Joanna About this site

About

How this site is organized and what it's for

Weblog start page

The start page contains the most recent 15 articles.

Home page
The main home page of my website, not my weblog. Currently not used.
------------------
Articles by month
Click here to get all the articles for a particular month.
This month's articles (if any)
Current month
Today's articles (if any)
Articles dated 2008/10/12 only

------------------
Subtopics
PC

------------------
My email address
Site map
Search my weblog
Search for text on this site
You may have to use search
if I move files around!
Listing of all articles by date
Moving man
Flavours
There's more than one way to view this weblog; these links display the current page in other formats.
External links
These are a few of my favourite sites.
T E S T
Slashdot yesterday

Copyright © 2003-2007 Alternate Worlds Publishing, Boston MA USA


powered by blosxom -- www.blosxom.com
Wenhua dageming de zhongyao jiaoxun shi bixu fandui geren mixin
If I have been able to see further, it is because I am surrounded by midgets.
Never ascribe to stupidity that which can adequately be explained by malice.
"Your argument's repugnant and intriguing." "That's kinda my thing."

Danny's Weblog

2006 Apr 26 [ Wed ]

Another reason for my laptop overheating problem?

A few days ago my laptop abruptly died during a video game. I figured it was overheating again. As I folded it up to put it away I noticed that, as usual, the steel tabletop under it was quite hot.

Then it occurred to me that the little rubber feet originally stuck to the bottom of the case had fallen off a year or two ago. Hmm. I tried sticking some domestic rubber feet on – after cutting them to size with scissors – and the laptop now sits *much* higher above the table surface. Now you might think that reduces heat transfer to the table – lose – but the intake port for the fan is *under* the laptop, so win, big. The outlet airflow feels distinctly faster.

Now I feel a little silly for not thinking of this before. At the time I first noticed the original feet were dropping off I vaguely figured the separation was defined by some other plastic protrusions in the case molding but now I look at it they are clearly inadequate.

2006 Jan 06 [ Fri ]

My laptop woes abate

My Atec laptop has had various problems, although it's never needed an expensive repair.

Recently it started shutting down very rapidly, amid apparently bogus "low-battery" warnings. In fear and trembling, I disassembled it. I was expecting to find the fan compartment was clogged with dust, but it was almost clean. However, there was almost no heatsink compound detectable and what was visible had hardened. I figured this matched up with the fact that the case of the machine got red-hot but the exhaust air was scarcely warm, so I went looking for heatsink compound.

I had assumed this would be available because there are a ton of places that assemble PCs, but I had no luck. Eventually a tech at PTC on Monivong took pity on me and gave me a little hypodermic for free (hmmm... I wonder if he filled it with toothpaste... nah...).

I opened up the laptop again and reassembled it with the heatsink compound. (I have at least a dozen screws left over – no joke.)

Then I didn't get the fast bogus low-battery shutdowns; now I was getting *instantaneous* shutdowns.

I figured that when I reassembled the case I had left something loose enough that a slight flexing could short out the power supply. I experimented with using an external mouse and keybd, and that worked very well.

After a while I tried running without the external devices, and the laptop *still* worked. Then it occurred to me that the period when I suffered the instananeous shutdowns was exceptionally cold and dry, with nighttime temperatures going as low as 70 F; and right around the start of the problem I had actually felt a static zap in an air-conditioned store. Hmmm! I guess what I had actually done was loosened one of the earthing straps which, when you open up a laptop, you can always see were added in desperation at the last moment before the ship date, and get in your way until the tacked-on solder joint snaps. Static zap: foom. I guess I was lucky it just reset the machine rather than burning something out.

Anyhow, now it seems to be working fine... touch wood. I'm hoping the earthing wire got jiggled back in place; but it's probably just hot and humid now.

2005 Jul 07 [ Thu ]

Monitoring temperature etc in your data center

Slashdot discussion: ask.slashdot.org [http://ask.slashdot.org/askslashdot/05/06/30/1643255.shtml]

Summary:

1. There are standalone boxes which connect to the phone line and you set then up to call you if a threshold is tripped. Not cheap, and not automatable, but for *emergency* monitoring you probably want to consider a solution that does not *depend* on something you're trying to monitor.

2. APC and other UPS hardware includes (at least in their pricey models) a temperature sensor which can be accessed by their (pricey) monitoring software

3. Various places do sensor gizmos that plug into a serial or Ethernet port and tell you humidity, windspeed etc

4. Several posters recommended Nagios, free software for network monitoring *for which many commercial hardware devices are available*: http://www.nagios.org

5. If you are very cheap you can wire up an ordinary wall thermostat to a serial port to detect the contact closure

6. Your building may already have a fire alarm monitoring system which could be wired up to a contact closure.

2005 May 10 [ Tue ]

Interesting phoneline/VoIP devices

In the course of a Slashdot discussion on the SPA 3000 home PBX system – and inevitably the Asterisk free-software PBX system – there were a couple of links to fascinating and inexpensive hardware: it.slashdot.org [http://it.slashdot.org/it/05/05/09/1715243.shtml]

Sipura sells little boxes for less than 100 USD that connect up your analog and VoIP components: www.voipsupply.com [http://www.voipsupply.com/index.php?manufacturers_id=14]

Review of a Sipura box. It's just 59.95 USD but you set it up via a web interface!: www.geekgazette.com [http://www.geekgazette.com/index.php?option=com_content&task=view&id=28&Itemid=26]

Later on in the above article they talk about setting it up to talk to an Asterisk controller but I couldn't quite grok whether they were just using the Asterisk system for *test* purposes or whether the intention of the Sipura box is to function as a hardware interface between an ethernet-only Asterisk box and a POTS phone. My apologies.

Globalinx have nifty little adapters which interconnect your computer, analog phone, fax etc. The one shown here doesn't have VoIP support but man it has ports for six different devices! www.digitone.com [http://www.digitone.com/]

I sure wish something like this were available in Cambodia. So far all I've seen is a two-extension adapter at Phnom Penh General Electric on Monivong (it has a very low-tech look, as if it works off two 2N2926's), and some slightly cheesy VoIP boxes at PTC on Monivong.

2005 Apr 03 [ Sun ]

All new ATA hard drives include little-known password feature

The German magazine c/t has a good article on this (in not very mangled English): www.heise.de [http://www.heise.de/ct/english/05/08/172/]

The intention was to make it impossible to access the drive until you put in a user-selectable password – which is good, although specialist data retrieval companies have developed ways to get around it. It's apparently been a part of all 3.5 and 2.5 drives for years.

c/t however points out that it is theoretically possible for malware to set the password. Conceivably you would then need pay the malware writer a ransom to get access to your own data!

The article says that this feature is more often accessible on laptops (because they are more liable to theft, presumably) than on desktop PCs. But at a minimum, the BIOS *should* set the feature to disabled until the next cold boot.

I should certainly check to see if this is a BIOS option (I don't think so, but BIOS options are often impossible to understand unless you already know about the feature.) The c/t article has links to various utilities, but it sounds like you should try them out on a test disk before risking your own data. In the Slashdot discussion I found the c/t link in, one poster said that some Dell models allowed you to *set* a password using some punctuation characters, but then *rejected* those characters when you tried to *use* the password! Yipe! (I've run into a similar problem with Unix passwords: you can enter a pw of arbitrary length, but it only *reads* eight characters, and if you enter the *full original* password it is rejected!)

c/t includes this link to unlocking an ata drive which gets locked: www.rockbox.org [http://www.rockbox.org/lock.html] You may need to issue the "security erase" command. Gulp!

2005 Mar 27 [ Sun ]

Mounting speakers in walls

Many years ago I was interested in hi-fi and learned all about speaker design: impedance curves, crossover networks, resonance, etc.

