This is essential if you are looking to delegate control of users within an OU to a different group of select logins to allow others to perform basics tasks like changing passwords without needing the domain administrator account.

The Snap-In required is "Active Directory Users and Computers". This would be set to "Enabled"  along with the main policy to restrict all other snap-ins allowing the specified users access to the Active Directory structure and the group/users they are controlling.

Attempting to login as a user delegated to perform password resets and attempting to launch the required snap-in as a restricted user may result in the following error:

The snap-in below, referenced in this document has been restricted by policy. Contact your administrator for details. Folder.

Clicking past this error message will result in the Snap-In loading and functioning correctly, but its disconcerting to the user and annoying.

It appears to reference a Snap-In "Folder" which has been restricted by the Group Policy set as mentioned above (remember all Snap-In's except for Active Directory Users and Computers are restricted). However, there is no reference to any such Snap-In in the GPO Editor.

Creating a Custom Admin Template

We need to add an entry to the Group Policy that enables this Snap-In. This is done by added a Custom Admin Template that references the Snap-In's GUID. Assuming you have created and saved an .msc file containing only the Active Directory Users and Computers Snap-In (done via MMC), this can be viewed in a text editor to find the GUID. For further streamlining for pure password reset application, you could create a Custom Taskpad .msc but I won't go into that here.

There will likely be several GUID references in this file but the important one is

{C96401CC-0E17-11D3-885B-00C04F72C717}

You now need to create a Custom Admin Template to allow this Snap-In. Open a blank text document and copy the following:

CLASS USER

CLASS USER

CATEGORY "Windows Components"

CATEGORY "Microsoft Management Console"

CATEGORY "Custom Settings"

POLICY "Microsoft Folder Snap-In"

KEYNAME "Software\Policies\Microsoft\MMC\{C96401CC-0E17-11D3-885B-00C04F72C717}"

EXPLAIN "Permits or prohibits use of this snap-in."

VALUENAME "Restrict_Run"

VALUEON NUMERIC 0

VALUEOFF NUMERIC 1

END POLICY

END CATEGORY

END CATEGORY

END CATEGORY

Save this file with the extension .adm in C:\Windows\inf

In the GPO Editor for the policy in question, navigate through the tree to

User Configuration

Right click on

Administrative Templates

And add the .adm file you just created.

Now navigate to

User Configuration
--->Windows Components
------->Microsoft Management Console
----------->Custom Settings

There should now be an entry named "Microsoft Folder Snap-In". Set this to "Enabled"

Re-apply your policy across the network and logging in as a delegated user and launching Active Directory Users and Computers should no longer give an error message.

There were a couple of problems I needed to overcome though - it would not be a case of simply swapping units over.

The first thing I needed to do was figure out a way to connect the EEE PC to my existing M2-ATX PSU so that the ignition was still able to control the PC turning on and off. On the existing system, this is done using a simple connection to the header on the motherboard. When the PSU recieves the ignition signal, it shorts these two pins together and the system powers on.

On the EEE PC, there is no header that I could connect such a lead to, so I had to rip things open and add one myself...

How to Get at the Internals:

- Remove all screws from the underside of the unit
- On the top edge of the keyboard, there are a couple of springy clips holding it in. Use a flat screwdriver to hold these in and prise the top edge of the keyboard out. The whole thing should slide upwards now and will be attached to the main chassis by a thin ribbon cable.

- You can unclip the ribbon cable to get the keyboard out of the way (carefully!)
- With the keyboard out of the way, remove the black screws holding the top half of the chassis down. You don't need to remove the two either side of the screen. There are also a couple of plastic clips on the rear - by the screen pivots. These need prising apart (again - either a flat screwdriver or case splitter will be fine). The cover should lift off, revealing the motherboard.

- There are a couple of retention clips on the bottom side of the motherboard which can be released by sliding the motherboard up slightly. It should pop out.
- The VGA port also needs to be levered slight to release the side of the main board. When this is done, it should come loose and if you remove the mini-plug for the fan connector, the board can be removed entirely.

