How I got root / superuser / su on my HTC Desire Android phone and made it like new again

I succeeded on my now-retired 2010 HTC Desire.

After much searching any many dead ends, i found Revolutionary

turn on USB debugging (settings | applications | development )
download and unzip revolutionary, cd to the directory it’s in
sudo ./revolutionary
(you have to run it as superuser because it switches some USB drivers in and out while the phone reboots)

I tried this on an HTC Wildfire, and didn’t quite succeed (maybe with another couple of hours, or another couple of updates??)
HTC Wildfire (buzz-1.01.0002, Android: 2.2.1, ROM version: 2.38.841.1)
First I had to downgrade to buzz-1.01.0001, which is the version without the Telstra (or whoever else) branding and special useless apps.
i followed the goldcard instructions here:

as superuser (because adb needs to load and unload modules)

run all the adb commands in step1.bat

you’ve already made a goldcard, so don’t worry about the GoldCardTool.exe step

run all the commands in step2.bat

it’ll fail on the second last one:

fastboot flash zip

because telstra have special software and special hardware and special branding and everything.  i even tried the official HTC rom with the older hboot:

fastboot flash zip ../../RUU_Buzz_Froyo_HTC_WWE_2.22.405.1_Radio_13.55.55.24H_3.35.20.10_release_160191_signed.exe
sending ‘zip’ (135564 KB)…
OKAY [ 22.908s]
writing ‘zip’…
(bootloader) signature checking…
FAILED (remote: 12 signature verify fail)


back to the official HTC developer site, resign myself to plug my linux-based phone into a windows machine, install the HTC sync drivers, download and run PG7612000_Marvel_hboot_1.09.0099blahblahblah0214.exe

it finds the phone and offers to update the ROM version from 2.38.841.1 to 1.52.841.2

but no.  ERROR (130) MODEL ID ERROR

next good idea:

do all the steps, upgrade the telstra firmware (the only one that works)

then see if it will let you get the token for htcdev


This year (2-13) WordPress turns ten. That means it’s ten years since Howie and Tom tried to convince me to dump my flat html website for a blog. And finally I’ve done it. See, I listen to you fellas.

It’s because of the social thing.

wordpress logo

getting a computer to write a novel

The two novels (if we can agree to call them that) that accompany this introduction are co-winners of the inaugural Hofstadter Prize for Machine-Written Narrative, awarded by the Society for Analytical Engines to the best computer-written novels of seventy-thousand words or more, as judged by a Committee of writers, literary critics, computer scientists, and ordinary humans not unlike yourself. The Bonehead Computer Museum and Bees, or the Floating Point Error, A Dissertation, (“Bonehead” and “Bees,” for short) represent the state of machine-written narratives in the year 1998.

One of the more startling developments in the entire process is that both winning entries were written not in LISP, the programming language generally preferred for artificial intelligence (AI) programs, but in APL (the letters stand for “a programming language”); not only that, they were written in a dialect of APL that runs only on Data General NOVA computers, a model last manufactured in 1982, and currently in use only in the on-board flight computers in Grumman-built AWACs, the military aircraft used for airborne battle command. The actual computer on which the two novels were “written” was obtained at auction of a government surplus, end-of-useful-life AWAC parts, and it is interesting to note, (given the subject of Bonehead) that this machine was in use over the Kasimiyah ammunition dump during the Gulf War.

After the computer was obtained, there still were some interesting problems in setting up the run-time environment for the storywriters. On the hardware side, constructing the NOVA’s information environment required some ingenuity, since NOVAs were largely obsolete before the Internet existed, and therefor there was no easy mating protocol to hook the CPU to the network card. On the software side, the Committee faced the crucial challenge of verifying that the programs behaved as advertised; that is, that they were not hoaxes, the software equivalent of the dwarf-in-the box chess-playing “machines” of the late 1800’s. Making this verification was no mean feat. APL is a language known for its concision, ability to manipulate symbols, and “power;” it is even more famous for being inscrutable even to those adept in programming it. APL was designed to use all the characters on the original “symbol” type-ball of the IBM selectric typewriter, and in appearance it more nearly resembles Egyptian hieroglyphics than any other language. (APL is called a “write-only language,” since nobody knows how to read it.) To make matters worse, the source to the APL compiler was encumbered when Fairchild Semiconductor won its notorious antitrust suit against Data General, therefor the only way to verify that the submitted programs actually “wrote” the novels that they claimed to was by disassembly of the MP/AOS pseudo-op pop code that the compiler produces as an intermediate step? laborious process akin to putting together paper documents that have gone through a shredder. If it were not for the stunning clarity of the MP/AOS assembly language programming manual, this present volume would not exist, and the Hofstadter prize would await its first claimant.

Complete APL sources to the programs that wrote Bonehead Computer Museum and Bees are included on the CD-ROM packaged with this book.

[i forget where i got this.  but it’s quoted on computer-programming-forums with a fantastic discussion including somebody who is the Author of the Chronicles of the Magic Jigsaw Puzzle ]