Mailing list archive search
1. What's this?
A full index of the Mailing list archive.
2. How to use it
- Type words to be searched in the search for box
- Select if you want to search the whole of each message, or just the subject lines
- If you are searching the whole message, you can also restrict yourself to specific years
- If you want to exclude a word, precede it with a tilde, eg ~system
- Click the search button
3. Didn't find what you were looking for?
There may be several reasons for that:
3.1. We haven't got it indexed
- There is a lag of a day or so before new messages are indexed
- Even with old messages, our definition of a word and yours may vary. We think this:
- We don't index words longer than 40 characters -- this eliminates many URLs
- We only index words containing letters and digits and a few symbols (FN and 24x7 are words but F[N] is see as F and N -- ie two one-letter words.
- On the other hand, there are no stop words. Every word is indexed.
3.2. It doesn't appear as a word in the message
Sadly, many messages are quite battered by the time they reach the Mailing list archive. Words may be badly split at the end of randomly broken lines....
That's the last time I travel on an untes= ted aircraft in a hurry
....will not index the word untested (we'll have untes and ted)
3.3. You've used operators we don't recognise, so we treat them as words to be searched:
carl or gregg
.... does not do a search for messages containing carl OR gregg. It searches for messages containing carl AND OR AND gregg
You used a tilde. We treat ~ as a NOT operator:
will be parsed as help AND ~me -- meaning a match must contain help and must not contain me
4. Other ways to search
4.1. Other indexes
Is a downloadable archive of the Mailing list. (Windows only)
Google has most of the Mailing list archive indexed, and is usually only a few weeks behind in indexing new messages.
Google isn't aware of the Mailing list archive as a separate part of the site, so its searches amy return more than just messages. You can partially restrict it to the Mailing list archive by adding some qualifiers to the search string, eg:
carl gregg site:www.rebol.org inurl:ml-display
Google also limits the search results to the first 1000 hits.
5.1. How do I exclude a word?
Type a tilde at the front of it, eg:
- carl sassenrath -- searchs for "carl" AND "sassenrath"
- carl ~sassenrath -- searchs for "carl" BUT NOT "sassenrath"
5.2. What about paths and things?
The indexing is designed to work with many words that are common within REBOL.
You can search for paths: e.g. system/words or system/schemes/http, but we do need a precise match on the path (system/schemes does not find system/schemes/http)
Similarly, ! and ? are treated as letters when they are at the end of a word:
- A search for object! produces different results to object? or object
- This has a perhaps unwanted side-effect:
- if a word is the last word in a question, it will have a ? and so will be indexed as a separate word.
- Try a Just subject search for windows? and then for windows to see the issue.
- We'll fix that when we add a layer to do more complex searches
5.3. Why do I sometimes see false hits?
This can happen when different threads have been merged into one. Each message is indexed under its original subject line. Example:
- Do a Just subject search for recursive functions
- There is an entry for 20-apr-2001 that does not have those words in its title
- Click on the thread, and then find and click on the link to message 9 -- ie the [9/10] link
- You'll see that it does have those words in its subject line, even though it is part of a thread in which most messages have a different subject line.
5.4. What about 1980 and 2024?
If you use the date index you can see some messages dated 1980 and 2024. But this search does not offer those years as options. You can search the strange years using LDS -- see search-rml in the LDS documentation.
That's deliberate. There are so few bad-year messages, that the search does not index them.
5.5. email addresses?
They are nor indexed because:
- we do not index the from: field in the email
- we do not recognise the @ symbol as a letter. It's treated as a word separator. So parts of email address that appear in the body of messages may be indexed, but not the whole string.
That's deliberate to make it a little harder for the harvesters.
Tell us about them! Please use the feedback form.