For the purpose of this guide, I was interested in the power button, located near the right hand side screen pivot point. The images below show the switch with what appears to be 5 soldered connectors. With a bit of testing and a flat screwdriver, I determined that the pins which needed shorting together to initiate turn on / switch off were the top two.

So I went about soldering some fly leads and threading them through the chassis, making them secure so they couldnt be pulled off easily.

Putting the EEE PC back together again and plugging it in / connecting the battery, allowed me to test that it all worked properly. Touching the two fly leads together (same as ignition ON) starts the PC up. Once booted, touching the leads together again initiates shutdown (same as ignition OFF).

Next problem is to sort out is the difference in voltage supplied by the regulated PSU (12V) and the input required by the EEE PC (9.5V). Stay tuned...

Wiring

The first job was to lay down all the wiring. This included power cable, speaker wire and extension leads for the screen.

In the above diagram, what is not shown, is that both the PC and Amp take their power from a thick power lead, fused, put through a distribution block, and run directly from the battery. Additionally, a switched positive connection is required for the M2-ATX to initiate remote startup and shutdown procedures. The easiest place to run this from, I found, was the existing headunit ISO loom.

The black, red and green connectors running to the screen in the diagram are the power supply and USB/VGA connectors. The power was spliced into a connection from the M2-ATX as the screen is designed to run from regulated 12V DC. These cables were run down the centre of the car.

Finally, as shown on the diagram, the audio connections run from the PC to the AMP via a 3.5mm audio jack adapter --> RCA/Phono connectors (blue), and then from the amp, I wired the speaker outputs to the old headunit ISO loom (purple wire pairs). All the speaker cable in the car was replaced with decent quality, shielded stuff. The picture below shows the cables being laid with the centre of the car bare as a work in progress.

Another note: When running the cables down the car, it is important to physically separate the power and data/speaker lines to prevent inteference which commonly manifests itself as alternator whine coming through the speakers. The easiest way to do this is to run the two sets of cables down either side of the centre console, or run them down either side of the car itself.

The above pic shows the cables running from the car interior, through the ski hatch and into the boot.

The Screen

With all the wiring in place, access to the wiring loom behind the dash was no longer required and so the screen was the next thing to fit. Obviously, prior to the final installation it had been tested for fitment so it was simply a case of refitting the centre console with the screen installed into the dash. The pic below shows what things look like behind the scenes on the dash:

The PC

And finally the PC was secured in the boot using metal mounting brackets beside the subwoofer and the wiring tidied up. All the USB data leads, VGA connection and the power and switched ignition leads were connected to the PC and Amplifier via the M2-ATX. Just to note that at this point, the amplifier being used is quite a basic one which is why it has not been installed 'cleanly'. When I get a proper amp, the boot installation will look a lot tidier and I will gain a bit more boot-space.

You can see the PC beside the subwoofer and it fits snugly in the gap.

The ultimate plan is to create a 'stealth' panel infront of where the PC is currenty fixed and fit the amplifier behind that as well. That should make things look a lot cleaner and give some more room (not that I use the boot a lot anyway).

The Finished Article

With everything done, here are a few pics of how it looks completed and installed:

Loading/resuming Windows XP at bootup:

The main frontend view (this is Centrafuse, but I usually use RoadRunner)

Sat Nav view

Media player currently playing Linkin Park

GPS and WiFi adapters installed out of the boot for better reception. They are hidden behind the seatbelt and secured to the plastic coving.

View from the rear seats.

Overall, the installation looks very OEM (to me at least! and boot excluded obviously lol). There is no evidence of the masses of cabling in the car and having the PC gives me a lot of functionality that I wouldn't get with a regular headunit or even an all in one headunit costing a lot more.

There is also a lot of room for potential upgrades and further features which I can implement at a later date - and I have hinted at a few plans throughout the course of these posts... so this is basically a never ending project!

I hope it has been a good read and things do make sense. If you have any comments or questions, either post away on any of the posts (parts 1- 4) or just get over to the forums for anything a bit more in-depth.

