[BETA] BoardTracker Search Mod v1.9.0 - complete board searc

[BoardTracker]

MOD Title: BoardTracker Search Engine Mod

MOD Description: This search hack replaces the built in phpbb search with boardtracker search and results are fully integrated with your board. Search results can be topics from your board or all boards or a select group of boards - your choice! All the search processing and indexing is done by boardtracker servers so this solves any load problems board searching often causes particularly on big boards.

MOD Version: 1.9.0

MOD Download: http://forums.boardtracker.com/download ... _1.9.0.zip

Demo Board: You can see it 'in action' on our own support forums here..

http://forums.boardtracker.com/searchbt.php

http://forums.boardtracker.com/searchbt ... =4&order=1

More details:

First of all its very easy to install!

Results can be formated in any way you want with some simple template variables. Results appear within your site - fully integrated!

You can still keep the existing phpbb search script and a link for 'Advanced Search' exists on the boardtracker search template leading there - just uncomment it to use. You can remove that link completely if you don't want to allow users to have any access to the built in search.

A control panel in the admin section allows you to switch the boardtracker search mod on or off at any time so you can switch between boardtracker and phpbb search whenever you want.

The following configuration options/defaults exist:

Switch search on/off

Switch 'safe filter' on/off

Board ID - you can limit search to just your board, or several boards you choose or ALL boards indexed by boardtracker. This allows you to setup traffic trade deals with other boards or just circulate traffic among your own boards if you have several.

Category ID - you can define a category on boardtracker and searches will only pull results from that category.

Default Sort Order - you can sort results by relevancy or by post date.

Default Search Scope - results can be limited to last day, week, month, year or all time.

Default Search Query - a search term that will be executed whenever the boardtracker search script is called directly with no params. For example you could set it to "blah" and it will just display all threads about "blah"

Stopwords - specify a list of words (space separated) which you want to stop people searching for.

User selections on the search form override the defaults.

New in version 1.9:

This new version (will be released as v2 if/when the moddb opens for submissions again) now features several new search options including: author search, forum and related forums search, BT category search, global search, rss feed of search results, stopwords and more.

Note that the search template in earlier versions is not compatible with this version which is designed for the v2 search script on boardtracker so make sure to use the new version.

If your board is not yet indexed by BoardTracker, make sure to submit it!


8)

[drathbun]

What is the lead time, meaning if I enter a post now how soon will it be indexed? Can you / do you guarantee accurate results after some amount of lead time?

Can I search for an exact phrase?

Are there advertisements embedded in the search results?

Can you search within search results, meaning after I run a search can I further refine that search without starting completely over?

Searching is such an important part of my board (search.php is the 4th most hit page) that I need to know exactly what's going on before I consider changing how it works. But if you have some of the features I've mentioned I will have to seriously consider looking at it.

[Paul]

Wich things will be saved at the BoardTracker server? usersnames, or only user_id's? And how is this information protected?

[Kalipo]

Ha! I asked about this in their forum. VERY happy to see this MOD.

[BoardTracker]

What is the lead time, meaning if I enter a post now how soon will it be indexed? Can you / do you guarantee accurate results after some amount of lead time?

In usually takes between 15-60 minutes for a thread to be indexed. The more active forums in a board are more frequently scanned for new threads. It could take longer for the slower forums in a board.
We are working on a mechanism that will syncronize adding new threads in near realtime.

Can I search for an exact phrase?

Yes.

Are there advertisements embedded in the search results?

At the moment there is only the "powered by BoardTracker" icon.

Can you search within search results, meaning after I run a search can I further refine that search without starting completely over?

BoardTracker works like a standard search engine. If you want to search within the search result, simply update the phrase in the search box and add additional words you want to filter things by.

Searching is such an important part of my board (search.php is the 4th most hit page) that I need to know exactly what's going on before I consider changing how it works. But if you have some of the features I've mentioned I will have to seriously consider looking at it.

We understand. We think searching in a message board is extremely important and for that reason (combined with the fact that the board software's built in search is greatly lacking in most cases and causes many times high server loads) we created this mod.
We are always working on improving the system and search mechanism.
Any input will be greatly appreciated.

[drathbun]

Are there advertisements embedded in the search results?

At the moment there is only the "powered by BoardTracker" icon.

