The new totspCode is out, but I have changed a plethora of stuff since the release. I plan to finish up some major new functionality in the next day or so and re-release.
I also have a set of things to complete for another later release.
Here are the latest priorities:
FOR NEXT DAY OR SO
- Performance: Build of site local RSS file and pulling of remote site "source" RSS feeds should be set to only check/happen when a request param such as "UPDATE=true" is present on URL. This way a cron job can be setup to run background and update sources and live users do not have to wait on this update (and note that even for live users remote source files are cached for 60 minutes, but the addition of the request param would avoid even the CHECK if the cache needs to be updated).
FOR NEXT FEW WEEKS
- User Preferences: Allow user to set how they want storyComments view to sort by default, and how many items per source they want to see.
- Make comments threaded. (had to push this one out due to upcoming vacation)
- Make comments editable by posting user (if authenticated).
- Work on font sizes and colors (its a bit hard to read right now, I like the colors overall, but needs some font help).
ANY NEW BUGS/ISSUES/PROBLEMS/SUGGESTIONS and so on, please add as comments to this article, keeping track here (other posts relating to new totspCode are deprecated as locations to store issues).
Comments
RE: Penguin 5 "Azure" TODO
Here's a suggestion: can you post the time a feed was last updated? The LinuxToday one doesn't seem to change all that often and when it does it looks pretty much the same, so it may be nice to see how old it is.
RE: Penguin 5 "Azure" TODO
COMPLETED the Performance steps with update_rss.php and update_sources.php created and available to call via external cron job. No longer is/are local rss file or remote source files checked during actual execution of site (did not make a dramatic perf difference, but a little).
blockcipher suggestion response
yeah i can do that, i will put in soon.
note though that i think the linuxtoday feed has serious issues. it used to be TXT only (not RSS) and i have quit bothering with text only crap. i need to look into that specific feed a bit more. putting that on my list too.
RE: Penguin 5 "Azure" TODO
COMPELETED checkSiteURL, wasnt on this list but was needed. if user has bookmarked "screaming-penguin.com" or "totsp.com" (or any valid alternative domain name that results in request to totspCode based site) then sessions can get munged in PHP based on new cookie created for different domain name.
new function checkSiteURL is on main.php and will redirect via 302 any request for something that does not match SITE_URL as defined in config.php (in the case of penguin thats www.screaming-penguin.com).
Penguin 5 "Azure" CODE UPDATE - RERELEASE
I will be repackaging all the changes to the code since the released version and updating the said released version in the next 24 hours.
I will post here when I complete the update and will just update the existing totspCode release projects story.
blockcipher suggestion done
sources now display the ctime of the local source file (which is the last time the file was updated).
LinuxToday feed?
LinuxToday got bought up some time back by Internet.com and I cant tell if they have a valid feed or not (I prefer direct feeds over aggregators).
The only linuxToday feed has been disabled because it was broken. If someone knows a valid RSS URL for linuxToday let me know and I will update it, other than that linuxToday is dead meat around here.
RE: Penguin 5 "Azure" TODO
Need to add comments to the search.
RE: Penguin 5 "Azure" TODO
add the total number of comments for each story on the main page next to each title
RE: Penguin 5 "Azure" TODO
cleaned up the displaying of RSS feeds from file. this seems to be what takes the most time. displaying a penguin story happens instantly, displaying a "home" page with 10 or more custom sources takes a few seconds. cleaned this up so there are far fewer database trips, but still seems a bit slow. may simply be due to nature of totsp servers setup with filesystem NFS and 3 load balanced web servers.
Whoops, sorry.
It wasn't LinuxToday, it's LinuxMagazine. Sorry about that. Otherwise, thanks! Great job on the site.
RE: Penguin 5 "Azure" TODO
REMOVED MAGPIE FEED PARSER. Magpie is a great sourceforge project and it works like a champ, it was just slower than my old arse manual open the file and parse it stuff which uses rssparse.php3 (http://nwow.org/code.php3). This code is ancient, 2000, so it doesnt have all the bells and whistles of magpie or others and may have caveats in the future, but for now its what penguin and totspCode will use.
Its about 50% faster in rudimentary tests (the view all source page and a watch).
re linux magazine
checked that one, appears to be good rss feed and appears to working now on penguin.
RE: Penguin 5 "Azure" TODO
new version with all latest updates (which were quite a few) posted to projects page totspCode.
RE: Penguin 5 "Azure" TODO
i updated the site keywords today and it seems to HAVE WORKED! that is the google ads appear to make a bit more sense in relation to the site for the home page. (Articles were already making some sense, now the home does also.)
The new rev of totspCode includes a few new "keywords" features. First off the keywords for the entire site are defined in the config file (rather than hard coded) and secondly each article has the title added as keywords, each word separated by comma.
I will work more on optimizing the google stuff as I go, but I already think the new code IS an improvement.
RE: Penguin 5 "Azure" TODO
I got some wierdness on the storyComments view. Its not showing a comment I posted about your mailbox migration perl script.
RE: Penguin 5 "Azure" TODO
New suggestions:
1. On the "view all front page" stories make a link or some means to drill down into the comments for each story.
2. Investigate the login changes as now when linking to the penguin with correct GET params (user/pass) it does not login, it still requires a click.
RE: Penguin 5 "Azure" TODO
Completed the comments drill down (item 1 on the prev post) and have not yet had a chance to look at item 2.