Cheers.

Now that the PC itself was up and running and the screen and mount were completed, it was time to work on the insides of the beast, so to speak: the software - the thing that makes everything work!

The Backend

The PC itself was based on an extremely cut down version of Windows XP SP2. I removed unecessary features, slipstreamed hotfixes and drivers, and pre-installed some tweaks making a totally unattended installation that automatically installed everything I needed. This created a small install footprint and made the OS very quick to boot from cold / resume from hibernation.

I did experiment with using a 4GB compact flash card and CF -> IDE adapter as a boot drive mainly for the speed advantage (and solid state advantage), but due to the limited write cycles on CF cards, installing a standard version of XP (dynamically) would severely reduce the lifetime of the card. There are solutions to this (eg, using XP Embedded from Microsoft), but that involves a lot of hassle, so I opted to go for the old fashioned mechanical hard drive.

The entire system was tested indoors using a regular ATX power supply for ease of troubleshooting any problems - especially with the multitude of USB devices being used. Bluetooth partnerships were created with headsets and phones and COM ports configured for the GPS reciever and FM radio. When the system was found to be stable and with everything fully functional in it's barebones OS state, it was time to install and configure the frontend.

The Frontend

The frontend is essentially what you see when you look at the screen. It is the software that is running and gives the user the functionality required from the PC in the car environment. There are several front ends available for use - some free to use, some not. Which frontend you choose is down to personal preference and before making a decision I tested two of the biggest ones on the system:

RoadRunner

This front end is free to use and is customisable both in terms of plug-ins and skins. It links to other external applications to handle various functions. Eg. Audio playback is controlled by winamp, GPS navigation can be controlled by Freedrive/Mapmonkey (maps are required from "Destinator" but need to be purchased) and DVD playback could be handled by PowerDVD, for example. It has a large userbase, because it is free and does the job very well.

Centrafuse

The other frontend I tested out was Centrafuse. This one is not free to use (there is a trial available), and comes in several different versions - the price of each is determined by it's feature set. Again, very easy to work with - perhaps more so than RoadRunner, and there are a lot of plug ins and a few skins available. Centrafuse handles the majority of functions independently (as far as I can tell) as it doesn't seem to call on other external programs as much as RoadRunner. One point to note is that Centrafuse does seem a little more resource hungry compared to RR.

Each frontend is very easy to set up and get working for basic functionality but things like phone control and GPS are a little more involved.

Now that system is fully operational, all that remained was to transfer some multimedia (audio and video) onto the hard drive so that there would actually be something to play! When doing this, make sure all files have correct ID3 tags otherwise when the frontend reads them, if you try searching for something by listing albums or artists, it will be a mess...

Transferring material to the PC in my case was simply a case of connecting to the wireless network and accessing network shares containing the albums etc that I wanted. This method will also be used when the PC is installed as the wireless link works from the driveway without any issues.

With everything as I wanted it, the final stage was to install the whole lot in the car. See Part 4!

Following on from Part 1, this section is going to detail what I did regarding the most important thing in the project: the screen.

The screen is the one thing that you and everyone else who uses the PC looks at all the time so it needs to be visible - but at the same not distracting to the driver. It also needs to be accessible so the PC can be operated via the touch panel and ideally needs to fit in with the rest of the interior (I prefer things that look OEM and subtle).

What I decided to do was fabricate a mount for the screen into the centre console in place of the existing headunit. Rather than butcher the original console, I picked up another from a car of the same model as my own which was being broken for spare parts.

The Process

That done, I took the screen apart, separating the case from the actual panel and electronics inside, and provisionally secured the casing to the console using small blocks of MDF and screws:

With the main construct in place, it was time to start playing with the fibreglass.

