View Full Version : FREEBS Feature and Functionality Suggestions
Please post any features or functionality that you desire to be included in FREEBS in this thread.
Thanks
jhughes
04-12-2005, 01:27 PM
Just keeping feature requests from me on the correct thread
* Remove duplicate ASIN
* Remove ASIN when price is "Too Low to Display"
* Remove ASIN when "Item not available"
* Nice SearchBrowse by keywords
* Easy way to grab and populate best selling items
* Cache for Images and descriptions to reduce load on Amazon
* Cache AutoDelete daily or after n000 hits
Just keeping feature requests from me on the correct thread
* Remove duplicate ASIN
* Remove ASIN when price is "Too Low to Display"
* Remove ASIN when "Item not available"
* Nice SearchBrowse by keywords
* Easy way to grab and populate best selling items
* Cache for Images and descriptions to reduce load on Amazon
* Cache AutoDelete daily or after n000 hits
Dupes are covered.
Have not encounted a too low to display yet to write code for that.
Items not available are excluded.
Search is later, but something will happen, the API is too easy to use.
Not sure on the best selling part. Tell me more.
The book medium image is the only image pulled from amazon.
Not sure what ya mean on the autodelete, tell me more.
The demo is up Genealogy Book Store (http://www.familyseekers.org/genealogy-books/)
jhughes
04-14-2005, 01:52 PM
Have not encounted a too low to display yet to write code for that.
Example: Look at the price on this product
http://www.amazon.com/exec/obidos/tg/detail/-/B0007KX4SS/104-2078014-4361551
Search is later, but something will happen, the API is too easy to use.
I am glad
Not sure on the best selling part. Tell me more.
Every category has a Top Seller or Best Seller list.
It's every updated every hour for electronics and slower on some other categories (I think)
If you display this category or product from this category the chance of
the item missing (404 error) or the price incorrect is very high.
Something like a cron - job update or cache empty of special categories such as this.
Here is an example.
http://www.amazon.com/exec/obidos/tg/new-for-you/top-sellers/-/electronics/226184/
Not sure what ya mean on the autodelete, tell me more.
Related to above I guess if you get a 404 then to be able to handle it gracefully. If you keep items in the cache and someone clicks on an item no longer at Amazon it would be nice to
a.) Delete the item from the cache
b.) Take them to "Similar" products using the "category name" in a search
to browse similar items
c.) Alternatively use a cron job to remove cache files at a timed interval but if your site gets a lot of hits you can set it to every 2000 unique visits - just an iidea
Thank you very very much for putting the time and sharing the code :D
If you mean a hot items page in general, we can do that, maybe. However this is only supposed to be books currently. That's getting closer to store functionality. However I will look at the API calls available for that type search.
So far the book.php page does one SOAP call to the amazon AWS server. Then the client calls the amazon image server for the books medium image. I have added small images to similars. That's 5 more calls from the client to the amazon image server. But each image is about 2K, I don't see these fetches as overhead really. Especially compared to other store pages I've seen.
When I run the p3.php file, I get a 404 error and see ...\ap3.php in the address bar. This is only in 1.05. 1.04 seemed to work fine.
When I run the p3.php file, I get a 404 error and see ...\ap3.php in the address bar. This is only in 1.05. 1.04 seemed to work fine.
Oh damn I see what I did. I'll fix it and issue a new release.
To fix it in the mean time, open p3.php and on line 27 change...
echo '<form action="ap3.php" method="post">
to...
echo '<form action="p3.php" method="post">
1.06 is posted.
Sorry for the problems and the post edit, I hit edit instead of quote.
p3.php has become too powerful. You can and if you're like me, you're trying to popluate the database with as many books as possible. Keep the Amazon TOS in mind when doing so. That's on you.
However this has brought up another issue. Book List pages with hundreds of books. These pages have too many links and they are well over 50k and even 100K in some cases. This is not good search engine food.
So, we have pagination. Check my FREEBS book stores for the example and let's hear some feedback.