I take that to mean that there may be advertisements at some point in the future.

Can you search within search results, meaning after I run a search can I further refine that search without starting completely over?

BoardTracker works like a standard search engine. If you want to search within the search result, simply update the phrase in the search box and add additional words you want to filter things by.

So then it does not reuse the existing search results, but simply re-executes the search across the entire board once again, correct?

[BoardTracker]

Its possible sometime in the future but its not decided and either way it wont ever be without providing advance notice so there will be no surprises.

It will do the search again on the whole system, however remember its done on boardtracker resources so you can do as many searches as you want to find what you need. Hopefully a search will return the best results first time anyway..

Try it out and see how you and your members like it.

[drathbun]

At some point there is a server load on my board due to your spidering process, yes?

I might try it, but I would try it out on a test board first. My biggest board has a quarter-million posts, and I've already made some adjustments to the search program myself.

I'm not trying to be a jerk here but I have to ask, what's in it for you? Why provide the resources etc. if you're not planning to put advertisements on the results eventually? Can your systems handle a board with a volume like mine? What about boards with a million or more posts?

Do you have a throttle on the number of searches that can be performed? My board runs over 150,000 search queries a month.

What about options like "search title only" or "search message only", are they there?

Here's where a lot of these questions are coming from. I have customized the search process already, and am in the process of doing more. I also offer the google search option in the footer of my board. The primary reason for offering the google search link is that phpBB by default does not index "short" words and people are always wanting to search for version numbers, which are 5.1.4 or 6.5.2 and so on. Those do not get indexed by phpBB.

I am assuming that your process works like google, where you scan the board via some sort of spider-like process, and store the text of the topics / posts in a database. Then the search process scans your text database rather than using the word / post index structure provided by phpBB. Is that a reasonable assumption?

What about signatures? One of the issues with google search is that it includes all of the content on the page, including user signatures. So when people search for version numbers (as above) they often get more posts from users that are using that specific version and have it mentioned in their signature, rather than posts about that version instead.

What about standard search options like "View unanswered posts" and "View posts since last visit", are those redirected to your site, or are they still performed by the local copy of search.php? The reason I ask this is that if there is a 15-60 minute delay in spidering new activity, the "View posts since last visit" loses most of it's functionality.

[BoardTracker]

Our spidering process is unlike any other search engine spider since it is designed only for message boards and it recognizes and only reads new topics. It doesn't scan every page on your site and doesn't scan the same things over and over like google or yahoo for example. So our spidering activity will not cause any noticeable load on your servers, basically not much more than a single user reading each new topic once.

It also never reads sigs - text from sigs are simply ignored and are not in our db at all, neither is any other surrounding text from the page which is not part of the topic itself - this ensures our search results are the most relevant.

Whats in it for us is to bring what we believe to be a great product to the attention and benefit of all board owners and users out there. We have a whole range of mods planned (and already existing like our tagging and tag cloud mods) which will bring the best aspects of "web2.0" to the boardscape in general and further promote growth and user activity among all boards, communities and boardtracker itself. BoardTracker also provides 'persistent search' capabilities allowing our users to be alerted (by email, IM, RSS) anytime a thread is posted on any board which matches their predefined search terms - this is a really useful feature and helps bring more people to the boards.

We don't throttle searches at all. We currently index over 25 million topics and will reach 100M in the not too distant future. We can certainly handle boards with a million topics and will continue to add new hardware as and when needed for future expansion.

My guess is that many of those 150k searches on your board are the result of not finding what was searched for the first (or even second or third) time. My own experience with searching on boards running phpbb, vbulletin, invision or others shows that finding relevant results, particularly on a big board, is no easy task - these programs are great message boards but they are not search engines. Its likely that using boardtracker will result in fewer searches performed since they will find what they look for sooner which will result in happier members.

Searches on boardtracker are done on the title and message combined, there's no real need to separate them.

Things like "View unanswered posts" and "View posts since last visit" are not done by boardtracker, for those you would still use standard phpbb functions. These are not full text searches anyway and should be easy enough for phpbb to handle. Presumably some of those 150k searches are of this type also so you would still have those. Also the built in search would basically become the 'advanced search' for users who need to search for other things besides threads, such as author search or specific forums etc.