Fibreglass Application

Fibreglass kits consisting of sheets of dry glass fibre, resin and catalyst are readily available at most DIY stores and it is commonly used in automotive applications for repairing/modifying bodywork. The fibreglass was applied in two stages: Firstly, a flexible sheet of dry glass fibre was placed over the area(s) which needed moulding to the console. The desired effect was that the fibreglass would form a solid framework to build on:

To mould the fibreglass to the required shape, the resin, mixed with a small amount of catalyst "hardner" was brushed into the sheet. This was all soaked in and the shape began to emerge:

It was important to remember not to cover the rear of the original screen casing so that the screen could actually be refitted!

A word of warning: The resin and hardner have a VERY strong chemical smell which lingers even when they are drying and this is something that really needs to be done outdoors or you will actually pass out from the fumes...

Filling and Sanding

After leaving the fibreglassed console to dry overnight, I was greeted with a rock solid finish which as you can see from the above picture is quite rough looking. To smooth things out, I used a product called Isopon P38. This is an easy sanding filler, commonly used in car body repairs - which again, comes in two parts - the putty and a hardner. This was spread liberally over pretty much the entire surface of the console which had been fibreglassed in order to fill the cracks and dips where the material had set.

At this point, the finish was irrelavent as when dry, the entire console was sanded down to give a smooth surface. Starting with coarse sandpaper and working through the different grades right down to ultra fine wet and dry paper, the whole thing was sanded blending the filler and fibreglass into the original console plastic. In my case, after the initial sanding, I wasn't 100% happy with the finish and applied a bit more filler to certain areas of the console and re-sanded them.

Another word of warning: As with the fibreglass, the filler/hardner does have a strong chemical smell and when sanding, produces A LOT of extremely fine dust particles. Again, best done outside.

The Finished Article

The only thing remaining at this point for me, was to spray the console black again and fit the screen back into it's casing. I used some standard grey primer on a couple of coats, followed by a few coats of black interspersed with some more wet and dry sanding to give a smooth a finish as possible.

It was then simply a case of popping the gear stick gaiter and screen and associated electronics back in:

The end result is a customised, one-off centre console which is designed perfectly for that screen. I think it looks OEM and sits quite well in the dash. It's also easily accessible - being right infront of the gear stick.

Coming soon, Part 3: where I will talk about software this time....as always leave comments and stay tuned!

After my exams had come to an end, I decided I needed a mini project to keep myself occupied for at least a few weeks, so I decided as any normal, sane and non-geeky person would, to put a PC in my car.

I spent countless hours researching various components and determining what the best route to go down would be. Installing a computer in a harsh environment such as a car needs a bit more thought than your average 'buy-the-components-and-slap-them-together-bob's-your-uncle-fanny's-your-aunt' kinda thought process.

- There are space considerations. The entire computer needs to fit comfortably into a small area without sacrificing functionality.
- There are operating constraints. The computer should be sufficiently cooled to prevent it melting and destroying your car and perhaps most importantly, it needs to be low power so as not to drain the battery when working or when idle.

Specifications

Taking this all into account, the core specifications for the computer were as follows:

- CPU/Motherboard/RAM:

VIA EPIA M10000 'Nehemia' This board is based on the mini-itx form factor (it's really, really tiny), has an integrated 1GHz processor, graphics, sound, USB and LAN. It is extremely low power and as such runs quite cool. The integrated graphics is more than sufficient for in car purposes as are the other onboard components. Coupled with this platform, I chose to fit 512MB of RAM (mainly because that's what I had lying around at the time - yes, I have have RAM lying around, ok?)

- Storage:

I chose to use a 2.5" laptop hard drive as they tend to be a bit more resiliant towards shocks which will be ever present in an automotive environment. Alternatives include booting from a sizeable Compact Flash card and appropiate adapter but due to the limited write cycles as well as other limitations, this method, although fast, requires much more work on the software.

- Power:

To provide the computer with power I used an M2-ATX automotive PSU. Why not take a 12V DC supply straight from the battery I hear you say!? Well, besides from the obvious hassle of having to install a manual switch to turn the computer on and off, running a line direct from an in-car 12V source leaves sensitive PC components open to large voltage fluxuations. For example, the battery voltage can vary from 11-14V depending on the operating status and during engine cranking can jump up to 10's of Volts higher.

