Thursday, April 25, 2002

The joys of struggling with Frontier's Manila plugin arch. #

I've been continuing work on my Citation Management Plugin (a Manila plugin to allow me to import, manipulate, and export citations/URLs).  Part of the challenge I'm facing is that I don't know well the Frontier/Manila/Manila Plugin frameworks in which I'm programming.  Documentation is very good in spots (I'm so happy for Matt Neuberg's book and Dr. Matt pages) -- but lacks any overarching organization to allow Frontier newbies like me to get up to speed easily.  (At least, that's my take.) For example, let me share my latest struggle as a way of venting my frustration and also putting it out there in hope of getting help.  I've written a MOA2 converter in my the plugin and am now trying to display the results as DHTML tree.  I found WebFx' xTree, for which I've worked out a static HTML mockup.  I'm trying to move that mockup into my plugin and have run into a problem.  How do I place place the javascript file (xtree.js) into my plugin and have it served from Frontier -- without any interpretation by the mainresponder framework? The way that I've been able to send an HTML file out without having mainresponder do anything to it s to pre-empt  the framework by code segments like this: local (pta = html.getPageTableAddress ()); pta^.responseHeaders.["Content-Type"] = "text/html"; pta^.responseBody = string(CitationManagerSuite.pages.myTreeDemoText); scriptError ("!return") which allows me to serve up the HTML I have stored in CitationManagerSuite.pages.myTreeDemoText I can obviously use the same idea to serve my javascript -- but there must be a better way.  (But I don't know the better way because I don't understand fully things like "the walk"  (I'm linking to the google cache of the page because the original page is down at the moment... Actually, as Lawrence Lee kindly pointed out to me by email, archive.org is a better source of caching for the Dr. Matt Pages.)    Posted by Raymond Yee on 4/25/02; 11:14:02 AM from the Web Technology dept. Discuss (4 responses)

The "backlash" against SOAP #

Mark Pilgrim:  " Joel wants something that doesn't exist (and may never exist): an enterprise-level distributed object system that's platform-neutral, vendor-neutral, language-neutral, patent-free, and that enjoys wide industry support. Lots of luck with that. Meanwhile, pundits will keep having these pointless little flame wars about wire protocols, and programmers will keep scraping HTML pages because 99.99% of the information we want is only exposed through web pages anyway." I generally agree with what Mark wrote. Nevertheless, I'm still excited that google has offered a programmer interface to its search engine at all.  Granted, as Paul Prescod argues, having a URI that yields XML is potentially more powerful than just providing a SOAP interface.  (It would be wonderful to be able to give someone a URL (to get XML in return) rather than directing others to a programming interface.)  Still, a SOAP interface is better than no sanctioned programming interface at all (yes, one can scrape the HTML from google but one might be violating google's term of service doing so....the excitement about google's API is the fact that it's inviting programmers to use it openly.) Ed Dumbill is concerned that W3C is loosing control over the whole web services field -- and that the big companies are pushing the whole ball of wax into proprietary systems.  Yes, that might be right.  But the W3C better get into the game with the "big boys" to protect the interests of the Web...How will it do so?   Posted by Raymond Yee on 4/25/02; 9:36:15 AM from the Web Technology dept. Discuss