spam filtering as a proxy for search market share
why is it that the most basic spams and 419 scams make it past yahoo's spam filters into my yahoo inbox? i was willing to go give them a fair shot with their new ui, but their spam filtering is beyond bad, and makes their new mail beta just as unusable as the old one. almost makes me wonder if they have commercial reasons for letting a lot of spam through to their userbase. not having a capable contextual advertising platform must put them in some tight spots when revenue maximization time inevitably rolls around, and spam thresholds are early victims, i suppose. even more so at MSN, whose hotmail is even worse. it is the rare event, however, maybe once a month, that spam makes it to my gmail inbox.
the less search marketshare your email provider has, the more spam you can expect in your inbox.
update: the new york times reports that gmail spam filtering is getting even better. meanwhile, the obvious spam in my yahoo inbox continues.
looks like yahoo can't even afford a SSL certificate for their mail domain.. oy. plus they insist to show you a spammy 'start page' instead of your inbox. someone getting desperate in the monetization department?
what is especially interesting to me is that sakku is a swiss startup. i have outlined before why switzerland is going down the drain. in fact, i sometimes tell smug swiss people that i am an economic refugee. all the more uplifting if a startup is ignoring the naysayers and delivering a cool product.
an inconvenient truth
went to see the most important film of the year today. very well done, with a data-driven approach to hammer the reality of global warming home. the only nit i have is that the call to action comes too late in the movie (in the last five minutes, and in the credits). also, there was not enough coverage of solutions beyond vague calls to save energy. no advice how to short a basket of detroit companies and buy efficient cars with the proceeds, no talk of emissions trading, no call for new nuclear power plants. instead, we got the advice to pray
still, go see it even if you are familiar with all it's points. longer playing times matter.
has last.fm jumped the shark?
opensourcing the google os
it has long been argued that the google os, particularly MapReduce and GFS, is google's real competitive strength. yahoo, meanwhile, is paying developers to develop clones of these. with seeming consolidation on a common computing platform, and ever-rising data center expenses, you gotta wonder how much sense it makes for the big three to duplicate all that CAPEX. they might be better off outsourcing their datacenters, and sharing some base datasets, such as a crawler cache (kinda like the feedmesh network).
the outsourced company, on the other hand, would end up running a grid with several million nodes and could optimize running costs overall, by using very low power servers, running on an opensourced processor architecture.
the call for papers for OSCOM III is now out. here is the intro paragraph:
Open Source CMS Conference III
April 16-18, 2003 -- , Boston, MA
Call for Participation
Proposals Due: January 15, 2003
The theme of the next OSCOM is "Leveraging content with CMS: authoring and syndication towards the semantic web"
Content management treats content as something static more often than not. Content can be much more than that, though. OSCOM III will show attendees how to make more out of the content that is stored in their CMS, how to make content discoverable, provide annotation and other advanced services. OSCOM III will also explore the frontiers with a look at semantic web technologies, and what they might add to content management in the future.
The theme has several aspects:
the recent postnuke forks might be a good thing, if the following two things materialize.
postnuke: The Embedded PN initiative will allow us all to leverage all the amazing pieces of php code out there in a fast and reliable way, by allowing client applications to work under PostNuke seemingly. This will hopefully become the PN applications server framework.
envolution: This whould more or less be a direct replacement for the current hooks (autolinks, wiki, and ratings) but would not replace hooks. It would allow for direct api calls to classes that provide functionality for any module to call if it is turned on by the site administrator. So for instance polls and comments could be loaded dymaicly and used within any module as a direct api call. This would allow the wite administrator to add polls directly to a news article. This would also allow comments to be used by any module that you want to allow to use them. For instance you could allow comments on your downloads. This is a small example of the functionality this could provide.
guess where those features will end up.
I have written a master thesis that aims to establish a Open Source Projects Framework.
The historical roots of Open Source are outlined. A comparison between Open Source projects and classical projects highlights strengths and weaknesses of both, and defines their attributes. Existing Open Source theories are evaluated, and the requirements for a framework for Open Source projects are determined. The framework introduces the notions of actors, roles, areas, processes and tools, and depicts their interrelationships in a matrix. Each aspect of the framework is then further developed to serve both as a conceptual foundation for Open Source and a help for organizing and managing Open Source projects.
I would appreciate any feedback, comments, suggestions, rants, corrections at rothfuss at abstrakt ch
New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.
Keep up with the latest Advogato features by reading the Advogato status blog.
If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!