The M2-ATX, which is rated at 160W, protects against these surges and also allows remote turn on of the PC via a switched 12V ignition feed. That is, when the key is turned in the ignition, the computer begins to boot.

- External Devices:
Nearly all the periphals connected to the computer are via USB. A mains powered hub was used and modified to work with the M2-ATX. Essentially, the 5V USB standby voltage was isolated within the USB terminal so that rather than drawing power from both the USB port and external power, the hub was exclusively external powered. Accessories included: USB GPS reciever, Bluetooth, FM Radio, Wi-Fi, USB touchscreen.

Stick it all in a well ventilated, home made case (containing custom hard drive mounts made from picture frame hooks!) and you have 1x Car PC.

Part 2 coming soon. That goes into details regarding the software used both behind the scenes and as a front end.

The alteration is performed by a little black box which sits at some point between the Vehicle Speed Sensor (VSS) in the engine bay and the Speedometer on the dash. With this little box in place and altering the signal, the speedometer will now display measurements in Miles Per Hour and the odometer will also begin to clock in miles rather than kilometers.

If you want to know how the little black box works, read on. If not, then skip the next paragraph

The VSS sends a series of pulses to the Speedometer and the frequency of these pulses is used to determine how fast the vehicle is travelling. What the little black box does, is alter the frequency of the pulses to produce a resultant difference of a factor of 0.6, thus reducing the reading that the speedometer gives - i.e. it displays the speed effectively in mph.

Welcome back...

Anyway, all is well and good, unless you have a Honda VTEC engine - in which case a little more thought needs to go into the placement of the converter. VTEC gives you a boost of power at a certain RPM and activation is controlled by the ECU. Based on my Honda Prelude, there are three conditions which must be met in order for VTEC to engage:
- The car must be up to normal operating temperature
- The car must be travelling faster than 14kmh (or mph)
- Oil pressure must be high enough

The car is setup as follows:
VSS --> ECU --> Speedo

In theory, you can put the converter inbetween either the VSS and the ECU or the ECU and the Speedo. Either way will produce a reading in mph, but the position used will produce different side effects, especially on Honda VTEC engines.

Method 1:

Fit the converter between the VSS and ECU, in the engine bay.

The VSS is located just on the side of the engine block towards the top of the gearbox. It has three wires coming from it: +ve (yellow/black striped), -ve (black) and the speed signal (orange). The converter needs to fit in-line with these wires.

End result:
-Speedometer reads in miles per hour
-112mph limiter removed
-No VTEC in 1st gear.

Method 2:

Fit the converter between the ECU and the Speedo unit, just behind the actual speedo.

The wires required are located on the green, 10-pin connector towards the top left of the speedometer unit (as you look at it). The wires required are +ve (yellow), -ve (black) and the speed signal, which again is orange (sometimes orange with silver dots for some reason).

End result:
-Speedometer reads in mph
-VTEC in 1st gear
-112mph limiter still in place.

Why does the position matter so much?

Imagine this scenario:
You are accelerating in 1st gear, and given that all the above conditions are met, VTEC will engage at the correct revs. However consider that now you have the MPH converter in place. This is effectively turning the 14mph condition into a limit that is either out of range of 1st gear or right at the top of the rev range where VTEC engagement would be pointless for such a short time.

In other words, by fitting the converter before the ECU, all the subsequent signals related to the vehicle speed are now in MPH internally, but the ECU expects them to be in KMH. By fitting the converter after the ECU, the internals stay in KMH and only the visual output (i.e. the dials) are changed to MPH.

As for the 112mph limiter...well, thats another story

The main point of this post is to hopefully help anyone with a Honda import vehicle who is unable to get VTEC in 1st gear. For more information check out PreludeUK or other Honda specific forums.

A while ago, I built a PC and used an All-in-Wonder Radeon 9700 Pro graphics card. For those who aren't too familiar with them, read here. They were one of the first cards to require an external power connector to provide extra juice that couldn't be drawn from the AGP bus.

