Or vice versa: on restoring an antique baby changing table for our newborn.

As we were getting the room ready for our first child, we came across this interesting piece of furniture. The previous owner told us that the original chest of drawers must have been built in the 50s or 60s. It had been repurposed with the tabletop at some point and used in the family ever since.

We took it home on the spot.

I wish I had taken a shot before cleaning it for a more dramatic before/after effect.

In any case, the troubles lay ahead on the inside. Decades of wear and some crude attempts at mending it had made opening and closing the middle drawer an interesting challenge. Swearing at it did not help.

The top drawers worked okay-ish, although one of them had started to eat into the divider frame, generally not a good thing.

Also, at 75 cm high (29 ½") it was too low for either of us to use, so I thought there was potential for improvement there too.

And so the story of this project begins.

Fixing the runners

During the first inspection, it was clear that the runners, the pieces where the drawers slide on, would have to be repaired. The abrasion wear had caused two effects: it carved a track on the runners themselves, and removed part of the drawers' base.

Most of the work here was to remove enough material from the runners to make them flat, so that new runners could be glued on top. Similarly for the drawer bases.

This gave me the opportunity to use one of my favourite planes: the Preston No. 2579, a bullnose plane that can double up as a chisel plane. Ideal for awkward-to-get-to places such as the interior of the chest of drawers. I also learn that you can plane on the reverse with it, by pulling grip instead of pushing.

One does not...

Use nails on the runners.

When at some point the wood surrounding nails erodes and disappears, only the iron is left, scratching and carving into the drawers. Which ruins them in no time.

This very unexciting step was all about removing the nails before planing. I was not happy about missing the last one and thus getting a freshly sharpened chisel nicked.

Middle drawer reconstruction

Due to the massive amount of wood missing from the middle drawer's base, I realized I would not be able to repair it in place. So I decided to take it apart.

It made it easier that it was built at a time when hide glue was still common. I used hot water applied with a syringe to soften the glue at the joints. With some persuasion and a nylon hammer, the side came off much easier than anticipated.

The rest was all about getting the measurements right, building and glueing up the replacement for the missing material, including the half dovetail. And as usual, checking for square.

Next stop: recreating the groove for the drawer bottom.

This allowed me to use another plane that does not come into play often: a wooden plough plane. It worked perfectly to create the groove. One only has to remember to start from one end and then progressively work your way backwards until reaching out the other end and the adjusted depth.

The rest was about doing the same for the other side of the drawer, which was much easier and could be done in place. The Preston plane was invaluable here to do the final planing on the drawer bases, right up to the drawer front.

To plane the bases, I had to get creative with holding the drawer on the workbench.

Break a leg

Just as I thought the major repairs had been finished, I noticed one of the legs had a major split. A previous woodworm infestation was to blame there.

Nothing that some hide glue could not fix –plus a wood nail for good measure. The feet extensions were tested in parallel and fitted perfectly.

With the new shoes on

I was really pleased with the feet extensions. Not only they do their job nicely, but they look great too. They raise the original table up to 15 cm (6 inches) more, which makes a world of difference. Our backs will welcome it.

My main concern was that the extended feet might look out of place. But seeing the finished table, I felt they even gave an air of novelty to an otherwise utilitarian piece of furniture.

I did not get a perfect colour match with the stain, but it was close enough for my taste. It has a slight redder tint when compared to the original feet, due to the homemade, rust-based stain. I could probably make it darker, but I'm happy with the way it is now.

New runners

To repair the runners, I glued strips of hardwood of roughly the right height to lift the drawers, enough to slide and close properly again. I had oak and beech at hand, which should extend the chest of drawers' life for many more years of service.

What the pictures do not capture is that this was the most tedious part of the restoration.

Every drawer involved an excruciating trial and error dance, in which the fit would be tested, then a few mils from the runner strip would be planed off, and then the fit tested again. Rinse and repeat, until the drawer would slide smoothly without binding.

The middle stretcher that acts as the runner for the two upper drawers can also be seen in these pictures. The original was bent and worn in such a way that I had to discard it and build a new one. It is fitted to the front section with a mortise and tenon joint, and a screw at the back. I'm not sure why the back was not also a mortise and tenon for increased strength. I'm guessing it was because it was fitted after the main carcass was put together

Top drawers

Fixing the top small drawers was much easier than their bigger sibling. One of them had just minimal wear. That was dealt with with a thin strip of pine, planed to size to restore the original drawer height.

The other drawer had lost a fair chunk of the middle section on its bottom, which required some major surgery. I sawed off the bottom at the highest part of the curvature of wear, planed it square, glued oak strips and planed them to achieve the right eight. The Preston 2579, used as a chisel plane, was perfect for planing up to the front piece.

Finish

Applying the finish is always a rewarding moment. It generally marks the point when the hard work is complete and you can sit back and enjoy the result. I applied three coats of blonde shellac for the contemporary look.

And so the restoration was done.

All in all, what could have been a few days of work full time took a span of months, with time stolen from weekends and evenings. Most importantly, still finished ahead of the deadline.

Done

Slides

Reactions

Making a French cleat for a solid and portable tool rack.

A couple of years ago I built a rather spartan hand tool rack, also known as chisel rack or till. I might write more about it at some point, but for what relates to this post, I attached directly to the wall at our old place.

The house was an Art Nouveau building from 1915, with solid brick walls. I used wall plugs and long slotted brass screws for the classic look. Simple and effective. The downside was quite a few holes in the wall, which were a bit of a pain to restore when we repainted the wall before moving out.

The new place we moved into also has solid walls, but unfortunately behind 3 cm (1 ¼") of mortar, which start to crumble as soon as the word "drill bit" is spoken.

Generally, securing objects to the wall here requires careful planning and very long wall plugs. So I thought a different approach was needed. Although I knew about the French cleat as a neat and solid option, I had not used it before. Time to test it.

I used two long strips of beech, one for each side of the cleat, which I then planed to 45° matching angles. The process was easy: measure and scribe the 45° line with the marking gauge, and then keep planing at an angle.

To make it easy to hold the cleat and be able to plane up to its full depth, I temporarily attached it to a long pine plank I had lying around. The pictures show only the wall side of the cleat, but the toolrack side was created in exactly the same way.

I was rather pleased with how well the two pieces matched together, and the degree of accuracy that you can get with hand tools. I'm a convert now. I french-cleat all the things.

All engines in Volkswagen Type 2 busses from 1968 to 1979 were stamped with a serial number, including a two-letter code. Here's a vintage bus owner's guide on how to decypher the engine code.

Quick links

• ⏩ Take me to the engine codes list
• ℹ️ Why do I need the engine code?
• ℹ️ Where do I find the engine code?
• ℹ️ How do I read the engine code?

Engine types

Volkswagen vehicles of the era were classified into platforms that described their type. Type 1 (Beetle) and Type 2 (Transporter) were two of the most popular ones.

The engine designations were the same as the platform names of the vehicles they were first introduced in. However, there weren't Type 2 or Type 3 engines as such. Those vehicles did not feature new engine designs when introduced and used the already available Type 1 and Type 4 engines.

Type 2 busses from 1968 to 1979 were powered by one of two kinds of engines depending on the production date (and destination): the Type 1 and Type 4 engines.

Type 1 engine (a.k.a. upright): the original engine that was introduced with the VW Beetle, and thus inherited its type designation. Up to 50 BHP and 1.6l displacement. Fuel delivery was performed via a single carburetor.

Type 4 engine (a.k.a. flat): first introduced with the Type 4 model and also inheriting its name, this bigger (and flatter) engine was provided more power and higher displacements with up to 70 bhp in 1.7, 1.8 and 2.0 litre configurations. Fuel delivery was initially performed with double carburetors, and starting with model year '75 in selected markets via Bosch's L-Jetronic fuel injection. Also used in the Porsche 914, the Porsche 912E and the first generation of (aircooled) Vanagons.

Within each class of engine there were multiple variations: displacement, fuel induction system, and automatic vs. manual transmission, to name a few.

Engine codes

Quick reference

This overview can be used as a quick reference for bus engines. For more details and compatible engines in other vehicles, see the full tables below.

Early bus ('68 to '71)

Late bus ('72 to '79)

Other models

Other VW and Porsche models featured the same Type 1 and Type 4 engines as buses, with some differences in the crankcase (e.g. location of oil dipstick) and fuel delivery (e.g. D-Jetronic vs. L-Jetronic). In some cases, these engines can either be adapted or used as a drop-in replacement in a bus.

Air-cooled Vanagon ('80 to '83)

Type 4 ('68 to '74)

Porsche 914 ('70 to '76)

Porsche 912E ('75 to '76)

Beetle

TBD

Notes

1. M-code 157: exhaust emission control for the US market. This code had different meanings depending on the engine and the model year. For instance, on the 1700 engine from Aug. '71 it meant turning off vacuum advance on 1st, 2nd and 3rd gear.
2. M-code 240: engine with dished pistons (low compression) for countries with low octane (lower quality) fuel.
3. M-code 252: smaller 1.3 l engine for Italy.
4. M-code 252: 2.0 liter with 70 HP CJ/GD/GE instead of 1.6 liter motor with 50 HP (USA/CDN, S, AUS). The meaning change for the M code was from 1976 onwards.
5. The GE engine was the first one with hydraulic lifters and a single oil relief system. Model years '77 and '78 featured cylinder heads with oval exhaust ports as their predecessors. Model year '79 introduced the one-year-only cylinder heads with square exhaust ports and the fuel-injected air/fuel ratio closed-loop control for buses sold in California, which had stricter emission regulations.
   • 08.76 - GE 000 001 through GE 007 082 (7082 buses produced) most probably introduced in Sweden and Australia.
   • 08.77 - GE 007 083 through GE 039 331 (32249 buses produced) introduced in US/CDN.
   • 08.78 - GE 039 332 through GE 049 407 (10076 buses produced) cylinder heads with square exhaust ports, California fuel injection system introduced, along with continued production of the Federal fuel injection.
6. Vanagon-style: square exhaust port heads (029 101 371 on Bay Window Bus, 071 101 371 on Vanagon), hydraulic lifters

ℹ️ Early bus engines are Type 1, single-carburetted

ℹ️ Late bus engines are Type 4 in the USA, Canada, Australia and Sweden; dual-carburetted or fuel-injected. In most of Europe and other countries, they are generally Type 4, dual-carburetted, with some Type 1 exceptions.

ℹ️ Late bus engines with carburettors have a fuel pump hole in the lower right case half, whereas fuel-injected ones do not. The hole can be blocked for a carburettor to fuel-injection conversion, but not the other way round.

ℹ️ The air-cooled Vanagons also featured engines with codes CT, CS and CZ. These were neither Type 1 nor Type 4 engines and never used on T2 buses, thus they are not on the list. They were a newly developed inline motor, which shared some similarities to its predecessors (oil filter, similar cooling, etc).

Why do I need the engine code?

Knowing the engine code is not strictly necessary. However, it can be quite helpful in a number of situations. For instance, here's what the engine code will tell you –

If you are buying a new motor:

• The displacement (unless a previous owner has modified it)
• The original fuel induction system and if it has been changed from the factory
• The original application: whether it came from a bus and which year, or from another vehicle (Beetle, 411, Porsche, Vanagon)

If you are replacing or installing a new motor:

• The mount options (for instance air-cooled Vanagon mounts are different)
• The EGR and fuel induction system compatibility with your bus (location of components, different engine tins, etc)
• Exhaust system compatibility (e.g. cylinder head exhaust port type vs heat exchangers)

If you are a new bus owner:

• A bit more about the history of the bus –whether the current engine matches the model year and serial number, and whether it's a factory or replacement engine

Ultimately the engine code is simply another tool to know better a particular motor and help taking more informed decisions about maintenance, prospective work, installation and purchase.

Where do I find the engine code?

Type 1 engine number location

Type 4 engine number location

How do I read the engine code?

The engine number takes the following form:

[R] CC SSS SSS [X] [V]

• R: optional "recycled" pictogram. Only present in replacement motors. See "X" below.
• CC: generally a two-letter code that identifies the engine generation. In some cases, it can also be a letter and a digit (e.g. B0, B5, H0, H5)
• SSS SSS: 6-digit sequential serial number, starting at 000001 for each two-letter engine code
• X: replacement engine. Optionally an X letter indicates that a particular engine has been rebuilt from VW and retrofitted in a bus, replacing the original factory engine. Remanufactured engines also featured a "recycle" pictogram preceding the two-letter code.
• V: optional VW dot logo

Engine code exceptions

💡 There is an exception to this scheme in the Porsche 912E engines: the engine number has no leading letters, it is instead a 7-digit number with an 8-pointed star pictogram on each side. The number always starts with 406 and the four subsequent digits are the sequential serial number of the 2099 Porsche 912E vehicles that were ever made. The earlier Porsche 914 engines followed the standard VW engine code scheme.

💡 Another exception can be found on replacement engine cases as opposed to complete replacement engines, whereby the original engine number would have been milled away and only the "recycle" pictogram and the two-letter engine code were stamped. That is, the original serial number was lost.

💡 Some aftermarket engine rebuilders replace the stock VW engine code with their own numbering scheme. An example is, reportedly, the Dutch engine remanufacturer VEGE, who replace the original code with a riveted plate and a custom code.

Sources

The engine code table has been collated from multiple sources. All codes previously scattered through multiple locations have been aggregated into the table for easy reference, regardless of the engine's export country.

1. Engine code letters, from the Volkswagen Parts Service. Only US and Canada engine codes
2. The Samba forum engine codes list. Mostly US engine codes
3. Rolf-Stefan Badura's engine codes list. Mostly European engines.
4. Motorkennzeichnungen für luftgekühlte VW-Motoren by Michael Knappmann. Includes all aircooled VW engine codes, not only those for the Type 2.
5. VW Transporter and Microbus Specification Guide 1967-1979, by V. Molenaar and A. Prinz
6. VAG vehicle codes. Some data regarding years of manufacture seems not to be accurate.
7. Air-Cooled VW Engine Identification Letter I.D. Codes & Displacement Chart at volksbolts.com
8. Bosch Classic Service
9. How to rebuild your VW air-cooled engine, by Tom Wilson
10. Reparaturleitfaden Typ 1, Typ 2 bis 6.79 - Instandhaltung genau genommen, by V.A.G. Service. Scanned by the late Michael Knappmann.

A quick guide on removing air from hydraulic valve lifters (bleeding), such as those used on the fuel-injected VW Type 2 bus.

Why do hydraulic lifters need bleeding?

The main goal is to ensure that before installation, hydraulic lifters are properly filled with oil and oil only. By bleeding them, unwanted air is removed from the lifters.

The uncompressible oil content in their pressure chamber is what makes them effectively work as solid lifters during operation. If air (compressible gas) comes in, the mixture will create a compressible cushion in the pressure chamber. As a result, the lifters will be springy and will no longer act as solid ones. If the lifters are fully functional, they should eventually refill/bleed while driving the bus automatically, but it might take a while and they will be noisier until then.

💡 You can easily test if some air (or actually only air) is present in a hydraulic lifter. Simply push on the pushrod socket underneath the lock ring with your thumb. If you can depress it, it needs bleeding out the air inside. If you cannot depress it, then the lifter is fine.

This is not an essential maintenance procedure, though. In fact, you can bleed the lifters on the vehicle. There are multiple sources with instructions on how to do it (that ratwell article is one for instance), but it's beyond the scope of this guide. Bench bleeding lifters will help you with setting their preload when adjusting the valves, as you will be able to better feel the point of solid contact with the pushrod. Empty or partially air-filled lifters will have a spongy contact point, which generally requires more experience to find.

Also, it's never a bad idea to clean them up after 40 years of use.

How they're put together

Here are some diagrams to understand the different parts of a lifter and how they are assembled:

Notice that:

• the plunger and pushrod socket do not actually contact the lifter body. It's a precision fit that allows for controlled oil bleeding and enables the self-adjusting function of the lifter;
• the pushrod oil passage is exclusively fed by oil bleeding. That surprised me, as I thought a mechanism that allowed more oil flow would be needed. That's a topic for another thread though;
• the pressure chamber is refilled via the four oil feed metering grooves on the pushrod socket, when the ball check valve is open.

And my actual lifters, disassembled. Notice that the ball check valve is part of the plunger assembly. There is no need to take it apart. Also notice the oil metering grooves at the base of the pushrod socket.

The tools

• Bench vise
• Tweezers
• Engine oil
• Hydraulic valve lifter to bleed
• Hardwood dowel

Tweezers – Use your choice of tool instead of tweezers if you will. The reason I like them is that they let me remove the lock ring, extract the plunger and push the ball check valve when either removing or inserting the plunger. All in one tool.

Dowel – The Bentley manual recommends using an old valve guide or sawn-off pushrod to push the socket back in place to insert the lock ring while on the vise. I did not have an old pushrod laying around, and even then, I doubt I would have sacrificed it unless it were bent. A hardwood dowel (e.g. beech or oak) does the job just fine and it's easier to saw. Just make sure that it is long enough to give you room to insert the lock ring while on the vise, but short enough not to snap out of the vise. You can use a pencil sharpener to thin and round the end that goes into the socket for a better fit. Also, if I were to do this often, I would probably fabricate a small base for the dowel for it to be perpendicular and have better stability on the vise.

Engine oil – I used a small bottle with a pipette. Not strictly necessary. I used it more for the fact that there is less chance of oil spilling than with a big container. However, the pipette proved quite useful for precisely filling up to the bleed hole.

The procedure

1. Pry out the lock ring
2. Remove pushrod socket, plunger assembly and plunger spring
   👉 While you are at it, do clean the lifter with your solvent of choice. Small particles of metal like to collect at the bottom of the pressure chamber, the ball check valve can also stick with varnish.

3. Fill valve lifter body with oil up to bleed hole

4. Insert plunger spring

5. Install plunger assembly and push downward; at the same time, open ball check valve with scribe

👉 Notice oil starting to bleed out of the hole in the lifter body. Keep pushing downward until you meet some resistance.  This will generally coincide with the plunger top just starting to be submerged.

6. Insert pushrod in socket and slowly press together in vise (bore must face upward) until lock ring can be installed.
   👉 Notice how the socket will protrude a bit, as you cannot push it further down with your fingers. That's what the vise will be for.

👉 Make sure the dowel is truly perpendicular to the lifter and the vise grip. Really, check it out. If it's not, both dowel and lifter might snap out of the vise under pressure. You don't want to damage the lifter, or most importantly, yourself.

💡 Notice how oil is seeping out of the bleed/feed hole as you push. I'd recommend pressing slowly, waiting for oil to come out a bit, then cleaning up, then turning the vise again.

💡 Oil will also seep out from the top of the pushrod socket. Remember controlled bleeding is also happening there to send oil up the hollow pushrod. As our pushrod substitute here is not hollow, oil will escape through the sides of the socket, as the oil has nowhere else to go to otherwise.

💡 You are effectively doing the same action as when adjusting the lifter's preload on the bus. Only that here you will go as far as the lock ring groove, not further.

7. Install lock ring

8. You're done

Reference

1. Hydraulic lifters by Richard Atwell, for Vanagon lifters
2. Bleeding and cleaning hydraulic lifters by Christopher Schimke, for Vanagon lifters
3. Adjusting hydraulic lifters after a rebuild

The motivation for this guide was to provide easy, step-by-step instructions for the procedure, specific to the Type 2 bus.

Other than the fact that there was no such written material, there seemed to be some confusion about the procedure, particularly because:

• Richard Atwell has an extensive and excellent article about hydraulic lifters, but it does not mention the manual bleeding procedure. It also shows lifters with a different design than the ones used in 78-79 buses.
• The Type 2 (Bay Window) Bentley manual describes a procedure that is fairly cumbersome, involving submerging the lifters in oil and using a press, which most hobbyists will not have.

Reportedly, the Waterboxer Vanagon Bentley Manual instructions were easier to follow and make more sense than the Bay Window Bentley ones. I tried them, and they worked beautifully. This guide is the result of using that method specifically for the hydraulic valve lifters used in the late bus Type 4 engine.

This post is currently a draft

Quickstart

Pick the timing method you want to perform to see how it's done. Alternatively, read on for more information on timing a Type 4 engine.

On timing

Ignition timing of the engine, or simply timing in short, is the process of manually setting the point in time where the spark will occur to ignite the fuel/air mixture.

Rather than working with time units, the ignition point is measured as an angle relative to Top Dead Center for cylinder no. 1 at Compression time (TDC or TDCC). As such, the ignition is said to be:

• Advanced: when the spark occurs before TDC. The timing angle will be a positive value with a "BTDC" prefix. E.g. 7.5° BTDC
• Retarded: when the spark occurs after TDC. The timing angle will be a positive value with a "BTDC" prefix. E.g. 5° ATDC

[X] Advanced vs retarded timing

Timing is a maintenance procedure that needs to be performed whenever an engine is started for the first time (e.g. after a rebuild), or after adjusting the distributor point dwell.

It is important in the sense that the spark needs to occur at the right point in time to fully ignite the mixture and take advantage of the full energy of the combustion.

If your engine has not been adjusted as per the timing specs, here is what can happen:

• Ignition too advanced: spark occurs too early, generally before the piston has reached TDC and within the compression stroke. TODO: effects
• Ignition too retarded: spark occurs too late, generally after the piston has reached TDC and within the power stroke. TODO: effects

On timing methods

Static timing is the most basic of all. It is a rather crude but still valid means to either set the ignition timing to an adjustment good enough to start and run an engine satisfactorily for the first time, or to time an engine when a stroboscopic lamp is not available. It can also be used when installing a new distributor. Its strongest point lies in its simplicity. The weakest point is that it is simply an approximation of real world conditions: with the engine stopped, running conditions cannot be faithfully reproduced. Accuracy is also an issue, mainly due to the accummulated error of the mechanical plays involved and also due to the use of a test lamp as a measuring instrument. All that said, the result can be close enough to the timing achieved by other methods. Generally, it is done only once, and subsequent timing adjustments are of the dynamic kind.

Dynamic timing is the next step and the recommendation from VW and the ACVW community to do the adjustment under real world conditions. That is, with the engine running. There are two methods available which are essentially the same: they differ only in the point (i.e. engine condition) where the adjustment is being made. In both, we choose a characteristic condition as an anchor point to set the timing, and that point will be the baseline upon which the timing will be applied throughout the whole range. Generally, this anchor point will translate into a timing advance figure that will be added/substracted to a distributor's advance curve.

The actual measurement is performed with a stroboscopic lamp pointed to the crankcase scale, which is synchronized to lit at every spark plug #1 fire event. This ensures an accurate representation of the point in time when the spark is actually taking place (i.e. the firing point can actually be seen on the scale). The cyclic nature of the measurement also provides an intrinsic averaging of the readings.

• Dynamic timing at idle: if the chosen condition for the dynamic adjustment is at idle speed, we will be performing an idle timing adjustment as per factory recommendations. This is the best and easiest choice for an initial tuneup of a new engine, a low mileage one, or one that has been well maintained along with its distributor. This timing will be then carried out according to the distributor curve to the higher RPM ranges.
• Dynamic timing at full mechanical advance: given that new engine condition cannot be assumed for most buses on the road, and from empirical testing and accummulated experience, the ACVW community came up with a variation of the idle timing method that takes into account an aged engine/distributor and the most optimal setting for the life of a bus engine. The idea is to choose full throttle as the point of adjustment, which is where the engine will be doing most of its work. At full throttle (or as soon as ca. 3400 RPM are reached), the distributor's mechanical advance mechanism will be fully deployed. This gives the method its name: full mechanical advance, and it implies that the advance

Static timing

[[/vw-bus/timing/Distributor-wiring-3.jpg]]

What you will be doing:

• Setting the ignition point under static conditions with the engine stopped, as a means to be able to start the bus.

You will need:

• 10 mm wrench (to loosen/tighten distributor clamp nut)
• 22 mm wrench (to rotate the engine by hand)
• Test lamp or multimeter with continuity tester

Assumptions:

• Bus in neutral and e-brake set.
• The timing has either not been set before or cannot be guaranteed to be correct
• The TDC mark on the fan pulley has been found and optionally marked with some paint
• The distributor driveshaft has been installed and oriented to specs (12° from the case parting line)
• The crankcase is at TDC compression (TDCC) for cylinder #1 (distributor rotor pointing towards the engine compartment lid)

Steps:

1. Look up and note down the idle timing advance for your engine. E.g. 7.5° BTDC
2. Connect one end of the test lamp to one of the tabs of the coil's #1 terminal, also marked as "-".
3. Connect the other end of the test lamp to a chassis ground point (e.g. the engine case). The test lamp has no polarity, so it does not matter which way round you connect it. The lamp will be lit only when the points are closed (i.e. during dwell), and it will be off when the points are open (i.e. during the spark event)
   [./] Image shows test lamp connected
   [ X] Image shows primary circuit
4. Remove the distributor cap, so that you can better watch the rotor.
5. Also remove the main high tension cable in the middle of the cap and ground it while you are at it. More experienced folks will keep the distributor cap on, but we want to watch the rotor and the cam.
6. Turn the ignition on, but do not start the engine. The lamp may or may not be lighting at this point, depending on the position of the points.
7. Rotate the engine clockwise using the 22 mm wrench on the alternator pulley's nut, until the notch on the fan pulley aligns with the angle on the scale that you noted down on the previous step. If the crankcase is at TDCC (0° on the scale) and your idle timing advance spec is before TDCC, you will need nearly two clockwise revolutions of the crankcase to ensure you are at TDCC. You can optionally remove the cap of the alternator and watch where the rotor is. It will always be pointing towards the engine compartment lid at TDCC (TODO: single carb?). [X] Image shows scale at 7.5° (with an arrow) and alternator nut with wrench. A second arrow shows clockwise motion
   Note: it's best to rotate the engine clockwise for two reasons: a) the engine actually turns clockwise. If you do the same manually, you don't have to account for the distributor gear play and your timing will be more accurate; b) you don't run the risk of loosening the alternator pulley nut while rotating it with the wrench. However, bearing all that in mind, you can actually turn the engine counterclockwise, simply be careful to not loose the alternator nut, and most importantly, when you do turn CCW, turn a bit more than you need (e.g. a quarter of a revolution) and then get to your intended position turning CW. This way you will have overcome the distributor play long before reaching the desired position, making your measurement accurate again.
8. Loosen up the distributor clamp nut using the 10 mm wrench. It should still be tight, but loose enough that you can rotate the distributor case by hand.
9. Turn the distributor case in the direction that you need so that the notch on the case is aligned with the rotor pointer and the lamp lights up. That is roughly the point of maximum aperture of the points, which you can also see on the distributor cam. You are now in the ballpark, but we want to be more accurate.
10. Turn the distributor case again, but this time CW until the test light goes off. We are now starting to do the fine adjustment.
11. Turn the distributor case once more, this time CCW and aiming for the point where the light just barely goes on. If you've found it, move on to the next step. If not, iterate through this and the previous step until you hit the sweet spot moving the case CCW where the light is on, but just barely. We are aiming for the point where the points start opening and the spark is being fired. This is a simple process that appears long-winded when explained in words: you can watch this video on distributor case adjustment for static timing to easily grasp it visually. The video shows a Type 1 engine, but this part is exactly the same for a Type 4 engine.
12. Tighten up the distributor clamp.
13. Double-check the timing point with crankcase rotation: turn the crankcase 1/4 of a revolution CCW and then turn it CW until you reach the timing point in the scale.
14. If the lamp is lit, you are all set. If not, repeat steps 6-11 until you achieve the finer setting.
15. Replace the distributor cap and its connections.
16. You should now be able to start your engine. Static timing is just an ancillary method to first start the engine. Do time it using one of the dynamic methods for a more accurate tuneup.

Dynamic timing at idle

What you will be doing:

• Setting the ignition point under dynamic conditions with the engine running at idle.

You will need:

• 10 mm wrench (to loosen/tighten distributor clamp nut)
• Stroboscopic lamp

Assumptions:

• Bus in neutral and e-brake set.
• Engine can be started.
• The TDC mark on the fan pulley has been found and optionally marked with some paint
• The distributor driveshaft has been installed and oriented to specs (12° from the case parting line)

Steps:

1. Look up and note down the idle timing advance for your engine. E.g. 7.5° BTDC
2. Connect the stroboscopic lamp as per its instructions. Generally the red wire is clipped to battery positive and the black wire to chassis ground. The third wire has a clamp where the bus' ignition wire for spark plug #1 goes through, in the direction of current marked on the clamp. Do check your lamp's instructions to be certain.
3. Loosen up the distributor clamp nut using the 10 mm wrench. It should still be tight, but loose enough that you can rotate the distributor case by hand.

You will need:

• 10 mm wrench (to loosen/tighten distributor clamp nut)
• Stroboscopic lamp

Assumptions:

• Bus in neutral and e-brake set.
• Engine can be started.
• The TDC mark on the fan pulley has been found and optionally marked with some paint
• The distributor driveshaft has been installed and oriented to specs (12° from the case parting line)

Steps:

1. Look up and note down the idle timing advance for your engine. E.g. 7.5° BTDC
2. Connect the stroboscopic lamp as per its instructions. Generally the red wire is clipped to battery positive and the black wire to chassis ground. The third wire has a clamp where the bus' ignition wire for spark plug #1 goes through, in the direction of current marked on the clamp. Do check your lamp's instructions to be certain.
3. Loosen up the distributor clamp nut using the 10 mm wrench. It should still be tight, but loose enough that you can rotate the distributor case by hand.

Dynamic timing at full mechanical advance

What you will be doing:

• Setting the ignition point under dynamic conditions with the engine running at full throttle.

You will need:

• A second person to either press the gas pedal or open the throttle
• 10 mm wrench (to loosen/tighten distributor clamp nut)
• Stroboscopic lamp
• Optionally a tach meter

Assumptions:

• Bus in neutral and e-brake set.
• Engine can be started.
• The TDC mark on the fan pulley has been found and optionally marked with some paint
• The distributor driveshaft has been installed and oriented to specs (12° from the case parting line)

Steps:

1. Look up and note down the idle timing advance for your engine. E.g. 7.5° BTDC
2. Connect the stroboscopic lamp as per its instructions. Generally the red wire is clipped to battery positive and the black wire to chassis ground. The third wire has a clamp where the bus' ignition wire for spark plug #1 goes through, in the direction of current marked on the clamp. Do check your lamp's instructions to be certain.
3. Loosen up the distributor clamp nut using the 10 mm wrench. It should still be tight, but loose enough that you can rotate the distributor case by hand.

Appendix: finding TDC

With engine in the bus

[X] Image shows driveshaft 12°
[X] Image shows valve train status

With engine removed (timing scale check)

Starting with model year 1975, VW busses of the era featured fuel injection via Bosch's L-Jetronic system, with an analog Electronic Control Unit (ECU) to manage fuel induction. Over the years Bosch and VW produced different types of ECUs, with a plethora of part numbers. This guide seeks to provide a reference to their application and interchangeability.

Part numbers and interchangeability

ECU capabilities

Most of the L-Jetronic ECUs in VW busses operate in an open control loop, by sensing engine speed and load as the essential inputs, and converting them to injection time (t_i) as output for each driving condition. The last model year (1979) featured a closed control loop with a lambda sensor to sense and feed back the air/fuel mixture density.

The output stage drives the fuel injectors directly, by opening their valves with a pulse width equivalent to the calculated injection time.

It is a fully analog control unit, with no microcontroller or firmware. It is mostly built with discrete electronic components and generally 3 custom Bosch integrated circuits that perform the Pulse shaper/divider, Division control multivibrator and Multiplier functions.

In addition to the fundamental engine speed and load inputs, other variables also play a role in adjusting the final injection time. These are listed here in terms of ECU capabilities that were added or removed depending on the application:

Post-start enrichment – After the motor has started, an additional quantity of fuel will be injected as means of enrichment, that will be dependent on time and engine temperature (TS2, ECU).

Intake air correction – In order to compensate for the physically induced measuring error of the air flow meter, the injection time is corrected as a function of the intake air (Air flow, TS1, ECU)

Full load enrichment – when a full load event is detected via the Wide Open Throttle (WOT) switch, the injector time is extended to enrich the mixture (WOT switch, ECU). Only busses from model years '75-'76 and '79 California featured a WOT switch.

RPM limiter – fuel injector pulses are stopped at engine speed values higher than 5400 RPM (ECU).

Acceleration enrichment – Upon acceleration with the gas pedal, a temperature- and time-controlled fuel enrichment takes place (TS2, ECU)

Anti-bucking – When bucking occurs, it is assumed that the load changes are smaller than in an intended acceleration. If a load change occurs below a certain threshold value, the control variable will not be the newly calculated value of the slightly changed load, but instead it will be an approximation of this new value (Air flow, engine RPM, ECU).

Interchangeability

Almost all VW Type 2 ECUs are interchangeable with either moderate or no extra effort. The only exception is the '79 CA ECU, which would require major harness modifications.

There is no difference in manual vs. automatic from the ECU point of view.

In more detail:

• RPM Limiter function (interchangeable with all ECUs except '79 CA). A distributor rotor with mechanical RPM limiter is needed if the ECU does not have the electronic function.
• Wide Open Throttle switch support (interchangeable with all ECUs except '79 CA). Reportedly, there isn't a noticeable effect in running an ECU with that capability, but without the actual WOT switch.
• Air temperature sensor support (interchangeable between '75 and early '76 ECUs). The '75 and early '76 busses AFM did not have an air temperature sensor output (TS1). The AFM can be upgraded to a 7-pin one and a later ECU can be used for replacement, but not the other way round (unless you downgrade to a 6-pin AFM).
• Lambda sensor support (only '79 CA). With the right harness to account for the different double-relay and removing the injector series resistors, it might be possible that this ECU can operate in open loop without the O2 sensor. Probably too much of an effort to test this interchangeability option.

Additional notes

Some things to notice on compiling this overview:

• The Bosch documentation is not very consistent in describing which exact capabilities an ECU part number had. Also some key functons such as injector current control of the '79 CA ECUs were never listed.
• There were separate part numbers for ECUs going to the Swedish market too.
• From the descriptions, some degree of interchangeability can already be inferred.
• The first ever Type 2 ECU (1975, CA Manual) seemed to have all available functions except for Lambda control (which had not yet been on the market). