Diag| Considering item [tag:pappp.net,2025-09-03:/2412311] "The worst possible antitrust outcome"
Diag| Considering item [tag:pappp.net,2025-09-03:/2412267] "Mis-issued certificates for 1.1.1.1 DNS service pose a threat to the Internet"
Diag| Considering item [tag:pappp.net,2025-09-02:/2412032] "Judge: Google can keep Chrome, must share search data with “qualified competitors”"
Diag| Considering item [tag:pappp.net,2025-09-02:/2411957] "This ultra-rare ’90s LaserDisc game console can finally be emulated on a PC"
Diag| Considering item [tag:pappp.net,2025-09-02:/2411847] "Imgur's community was in revolt"
Diag| Considering item [tag:pappp.net,2025-09-01:/2411633] "Intel Patents 'Software Defined Supercore'"
Diag| Considering item [tag:pappp.net,2025-08-26:/2410252] "Doge uploaded live copy of Social Security database to 'vulnerable' cloud server"
Diag| Considering item [tag:pappp.net,2025-08-26:/2410211] "We regret but have to temporary suspend the shipments to USA"
Diag| Considering item [tag:pappp.net,2025-08-26:/2410193] "Troubled USB Device? This Tool Can Help"
Diag| Considering item [tag:pappp.net,2025-08-26:/2410135] "The size of Adobe Reader installers through the years"
Diag| Considering item [tag:pappp.net,2025-08-25:/2409944] "Google will allow only apps from verified developers to be installed on Android"
Diag| Considering item [tag:pappp.net,2025-08-24:/2409578] "Picking an Old Operating System"
Diag| Considering item [tag:pappp.net,2025-08-22:/2409290] "US government takes 10 percent stake in Intel in exchange for money it was already on the hook for"
Diag| Considering item [tag:pappp.net,2025-08-22:/2409328] "Nitro: A tiny but flexible init system and process supervisor"
Diag| Considering item [tag:pappp.net,2025-08-21:/2408908] "James Dobson, Influential Leader of the Religious Right, Dies at 89"
Diag| Considering item [tag:pappp.net,2025-08-20:/2408610] "FDA warns public to throw out potentially radioactive shrimp"
Diag| Considering item [tag:pappp.net,2025-08-15:/2407469] "HTTP/1.1 must die: the desync endgame"
Diag| Considering item [tag:pappp.net,2025-08-15:/2407392] "Open hardware desktop 3D printing is dead – you just don’t know it yet"
Diag| Considering item [tag:pappp.net,2025-08-15:/2407386] "What kids told us about how to get them off their phones"
Diag| Considering item [tag:pappp.net,2025-08-14:/2407170] ""Privacy preserving age verification" is bullshit"
Diag| Considering item [tag:pappp.net,2025-08-12:/2406472] "Firefox’ new “AI” features cause CPU spikes and battery drain"
Diag| Considering item [tag:pappp.net,2025-08-11:/2406239] "Reddit will block the Internet Archive"
Diag| Considering item [tag:pappp.net,2025-08-11:/2406191] "GitHub is no longer independent at Microsoft after CEO resignation"
Diag| Considering item [tag:pappp.net,2025-08-11:/2406033] "Vanishing from Hyundai’s data network"
Diag| Considering item [tag:pappp.net,2025-08-13:/2406677] "QNX: The Incredible 1.44M Demo"
Diag| Considering item [tag:pappp.net,2025-08-08:/2405403] "New executive order puts all grants under political control"
Diag| Considering item [tag:pappp.net,2025-08-06:/2404914] "The Real Origin of Cisco Systems (1999)"
Diag| Considering item [tag:pappp.net,2025-08-01:/2403752] "Tesla loses Autopilot wrongful death case in $329 million verdict"
Diag| Considering item [tag:pappp.net,2025-08-01:/2403745] "Tested: Microsoft Recall can still capture credit cards and passwords, a treasure trove for crooks"
Diag| Considering item [tag:pappp.net,2025-07-31:/2403404] "Epic just won its Google lawsuit again, and Android may never be the same"
Diag| Considering item [tag:pappp.net,2025-07-29:/2402644] "RP2350 A4, RP2354, and a New Hacking Challenge"
Diag| Considering item [tag:pappp.net,2025-07-28:/2402376] "Windows 11 is a minefield of micro-aggressions in the shipping lane of progressl"
Diag| Considering item [tag:pappp.net,2025-07-24:/2401424] "Efficient Computer's Electron E1 CPU"
Diag| Considering item [tag:pappp.net,2025-07-22:/2400766] "A Professor’s Search for the MingKwai, a Lost Chinese Typewriter"
Diag| Considering item [tag:pappp.net,2025-07-21:/2400386] "Vibe coding service Replit deleted user’s production database, faked data, told fibs galore"
Diag| Considering item [tag:pappp.net,2025-07-21:/2400437] "NIH limits scientists to six applications per year"
Diag| Considering item [tag:pappp.net,2025-07-21:/2400421] "An artificially complex XML schema as a lock-in tool"
Diag| Considering item [tag:pappp.net,2025-07-22:/2400779] "MakeShift: Security Analysis of Shimano Di2 Wireless Gear Shifting in Bicycles"
Diag| Considering item [tag:pappp.net,2025-07-18:/2399996] "Replication of Quantum Factorisation Records with a VIC-20, an Abacus, and a Dog"