Coupled with this card, I used an Enermax 435W PSU, which was easily sufficient in terms of power output and was a good, reliable brand. However, I ran into problems with the computer booting from cold.

The fans would start spinning for a brief second before everything died. In my fault finding, I removed the power connector from the graphics card and the system booted up fine and displayed a message telling me to plug the card connector in. Voila, everything works!

Well, except that everytime I want to turn the computer on, I have to unplug the card...not so good.

I came to the conclusion that it was the power supply's surge protection of some sort kicking in and preventing the card from drawing the power it needed at boot up. (I tried a smaller, no name PSU and it worked fine. However, not wanting to destroy my hardware by using a cheap PSU, I needed a different solution).

The Delay

Since the system worked fine when the power was supplied to the card a few seconds after a cold boot, I figured the only sensible thing would be to simulate this delay electronically. In other words, have a system that sits between the card and the power connector that creates a delay between when I press the power button on the PC to power getting to the card.

After a bit of researching and developing and testing, I came up with the following circuit:

Components:

(1x) TIP122 Darlington Pair Transistor
(1x) 47K Variable Linear Pot Resistor
(1x) 100uF 16V Electrolytic Capacitor
(1x) 5.1K Resistor
(1x) 1K Ressitor
(1x) 6V8 Zener Diode
(1x) 1N4001 Diode
(1x) 12V 6A (Double Pole Double Throw) DPDT Relay

Putting It Together

Assemble the circuit using either a bit of stripboard, or if you have access to PCB fabriaction facilities, feel free to use the following mask, but you'll have to work out where things go for yourself...

Connect everything up and run a test just to make sure it's all working as it should. Remember, Yellow = +12V, Red=+5V, Black =0V. Connect it to the PSU and you should hear a definite 'click' as the relay latches and turns on. Make sure the transistors aren't melting and that the capacitors aren't exploding. Test it with nothing connected to the output just yet - use a multimeter to measure the outputs - they should read 12V and 5V when on.

When you are confident that it is working as it should be, connect it to your graphics card and enjoy never having to mess about with the power connector ever again!

How It Works

When the power button is pressed on the front of the computer (referred to as S1 in the schematic above), the 1k and the variable 47k resistor network allows some current to flow through and get stored in the capacitor.

As this charge builds up, the voltage at the Zener diode begins to rise. When it reaches 6.8V, the diode immediately begins to conduct, turning on the Darlington Pair Transistor.

This transistor pair is isolated from the main power circuitry using a power relay, which is in turn, switched on - causing the main 12V and 5V power lines to effectively connect straight to the graphics card via points A and B marked on the schematic. For reference, the additional diode is in place to protect the transistors.

The delay period can be varied to suit the application using the variable resistor.

Disclaimer

This is only a guide.

You may use the information in the article to help you solve problems you are having with your graphics cards or other related issues, but I AM NOT responsible for anything you blow up or damage. All the work you do is at your own risk. Remember: research and understand what you are doing. If in doubt, don't do it. If you have any questions, post here or in the forum and I can try and help.

Cheers.

Amongst all the regular fluid checks like coolant, brake and steering etc, the most important is the oil. The oil stops your car from grinding (literally) to a halt and making horrible screeching and sounds of metal against metal... As well as lubricating everything, it also traps road dirt and filters it out. Basically, its quite important.

The interval between changes will depend on your car (it is usually a good few thousand miles). On my 306, the interval is 6000 miles (diesels like to have more frequent changes - a petrol engine will probably have a longer cycle).

What you need

- Fresh oil. By fresh, I don't mean go and bore into the earth and refine your own, I mean go and buy some oil of the proper grade viscosity for your engine. The viscosity is denoted by the big writing on the side of the oil container in the form 10W40, 5W30, 20W50 etc. Find the rating for your engine by referring to a Haynes manual or your owner book. Also, make sure you get petrol or diesel oil accordingly.

