Beta: The CG-PowerPack of WordPress plugins

davidchait
(@davidchait)

19 years, 9 months ago

I wanted to get the word out to my more avid testers and other interest parties. Rather than generate a ton of individual archives, for the initial testing of my new CG stuff, I’m packing up ALL of my ‘libraries’, plus enabler ‘plugin stubs’ for WP1.2+, in one archive.
This includes my existing libraries:
1. CG-Amazon (amazon data feeds)
– added keyword searches, new Find feature to add to db, default to external links (and not internal to cgaindex), fixed cgaindex ‘bot thrashing’, and created CG-Inline for inlining Amazon products in articles. New XMLParser from Feedread.
2. CG-Feedread (RSS/Atom aggregation)
– New XMLParser, uses custom HTTP connection code to reduce ‘lockups’ due to bad/unavailable XML services. Fixed some encapsulation and output cases.
3. CG-Referer (referral, bot, and query/search tracking)
– Updated blacklists. Overhauled bot code. New bot and query detection.
4. CG-Samecat (listing posts within a category)
– No major changes.
… plus my new additions:
5. CG-Inline (embedded macros within posts for images, amazon items, and post links)
6. CG-Related (assisted process for building table of related posts and listing them)
7. CG-WhatTunes (tying music player/watcher http POST updates to CG-Amazon, to show the current music you are listening to)
… and associated helper files and functions.
Many of the above are completely missing documentation at the moment — but I wanted to get the new Amazon and Feedread stuff out, and had people interested in WhatTunes and Related, and didn’t make sense to go pulling them into seven different packages at the moment. 😉
Anyway… for those brave souls, go grab the CG-PowerPack, and let me know as you have issues. Feel free to start new Topics here for different plugins, so that they don’t all get junked in one post. 😉 For general issues or Qs, feel free to respond to this post of course!
FINAL NOTES:
– All CG-Amazon users really should update, and CLEAR their cache_amazon folder, as bots may have flooded it with junk.
– Feedread users, note that the cache folder name has been changed from cache_rssfeeds to cache_feedread to note ‘who owns it’. 😉
-d

Viewing 15 replies - 46 through 60 (of 145 total)

← 1 2 3 4 5 … 8 9 10 →

Thread Starter davidchait
(@davidchait)

19 years, 9 months ago

New CG-PowerPack 1.0b12 is being uploaded right now. Fixes the ‘single RSS entry’ problem, AGAIN, and doesn’t break normal feeds. Phun.
Results should be returned in seconds. I’ve had pauses of maybe 10-12s on busy days. If it runs really long, something bad is going on.
I wouldn’t want to make the timeouts longer, as Amazon should be returning results quickly — if they can’t, the server is already overloaded or down. In fact, it shouldn’t wait 30s… I should pop out in 10-15 probably on average case.
I usual use, the system is ‘re-looking-up’ data for an already existing cached item. So if it fails, it can look again next pageload. Rather have it fail quickly. 😉
In the future, I’m going to add a cron-ready command that can be run ‘offline’ to batch update your table, your wishlist, whatever, so that people don’t randomly hit reload-waits…
apakuni – I love all the help/feedback, but I’m really not looking to confuse users by having more places/versions of the code to download… It’s bad enough that I MYSELF have two versions (the powerpack and the individual archives), and that’s confusing people greatly already!
Give me a few days to look at your tweaks, figure out what/how to integrate. Again, appreciate people who are giving more feedback on things to make the CG stuff better for everyone! 😉
And let me know if you’re still having problems with timeouts, what time of day, what searches, etc.
-d

Thread Starter davidchait
(@davidchait)

19 years, 9 months ago

Hot on the heels, I’ve added minor tweaks to CG-Feedread and uploaded b13. The main changes here were:
– proper decoding of the timestamps, so that showTime does something useful and consistant.
– fix for Atom ‘summary’ field use for entry description, which was causing blank descriptions (well, actually repeated site-description).
feedread.php is the only file changed since b12.
-d

apakuni
(@apakuni)

19 years, 9 months ago

Beel – I thik I have an older codebase installed on my test box. Will move the code up and should fix my issue.
David – It was never my intention to maintain a seperate branch. I needed a place to put it so you could get it and maybe some others who could look at it and tell you if they thought it was useful.
Let me know when/if you elect to use anything I did and I will pull it down.

apakuni
(@apakuni)

19 years, 9 months ago

On the timeout issue … would it be useful to have that as a variable option in the search? If it is available, one could let the search run deeper for items with lots of possible matches. If the var is not set, it defaults to 15s.
Just a thought.

