There were quite a lot bugs from older version added in sourceforge. Now we fixed those bugs and also feature requests and put them to 2.90. So 2.90 should have less bugs and more features than 2.80.
And it's available for testing NOW!
What has changed?
*Support for flaging articles
*Filter support has gotten a lot more powerful (still much to do to get it to decent level)
*search box moved to better place
*support for searching articles by flags
*Feedreader Webaccess is skinnable.
*Possibility to limit Feedreader WebAccess to some feeds
*Better favicon support
*Possibility to pause Feedreader update process.
*Custom notification sound support.
*Delete support fo items in listview
*Copy to folder functionality changed.
*GREATEST THING IN MY OPINION: "Show only new headlines then selecting folder". Enable this in Feedreader properties and look what happens, if you select folder.
Where to get it?
Where should I report bugs?
Please report bugs in this thread. And ONLY bugs not feature requests.
WRT the new webaccess bits, it looks like it's prettier; however the default needs a little update or it doesn't work well with custom stylesheets.
At the very least there needs to be a body with a set of colour attributes on it so that text appears visible.
I changed the default setting for the form to :
<body bgcolor="#ffffff" text="#000000" link="#0000ff" vlink="#000070" alink="#d1afff">
<td bgcolor=#FF7F00 colspan=2>Feedreader WebAccess</tD></tr>
<td bgcolor=#1F558B valign=top>[CHANNELLIST]</td>
<td bgcolor=#ECEBE3 >
and this appeared to fix things. The colours might not be exactly suitable, but they're good enough. If you've got better colours replace them on the body tag. At least having that updated will make the process a little easier.
As an aside to this, Ctrl-A does not select all the text on the web page template; it probably should.
The favicon support does seem to be /much/ better in this version; cool :-)
Two things I noticed, though...
First is that I'd already added a 'Update favicons' button to the toolbar and on upgrading to 2.90 the entire toolbar was reset to its defaults so I had to reorganise the buttons again to how I like them and re-add the 'update favicons' button.
Second is that on pressing the 'update favicons' button, that is whilst the update is taking place, the toolbar buttons don't highlight properly. Previously it was possible to trigger but I could never get it to be consistent - now I can. Click on the 'update favicons' button on the toolbar. Click on the 'Ok' button on the dialogue that appears. Now float the mouse over the other toolbar buttons; notice that they highlight with blue, but don't un-highlight when you move off. The un-highlight only happens after the favicons update has completed.
WRT Filtering... how's this work ?
I want to delete any articles that end in ' - Sponsored link' or start with 'Ad - '.
I tried using just '- Sponsored link' as the text, or '* - Sponsored link' and nothing seemed to happen. Is it full regular expressions, or just substring matching ? Is it case sensitive ? Does it affect the pre-existing articles, or only new articles ? (ie can I expect to see articles vanish as I add new filters?)
Tiny bug I noticed which is easy to trigger...
Enter 'blah' as the text.
Select 'Delete' from the drop down menu.
Notice that the colour box vanishes (wouldn't it be better to just fade this, rather than remove it completely ?)
Click Add new filter.
Click Ok to accept the name.
Re-open the filters window from the menu.
Notice that the drop down menu now shows 'Highlight' but the colour box has not been restored.
Could double-clicking on a filter just copy its details into the lower fields ? Then you could just edit the entries and re-add the filter to replace the old one (if the name was the same) or provide a new rule (if the name differed).
This can further be checked by using the following RSS file - the script execution tests all show that the browser is attempting to execute scripts against the settings of the flag.
WebAccess for deleted feeds is still possible. This is an earlier reported bug but appears to be still be present.
1. Select a feed in FeedReader.
2. Enable its web access in the properties, and click ok.
3. Open the web browser view by clicking on the mini-world icon on the status bar.
4. See that the feed is in the web list of feeds.
5. Select the feed and delete it in FeedReader.
6. Reload the web browser view (with the reload button, or similar).
7. See that the feed which was deleted is still in the list of feeds.
Great release. I have uninstalled last release because some coding and time problems. Then I have tried some of other rss reader project. They were so buggy that I decided to go back to FeedReader. And this release surprised me a lot - everything seems to be ok now.
Sorry, this is not exactly a bug report but it is specific to this release.
I really like the "Show only new headlines then selecting folder" option. But with the "Show only new headlines then selecting folder" option, it would be better if there is a separate column for feed title along each headlines for easy identification of origins of headlines.
Thanks and great job for the 2.9 release
Is there a way to have the results stored into a sql database instead of an xml file?
Is there going to be an interface for searching via the web?
1. Horizontal scrolling of feed list is flawed. The number of unread headlines doesn't get redrawn properly.
2. Clicking on the feedreader's icon on the taskbar notification area doesn't spawn a taskbar button even though the window is maximised.
3. Column width isn't fixed. It may be reset to a very narrow width.
2.90 has now been released through SF; faults remaining from the RC1...
1. Ctrl-A in the web access template doesn't select all text.
2. Whilst a favicon update is taking place the toolbar icons fail to unhighlight.
3. Unable to check filtering because still don't know how they work.
4. Colour box not restored after adding a non-colour filter and reopening filters box.
6. Deleted feeds are still displayed in the web access box.
7. Horizontal scrolling of feeds still renders wrongly.
8. Column width isn't fixed and may result in tiny Title columns.
The only difference I noticed to 2.90 was that the Manage Extensions dialogue is now populated.
2 more bugs to add to that:
1. When the "Start minimized to tray" option is selected, after clicking the tray icon to open the program, there is no icon in the taskbar (not the system tray, but the list of running programs).
2. The titlebar still says "Feedreader 2.90 RC1"
Currently it is showing a parse error on this feed http://www.bureau42.com/rdf/index.php
Its a fairly small feed as well: Is it because there is a docttype at the top?
<!DOCTYPE rss PUBLIC "-//Netscape Communications//DTD RSS 0.91//EN"
<description>Inspired Weirdness Is Our Business</description>
<copyright>Copyright 1998-2003 bureau42.com</copyright>
<title>Surface Discussion - "101" (Series premiere) (0)</title>
<title>Emmy Award Winners (0)</title>
<title>All-Star Batman and Robin, #1-2 (0)</title>
<title>Weekly DVD Picks - Tuesday, September 20, 2005 (1)</title>
<title>Lost co-creator to work on Ultimate Hulk vs Wolverine (1)</title>
<title>Battlestar Galactica: Flight of the Phoenix (11)</title>
<title>Review: Threshold"Trees Made of Glass" (pilot) (8)</title>
When fetching news from Reuters:
I'm told there are 3 new items, but when I then look at the list your program highlights 3 items a week ago.
And no, its not me, I have marked them all as read, it keep happening over and over - some messages that should be old are suddenly highlighted.
Feedreader displays this feed: http://www.aintitcool.com/index_scifi.xml
But when I double click on the headlines nothing happens. If I rightclick on a subject and select "Open in new window" Feedreader opens a normale Explorer (not Internet) window with the path pointing to its install dir!
It can't parse this feed: http://www.podfeed.net/rss/1.xml
Says there isn't one at the address, but there is.
When I started it today I got an error message saying there was a parse error on reading the config file. (feeds)
I haven't changed the files, so apparently it has saved something it can't read.