The big problem with speakers has always been reflections. Whether from the internal surfaces of the speaker itself, or the surroundings, the reflections cause undesired resonance, impedance anomalies, etc.

I was vaguely musing about this a few weeks ago and it occurred to me that many of these problems are alleviated by mounting speakers in the wall and allowing the rear of the speakers to remain open into a large, relatively non-reflecting room. (Not my original idea!)

Today I Googled for that idea and was unable to find it. In particular, when I looked for wall-mounted speaker info, the manufacturers made no recommendations that I could find for making sure that the speakers were mounted in a repeatable way – that is that ensures that the required specifications for frequency response etc can possibly be met.

Likewise, some happytalk pages about installing wall-mount speakers (aimed at people moving into a new house) made no mention of verifying that the response of the speakers would be adequate, and gave the impression that they could be covered (concealed) with various fabrics at the whim of the homeowner. (Ie, the effect on frequency response of high-frequency losses in the fabric – or anything else – was ignored.)

Many commercial speakers (in cabinets) sound rotten even after (presumably) someone verifying the design. What is the chance that speakers will have any fidelity at all after installation by a *builder*? This reminds me of the way humanity has lost manned spaceflight capability; I suppose it's hardly as important, but it's just as significant, that what was once *common knowledge* about acoustic design has disappeared from the culture.

Incidentally, in Googling for the idea above, I was curious what solutions had been found for maintaining phase coherence between high and low frequencies if the high-frequency radiators are mounted separately from the low-frequency (allowing the high-frequency radiators to be swivelled inwards for a larger "sweet spot"). I was expecting digital delay lines and whatnot... but no....

This guy sounds like he knows more than I ever did, so I guess I can't say audio knowledge has disappeared from the culture *entirely*: www.helsinki.fi [http://www.helsinki.fi/~ssyreeni/dsound/dsound-a-03] (it's a long article about sound synthesis, and discusses eg "matrixing subchannel coefficients" *before* the "Advanced Topics" section).

On the other hand, it seems to me that three decades ago a lot of the webpages one sees advertising audio would have been viewed as preposterously ill-educated by most audio consumers.

2005 Mar 22 [ Tue ]

A list of apps you can run directly from your USB key without needing to install

loosewire.typepad.com [http://loosewire.typepad.com/blog/2005/03/a_directory_of_.html]

His actual list is not very long, but it's interesting to see them organized together, and he has a lot of comments in his blog with more info.

One app that is not on his list is Putty, or anything like it. Putty is the ssh client that I use; it includes normal ssh features like command-line scp, and a gui to set up tunnelling. Putty is great, and it is possible to run it from usb without installing the executables, but for some reason it saves its preferences in thew Windows registry, without even alerting you. Every time you use Putty on a machine, you are leaving traces which can lead a hacker to your account. Oh well.

Slashdot discussion: slashdot.org [http://slashdot.org/articles/05/03/21/2220255.shtml

] 2005 Feb 23 [ Wed ]

How to deal with a recalcitrant Epson printer

I used to have a cheap Epson printer. I was quite pleased with it, but everything about changing cartridges was a pain. In particular, it seemed to want to prevent me from changing the cartridge sometimes and I never figured out why.

This guy offers the answer:

That doesn't stop Epson from selling ink (Score:4, Insightful) by Webmoth (75878) on Tuesday February 22, @08:58PM (#11751641) (slashdot.org [http://slashdot.org/)] yro.slashdot.org [http://yro.slashdot.org/comments.pl?sid=140311&cid=11751641]

So, my Epson R300 has ink cartridges that are just ink. However, it meters use, and assumes that a cartridge lasts for so many seconds at such-and-such coverage. Therefore, the cartridge can still have a fair amount of ink in it before it tells you to replace it, OR it can go empty before it tells you to replace it. I'm guessing they build in a pretty good fudge factor to ensure that it never goes dry.

Here's the stinker: most Epson printers will NOT let you replace the cartridge until it says it needs to be replaced. So if it tells you it needs to be replaced, and you just pop the old cart out and put it right back in, it will assume that a new, full cart is installed. Then when it DOES run dry, it won't let you replace it because it doesn't think that it's empty.

There's a workaround though: turn off the printer. Then look under the printhead carriage, there'll be a plastic tab that prevents you from sliding the carriage out to where you can change the cart. Just flip this tab forward, and replace the cart. Slide the carriage back, and turn on the printer. It won't even know that you've just changed the cartridge.

Since the cart is separate from the head, and the head isn't replaceable, it's probably a good idea to NOT let it run truly empty, as then you'll end up with air in the head that you'll have to purge.

I've got another gripe about inkjets, and they all seem to do this. If, say, your cyan has a blocked head, you can't just clean the cyan. You have to clean them all. This wastes ink from colors that don't need to be cleaned!

It's not cleaning the heads, it's cleaning out your wallet.

The cleaning thing irritated me too, but I assumed it was just laziness in the software writer... but now I can see that it's all part of the same picture.

Here's another interesting post about Epsons. I never heard of this:

Re:Epson printers... (Score:5, Informative) by Nogami_Saeko (466595) on Tuesday February 22, @06:40PM (#11750469) yro.slashdot.org [http://yro.slashdot.org/comments.pl?sid=140311&cid=11750469]

Not exactly the same. Epsons will stop printing when the ink recovery pad in the bottom of the printer is saturated (according to whatever calculation they use). The upside of this is that you don't have to worry about ink pouring out of the bottom of your printer because the pad overflows.

It's possible to reset the printer by pressing a combination of keys on the front panel. Of course, it's recommended that you remove and clean the ink sponge first (there are websites that show how to do this).

N.

2005 Feb 05 [ Sat ]

Can you believe the "Media Manufacturer" and "Dye Data" fields?

CDs have various pieces of information encoded in the "ATIP" region. In a posting in a thread about CD longevity: slashdot.org [http://slashdot.org/articles/05/02/05/0024258.shtml] someone quoted an interesting statement from the CDR FAQ:

Re:I've not seen the silver/gold (Score:4, Informative) by The Vulture (248871) on Saturday February 05, @12:22AM (#11580464) (www.vulturesnest.net [http://www.vulturesnest.net/)] slashdot.org [http://slashdot.org/comments.pl?sid=138391&cid=11580464]

I have to wonder how true the results of Alcohol 120 really are though. (Note that I've not used it, so I can't really say).

From the CD-Recordable FAQ [cdrfaq.org]: www.cdrfaq.org [http://www.cdrfaq.org/faq02.html#S2-33]

"Two components that many users of these programs always take as gospel are Media Manufacturer and Dye Data. These two readings are next to worthless.

The reason for this is that many CD-R manufacturers (like CD- Recordable.com) purchase their stampers (the nickel die that all CD-R substrates are molded from) from 3rd party sources. These 3rd party sources (either other disc manufacturers, or mastering houses) encode the data that these 'Identification' programs read, at the time that the original glass master is encoded. The 'Manufacturer' information that is encoded is usually the name of the company that made the master. Since stampers made from that master will be sold to disc manufacturers the world over, all of discs that those manufacturers produce from those stampers will contain the same 'Manufacturer' information. Information which is obviously quite erroneous and irrelevant. Very seldom will the 'manufacturer' information encoded on a CD-R actually tell you anything other than who made the original master. [...]

The second piece of data (the dye type) is also dubious. Because most master/stamper configurations are designed to be matched to specific dye types (Phthalocyanine, Cyanine, Azo, Etc), the 'Dye' information that is encoded when the master is produced indicates the type of dye that the master was designed for. This of course, does not assure that the manufacturer that buys and uses this stamper will be using it with the dye that it has been designed for. It is quite possible that a stamper/dye combination is used by a CD-R manufacturer that contradicts the 'dye' information encoded on the master. Therefore that information becomes as potentially misleading as the 'Manufacturer' data discussed earlier."