Diag| Considering item [tag:pappp.net,2025-07-15:/2399026] "Marc Andreessen Is a Traitor"
Diag| Considering item [tag:pappp.net,2025-07-08:/2397262] "GlobalFoundries to Acquire MIPS"
Diag| Considering item [tag:pappp.net,2025-07-04:/2396489] "Nvidia won, we all lost"
Diag| Considering item [tag:pappp.net,2025-07-04:/2396473] "Ubuntu 25.10 to drop support for effectively all existing RISC-V hardware, focuses on future RISC-V hardware instead"
Diag| Considering item [tag:pappp.net,2025-07-03:/2396128] "AI note takers are flooding Zoom calls as workers opt to skip meetings"
Diag| Considering item [tag:pappp.net,2025-07-03:/2396110] "Surveillance Used by a Drug Cartel"
Diag| Considering item [tag:pappp.net,2025-07-02:/2395852] "Why The Latest Linux Kernel Won’t Run On Your 486 And 586 Anymore"
Diag| Considering item [tag:pappp.net,2025-07-01:/2395685] "Donkey Kong Country 2 and open bus"
Diag| Considering item [tag:pappp.net,2025-06-29:/2395064] "Wayback: experimental layer to run X desktop environments on Wayland"
Diag| Considering item [tag:pappp.net,2025-06-29:/2395059] "“I want a good parallel computer”"
Diag| Considering item [tag:pappp.net,2025-06-27:/2394461] "Apple Just Patented an Image Sensor with 20 Stops of Dynamic Range"
My research group will be building our new (smallish) research supercomputer NAK:(NVIDIA Athlon XP cluster in Kentucky) on Friday, April 16, 2010 from 10A to 4P in FPAT672. UK students and other interested Lexingtonians are invited to come help with the build, so if you would like to play with the guts of a big cluster, you will be welcome at the (re)Build Party.
If you can come up with a better phrase (with a better acronym) for the “NoBuPAG” principle discussed in the machine description, that will be really welcome too.
NAK will provide a testbed for continuing research into building tools for performing useful compute work on GPUs. It presents a different model than the conventional GPU as an attached co-processor to powerful compute nodes model, which has thus far proven impractical to program for. Instead, NAK treats the nodes as “Nothing But Power And Ground” (and a network interface…), and will be running all of the heavy compute on the GPUs themselves, through a mechanism extended from our MOG project.
I’m getting my Buildycrunk on, so are all these people. You should be too.
I’ve seen coding, grading, knitting, antenna fabrication, card games, board games, and all kinds of awesomeness.
Last time was amazing, and many excellent things were accomplished. Everyone rest up, come up with something to hack on, and show up; a grand time will be had by all.
There will probably be more detailed information online shortly; I am slightly out of the loop because I missed last week’s Collexion Lunch (due to allergies), and this week’s Wednesday Collexion meeting due to to sticking around, keeping the lab open, and helping some of my kids finish up tedium lab week in EE281 (where we make the kids design and build something that takes at least 4 TTL chips, by hand).
I’ve decided I need to start using a reference manager utility. My old system of keeping a text file full of BibTeX entries in a folder with pdf’s, with an extra “file:” field for the file name of the document is a little crude, and starting to break down as I get large piles of documents for some topics.
Because my PSY562 class this semester is largely using readings from HFES journals, I’m going to use the pile it generates as a testbed to find something I can use generally.
My requirements for a reference manager include:
* Accepts and Emits BibTeX Citations
* Capable of linking citations to files
* Storage format which is (roughly) human readable
* Easily transported database
* FOSS
* Works on Linux
* Limited dependencies (I’d prefer to avoid Java or Qt)
* Works without network connection
The most widely used solution, EndNote, fulfills very few of the above (plus, interoperability dickishness, but fortunately there are lots of projects to make reference managers floating around the ‘net that seem promising. Unfortunately, most of the promising ones are dead. The best of the actively-developed bunch seems to be Referencer. Referencer is a C++/gtkmm app (so it plays nice with my XFCE4 environment), with a Python plug-in system (Should I ever choose to use it), reasonably limited ties to the various non-GTK gnome libraries, and stores it’s records in XML. It looks remarkably close to what I want, and has some features I didn’t know I wanted (preview icons, tags) that are pretty useful. I think it’s going to be a keeper, but would love to hear what other people are using.
