i contacted steven rainwater about the ratings-thingy. the predominance of 1s is because of the spammers: everybody kept hammering away at those 1s :)
also, the diary ratings contain old accounts. steven daren't touch the diary code because ... well... in my mind, it looks like pure spaghetti :)
so a bit more hunting to do to get to the bottom of _that_ one...
but - i have to say: i am disappointed (with the people who rated, and with myself) at the large predominance of 1s on _my_ profile.
whoa. ok. nymia has a link to the report page for his statistics.
i decided "hmmmm.... that looks iinnntereesstiiing" to do an analysis on.
the program i wrote - advogato_report.py takes one argument - the advogato username that you want to see a histogram on, of their diary-thingies time advogato rating. ok - it's important that i say that i don't really know _exactly_ how the "rating" thing works - but i believe it takes diary ratings and advogato metrics into account (it may not even take diary ratings into account: i don't really know).
the principle is that i create a histogram - but, instead of an ordinary histogram (count the number of items) i add the "confidence" statistic instead of "1". so it's a kinda 2D map. also i haven't included the "spread" +/- but i have a way that i think i might be able to do that... we'll see.
also, in the output, i deliberately excluded "1" ratings because they seem to massively predominant, making the rest of the histograms utterly irrelevant. again - i don't understand the maths, and haven't looked at the advogato source code _to_ understand the maths.
but.... it is fascinating.
here is nymia's "histogram":
1.5- 2.0 ********* 2.0- 2.5 *** 2.5- 3.0 ******* 3.0- 3.5 ******** 3.5- 4.0 ********** 4.0- 4.5 ****** 4.5- 5.0 ********* 5.0- 5.5 ************* 5.5- 6.0 ********** 6.0- 6.5 ************ 6.5- 7.0 *************** 7.0- 7.5 ************* 7.5- 8.0 *********** 8.0- 8.5 **************** 8.5- 9.0 *********** 9.0- 9.5 ***************************************************************** 9.5-10.0 *******
nice. clean. lots of 9.5s. the proportion of 9.5s is far greater than any other.
this is titus:
1.5- 2.0 *** 2.0- 2.5 ********** 2.5- 3.0 ********** 3.0- 3.5 ********* 3.5- 4.0 ************** 4.0- 4.5 ********* 4.5- 5.0 ******************* 5.0- 5.5 ************************ 5.5- 6.0 ******************************* 6.0- 6.5 ***************************** 6.5- 7.0 ************************************ 7.0- 7.5 ********************* 7.5- 8.0 ********************** 8.0- 8.5 ********************************** 8.5- 9.0 ****************** 9.0- 9.5 ************************ 9.5-10.0 *****************************************************************
again - lots of 9s and 10s - but there is a skew towards the 6s to 8s, and there are proportionately _less_ 9s and 10s as compared to nymia's histogram.
berend's:
1.5- 2.0 * 2.0- 2.5 *************** 2.5- 3.0 ******** 3.0- 3.5 ******** 3.5- 4.0 ******* 4.0- 4.5 ** 4.5- 5.0 ****** 5.0- 5.5 ******* 5.5- 6.0 ******** 6.0- 6.5 **** 6.5- 7.0 ************ 7.0- 7.5 ******* 7.5- 8.0 ****** 8.0- 8.5 ***************************************************************** 8.5- 9.0 ****** 9.0- 9.5 ***************** 9.5-10.0 ********************************
looottts of 8s, lots of 10s (weirdly fewer 9s) and not much else. even compared to nymia's graph, the predominance of 8s and 10s is _really_ sharp.
hypatia:
1.5- 2.0 **** 2.0- 2.5 ******************** 2.5- 3.0 ******** 3.0- 3.5 ******* 3.5- 4.0 ******** 4.0- 4.5 ******** 4.5- 5.0 ********* 5.0- 5.5 *********** 5.5- 6.0 ************ 6.0- 6.5 ********* 6.5- 7.0 ***************************************************************** 7.0- 7.5 ********* 7.5- 8.0 ***************** 8.0- 8.5 ******************************** 8.5- 9.0 ************* 9.0- 9.5 ******** 9.5-10.0 ****
7s, 8s, and 2s. _bizarre_.
now we get to a couple of _really_ interesting ones.
