tag:blogger.com,1999:blog-21167013.post5958787001044863135..comments2008-04-14T19:21:18.846-07:00matt_gigerhttp://www.blogger.com/profile/08948284981378368476noreply@blogger.comBlogger5125tag:blogger.com,1999:blog-21167013.post-28199564419605372572008-04-14T19:21:00.000-07:002008-04-14T19:21:00.000-07:00dude - your all backwards.. If KML is a standard, why not have a reference implementation to write it out.. If its reasonably well done, what's the downside? Modern sfwr is <I>all</I> libraries.. <BR/><BR/>As for reading.. why do you want to write parsing stuff? The 0.1 libkml is not great for reading.. but its uh, 0.1. <BR/><BR/>This is made for you...<BR/><BR/>However, you are the clearly opinionated, go-it-alone individualist that is writing your own browser.. gotta like that.. on the whole, confusingdarkblue_bhttp://isde5.pbwiki.comnoreply@blogger.comtag:blogger.com,1999:blog-21167013.post-12139102814153694332008-04-02T01:00:00.000-07:002008-04-02T01:00:00.000-07:00Google is candidating KML as an ISO standard. As such, an opensource implementation is a almost a must.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-21167013.post-57105656703785122972007-10-25T22:59:00.000-07:002007-10-25T22:59:00.000-07:00I'd like to see libkml available under <A HREF="http://www.boost.org/more/license_info.html" REL="nofollow">Boost License</A>. BL seems to be most reasonable license for an Open Source library. It's well thought.Mateusz Loskothttp://www.blogger.com/profile/12641441234340682916noreply@blogger.comtag:blogger.com,1999:blog-21167013.post-16590914294252607542007-09-20T20:24:00.000-07:002007-09-20T20:24:00.000-07:00You sound like you know something about it, and it sounds like it is for content consumers. Great! <BR/><BR/>I'd love to not have to worry about doing all of that in order to support KML in EarthBrowser. I agree that determining visibility for a large sets of data is exceedingly difficult to get right. <BR/><BR/>Not knowing anything about what the library supports beyond vague generalizations, I'll hold off judgement. I just hope that the library doesn't take it upon itself to create threads, access the network or other unpredictable behavior. I've been burned many times with those sorts of integration issues.<BR/><BR/>Please give details? What OS license will it be for one?matt_gigerhttp://www.blogger.com/profile/08948284981378368476noreply@blogger.comtag:blogger.com,1999:blog-21167013.post-91546801875097416572007-09-20T20:07:00.000-07:002007-09-20T20:07:00.000-07:00Because KML is more complex than you think it is.<BR/><BR/>Do you understand how shared and inline styles are merged? How &lt;Region&gt; really works? How &lt;Update&gt; functions? Do you know the caching behavior of the many different ways in which content can be fetched? <BR/><BR/>Furthermore, do you know how to hold several thousand features in memory and traverse their structure 60 times per second to query for what is visible at that instant? The 2D case, while not thousands of potentially active drawables, is still not trivial. <BR/><BR/>The goal could be to allow other implementations of KML (NASA World Wind, ESRI ArcGIS Explorer, etc.) to have an easier time of doing this. It turns out that KML 2.2 is actually pretty darned complex and subtle and hard to implement correctly.Anonymousnoreply@blogger.com