The only reliable piece of information in the "ATIP" region is the disc length. See section (2-38) for further remarks."

– Joe

2005 Jan 20 [ Thu ]

Solutions to the tangled-headphone-cord problem

Several moons ago I mentioned that an MP3 player I had bought, like all such devices, has an irritating tendency to get the headphone cord tangled: www.panix.com [http://www.panix.com/~dannyw/weblog/Reviews/Hardware/pocki02.html]

Here's an interesting design to keep the cord neatly retained in storage (Sumajin Smartwrap): schwarztech.us [http://schwarztech.us/reviews/smartwrap.shtml]

It appeals to me more than this gizmo which provides a self-retracting cord, like you've seen for modem-phone cables, or for mice (JavoEdge Ear Buds):

www.pocketpcthoughts.com [http://www.pocketpcthoughts.com/articles.php?action=expand,24574:dont_paginate,1]

www.davespda.com [http://www.davespda.com/reviews/other/javoebuds.htm]

As the second link shows, the drawback of the JavoEdge EBuds is that you are stuck with their headphones. Most people like to have a choice of headphones. I was very struck from the table in the second link that the Koss KSC-19 headphones have *20 dB* more sensitivity than the JavoEdge! Plus more than *two octaves* of extra low end.

Also, if I were designing the product, I would have made something to retain and protect the "buds" themselves. Making a nice rugged package to protect the cable but allowing the buds to twist in the wind just seems all wrong. I would use both products with a case.

Anyway, I don't understand why *all* headphones don't come with some simple retainer along the line of the Smartwrap. Humanity probably loses a million hours a day while people fumingly disentangle their headphone cords. Something like this (although it looks a little flaky): www.cableorganizer.com [http://www.cableorganizer.com/cord-it/]

2005 Jan 17 [ Mon ]

Mobile phones and their software

I have been quite interested in the latest advances in mobile phones, but I still use a basic model because my impression is that the multimedia etc features are poorly standardized: if you try to send even a .jpg to your buddy, it will not be supported by your phone, your network, his network, his phone, or software.

Additionally, the little convenience features are never spelled out, and on a device with such a lousy user interface the convenience aspects are crucial. For instance, my current phone has a big *inconvenience* feature: if you hold the backspace key down for too long (perhaps a second!) it deletes the entire SMS message you've been keying in! What maroon thought of that?

Anyway, I know a girl who has a Nokia 3200, an early camera phone, and tried to charm her by figuring out some of the features. I had previously tried to download Nokia sw and run into some peculiar provisions on the Nokia websites. In particular they required you to register your phone in order to make a download. Wtf?

Seemingly they have changed their policy. The following file was easy to find and download. It works for umpteen models (some restrictions on the features of course for different phones) for W2K and WXP.

nds2.nokia.com [http://nds2.nokia.com/files/support/global/phones/software/Nokia_PC_Suite_641_] en.exe

I was also able to download the manual as a .pdf; I don't have the link handy.

I had no problems it all making it work under W2K with the infrared port though it's dog-slow; also, when I load new sw that uses my infrared port it tends to kill previous sw which was trying to use that port, so I'm thinking doomily about my Palm IR link.

The interface is a little strange. I was able to ctrl-a select all the files in a folder, but then when I tried to copy/paste them or drag them nothing happened. it turns out that you need to use not Nokia's nifty application file windows, but look under "My computer" for the Nokia device. (Make sure you look for the actual device, not some sort of sync folder arrangement which I haven't figured out yet: if the phone is disconnected, the actual device disappears from My Computer.)

Once you've figured out the general arrangment of the sw, there are a ton of little convenience features. For example, as it knows what kind of phone you're going to send to, it automagically picks the right file format. So you can click on "portrait" or "wallpaper" and it immediately picks the right pixel dimensions for your phone.

It also has the feature of previewing a sound as it will be played on the phone, which was quite helpful. (For some reason the quality of a paid music download was a lot better than one I made myself, but I'm not sure if that's a real issue.)

On the other hand, there were some worrying remarks in the manual about issues that did not concern me too much just yet. For instance, they recommend using a special data cable (for higher transfer speed) but for some reason it needs a special driver install, and the install procedure seemed to have a million known problems.

Likewise, I was not at all happy with the setup for using the phone as a modem. I did not play with that at all as it sounded as though you could easily mess up the basic connection.

Another problem area is transferring multimedia between phones. As I said above, the compatibility issues are still there; you need to have another compatible phone to check out multimedia setup, but most people don't have that or are trying to set up two phones at once. I left that alone, even though the girl would really like that.

Anyway, overall I was quite impressed. The 3200 is an early model, too, with few features: no Bluetooth, MP3, memory cards, and very little memory (2 MB, I think). Presumably more recent models like the N-Gage will have additional convenience features. But I was really quite charmed with the simple feature of being able to easily associate a snapshot with a caller.

2004 Dec 16 [ Thu ]

MP3 players which can record from audio inputs to MP3

I have been looking for an inexpensive device with this feature for a long time. Many such devices have a "voice recording" feature, but (although they try to conceal this by not referring to it in their specs) it uses some godawful abysmoquality compression scheme, and is in any case mono and has no line-in or even mic-in.

The following link shows that "wewa" has multiple products with this feature, both with hard-disk and flash memory: www.wewa.com.hk [http://www.wewa.com.hk/products/products.jsp?mode=second&groupid=46]

They refer to it as "e-mp3". Regrettably they still don't give very full specs on this feature: eg input sensitivity or noise floor with the various bitrates (commendably they give a wide choice of bitrates up to 192 kbits).

I happened on this site not via a web search but because Anana Computer is selling a wewa model "Dolphin wmp-317e plus" for 120 USD. Oddly enough the website lists the "Dolphin wmp-317 plus" *without* mentioning the "e-mp3" feature. I assume it is a separate model because the packaging on the unit at Anana certainly seemed to be talking about it.

Here is a page discussing some of these products for import into the US: www.cirago.com [http://www.cirago.com/electronics/company/news.html] (The page is aimed at retailers and distributors.)

I see from this (I had not noticed on wewa's own page) that the Omnipotence WMP402+ model includes an SD card slot. Regrettably the feature list does not include supporting folders (when I'm listening to my music tracks I don't want to occasionally hit Cambodian language study tracks, and vice versa).

2004 Dec 14 [ Tue ]

Booting a USB key with *multiple* operating systems

Slashdot discussion: ask.slashdot.org [http://ask.slashdot.org/askslashdot/04/12/10/239239.shtml]

This posting looks good:

Combine grub and isolinux (Score:2) by JeffL (5070) on Saturday December 11, @12:15AM (#11058369) (ibgwww.colorado.edu [http://ibgwww.colorado.edu/~lessem/)] ask.slashdot.org [http://ask.slashdot.org/comments.pl?sid=132388&cid=11058369]

I boot multiple disk images from my usb key using grub as the menu, and memdisk from the syslinux package. In grub/menu.lst I have something like:

title Hitachi Drive test utility
root (hd0,0)
kernel /boot/memdisk
initrd=/boot/dft377_29.img 

Some things boot fine, like the above referenced Hitachi DFT, but other disk images don't seem to work. I've successfully booted DOS/Windows floppy images for doing BIOS upgrades, etc.

Memtest86 can be booted directly:

title Memtest 86+
root (hd0,0)
kernel /boot/memtest86+.bin  

2004 Nov 29 [ Mon ]

Commercial interface available between PC and switches

Many interesting business or technical applications of PCs really need a new operator panel to be provided, so that the most frequently used controls are logically laid out, and industrial-strength switches, rated for heavy use, harsh environments, gloved hands, milspec low-illumination "condition red" switching, etc etc can be used.