Thanks
I think you've addressed this but wanted to bring it up. The "Similars" listings show external links to Amazon if the books aren't in that Bookshop. Could it be coded to show ALL "Similars" as internal links regardless of whether or not it's currently in our particular bookstore?
I think you've addressed this but wanted to bring it up. The "Similars" listings show external links to Amazon. Could it be coded to show internal links instead?
Where we have the book on file we show a local link now. The image links are out going though.
The code to resolve similars not on file when displaying the book page could be quite a bit, so I doubt I'll do that. (ie. Check if sim is on file, look it up, store it, store it's similars and comments.)
Who knows, maybe one day. That would be pretty dynamic too. Let me think it out.
Where we have the book on file we show a local link now. The image links are out going though.
The code to resolve similars not on file when displaying the book page could be quite a bit, so I doubt I'll do that. (ie. Check if sim is on file, look it up, store it, store it's similars and comments.)
Who knows, maybe one day. That would be pretty dynamic too. Let me think it out.
Just a thought. It would add several hundred (if not thousands) of spiderable pages to the site as a bonus. Plus, it would cut down on the number of outbound links.
Just a thought. It would add several hundred (if not thousands) of spiderable pages to the site as a bonus. Plus, it would cut down on the number of outbound links.
I agree. I looked at it once and the loop was very busy and I balked. I'll rethink it. While there are quite a few queries, they are indexed one record lookups and inserts. There are no record pools pulled and sorted, so the overhead may not be as bad as the number of queries would lead one to believe.
Thanks
Check if on file. (already doing it.)
If not, store the item record, up to 3 comments and up to 5 similars X 5 possible unresolved similars.
That adds up fast.
I just can't do it on the fly. I forgot the 5 SOAP calls necessary too.
If I did the multiple inserts in one query, we could have 20 queries for that loop. Plus a total of up to 6 SOAP calls. Too much to put on that page and have it display in any reasonable time.
I'll look into caching and then batching them in or something. Maybe those not on file go to a cache and then when p3.php is run it first pulls and resolves the similars and then gets new books. May just another option all together.
Ok, I figured it out. I will make the URL internal anyway. Then when the user clicks it, if it's not on file, we will get it and store it and have it on file. This would be 5 fast queries and only one soap call.
That's the ticket!
I knew you could do it!
This cold has kept me in most of the week and weekend. I tried to trim trees and rake today, but just got sicker. So I'll sit here quietly and code some more.
I may release 1.08 of there are no complaints from the testers and then make this mod for 1.09. That way it will be a php file replace for the upgrade only and I'm not putting out too many new functions at once.
Thanks for your input.
jhughes
04-18-2005, 08:17 AM
This cold has kept me in most of the week and weekend. I tried to trim trees and rake today, but just got sicker. So I'll sit here quietly and code some more.
Wish you quick recovery :D
I have not installed it yet due to lack of time, but eagerly looking forward to installing FreeBS.
Wish you quick recovery :D
I have not installed it yet due to lack of time, but eagerly looking forward to installing FreeBS.
Cool, unless a bug is reported on the new additions, 1.10, we got us a book store.
Enjoy your never ending book store, you now have the ability to display any book on file at amazon if you cache it or not.
Even if you purge your database after being spidered and have no books on file locally, you will still display any book still on file at amazon properly if the URL saved by the SE is used.
rocket98
05-02-2005, 02:21 AM
Hi,
I just installed your script... works great.
? On the initial results page how can I display a image beside book listings. Right now only images I see are on a further detail page and for similiar items. These image URL's are already stored in DB [yes/no] so it could be added?
Works great
Thanks.
pgwrink
05-02-2005, 05:48 AM
Noppid you are starting to scare me. You are really running this site very well. No fighting or bickering no ego oriented posts. Waiting for other shoe to drop. hehe
vBulletin® v3.8.5, Copyright ©2000-2010, Jelsoft Enterprises Ltd.