myth that is 4.0...
What I would like to see is it (TS) released as open source that so some one else could take up the good work and continue on for us TS'ers.
Ralph
Posted by: Ralph Hall | Wednesday, December 14, 2005 at 05:43 PM
Nick this will finally make FeedDemon (IMHO) perfect. THANK YOU for your continued development efforts. This looks GREAT!!!
Posted by: David Orriss Jr | Wednesday, December 14, 2005 at 08:34 PM
Excellent news! I'm glad to see the removal of the subscription tie. Bravo.
Posted by: Steve | Thursday, December 15, 2005 at 12:39 PM
A few comments (without actually having used the software or knowing Windows UI conventions!):
I'd suggest moving the disclosure toggle widgets for the sections in the main content pane to the left, for the sake of consistency and clarity.
How do the search fields differ? If the bottom one searches just the subscriptions in the pane, the label should make that clear.
Both search fields could stand to be a little wider. Editing query strings in narrow fields is a pain.
Posted by: Daniel J. Wilson | Friday, December 16, 2005 at 01:15 AM
I think you should also add sub categories.
For Example, I like to read NFL news, but I'd like to categorize this into sections such as;
NFL
- Team News
-- Dolphins
-- Jets
- General News
-- Site 1
-- Site 2
Posted by: Charlie | Thursday, December 22, 2005 at 10:59 AM
Charlie, I agree that subfolders would be nice. Although FD2.0 will only offer a single-level treeview (ie: no subfolders), this will be added at a later date.
Posted by: Nick Bradbury | Thursday, December 22, 2005 at 11:08 AM