Earlier tonight I attended a sort of open-access seminar on multi-touch user interfaces, catalyzed by an EE senior design group working with Awesome Inc. to create a large multi-touch wall for the outside of their space. The attendece was wonderfully diverse on account of the announcement hitting the professional, academic, and hobbyist communities in Lexington (more events need to propagate like that!). Attendees included several members from the Lexington IEEE chapter (co-opted as a chapter meeting), a number of local creative types, many students from UK, and several other interesting folks.
The discussion centered around the CCV effort of the NUI group, and was fairly solid, although I would have appreciated a bit more technical depth. I actually ended up dominating a couple conversations on account of being better read in the area than most of the other attendees, and don’t consider myself particularly well-versed in multi-touch display technology. I still don’t really understand TUIO, I was rather hoping someone there would be able to explain it.
One small downside, I still get the uncomfortable feeling that everything said or done at Awesome is being sucked in and analyzed as a potential source of financial gain. The culture there always seems exploitative (or at the very least commercial) instead of communal, which is very unfortunate, as they have set up a nice space, and seem to be attracting interesting events.
I really am attracted to open-access, discussion based topic seminars like this, and would love to see more of them happen. There are lots of good efforts to bring that sort of thing to Lexington, from a seminar series Dr. Finkel is attempting to arrange this semester (based on students and faculty giving short presentations on neat things they have found), to Collexion and Dorkbot’s regular meetings. This is not the first time I’ve been after this sort of thing either, for a while the UK LUG was running some decent events in this vein, in particular I remember a successful LUG event on PyGTK, but the LUG is several years defunct…again…because those of us who were active didn’t have time to keep it going on our on (and be students), and no one else stepped up. I think some of the short-form lecture series like ignite have had events here as well, but those have never seemed as useful to me.
(This is very much an example of one of the little manic episodes that make me a good generalist/appear high functioning)
The left touchpad button on my laptop (Thinkpad T60p, hostname Monolith) has been “limp” for a while. It bothers other people who use my machine, because (objectively) it really does feel very wrong, but it had broken gradually and I had acclimated enough that it didn’t bother me. Last night I started paying attention to the problem, and it became maddening, so I decided to see if I could fix it. I looked at the problem last time I had the machine apart, so I knew there was a torn plastic tactile dome to blame. It is (as best as I can make out) impossible to order just the appropriate domes, and a whole new touchpad is 1. defeatist, 2. about $12 from shady ebay sellers, and 3. requires waiting for it to be shipped. I decided a better (ie. creative, free, immediate, and credit-card-fraud free) solution would be to go rummage in the parts bins, find a sufficiently similar tactile dome in something dead, and install it. The closest match I could find was the keyboard domes from the corpse of my old VPR Matrix 180B5 (The worst made laptop I have ever encountered. Every bit as fragile as one would expect something made by a Best Buy house brand to be, even though it was basically a re-badged Samsung P10. Polystyrene is not chassis material.) I now have a partial match (it’s a little too weak, and not “snappy” enough) installed, which is good enough to keep it from being bothersome.
Thinking about tactile domes reminded me of a fabulous article I read (I thought) about them several years ago. It turns out it was a much more general article about handheld devices, but it really was fabulous. The article is “Handhelds of Tomorrow” from the April 2002 issue of Technology Review. Ideo has a PDF available outside a paywall. The part about the tactile bubbles was one little subsection about Peter Skillman, who was “the hardware guy” at palm/handspring (weird corporate history).
The search for the article reminded me of a previous kick on the work of one of the other important palm/handspring people, Jeff Hawkins, who in addition to being a founder of both companies, is doing amazing work in neuroscience as it relates to computing, and has written a book On Intelligence and given a awesome TED talk on the topic.
Hurrah for (hypo)manic episodes?
I scored a (at least mostly) working Point of Sale terminal from the trash at the nearby location of a fast-food sandwich chain named for a mode of mass transit. It’s a pretty nifty little piece of hardware, a Micros Eclipse 400498, based on a 1Ghz Pentium 3, 512Mb of RAM, and normal (ish) PC-BIOS. It has a broken WinXP install on it now, I think it shipped with an older NT version, because the drivers are all fucked up. The fun part is the attached goodies: a 3-track magnetic stripe reader (credit cards, etc.), 2-line VFD display, and a 12” touchscreen (only 800×600, but pretty crisp and good colors).
I’m thinking it will make a bitchin’ jukebox. Scroll the track information on the VFD, put a touch-flow-esque interface on the touchscreen, hook up some speakers, etc. One of the housemates suggested a barmonkey (it could even process payments), but that is a WAY more involved project because of the valve rigs, and I have plenty of unfinished involved projects right now.
It is however being obstinate about drivers and alternative boot devices, and the manufacturer (micros) seems to believe that not providing any drivers/manuals/support of any kind will enforce support plans/upgrades/create security by obscurity or something, because their website is supremely unforthcoming. There don’t appear to be any WinXP compatible drivers for the various hardware, so the best choice is probably to try for Linux. Its not wanting to boot off USB devices (despite the BIOS’ claims that it will), doesn’t seem to like having a 2-channel ATA cable attached, has no CD drive, and has repeatedly failed on the wubi-install-from-USB trick, so it will be exciting to get going. Maybe some sort of netboot stunt.
After missing a few due to travel and other obligations, I’m back to attending DorkbotLex. #9 was last Saturday (the 19th) and, as is traditional, I’ve posted up some pictures on flickr. This month’s presentations were both electronic music projects, one set of hardware hacks and one piece of prototype software. As always, very cool, and a great source of energy for creative endeavors.
The camera handling for my last post reminded me to check to see if progress had been made on the CHDK port to for the Canon SD770 (the little point-and-shoot I have). There was a promising effort about a year ago, but the developer disappeared without releasing the partly functioning code, and it was quickly abandoned. Apparently someone else stepped up in November; there is now a (roughly) fully functional boot image available for the SD770.
CHDK features provide at least partial fixes for all my major complaints about the 770; the exposure override settings allow the flash to be kept under manual control through power cycles, even when the other features are on automatic, the exposure behavior can be more precisely controlled to hide noise issues in low light, and all the CHDK toys are now on hand, so I’m not missing any expected features.
Right now I just have a spare small card set up, but the process is non-destructive and simple, especially on cards <4Gb, so I’ll probably set up the pair of 1GB cards I usually use with it shortly. Anyone with a Canon camera should go set up a card with the appropriate CHDK image, it really does improve the camera.
EDIT: There are a couple bugs, the only serious one being that the camera crashes when using auto white balance with the version of CHDK in the linked thread loaded. I don’t have time to get into another environment to fix it.
(eventually I’m going to have to hack up a light tent (for the glare) and camera mount(for the jitter) for this kind of thing, but it hardly seems necessary with my little point-and-shoot)
One of the more promising bits of schwag at SC09 was a little USB button thing handed out by Pionier. The basic premise on the button is that after connecting it to a computer via USB, it lights up with enticing shifting colors, and, when pressed, sends the computer to the pionier SC09 website. It does this by announcing itself as a USB HID device, emulating a keyboard, and typing “{control}rhttp://www.pionier.net.pl/webstarter/09scpo6r8q {newline}”, which will have the desired effect… from the desktop on a windows box. After seeing what it did on a laptop on the show floor, I grabbed a few extras thinking they would be fun to hack.
I pulled apart one of the buttons to read off what the chips are, and things look promising. (larger image linked)
The board is based around a MXT8208 USB 2.0 Flash Disk Controller (datasheet). The chip consists of a 80c51 microcontroller with hardware USB 2.0 PHY, I2C, a general purpose UART, some dedicated flash management features, and 28 GPIO pins (mostly overlapped with other functions) in a LQFP48 package. This means it SHOULD be susceptible to much better hacks than simply changing the output string; with a little luck it will be convertible into a darling little intermediary device for attaching projects to a computer via USB by serial or PIO. Since there is no flash chip hooked up, the majority of the GPIO pins are free, making it especially suitable for that kind of repurposing. By far the primary use for this chip is making little USB flash drives (and, based on the information around the ‘net, this one is particularly favored in making “fakes” that misreport their capacity, because it is more programmable than many of it’s competitors), but, assuming I am interpreting the chinglish datasheet correctly, the non-memory USB widget use case here is intended as well. Matching the suggestion in the datasheet, the software for the button behavior is (apparently) on a K24C64 64kbit Two-wire EEPROM (Datasheet) mounted on the other side of the board.
The other components aren’t terribly interesting: a pair of SMT push buttons, a 12Mhz crystal, 6 assorted SMT capacitors, 3 assorted SMT resistors, and a 2-lead RGB color phasing LED(unfortunately, 2 leads means it probably can’t have it’s behavior altered beyond on/off).
Software wise, it reports VID=0×2000, PID=0xbeba, which seems to be a made up ID written to the chip in software. There is a UdTools utility offered up by micov (after a little bit of google translate-foo), but it only seems to be able to tweak the flash-specific functions, not perform general reprogramming. This IS promising, as it implies everything should be writable from a host computer, rather than having to pull the EEPROM and program it separately to alter the behavior, but I haven’t figured out how to do so yet.
