Good morning FYI Under the SWIKI menu item on the left menu, are all 6610 pages of the squeak wiki. It renders the raw html as scraped. I am currently working on the presentation of that data in a clean way. I must say that having 6610 subclasses of DocletSwiki pounds the heck out of the Morphic Browser. Also, a change in the superclass forces a recompile of the subclasses...which is now up to the 4000's as I type this...but you have to wait several minutes for that to complete.* The .mcz are available via anonymous ftp at the menmachinesmaterials.com address for those interested. A quick rundown on some plans (I will be adding them to the formal documentation contained within the app later) After the presentation ... 1. fix the links so that they navigate internally to SeasideDoc 2. that is dependent on the initialRequest/updateURL infrastructure being in place to allow direct navigation to a component given its URL. I currently have that in place in the zurb (example here: http://menmachinesmaterials.com/zurb/BuildingBlocks/Containers/MarketingSiteContentSection ) so porting that to SeasideDoc should be relatively fast. 3. provide a link to the live swiki at the top of each page 4. Self organization of the objects to include 4.1 keywords stored in the superclass with association to pages that contain them 4.2 scan through all the pages and identify "this page is of this FOO type" and then create "FOO type" categories under DocletSwiki and then have those pages place themselves under those categories. 4.3 extract "by author" if possible and store an association so that "show me all pages by this author come up" 5. Have the classes refresh themselves from the wiki's "recent" page that shows what/when a page was recently updated. cheers. t *I am open to other persistence options later down the road. I want to play with an application that has many thousands of classes for a bit to get a hang of it. _______________________________________________ seaside mailing list [hidden email] http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside |
On Sun, Dec 02, 2018 at 01:33:54AM -0500, gettimothy wrote:
> Good morning FYI http://menmachinesmaterials.com/SeasideDoc Under the > SWIKI menu item on the left menu, are all 6610 pages of the squeak > wiki. It renders the raw html as scraped. I am currently working on > the presentation of that data in a clean way. I must say that having > 6610 subclasses of DocletSwiki pounds the heck out of the Morphic > Browser. Why not a single class for a swiki page, with content and related page-specific metadata in inst-vars? Pierce _______________________________________________ seaside mailing list [hidden email] http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside |
Hi Pierce That is what I have. Each Wiki Page is wrapped by a class. Also, I have refactored it a bit and now the left menu shows the page title in stead of 'pageNNN' and the content displays without the cruft. Also, I moved the scraped data to the class side and things like 'title' and 'pagecontent' are parsed from that raw html. There is other meta-data I will be extracting ...things like 'date created' and "author" if possible. Anyhoo...the Squeak Wiki is now in classes/objects which lends the power of Smalltalk to bend it to our needs. I must say, the Morphic is really chugging under the load though. I am hoping its just trying to catch up to me moving the swiki classes to their own category. ---- On Sun, 02 Dec 2018 03:27:52 -0500 Pierce Ng <[hidden email]> wrote ----
_______________________________________________ seaside mailing list [hidden email] http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside |
Free forum by Nabble | Edit this page |