shlomif:
1.5- 2.0 ****** 2.0- 2.5 ****************** 2.5- 3.0 ******************************************************** 3.0- 3.5 ********************* 3.5- 4.0 ********************************* 4.0- 4.5 ************************************ 4.5- 5.0 ********************************************* 5.0- 5.5 ***************************************************************** 5.5- 6.0 *********************************** 6.0- 6.5 ********************************** 6.5- 7.0 ********************************************************** 7.0- 7.5 ****************************** 7.5- 8.0 ********************************************* 8.0- 8.5 ******************************************************* 8.5- 9.0 **************************************************************** 9.0- 9.5 ******************************************* 9.5-10.0 ***********************************************
lkcl (me - this was the whole point of why i did this analysis, to find out about me:)
1.5- 2.0 **** 2.0- 2.5 ********************** 2.5- 3.0 ***************************************** 3.0- 3.5 ************************************* 3.5- 4.0 ************************************************* 4.0- 4.5 ********************** 4.5- 5.0 ***************************************************************** 5.0- 5.5 ************************************************ 5.5- 6.0 ********************************************************** 6.0- 6.5 ************************* 6.5- 7.0 **************************************************************** 7.0- 7.5 ****************************** 7.5- 8.0 ********************************* 8.0- 8.5 ************************************************************* 8.5- 9.0 *************************************** 9.0- 9.5 ***************************************** 9.5-10.0 *******
_whoa_, what's going on here???
bearing in mind that both shlomif and i have been around for at least six years, on advogato, and that the "diary rating" system came in ... abouuut... 3? 4? years ago, this says far more to me than any single "rating statistic".
apart from anything, it says that advogato _desperately_ needs "tagging", and, better, a "rating" associated with the "tag".
that's assuming that the "rating" _has_ anything to do with the diary rating.
but - from a personal perspective, what the massive spread says to me is: i'm challenging people. some of those people respond positively, and some react negatively. some react by going "i do not want to hear what this person has to say. FUCK OFF, BASTARD AAAAGH you caused me PAAAAIN where's that rating button AAAAAAGH godihavetogetthispersonoutofmyLIFEohgod-ohgod-ohgod ahhhhh thaaaaat's betterrrr my ears are now full of wax ahhh, such peace".
some react by going "hmmmmm....." :)
i'd like to find out what the deal is with the massive predominance of "1" ratings, because it says a great deal when people rate you with "1" but the number of "1s" is so extreme that i had to exclude them from the histogram printouts.
take a look on my wiki because i have been doing an enormous amount of research into semantic web stuff.
look up mediawiki with the graphviz extension, wysiwiki, wikiphone, leo, freemind and goldparser amongst other things, and you will go "ohhhhhhhhh" :)
_especially_ look up wysiwiki and wikiphone. note in the pdf that the researchers are particularly irritated that "Logo", a program which ran on an Apple II, and FORTH programs which people wrote for "Logo" back in the 70s and early 80s *don't work*! on modern technology!
i mean what the xxxx is up with that!
p.s. http://wiki.lkcl.net/GraphicalCommunication look up Inkscape - it's a vector editor program which has a real-time collaborative plugin, which rather annoyingly seems to need a server that isn't up - or i don't understand how to use it. needs some better collaborative infrastructure to locate people and projects. i.e. it needs.... *sigh*... it's a long story...
in early 2005, i started reverse-engineering MAPI, starting after an article i wrote. i posted on otlkcon-devel to let people know, and also a couple of other places like the evolution lists.
i was then contacted by someone working on openchange who expressed an interest. i helped guide them through the reverse-engineering process, as up until that time, openchange was going in the wrong direction: they were reverse-engineering the JET database (!!) not the over-the-wire data.
i then recently reviewed the web site, a couple of days ago. i was very surprised to find that there was no mention - at all - of my name, on the openchange web site. no mention at _all_ let alone saying "thank you".
i contacted the openchange people, to ask them about this curious lack. i received no response, so i got on to the irc channel.
what happened makes me very sad - for the people involved.
these logs show, basically, that samba team members are still totally unrepentant of their theft of knowledge and their desire to steal credit for work and techniques that they don't really understand.
at some point, however, the shit is seriously going to hit the fan.
theft has consequences.
robsta: apple has _always_ been worse than microsoft.
remember - they have been stealing free software code and benefitting from it for YEARS, why do you think they started with the BSD kernel??
it took me FOUR YEARS to find their patches repository, completely by mistake, under which they "grudgingly" fulfil their obligations to release GPL and LGPL source code modifications.
probably only because their lawyers said that they had to.
i keep emphasising to people that the core problem is the concept of a "Canceration" - a corporation whose sole and exclusive and pathological purpose is to maximise making money money money money money, and to hell with any people who are exploited, to make that happen.
so - apple is much worse than microsoft: at least microsoft is honest about it and openly goes after intellectual "property" (which is slavery in its most basic form - such arrogance to think that intelligence can be owned).