Thread Starter davidchait
(@davidchait)

19 years, 9 months ago

Understood — no rush to take it down, let’s see if we get any further feedback…
On timeout, certainly possible to add another option. The real problem is that it shouldn’t be taking so long for responses (note that on many servers, PHP has a 30s timeout in a script, which would nuke the process anyway…). It may be that there are better options that I can use in the initial search to make it more ‘lightweight’ so I get responses faster. Hmmmm…
-d

apakuni
(@apakuni)

19 years, 9 months ago

Updated my test distro to include CG-POWERPACK-1.0b14 code. Renamed to CG-POWERPACK-1.0b14-D.zip to stay in line with Davids changes.

Thread Starter davidchait
(@davidchait)

19 years, 9 months ago

Awesome, thanks!
Both of you, definitely let me know when you get to try the CGA Admin search system with the new code… You should notice the difference immediately — well, assuming Amazon isn’t being slow-as-molasses. 😉
-d

Beel
(@beel)

19 years, 9 months ago

A lookup for “testament” renders 10 items in 2 seconds and 50 in 8 – pretty impressive.
apakuni: Searching for phair on music returns 14 items and B00009OOH9 returns “Liz Phair (Liz Phair)”

Beel
(@beel)

19 years, 9 months ago

apakuni: I posted the above just to let you know that there apparently is something wrong with your set-up. I am hosting off my computer and don’t have any special modules enabled. Pasting the link above directly into my browser does not return an error. Maybe the site was having difficulties? Have you tried it again recently?

Thread Starter davidchait
(@davidchait)

19 years, 9 months ago

beel:
glad it all seems to be working. good sanity check.
apakuni:
– the requests are all tagged with my dev token and associate ID — the ‘rewriting’ of associate IDs happens in the output routine.
– that looks like a proper URL, though I’ve never seen the “error 11, Resource temporarily unavailable” thing. Sounds like it’s unable to open an outgoing HTTP request or something weird.
– Just needs basic PHP with XML functions (which are part of the default installs from what I know). No PEAR, no SOAP, no other things. Clean and simple.
Yeah, if you want to give me access, I’d be glad to help troubleshoot. Never heard of anyone having problems like this, so would be interesting to see phpinfo() on your box, and I can turn on heavy-debug-output mode and maybe it’ll tell us what’s croaking.
-d

apakuni
(@apakuni)

19 years, 9 months ago

David …
Shoot me an email (apakuni_AHT_ursastudios_DOHT_com) and let’s set up a time to troubleshoot. In the interim, phpinfo is here …
http://admin.kavido.com/phpinfo.php

apakuni
(@apakuni)

19 years, 9 months ago

David …
Are you ever in IRC? we could also hook up on AIM, ICQ, MSN or Yahoo.

TechGnome
(@techgnome)

19 years, 9 months ago

AAAAAAGH! David, I’ve got the latest CG-Amazon installed, but I’m having a bit of a problem…. it doesn’t validate….. also I had to make a small change. It kept wanting to output a target element (which is invalid in XHTML) even though I had set it to not use external links.
what I’m now having a problem with is the order of the ul, li, & a tags…. not sure how it should be, but having the a tag wrap around the li tags doesn’t seem to work. Well, it doesn’t work for the validator.
url: http://techgnome.anderson-website.net/index2.php
click the valid XHTML 1.0 under “badges” on the right.
Would like to get this to validate out properly…. I suspect that each li item is going to need it’s own a href…… ick!
Thoughts, comments, suggestions?
TG

TechGnome
(@techgnome)

19 years, 9 months ago

Beel – thanks… I’ll set it back to span and see what happens.
TG

TechGnome
(@techgnome)

19 years, 9 months ago

You rock Beel! It worked. I am once again XHTML 1.0 Strict compliant! Yessssss!
TG