- A new oil filter. These are smallish pot shaped metal containers (which come in various colours) and are screwed into the engine block and filter out all the crap the oil picks up through different grade filter paper within. The location of your filter may differ - the 306 filter was towards the front/bottom of the engine block shown below (its the white thing!):

- A tub or container. To collect the drained oil - make sure its large enough. The 306 takes 4.5 liters of oil.

Draining the Old Oil

It is best to do the change when the car is slightly warm - not too hot though as you don't want hot oil splashing everywhere. That would be bad.

First, locate the oil sump and drain plug under your car. It should look similar to the following image:

Find the correct size socket/spanner and undo the drain plug slightly. Get your container ready to catch the oil and remove the plug fully - it'll probably go all over your hand, but as long as you don't drop the plug in the oil container its ok....

Leave this to drain for a good few minutes, and in the mean time, remove the old filter - it should unscrew by hand but you can buy oil filter removal tools if you really need to. Some oil will fall from the filter once removed so make sure it doesn't drip on anything important. It's hard to imagine that this black stuff was once a smooth golden orange color....

Adding the New Oil

Once the oil has been draining for a while and has slowed to a slow drip or stopped altogether, you can replace the drain plug. It is good practice to use a new crush washer when tightening but it is not essential.

To fit the new filter, first rub some oil on the rubber seal washer which runs around the edge of the container (see image below), and screw it into place. Tighten as much as possible by hand - it should not move.

All that remains now, is to fill the engine up with the fresh oil. Locate the oil filler cap on the top of the engine (it is usually marked OIL or 710 *chuckle* I make myself laugh...). It best to use a funnel to prevent spillages and make sure you don't overfill!!!

Final Checks

Before starting the engine, check for any immediate signs of leaks and tighten the corresponding part appropriately...

Now start the engine and check once more for leaks while the engine is running. Stop and leave the car to rest for a minute, and ensuring it is on level ground, check the oil level using the dipstick. Top up if necessary, and you're done!

What you need

Well first of all, you need somewhere to host your images. They need to be in the same directory as each other, and you need control over the naming of the files (so somewhere like Imageshack won't work).

Your host also needs to be running PHP.

Method:

As you know, to place an image in your signature, you use [IMG] [/IMG] tags around the path of the image. Essentially, what you will be doing in this case, is rather than referencing an image file directly, you reference a .php file which calls a random image for your signature. With me so far?

Ok, first, name all your image files you want in your rotation logically (use 1.jpg, 2.jpg, 3.jpg etc.), then upload them all into the same folder on the server.

Open notepad or any other basic text editor and paste in the following code, courtesy of this site on a single line:

<?php $files = glob('{*.PNG,*.png,*.JPG,*.jpg,*.GIF,*.gif}', GLOB_BRACE); readfile($files[array_rand($files)]); ?>

Save the file as "rotate.php" (remember to change the file extension from .txt to .php), and upload this file to the same directory as your images.

Now alter your signature on the forum as [IMG]http://somehost/sigs/rotate.php[/IMG] you should have a fully functioning rotating signature!

Notes:

Some forums have restrictions on the type of file you can use as a signature image. Sometimes, the .php extension is restricted, which obviously causes problems when using the above method. If this is the case, use the following method: paste the code below into a blank notepad document and save the resulting file as "rand.jpg" :

<?php

Header('Cache-Control: no-cache');
Header('Pragma: no-cache');

$dh = opendir(".");
while (false !== ($file = readdir($dh)))
{
if (preg_match('/\.jpg$/i', $file) and $file != "rand.jpg")
{
$filelist[] = $file;
}
}

srand((double)microtime()*1000000);
$picnum = rand(0, sizeof($filelist) - 1);

header("Location: " . $filelist[$picnum]);

closedir($dh);
?>

Upload this file to the same directory as your signature images. This bit of code basically selects a random jpg image from the bunch you uploaded earlier (similarly to the first method), but since the file itself is technically a jpg (you don't want it to call itself during the random selection!), there is an exception statement included.

Again, use standard image tags to display your signature:

[IMG]http://somehost/sigs/rand.jpg[/IMG]

And Bob is indeed your Uncle. Any problems, post away.