whereas apple hangs on our every word, every line of code that we create, and drags us down by stealing from us and never giving anything back except by minimalist legal obligation.
and apple isn't the only corporation that does this.
the bottom line: we lead - corporations steal.
oooo dear. the hard drive has failed on my laptop. it's been 2 years: the machine has been pretty much on, constantly, 24x7, all that time (it's linux - why should i switch it off???)
thank god i bought a 250gb external USB drive and thank god i backed it up last month and thank god i managed to do one last final rsync before it totally failed....
now, i have a problem, though: i have all these machines around my house, but none of them as good as my laptop! 1.3 millllion files on it. 1 milllllion files and directories managed by debian. 16 thouuuusand debian packages installed.
i just... i just.... words nearly fail me! :)
p.s. the 8gb flash ide drive arrived: in the 700mhz hp omnibook 500, writing seems to work but there is a problem with reading... oops....
i've been doing two different sets of research. one of them in to PTSD, from which it's becoming increasingly clear that i suffer from, as a direct result of my time at linuxcare, and a combination of my treatment by andrew tridgell and jeremy allison during that time at linuxcare, and by linuxcare themselves (for example - no pay for ten weeks when about $USD 5,000 per month was going out _as well_ as living expenses in australia).
the second set of research is much more interesting (but is indirectly related). it's in intelligence, knowledge, languages (including programming languages) and how quantum mechanics links all those together.
the failures that occurred whilst i was at linuxcare were as a direct result of mismatch of intelligence, knowledge and language.
example: i am far more intelligent than andrew trigell or jeremy allison. i had much more knowledge than them about the workings of NT domains. i entirely lacked the language and the language skills to be able to communicate that knowledge to them [when someone with a pHD in parallel processing starts saying things similar to "surely it doesn't have to be this complicated" then 200dB flood-alarm-bells should be resonating your entire body not just your ears]. i entirely lacked the language and language skills to be able to negotiate a decent contract and to be able to explain to linuxcare that they were in danger of completely fucking me up, very badly.
the bottom line is this: it's becoming increasingly evident to me that it is absolutely 1000% essential that the free software community has a full-on intelligent semantic web development and communications framework.
one in which it is absolutely impossible to mistake someone's work as belonging to someone else, because there is bayesian-classifier "fraud" detection available.
one in which google is made irrelevant, because the tools available to hand, integrated into the desktop, the web environment and the programming environment, are infinitely better.
one in which the work you start doing can be halted immediately because you get a popup saying "hey, what you're doing - the doc strings and the stubs you've got 10% of the way through - looks verrrry similar to this this and this library, and if you run this tool i can convert the 2nd one from its original programming language, fortan 70, into your modern and current programming language that i _think_ you started typing in. did i get this right (yes/no), and if so, would you like me to help you, yes/no?"
the thing is - the components needed for all of the above are currently available.
the technology to achieve the above is already available.
the code, in free software, to achieve the above, already exists.
as usual, however, it's all part of "other projects", and it's "disjoint".
ironically, in this case, if we _had_ the tool i'm describing, then it would be easy to spot the technology components, in the way that i can, a mile off.
it's the chicken-and-egg thing again...
my laptop, already around 2 years old (again), and full of cracks in the plastic and score-lines in the keyboard...
i just kicked the USB handset out of it. the USB and the firewire sockets are nearly ripped out. the handset connector is trashed, and still contains bits of the USB socket.
i've given away a $2,500 fujitsu lifebook because the screen was only 1024x600.
... i'm screwed :)
oh wait - i have an 8-year-old HP 500 laptop, and i have an 8gb Flash IDE drive i want to try out.... hmmmm, i wonder if 0.5ms disk access time makes any difference when you only have a 733mhz processor and 256mb of memory...
well, i'll let you know! :)
hw6915 suspend/resume - might be fixed...
arg arg arg a post by paul psokolovsky on kernel-discuss@handhelds.org describes a horror-story debugging session in suspend/resume, where, it turned out, he hadn't converted _one_ device driver for the h4000 from a legacy struct device to the more up-to-date struct platform_device.
apparently you can't mix-and-match both in your hardware: you have to all struct device or all platform_device.
quick, quick, slow...
argh.
things were going _so_ well on the htc sable (ipaq hw6915) and then i ran into suspend/resume hell for over a week, went to holland for another week, and i think i left the charger there, so i can't carry on until i find it.
in the mean-time, i've been playing with other devices: sound on the blueangel, which is hell, and the s3c2442-based htc hermes, which is hell. all in all, i don't feel like i've actually achieved anything, for over two weeks. and it's pissing me off.
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!