But the PC does not have keyswitch inputs. Some general-purpose input adapters are available, but they are pricey and intended for lab equipment.

The following supplier provides various adapter circuits intended for turning a PC into a 70s-style arcade game machine. Although their devices are inexpensive, gamers can actually be quite demanding about real-time performance, so they handle things like multiple simultaneous keypresses far better than anything you can kludge up by wiring to a keyboard (but there is a pass-through port so you can have an actual keyboard continuously available but hidden in a drawer or something).

www.ultimarc.com [http://www.ultimarc.com/ipac5.html]

The link above is just for one product. They also have adapters which can handle rollerball-type inputs.

It occurs to me that this kind of thing is probably handy for movie-makers, too: the special-effects guys often have to whip up a control panel. Because the director often winds up changing the response spec for the panel on the fly to match script changes, there is often a special-effects guy crouching behind the panel holding a bundle of hand-wired switches in his hands and frantically trying to remember which three he has to flip on cue... but it would be a lot neater to interface to a computer and change everything in software, at least for props that are going to see some re-use.

Less glamorously, this might be just what you need to produce a cool-looking security system, or golf-course sprinkler system, or something, with nice clean back-panel wiring that can easily plug into a generic motherboard.

The following link may also be relevant: various hardware for arcade machine vendors (also interactive/kiosk machines, vending etc etc): www.happcontrols.com [http://www.happcontrols.com/]

Both links via a Slashdot discussion.

2004 Oct 06 [ Wed ]

Booting a computer off a Compact Flash card

I still haven't actually managed to boot off flash memory using the USB interface. Several people tell me it's easy, but I have never seen a good explanation of what the requirements are. For instance, presumably you need BIOS support, but one person tells me the BIOS did not *display* support until he plugged in a bootable memory module.

I recently saw a couple of USB memory modules whose packaging specifically states that they're bootable, and I may splurge on one.

Anyhow, I'm also interested in using USB memory to boot an older system which has perhaps no USB support, or no USB booting support. So I was interested to see a mention of IDE to Compact Flash adapters.

The Compact Flash memory modules (as opposed to other flash memory standards like Memory Stick and SD) all support multiple access standards. In particular, they are pin-compatible with the IDE standard. There's a good technical description here, as part of a page on connecting compact flash to a DSP processor: www.analog.com [http://www.analog.com/library/analogDialogue/archives/37-01/compact_flash.html]

Note that the above page, like many others I found on this topic, was written several years ago. Apparently most people consider it a done deal – it just wasn't for me (and hopefully you).

Since the CF modules are pin-compatible, a very simple adapter board can be offered. There are versions for a regular (desktop size) 3.5 drive connector (and power connector): www.acscontrol.com [http://www.acscontrol.com/Index_ACS.asp?Page=/Pages/Products/CompactFlash/IDE_To_CF_Adapter.htm]

and aloso for a laptop style "SFF" 44-pin connector, which includes the power connections: www.acscontrol.com [http://www.acscontrol.com/Index_ACS.asp?Page=/Pages/Products/CompactFlash/SFF_IDE_To_CF_Adapter.htm]

There is a note on these pages that the adapters do not support 3.3-V-only CF cards. I don't know if that's a big problem. I vageuly remember seeing that some CF vendors do specify that point, but I don't remember seeing it quoted on in-store packaging.

Here's a page which worries about another issue. The IDE spec for CF was issued when 4 GB was a lot, so the original CF spec is limited to 4 GB. CF version 2.0 apparently supports up to 137 GB. However, it occurs to me that a lot of the older systems you might want to dedicate to use with a CF adapter probably have older BIOSes which are limited to say 32 GB: www.fact-index.com [http://www.fact-index.com/c/co/compact_flash.html]

Here's a page which talks about some nitty-gritty details of using these adapters with MSDOS and Linux: www.tapr.org [http://www.tapr.org/tapr/html/Fcfa.html]

Their adapter price seems rather steep. Here's one for just 18 USD: www.logicsupply.com [http://www.logicsupply.com/product_info.php/products_id/100]

Well, it doesn't include LED indicators for power and activity.

Another oldish writeup (2000) which warns about the write lifetime (100,000 writes) being too low to support some operating systems: www.dansdata.com [http://www.dansdata.com/cfide.htm]

Both Windows and Linux default to using an on-"disk" swapfile which routinely swallows aged-out RAM contents. I noticed that Linux seems to work OK if I don't allocate a swap partition at all, although I haven't tried that recently (or on a CF system).

Here's a unit which has both a 40-pin and 44-pin interface: www.abiatech.com [http://www.abiatech.com/fb4617.htm]

2004 Sep 27 [ Mon ]

Slashdot discussion on resources for homebrew hardware hackers

slashdot.org [http://slashdot.org/articles/04/09/27/049228.shtml]

Various good links. For instance, I'd forgotten about Mouser: http://www.mouser.com

Boy, I really used to look forward to getting that catalog. (What a nerd.)

2004 Sep 18 [ Sat ]

Small, low-power servers -- hardware ideas

A Slashdot discussion had a lot of ideas on what hardware to use to set up a small, low-power server in the home: ask.slashdot.org [http://ask.slashdot.org/askslashdot/04/09/17/2221248.shtml]

A lot of respondents suggested using an old laptop, possibly with a dead battery or display.

Some links:

Soekris mini PCs: www.soekris.com [http://www.soekris.com/]

I took a look and they have several other interesting products: in particular a 4-port Ethernet card for PCI, and a low-profile Ethernet card with *two ports*. BSD, Linux and Windows drivers. www.soekris.com [http://www.soekris.com/lan16x1.htm]

Soekris also has a tenny-weeny 486-CPU board with a PCI slot that you can use with those Ethernet cards. It also has a serial port, compact flash, PXE and support for headless operation (using serial port). Price: 159 USD, board only qty 1: www.soekris.com [http://www.soekris.com/net4501.htm]

Pebble Linux (distro for WAN servers on eg Soekris): www.nycwireless.net [http://www.nycwireless.net/pebble/]

Soekris tutorial: www.davidcourtney.org [http://www.davidcourtney.org/soekris.html]

Mini-box ITX motherboard-compatilbe boxes; www.mini-box.com [http://www.mini-box.com/products.htm]

Mini-itx motherboards: www.mini-itx.com [http://www.mini-itx.com/]

Openbrick.org (small diskless PCs): openbrick.org [http://openbrick.org/]

Good short article on how much power regular PCs and monitors actually take (short answer: 80W during defragging; 60 W in use): www.techreviewer.com [http://www.techreviewer.com/viewpage.cfm/pi/29]

2004 Sep 14 [ Tue ]

Why is there so much noise on PC sound inputs?

I have always been disappointed with the S/N level of the microphone inputs on PCs. On the other hand, I could never understand how the audio level was not completely corrupted by all the RF interference from the digital circuitry inside the box. I used to look at the S/N specs on audio boards and think "on what input?". (The line-input port was usually OK with 60-70 dB, but the mic port was often around 20 dB S/N.)

I was unsurprised when my laptop had rotten S/N on the mic input, and after stewing about it for a long time I realized that if I disconnected the power supply the S/N improved a lot – to about 50 dB. Still, it's very inconvenient to be able to run your machine for only an hour or so at a time, so eventually I ponied up for a USB audio interface (I forget the name: it was intended for a Mac and has good reviews.) Much to my surprise however the audio S/N was *still* lousy when the power was connected.

