Some eagle eye description, trying to be unbiased:
http://diveintohtml5.info/storage.html

Did not find the most infurating sites again when quickscanning, but see some of the comments
https://hacks.mozilla.org/2010/06/beyond-html5-database-apis-and-the-road-to-indexeddb/comment-page-1/#comments
https://news.ycombinator.com/item?id=1395818
https://hacks.mozilla.org/2010/06/comparing-indexeddb-and-webdatabase/

Don't remember if he was the guy from Oracle with the quick implementation on top of Oracle tool (seems to have been ported to use SQLite in FF as back end engine, but disallowing the SQL API - sheesh) but his arguments are... shallow, as other vendors DB's can often be queried from a running DB and abstractions are plentyful (called ORM or ADO.Net or DBAPI or even ODBC or JDBC) - but read up and down that thread yourself ;-)

update: found a few more. read and vomit at own risk...

http://lists.w3.org/Archives/Public/public-webapps/2009AprJun/0137.html
http://lists.w3.org/Archives/Public/public-webapps/2009AprJun/0146.html
http://lists.w3.org/Archives/Public/public-webapps/2009AprJun/0307.html
http://lists.w3.org/Archives/Public/public-webapps/2009AprJun/1242.html
http://lists.w3.org/Archives/Public/public-webapps/2009AprJun/1283.html
http://lists.w3.org/Archives/Public/public-webapps/2009AprJun/1286.html
http://lists.w3.org/Archives/Public/public-webapps/2009AprJun/1319.html
http://lists.w3.org/Archives/Public/public-webapps/2009AprJun/1326.html
http://lists.w3.org/Archives/Public/public-webapps/2009AprJun/1333.html
http://lists.w3.org/Archives/Public/public-webapps/2009OctDec/0561.html


Perhaps you find more aggravating parts (still think on Mozilla somewhere, but probably moved or 404...) I missed tihs time around (did not want to dampen my day again...), but one guy posting back then on Mozilla did not show his Oracle adress that openly - therefore I googled him back then to verify.

HTH

thomas


>I see.
>Thanks
>>currently somewhat swamped, perhaps Al saved a direct link... Look for browser storage war html5 websql to find few some devs seeing the same chances for WebSQL,
>>look (AFAIR either firefox or html5 dev blogs) for the "decision" not to support SQLite as "standard", although
>>-FF itself uses it internally
>>-Thunderbird, another Mozilla product, would be unusable without it
>>-nearly all at least half-smart mobile platforms (including Symbian last decade, in WP7 you could add SQLIte, not sure about WP8) use it internally
>>
>>and then googe for the job profile of the guy "offering" some code to replace the WebSQL with some code around an Oracle FOSS leftover (AFAIR Sleepycat/BerkeleyDB, but only guessing at the moment) - was working for Oracle when he developed the replacement making sure the was no EASY local engine similar to RV's which could help exchange backends keeping the table/column structure identical on the front end.
>>Oracle could easyly have offered one of their FOSS MySQL engines as other implementation to SQLite, but if both largest SQL vendors decline, chances of another (relatively tiny) commercial vendor to offer his SQL workhorse just to have an alternative are small and PostgreSQL as other well tested FOSS vendor is definately built for small machines. Those 2 big DB vendors then scripting the arguments of "1 cannot be a standard" leaves a very bad taste over here. SQLite ***is*** FOSS, so why not fork it if you do not want to use your own product, the source can be used by everybody - if they want to keep their small embedded offerings closed source - fine, but either offer an implementation as "large vendor" or drop the argument that more than one implementation is needed.
>>
>>
>>
>>>Could you please elaborate or give us some links Thomas?
>>>Are you talking about this http://hacks.mozilla.org/2010/06/beyond-html5-database-apis-and-the-road-to-indexeddb/ ?
>>>>Nothing to fear... Actualy there was something VERY much closer to local cursors planned and already working in some browsers for HTML5: WebSQL to utilize SQLite. But that was killed by Oracle, which provided the code for alternative store implementation based on one of their bought up OS products and argued heavily against standardizing a relational table/column-based on only SQLIte and code already working across browsers based on GEARS to be maintained further. I am so glad Oracle protected me from the folly of trying to use local cursors based on SQLite that I mention that and the ridiculous Java/Android trial with a persistance similar to Craig warning about Google. Read up on the arguments leading Firefox dev to turn his back on WebSQL and to hail more modern object stores. Until you have read up that stuff, you have no idea where the satire is and what it does encompass...