BoardTracker search is designed to search topics and provide the most relevant results and being designed specifically for boards we think it does a pretty good job at that..

[drathbun]

Searches on boardtracker are done on the title and message combined, there's no real need to separate them.

I would disagree with that, I think. In many cases searching by title alone provides the opportunity for better search results because you're not outweighed by the post text, so there are fewer results returned. Olympus offers a "search first post only" option as well, which I just noticed the other day and started planning to see how to implement in the current search routine.

It makes sense that you would write your spider routine differently since you have a targetted audience.

BoardTracker also provides 'persistent search' capabilities allowing our users to be alerted (by email, IM, RSS) anytime a thread is posted on any board which matches their predefined search terms - this is a really useful feature and helps bring more people to the boards.

That's a service offered by / hosted by yourself? Or exists on my server via your MOD code?

My own experience with searching on boards running phpbb, vbulletin, invision or others shows that finding relevant results, particularly on a big board, is no easy task - these programs are great message boards but they are not search engines.

No arguments there!

I'm interested enough to take a look at your code now, at any rate. Thanks for sharing.

[BoardTracker]

BoardTracker also provides 'persistent search' capabilities allowing our users to be alerted (by email, IM, RSS) anytime a thread is posted on any board which matches their predefined search terms - this is a really useful feature and helps bring more people to the boards.

That's a service offered by / hosted by yourself? Or exists on my server via your MOD code?

Its on on our side, users have to register (free) on boardtracker to use the alerts service.

[Alexis Canver]

fix for admin panel, for 404 error.

Code: Select all

[ OPEN ]

admin/boardtracker_admin_search.php

[ FIND ]

                                'S_BOARDTRACKER_ACTION' => append_sid("boardtracker_search.$phpEx"),

[REPLACE WITH]

                                'S_BOARDTRACKER_ACTION' => append_sid("boardtracker_admin_search.$phpEx"),

[ FIND ]

                        $message = $lang['BoardTracker_search_updated'] . "<br /><br />" . sprintf($lang['Click_return_boardtracker_search_admin'], "<a href=\"" . append_sid("boardtracker_search.$phpEx") . "\">", "</a>") . "<br /><br />" . sprintf($lang['Click_return_admin_index'], "<a href=\"" . append_sid("index.$phpEx?pane=right") . "\">", "</a>");

[REPLACE WITH]

                        $message = $lang['BoardTracker_search_updated'] . "<br /><br />" . sprintf($lang['Click_return_boardtracker_search_admin'], "<a href=\"" . append_sid("boardtracker_admin_search.$phpEx") . "\">", "</a>") . "<br /><br />" . sprintf($lang['Click_return_admin_index'], "<a href=\"" . append_sid("index.$phpEx?pane=right") . "\">", "</a>");


[Peter77sx]

1. The board might be be added up to 30 days (or more) after receiving a request for adding

So... will we get notified when our board is successfully indexed?

[ZombieSlayer]

Does BoardTracker still register a member named "BoardTracker" to index member only posts?

That's the biggest difference between BoardTracker and Google. Google doesn't crawl member only areas.

[BoardTracker]

Does BoardTracker still register a member named "BoardTracker" to index member only posts?

That's the biggest difference between BoardTracker and Google. Google doesn't crawl member only areas.

BoardTracker generates the list of forums to index *as a guest* so does not index hidden forums unless you allow it to. Since we use a member account you can control exactly what parts of the board it can access using the normal phpbb member/group permissions system. You can also use robots.txt to tell the spider which specific forums within your board it should not index. So you can easily and fully control what BoardTracker accesses and indexes.

The biggest difference between BoardTracker and Google (in fact there are many) is that BoardTracker is built from the ground up to work with boards and so is highly optimized for that task and provides the most accurate, relevant results for board search. We also do persistant search which Google (in fact most others) does not and categorized search which others also do not plus we have tagging, tag clouds and other board specific functions and mods.

So... will we get notified when our board is successfully indexed?

When you submit your board it goes into the queue and when we process it you will be sent an email with instructions on how to access your control panel where you can categorize your forums and see which are being indexed along with some stats and charts. Once you finish categorizing we will start indexing soon after and you will see the BoardTracker spider on your board.

fix for admin panel, for 404 error.

Updated the mod. Thanks.