Incidentally it doesn't make much sense to me that disconnecting the power should help – that's why it took me a long time to try it. When the unit is operating off the battery, it still has to run a switching regulator to provide the multiple supply voltages needed: presumably 5 V, 12 V, 3.3 V, 2.2 V... And of course the digital circuitry is still runing also. Can there really be so much noise coming from the adapter and power line? I suppose so, but it's amazing that it overwhelms the internal noise. (Hmm... it just occurred to me that when it's running off the battery, it's not *charging* the battery. Maybe that's the culprit.)

Another irritating thing to me is that the software – as advertised, the gizmo worked using standard Windows drivers – does not allow the external audio gizmo to be used in parallel with the built-in sound hardware. If you have two audio cards in a regular PC, you certainly can, even with Windows 98.

Here's a Slashdot poster who points to a basic audio issue that was new to me:

Re:Who would buy intel? Who would use onboard... (Score:4, Interesting) by afidel (530433) on Monday September 13, @07:50PM (#10241595) slashdot.org [http://slashdot.org/comments.pl?sid=121714&cid=10241595]

Uh, this is a VERY well known problem with Via 686B Southbridges (among others) and Creative labs soundcards. Neither company knows how to follow the PCI 2.0 or 2.1 spec and so burst data transfers done by the sound card are corrupted. Some Firewire cards also have problems transfering to the iPod for the same reason. The problems were enough to put me off VIA permenantly. I now use SiS chipsets for my AMD systems and have had no problems (I don't need any of the expensive integrated stuff from a NForce board).

I certainly hear ticks and clicks on PC sound from MP3's. I thought it was just my own recordings, and then I thought it was everybody else's bad recordings, but now I wonder.

2004 Aug 30 [ Mon ]

Computers for travellers -- handtops or what?

While reading today's story on Slashdot about handtops – slashdot.org [http://slashdot.org/articles/04/08/29/148215.shtml] – devices almost as small as PDAs but containing a hard disk, standard operating system and perhaps a keyboard – I also thought about other recent ideas: the bootable Knoppix disk; and the "Globe Trotter", a 40-GB hard drive preloaded with Mandrake Linux that plugs into any PC with a USB port: linux.slashdot.org [http://linux.slashdot.org/article.pl?sid=04/08/28/2137245&tid=147&tid=198&tid=1&tid=218]

It occurs to me that while all the parts of a PC which normally fit into the case are naturally getting smaller and smaller – that is, cpu, RAM, hard disk and ports – the human interface stuff is naturally *not* getting smaller: if you make the keyboard or display much smaller it just does not work as well.

With that in mind, my own proposal is for something which does not work without cooperation, but would be great in an ideal world: places like airplanes, airports, internet cafes, hotels and so forth should offer, in addition to whatever they have now, *bare peripherals*: the display, keyboard, mouse and perhaps audio and webcam, plus an Ethernet connection, with some sort of standardized easy plug. Clearly it would be very easy to make the travelling PC: all you need is cooperation and standardization. It means travellers could have full security (assuming keyloggers weren't installed inside the keyboards) along with full ease of use. Oh, and a DC power port would be neat too.

I'd like to see a "standardized easy plug" for another reason: right now you can easily make a complete PC small enough to fit ten of them in a standard PC case, but there's no hardware to support that. Initially, all you really need to standardize is the plug; people can hack up their own trays to fit 5-1/4 drive bays. On the other side of that plug, you'd need some sort of distribution system and a KVM, I guess, but I'm sure that would appear once the basic system interface connector gets standardized.

2004 Aug 26 [ Thu ]

Why can't you set access privileges on your phone?

When I first got a digital phone I was disappointed to find that you couldn't easily disable (route to answering machine) calls from certain people. I had assumed that people who get a lot of calls (girls) would need it. Perhaps, since it is a feature that supports the callee not the caller, and in Thailand and Cambodia the caller is not the payer, the cellphone service companies don't want it.

As it happens I have encountered several girls who pestered me several times a day for weeks. (That's not boasting: it's just that here, I am the prey.)

Anyway, it seems to me that a cellphone is exactly analogous to a home directory in Windows. You should be able to set group and user rights; it would be cool to be able to set your "company I work for" group to access only 9-5 M-F, but give Bob, your drinking buddy who introduced you to the stripper, unrestricted rights, except you have had to disable him temporarily a couple of times when he did crystal at the weekend and he wanted to call you up and talk about how important all his friends are to him for hours and hours.

Of course, you would wind up with similar problems, like "Joe Jobs". Your ex-girlfriend would call you up and rant from some number, and at the end of the call you would add the number to your "blocked" list. Too late you would find out she had asked your boss to borrow his phone...

2004 Aug 18 [ Wed ]

Keyloggers -- actual product page

I have talked about keyloggers before, but had not tried searching for them as a product on the web.

A Slashdot posting today provided a nice informative webpage: www.keyghost.com [http://www.keyghost.com/]

Rememebr, I'm not suggesting you buy one. I'm suggesting you consider carefully how many people might want to record your keystrokes – and how little technical expertise, or money, they need to do it.

2004 Aug 07 [ Sat ]

Tiny Linux hardware

I've been thinking about doing a hardware project recently (even though in Phnom Penh I can't even find a *project box*) and this range of Linux boards looks interesting: the boards are just 20 x 80 x 6.3. The current consumption is high at 250 mA, but there's a "quiet mode" at 50 mA. Apparently they all have connector for SD non-volatile memory up to 512 MB.

gumstix.com [http://gumstix.com/features.html]

2004 Aug 03 [ Tue ]

New Sony Minidisc range offers audio *uploading* finally

An extremely irritating "feature" of the Sony (and other manufacturers') Minidisc products is that they are artificially crippled to prevent uploading audio tracks digitally back to the PC, even if you have recorded them yourself – of course, this introduces generation loss as the signal has to be re-digitized, as well as limiting the speed to the original recording time. Previous article: www.panix.com [http://www.panix.com/~dannyw/weblog/Computers/Hardware/recording01.html]

A couple of days ago I happened to check minidisc.org, which is announcing that the new Sony range of 1-GB Minidisc products *can* upload audio files: www.minidisc.org [http://www.minidisc.org/hi-md_faq.html]

I downloaded a PDF copy of the manual for the MZ-N900 (the top of the range) from Amazon (which still lists this model as not yet released, and lists another model only via the J&R site) and tried to find out what Sony actually promises on this feature... Aarghh! Sony explicitly set the document so you can't copy any of it to the clipboard! Wtf?? I'll give you the link to the pdf, but it looks like it may well fail: www.amazon.com [http://www.amazon.com/exec/obidos/redirect-to-external-url/ref=mt_mo_view_33014/002-8923125-7500865?path=http://media-server.amazon.com/media/mole/MANUAL000053235.pdf&append-uid=no]

Incidentally, I observe that Amazon says on a single page "Date first available at Amazon: February 25, 2004" and "item has not yet been released. You may order it now and we will ship it to you when it arrives". You should think about what that term "first available" seems to actually mean, when you see it in future.

Oh well, I'll have to keyboard it.

Importing tracks recorded on a Hi-MD Walkman to your computer

Tracks that have been recorded on a Hi-MD Walkman can be imported to My Library of SonicStage on your computer only once. Only tracks recorded in Hi-MD mode. [Note 1]

...

You cannot import tracks that were recorded in a Net MD mode to your computer, or tracks which have been recorded on a device which does not support Hi-MD mode. [Note 2]

...

When you transfer tracks recorded in linear PCM on the Hi-MD Walkman to a computer for editing (dividing or combining) in My Library, the editing may take a long time, depending on the lengths of the tracks and capabilities of your computer. This is due to limitations in the MiniDisc system. If you want to edit long tracks recorded in linear PCM, we recommend that you do the editing on the Hi-MD Walkman first, and then transfer the edited data to a computer afterwards. [Note 3]

Note 1: I don't understand what they're getting at with "can be imported... only once". Does that mean they can't be subsequently re-written back to another MD? How would they know? Also, the language is worryingly vague. I can then *output it* from their stinking heap-of-spit SonicStage application to .wav or something, right?

Note 2: I think minidisc.org did refer to this restriction, but it's strange. Maybe Sony feels that they had to support any *intentional* copy protection on their existing MD media. But how many people would have recorded audio to their *own* Minidiscs *intending* it to be copy-protected *against themselves*? I think Sony's PHB's wrote the spec again.

Note 3: Another new feature of the H-MD format is it supports linear, uncompressed recording, which is very nice. But it sounds like their development people ran into some horrible shortcoming in the Net-MD interface when you try to use the resulting very large files. It would help if they were less coy about what "long time" means in some cases. Does it mean that a 30-min track in linear mode takes 20 hours to upload? Or what?

Also: minidisc.org refers to the fact that an H-MD Minidisc unit can be used as a regular USB storage device, but (in "Storing non-audio data to a disc") the manual makes it clear that this feature cannot be used for audio files, even though the audio data on the disc is visible as the "file management" folders.

2004 Jul 27 [ Tue ]

Multiport network interface cards (NICs)

I have often wondered if people make multiport NICs, because they would make it easier to put together a router in a small form factor case. I've seen two-port NICs, but if I remember rightly they were so expensive that they made it silly to consider a router application.

Today I did another Google search, and found the Matrox "NS-QNIC/4", with four independent 10/100 duplex ports , and Linux support, at least for Red Hat and SuSE:

www.matrox.com [http://www.matrox.com/netweb/products/nsfnic4.htm]

It's available for just 70 USD: www.weirdstuff.com [http://www.weirdstuff.com/commerce/catalog/product.jsp?product_id=4172]

(Actually I'm a little worried that they may be discontinuing it. Apparently it was released in 2000, and I get the impression it's not a booming market segment.)

Test results: www.linuxtested.com [http://www.linuxtested.com/results_lan.html]

There was a link about it at whitepapers.itsj.com, but don't bother following it: after forcing you to click on a succession of pages, it leads you to a lengthy signup page. itsj.com, eat spit and die.

As well as a regular router, another application I was thinking of was a sort of firewall which would protect individual users on a small LAN from *each other* as well as the internet. I was thinking of using a six-PCI slot board to provide 24 ports, each of which would have firewall rules allowing it to connect to the Internet but not to each other. The reason, of course, is to limit vulnerability to worms which scan all the hosts on the local network for vulnerabilities.

2004 May 22 [ Sat ]

Microcontroller kits available

Slashdot discussion: ask.slashdot.org [http://ask.slashdot.org/askslashdot/04/05/21/0050256.shtml]

Digikey provides a wide range of small components like resistors, connectors and switches, as well as development kits and systems. Digikey accepts international orders and credit cards, but it may well be that credit cards are not accepted from SE Asia – I haven't tried it. (I should perhaps make the point that finding *any* parts in Phnom Penh is hard. Right now I can't find a shop that sells *resistors*.) dkc1.digikey.com [http://dkc1.digikey.com/US/MKT/Terms.html#Intl]

A microcontroller development kit using an ARM processor that connects to your PC using the USB port (and gets its power from there too): www.mpeltd.demon.co.uk [http://www.mpeltd.demon.co.uk/usbstamp.htm]

The AVR microcontroller: www.avrfreaks.com [http://www.avrfreaks.com/]

If you are located in the USA, you should be aware of Circuit Cellar: www.circuitcellar.com [http://www.circuitcellar.com/electronic-edition/] That URL is a link to their *paid-subscription* electronic version, which may work even outside the USA.

2004 May 11 [ Tue ]

Hardware mods to Nintendo Gameboy Advance

Apparently Nokia cellphone sales have been slumping lately because they don't have any clamshell-style models. Well, I have to say I like the clamshell format for *any* pocketable gizmo. My first PDA had that format and it was fine. It was totally reliable and I could keep it in my pocket without an extra case. Both my many Palm Pilots and my cellphones have needed an extra case, and still weren't reliable.

Gloomily looking in store windows, I have become aware that the Gameboy Advance SP is a great clamshell design. It doesn't have a keyboard, or a lot of memory, or a memory slot for flash, etc, but it looks good. And they're selling to little kids, so presumably the switches and connectors are pretty rugged.

So I've started wondering about adapting a Gamebody somehow: if not as a straight PDA, maybe for some special functions, like inventory, or some sort of short-range scanning (tricorder style).

Also, there's now a dual-screen model (called the DS) which has a lot of extra features: www.usatoday.com [http://www.usatoday.com/life/lifestyle/2004-05-11-nintendo-ds_x.htm] (warning: usatoday.com has popups when you enter or leave a site. You can use Proxomitron to disable all popups whatever browser you may be using.)

I haven't found a website which addresses this sort of modification/product explicitly, but here are some add-on gadget sites that give you an idea of what's possible... Oops! Here's a site explicitly for hw/sw developers: www.devrs.com [http://www.devrs.com/gba/]

And another site by an individual who offers plenty of GBA hacking info: darkfader.net [http://darkfader.net/gba/main.html]

Interesting gadgets for Advance: www.flash2advance.com [http://www.flash2advance.com/]

That site seems to be mainly about downloading games into your PC and then uiploading into a big Compact Flash card which is hardly what I need to do, but it's interesting to see what's possible just by plugging in a card.

The Flash2 Advance card and Flash2 Advance Linker cable seem to provide similar functionality on this site: www.lan-kwei.com [http://www.lan-kwei.com/gbaE/]

GBA BIOS info: www.work.de [http://www.work.de/nocash/gbabios.htm]

Another Flash Advance story, with review: www.cdrinfo.com [http://www.cdrinfo.com/Sections/Articles/Specific.asp?ArticleHeadline=GBA+FlashLink]

How to catch someone who steals your PC/laptop

Many thieves will imnmediately sell the laptop to someone who will immediately wipe the drive. But if the thief doesn't – perhaps he's a college student who just coveted your nice new laptop – maybe you can track him down by installing some sort of bugging system in the machine.

I tinkered with a pay service for this but it was very expensive and you had to install their (not very trustworthy) software to make it work.

If you have your own website and can run perl there are a lot of options. See this Slashdot discussion: ask.slashdot.org [http://ask.slashdot.org/article.pl?sid=04/05/07/2348245]

Remember that the police often don't respond to a violent crime in progress, much less some nerd griping about his laptop. Your chances of redress are probably better in a college/academy type of environment.

2004 Apr 15 [ Thu ]

The Nokia Ngage gets phased out by a new model

The Ngage didn't exactly set the world on fire when it was released. Now Nokia have a new model, the QD, which in many ways has worse features (eg no tri-band GSM, but at least you can swap the memory without having to remove the battery!).

The Nokia Ngage website: www.n-gage.com [http://www.n-gage.com/en-R1/gamedeck/ngage/]

I'm listing it here because a recent article on Slashdot listed some cool features and made me wonder if the Ngage will now be available at a low price. When it came out, I was so put off by the abysmally terrible styling and ergonomics that I never even looked at the specs.

But it has Bluetooth, an SD slot, compatibility with the Palm address book data, tri-band GSM (ie including US), downloadable Java software for eg an ICQ client, voice recorder with stereo input jack (although the format is regrettably AAC), handsfree speaker, WAP, GPRS, HSCSD, email protocols, MMS/multimedia messaging, MP3 player, FM radio...

Maybe I'll get one just to dink with all the features.

I probably wouldn't actually use it because I'd look so goofy holding the *edge*of a phone to my ear...

2004 Apr 13 [ Tue ]

Using datacenter-style racks

There's a good Slashdot discussion on how to set up multiple systems in a rack like datacenters use.

I would add the following:

1. I second the references to power consumption. A full-size rack can easily take enough equipment to overload the wall socket. It's worth checking the manual when you install each item and putting a sticker on its front panel that says how much power it takes.

2. Plan so that the *rear* of the rack is accessible without tremendous pain. (At one place, it was physically possible to get by, but the cables projected stiffly into the space, giving you the scary feeling that you were going to knock out the datacenter every time you tried to get by.)

3. I have never understood how designers decide whether to put certain connectors on the front or rear of equipment. I suppose power connectors mostly go on the rear and displays on the front, but not always. In most small (single-rack) systems, you have the patch panel (of course) facing front, but the Ethernet sockets for the equipment are at the rear. I've never seen a very impressive way of handling the transition.

4. Locate suppliers for miscellaneous hardware like cable trunks before you plan the layout. You may well find that a cable raceway with the features you want is just not available for the 2U slot you planned for it.

5. It's extremely handy to have a lifting table to support equipment as you install it. In general, equipment tends to be pretty heavy when you're just moving it around, but when you're trying to screw it into a rack it's torture. If your hands slip, you may well ding the rack or the equipment ears, even if you don't actually drop it. You will probably need a strong friend every time you move something. Remember also that moving stuff requires extra space at the front and rear (eg a 1U server in the middle of a rack will have to come out all the way absolutely horizontally).

6. Power supplies are *extremely* heavy and may well overload the fixing points.

7. Beware that the rack may entirely *tip over* while you are working on it. It's advisable to select a location where bolts can be drilled into the wall and a concrete floor. Use a metal detector first to check for power lines and rebar. If the rack tips over, it may well have sufficient momentum to punch straight through sheetrock. Even a wobble may well overstress solid-core network cables.

8. There are puzzling slight differences in rack mounting hardware. I get the impression the screw holes for computer equipment are usually a little different from telephone racks and a little different from music racks. Check this out carefully. In particular, phone racks tend to have built-in screw holes which are rather wimpy and easily bent, whereas computer cases have large square holes which take a more confidence-inducing snap-in screwhole (which reminds you where to put the equipment *back* after you've taken it out, too).

9. When all's said and done, racks are damned expensive if you don't *have* to use them. That free rack you picked off the dumpster may cost you $500 in miscellaneous hardware, and hours and hours of time searching for it.

2004 Apr 02 [ Fri ]

More info about the eMobile SIM-usb adapter

I have done an article about the eMobile before: www.panix.com [http://www.panix.com/~dannyw/weblog/Computers/Hardware/sim-usb03.html]

I tried installing SIMcard.exe on another PC today and it worked OK except for a few things which mainly mean the machine I used is old and slow.

1. The initial install screen shows a lot of question marks in the filenames which I assume are because Unicode 16-bit characters were used. (The programmer probably thought he was using English characters because they displayed that way on his screen, but the install sw doesn't know how to interpret them on a non-Chinese Windows.)

The adapter shows a blinking yellow led during install.

2. After the install, shutdown and subsequent reboot (presumably one time only) are extremely slow, presumably because of necessary updates to the Registry (a name which will forever live in infamy).

3. Even after the reboot, when I inserted the actual adapter and Windows said "found new hardware", Windows was achingly slow installing the driver: perhaps five minutes.

4. On the other hand the application software worked immediately the driver finished installing. The adapter shows a steady green light in operation.

5. The adapter does not show up as a device on the desktop, unlike a USB memory drive. It seems to do no harm to just pull it out, although I have not tried doing that while using the application software to read or write to the SIM.

2004 Mar 31 [ Wed ]

Some progress with SIM-usb adapters

I got fed up trying to make my eMobile SIM adapter work and bought a "SIM Master 4" with "SIM Magic Pro V7.0a" software.

Webpage showing the SIM Master 4 product: www.expansys-usa.com [http://www.expansys-usa.com/product.asp?code=INSIMM307&asource=ORG_6310I]

This installed and ran without much trouble except for a few things:

1. The backup app, dated 2003-09-22, doesn't seem to want to work via USB: it only has options for COM1 or COM3. A website suggested COM3, which is the default, which didn't work. Neither did COM1. There is a huge variety of bitrates available; I tried the slowest and fastest, which also did no good.

2. The "SIM Magic Editor v7.0", also dated 2003-09-22, more or less worked but failed to read several SMS message contents which I was able to read with the eMobile (see below)

3. The editor "sm.exe" crashes when I try to load a .mid file

Oh well... it established that the SIM was OK, at least.

I wondered if a later version of the sw was available for download, and was excited to find this link: www.inex-tw.net [http://www.inex-tw.net/download.htm]

However it turns out to be an *earlier* version, v4.3!

I then tried installing the eMobile software again on a *fresh* computer. Their website www.kinforce.com had sent me a link to a different software package to try, actually smaller: SIMCard.exe. This time it worked, although again the install process did not seem to go as the documentation suggested. In particular, after I rebooted and plugged in the adapter, there was a lengthy period in which Windows was working on installing the driver (which it managed to do by itself) where *no prompt was visible warning the process was still running; the disk activity light eventually went out after two or three minutes.

The application then worked: actually better than the "SIM Magic Editor", because it read the sms messages cleanly, including the date/time.

It would be nice to have a backup/restore app like the SIM Master.

I would like to give exact install instructions, but I'm confused by having problems on one machine and no problems on another. Also, my contact at www.kinforce.com told me the download link was just for me, so I can't post it here. If you are reading this because your eMobile doesn't work, all I can suggest is you look for the email contact address on their website as I did, and ask for SIMCard.exe.

2004 Mar 19 [ Fri ]

Download for SIM-USB adapter does not work

I eventually got around to downloading the English-language software for the "e-Mobile" SIM to USB adapter that I mentioned in a previous posting: www.panix.com [http://www.panix.com/~dannyw/weblog/Computers/Hardware/sim-usb01.html]

It's 14 MB. It was in .rar format which WinZip does not handle, so I downloaded WinRar (30-day shareware, although it includes unrar.exe, a command-line util which is apparently without a time restriction), but WinRar reported that most of the files had broken headers.

My guess is that the twerps who created it used a Chinese-language version of Windows which produced some sort of Unicode text strings which Rar running under a non-Chinese Windows can't handle – but that's just a guess. Fume, fume. I have to admit that the site I downloaded the program from probably have no particular relationship to the store in Cambodia which sold it to me; the store said "oh, sure the software will run in English – if it doesn't, just bring it back today" – and of course I didn't have time to check it that day before the store closed. Aargh. Oh well. The store doesn't know it but I was planning to spend a bundle of money there: not now.

2004 Mar 11 [ Thu ]

The e-Mobile SIM card reader with USB interface

This is available from eg www.kinforce.com. Their English download page has "e-mobile v. 1.0" (15M), as well as v 2.0, which is 30 M and says something about ".net framework1.1" which made me suspicious.

Here's the link to v 1.0: www.kinforce.com [http://www.kinforce.com/down/kinforce/GSM&CDMASIMCardExplorerEnglish.rar]

Hmmm... I wonder if Winzip does .rar. Oh well.

2004 Feb 28 [ Sat ]

Field Programmable Gate Arrays for funand no profit

No time to check out this link tonight...

www.fpga4fun.com [http://www.fpga4fun.com/]

If you ever picked up a soldering iron or a wirewrapgun, FPGAs should be on your list to check out.

2004 Feb 10 [ Tue ]

How to boot your PC off a USB disk

I haven't tried this stuff yet, but here are some links.

Slashdot discussion: ask.slashdot.org [http://ask.slashdot.org/article.pl?sid=04/02/06/2252202]

which linked to a long posting in a forum that looks promising (booting Linux and W98): forums.gentoo.org [http://forums.gentoo.org/viewtopic.php?t=78454]

I also found the homepage for "smart boot manager", an entirely free piece of software from Sourceforge. It doesn't sem to have USB features particularly, but it looks like if you can make a USB key boot at all, you can put "smart boot manager" on it and allow it to boot to multiple operating systems. btmgr.sourceforge.net [http://btmgr.sourceforge.net/]

I still haven't found any description of how to check whether your PC, BIOS or USB key actually supports booting. There were several postings that said they'd tried umpteen things to make one combination work but another combination worked first time.

2004 Feb 04 [ Wed ]

Who really made your laptop?

Almost all laptops are really made by a few Taiwanese manufacturers. The Atec brand in Thailand, for instance, is really reselling various Twinhead models.

I tried searching on these manufacturer names:

Twinhead

Quanta Computer

Compal Electronics

and found many interesting pages. This is a good overview: www.absolutelaptops.com [http://www.absolutelaptops.com/madebywho.asp]

and this at first drab-looking page links to various pages which allow you to search for your manufacturer by FCC ID or BIOS id: www.quest.co.nz [http://www.quest.co.nz/services/who.htm

] 2004 Feb 03 [ Tue ]

The three common PC case screws described

Here's a link to an OK guide to the common screws:

www.bunkermentality.net [http://www.bunkermentality.net/sg03.html]

The link I give is the last (answers) page of a three-page quiz; I thought I'd spare you the "quiz" aspect.

There was a Slashdot discussion about it that was mildly interesting:

slashdot.org [http://slashdot.org/articles/04/02/03/012247.shtml]

Personally, I don't think the guy gave a very comprehensive view. For instance, he doesn't refer to the widely-available plastic case screws intended to allow toolfree access.

Likewise, he doesn't give a good source for obtaining these screws. I've often seen miniscule packs on sale in real computer shops for several dollars.

Additionally, I have run into quite a few cases (heh) where slightly – frustratingly – different screws were used for the case and the card slots. In fact, I vaguely think that was the original standard.

The issue of clearance might have been addressed more. I have run into a lot of instances where I had screws which had the right thread but the head obstructed something. A lot of drive screws these days have an extremely flat profile for this reason, which was not shown in the images used for the quiz.

His general points about how the ludicrous proliferation of similar but non-interchangeable screws has blighted the lives of PC techs for 20 years were good.

2004 Jan 24 [ Sat ]

High-fidelity portable digital recording

A lot of PDAs and MP3 players have some recording features, but typically the audio *recording* (as opposed to MP3 playback) is poor: because the CPU has limited performance, the compression algorithm is crude and only low-fidelity recording is possible. (The brochures of course try to conceal this by eg stating frequency response and s/n specs which only are true for MP3 playback, and gloss over the fact that the *recording* is in another inferior format.)

A few devices do provide direct-to-MP3 recording, eg from iRiver: www.iriveramerica.com [http://www.iriveramerica.com/]

However, they still do not provide features you need for routine field recording: sturdy microphone inputs, switches etc.

"Core-sound" offers a range of audio gear and now a CF card which provides interfacing to many PDAs, some of which can run Linux. Their web page has very good explanations of many different combinations of gear: www.core-sound.com [http://www.core-sound.com/HighResRecorderNews.html]

The Core-sound products are definitely oriented towards high-fidelity portable recording. An inexpensive system can be assembled for around 1100 USD, including the PDA itself and various adapters.

Here's another site for Linux products which talks about Linux software for the "Core-sound" products: www.handhelds.org [http://www.handhelds.org/z/wiki/PDAudio]

This Linux site lists umpteen different Linux packages supporting many different portable music players: tuxmobil.org [http://tuxmobil.org/portable_players.html]

Many sites recommend Minidisc recorders for field recording. These are quite usable and the ATRAC digital recording format (especially in its more recent incarnations) provides quite good fidelity, but the relatively inexpensive pocket recorders provide no way to output the files in digital format: you can only transfer them for editing via the audio jack. This restriction is imposed arbitrarily to (it seems) discourage music copying, and has dissuaded me on several recent occasions from buying Sony products, even non-Minidisc.

If you do decide to go with Minidisc, be very careful to check that the model you are buying has an external mic jack. Few do. Likewise, do not be fooled by the presence of a USB port that this model can upload audio files: the USB port is crippled to allow download only. You will probably end up wanting to transfer audio digitally; you can do that with a rack-mount deck such as shown in the link below, but even when digital output is present it is an *optical* digital output: you still have to *play* the blasted audio at 1:1 speed to transfer it; as far as I can see there is still no deck which allows you to rapidly copy the *file*. www.aliceshop.com [http://www.aliceshop.com/acatalog/AliceShop_MiniDisc_42.html]

2004 Jan 22 [ Thu ]

Mixing Suns and PC hardware

I remember pushing for KVMs on all the PCs in the server room, and when it was all done nobody thanked me – they just thought all server rooms were nice and cool with plenty of desk space and a new, large, high-resolution monitor for you to use. Oh well.

It always niggled me that the Suns weren't hooked into the same system. Suns of course use, or always used to use, special hardware that didn't play well with PCs. (In particular there's a reset key on the keyboard which is (a) essential for some machines and (b) a nightmare if you drop a book on the keyboard.) Of course you could get special high-priced KVMs for Sun, but our Sun guru sneered that real men should just telnet into the box. Well "the heck with you! the heck with all of you!" (Gary Larson)

Or to put it another way, I was the poor sod who was tasked with installing the stuff, ie I always needed to make it work during the 1% of its life *before* you could rely on telnet. Or when telnet had just *stopped* working.

Anyway, I finally found a cheap adapter – I think it was a Belkin thing. It was surprisingly difficult to find the price for it – it was ton of those "if you have to ask the price you can't afford it" experiences.

Now it seem less expensive, perhaps because Sun is going over to full PC compatibility. The following Slashdot posting leads to an Ebay auction where such an adapter is going for less than $20. (I don't bother providing that link because it's probably very temporary.)

Not particularly well informed in some areas (Score:5, Informative) by elmegil (12001) on Wednesday January 21, @11:48AM (#8043957) (slashdot.org [http://slashdot.org/] | Last Journal: Friday January 16, @02:20PM)

> FTA: Unfortunately, you're stuck with Sun keyboards and mice, as the connector is Sun-specific, as well as certain specialty keys. There may be adapters, but I don't know how well they'll work with the specialized keys.

Not true. I've been using a Logitech TrackMan Pro for several years now, with the aid of a nifty box that converts PS/2 devices (has an input for a keyboard and for a mouse) into the Sun connector. It was a Sun part number, somewhat obscure, but definitely available and useful. It's called the "Sun Interface Converter" and the Sun part number is 595-3692 [sun.com]. I'd recommend you go looking for one [ebay.com] if you are having trouble coming up with Sun Keyboards & Mice or if you want to use your Sun system with a standard KVM switch (which is what I do at home).

The link to sun.com: sunsolve.sun.com [http://sunsolve.sun.com/handbook_pub/Devices/Input_Device/INPUT_Interface_Converter.html]

2003 Oct 23 [ Thu ]

X10 declares bankruptcy -- what to do?

X10 Inc, the source of all those annoying emails and web popup ads, is now in Chapter 11. But they do not own the X11 protocol! There are at least the following sites which provide products compatible with X11:

www.homedirector.com [http://www.homedirector.com/]

Home Director specializes in cabling and interface box hardware which allows you or a contractor to provide a neat, attractive installation which combines ethernet, video, audio and phone systems.

www.smarthome.com [http://www.smarthome.com/]

All kinds of plug-together gadgets, eg a carbon monoxide sensor that plugs into your alarm system; rain sensors.