FAQ: Browsing and Reading

How do I mark and retrieve all articles in certain groups?

There are two ways to do this:

If you want to do this on an ad-hoc basis after you've retrieved the headers, press Ctrl+A to select all of the headers and then select Online | Get Marked Article Bodies, or Article | Mark for Retrieval.

If you always want to get the bodies automatically for a particular group, select the group and use the Group | Properties option to open the Group Properties dialog. On the Retrieving page, check "Override default settings" and then check "Retrieve bodies for all new articles."

If I retrieve new headers, but I don't read any articles, how do I prevent the headers from being marked as read?

By selecting Group | Default Properties (or Properties if behavior is desired only in current group) and removing the X from the "New Article Headers" box in the "Marking Articles Read" category.

What does the "Catch Up" option do?

Here's the question that one of our beta testers sent us:

> I was wondering if I could get an explanation of the function of the
> catching-up choice in the online menu.  If, after being away, you want
> to "catch-up" what is the difference in using this choice and the mark
> all articles as read choice?

Agent remembers the number of the last article it has retrieved from the news server. When you "get new headers", Agent gets all higher-numbered articles. The Online | Catch Up With Selected Groups command queries the server to find out the highest number, but does not retrieve any headers, so that the next time you get headers, you'll only get articles posted since you did the catch up.

The Article | Mark All Articles Read command simply marks as read (changes them from red to black) all articles that you have already retrieved. It doesn't affect which articles Agent will retrieve the next time you contact the server.

BTW, another way to "catch up" is to open the Default Group Properties dialog and temporarily change the "Limit number of headers per session" value. That way, you can catch up by retrieving, say, no more than 50 of the most recent articles per group.

How do the auto-thread-expand options work?

There are two options on the Navigation tab of the Options | Preferences dialog box that determine whether threads are displayed in expanded or collapsed form:

Show threads initially Collapsed: If you check this option, all threads will be shown collapsed when you first open a newsgroup. If you un-check it, they will all be shown expanded.
Automatically expand threads containing new articles: If you check this option, threads will be expanded when you jump into them with the "Skip to/View next unread" commands. They're expanded only if you've already retrieved and read the first article in the thread -- the idea here is to avoid expanding threads that you've never seen before (and forcing you to step through them one article at a time). Only threads that you've previously shown an interest in are expanded.

This feature has been confusing a lot of our beta testers, and we're looking at ways to make it simpler. In addition, the option to automatically expand threads when you jump into them doesn't seem to work if you are in three-pane mode and have the focus in the body pane. Keep watching these pages.

Why does it jump when I Mark or Ignore an article?

This is caused by one of the new navigation features added in version 0.46. Agent automatically skips to the next unread article when you mark an article for retrieval or ignore a thread. If you do not like this behavior, you can disable it by selecting Options | Preferences and clicking on the Navigation tab. Now click on the Marking or ignoring an article skips to the next unread article option to remove the check mark and press OK to dismiss the dialog box.

Why doesn't the date appear in the article list?

Actually, the date does appear to the right of the article header. If you use the multi-paned view, it is probably off the right of your screen. Try using Zoom mode. Choose the Navigate | Zoom Window menu item, or click the maximize button in one of the individual panes. Agent is pretty smart about switching back and forth between the panes automatically, and you can switch manually by clicking on the tabs at the top of the window.

Why can't I make the status column smaller?

In addition to the line count, the status column has to display all of the article status icons (Mark, Watch, Ignore, Keep, Thread, Body). We've limited the minimum size of the column so that there is room for all of these. We found that if we didn't do that, new users would make the column so small that some of the information would disappear and then they'd get confused because it was missing. You can also use the Zoom command, which can be handy to make the article browser fill the entire window and give you maximum width for the subject column.

How do I suppress the header details when I'm reading an article?

Remove the checkmark from the Article | Show All Header Fields menu item (or use the right mouse button to bring up the speed menu and remove the checkmark from the Show All Header Fields item). This will start the article just after the header.

Note that this won't eliminate a header that somebody else quoted and included in the article. That makes it into part of the article body, where it is always displayed.

Why does some word-wrapped quoted text appear in the wrong color?

When you're viewing an article that contains quoted text, you may sometimes notice that the text is word-wrapped and that the wrapped lines appear in black instead of blue (or whatever color you've chosen for quoted text). This is actually a problem with the article itself, not with Agent.

This happens when the newsreader that created the article decides to word-wrap the quoted text and does not put a ">" at the beginning of the wrapped lines. Because they don't start with the quote character, Agent has no way of knowing that they were originally part of the quote, and so it displays them in black. (To prevent you from perpetrating erroneous quotations like this, Agent doesn't wrap quoted text. See the section on word-wrapping in the editor for more information about that.)

If Agent wraps part of a quoted line, it does not provide a ">" character but still displays it in blue. It can do this because it knows that the new line is really part of the quoted line above it.

How does word-wrap work when I'm reading an article?

The Article | Word Wrap menu item (or the Word Wrap item on the speed menu that you get by clicking the right mouse button while viewing an article) determines whether the article text is displayed word-wrapped. If word-wrap is on, all of the text will be visible no matter how small your display window is, but it may be hard to determine the article's original format. If word-wrap is off, the article will appear in its original format but you may have to use the horizontal scroll bar to see all of it.

Why do long articles appear to be truncated?

If you download a very large article, the display may only show the first part of it (the first 60,000 characters), followed by the message

[Sorry!
 This version of Agent cannot display articles longer than 60k bytes.
 We will remove this restriction in a later version.
 In the meantime, you can view the entire article by
 printing it or by saving it as a file.]

The entire article has actually been downloaded, and if you save or print it, it will appear in its entirety. The restriction affects only the display, and will be fixed in a later release.

Why do I get "This article is no longer available" errors?

This means that your news server no longer has the article that you just tried to download. This can easily happen if you have your purging settings defined to keep article headers for a fixed number of days or to never purge them at all (see the Purging tab on the Group | Properties dialog box). Because news servers keep articles for an unpredictable amount of time, the only safe settings are Purge All and Purge articles no longer available on server.

In earlier versions of Free Agent, misleading information from the server could also cause this problem even if you had the Purge articles no longer available on server option set. The following topic describes the reason for this and the fix that we implemented in Free Agent 1.0 / Agent .99.

Why do the latest versions retrieve headers more slowly?

One of Agent's purging options is "When the article is no longer available on the server". In previous versions (prior to Free Agent 1.0 / Agent .99), it would do this by checking the range of available articles reported by the server and purging all articles whose numbers were before the available range. For example, if the server reported that articles 204-566 were available, Agent would purge all articles with numbers 203 or less.

But our beta testing showed that in many circumstances there would be gaps in the reported range of articles. This would typically happen when one article was posted in a way that would cause it to remain on the server much longer than normal. This could happen if an article was posted with an explicit expiration date, or if an article was cross-posted. (In many cases, cross-posted articles are not deleted from any groups until they are deleted from all groups.)

These gaps would cause Agent to think that articles were still available when they really weren't. This would cause two problems for the user. First, you might try to retrieve the body for one of these articles and then have Agent report that the article was no longer available. Second, Agent would fail to purge large numbers of articles if there were really big gaps in the range.

In the latest release of Agent, we solved these problems by having Agent request a complete list of articles from the server. With this list, Agent is able to purge all articles no longer available, even if there are gaps in the numbering range. Agent gets the list of articles as part of the get-new-headers operation, after it has retrieved header information about new articles.

Our tests showed that retrieving a complete list of articles adds only a few seconds to the retrieval time for most groups. However, we have also found that some news servers are very slow to return the complete article list. If you've noticed that Agent pauses during get-new-headers, then you probably have one of those servers. You can eliminate the delays by turning off the Always get complete list of server's available articles option on the Group Properties | What to Purge dialog. This will make Agent work the way it used to. The downside to turning off the option is that the "no longer available" purging option may fail to purge all unavailable articles, leading to the problems discussed above.

What do I do about a "too many articles to retrieve" error?

This error message is displayed if the sum of the articles currently in Agent's database for the group and the total number of new articles that may be retrieved (i.e. the total number of articles in the range of new articles available from the server) is greater than 32,000. Of course, the total number of articles you end up with after the retrieval may be less than this. For example, Agent may end up purging some old articles. Also, there may be gaps in the range of available articles. But Agent has to prepare for the worst, which entails allocating enough space to store all old and newly-retrieved articles. Here are the workarounds:

Tell Agent to retrieve a smaller number of articles. You can do this with the Online | Sample... menu option, or by setting the Limit number of headers retrieved per session parameter on the Retrieving tab on the Group | Default Properties for All Groups dialog to a value like 15,000. This works well if this is the first time you have visited a group, and you are thus trying to retrieve all available articles. Once you're caught up, the number of articles retrieved in future sessions will be much smaller. (And besides, you're probably not going to read 32,000 articles anyway!)
Do a manual purge before getting new articles.

We plan to make Agent smarter about this in future releases, so that it displays the error only in the rare cases when there really will be 32,000 articles after the download.

[Previous FAQ Page] [FAQ Contents] [Next FAQ Page]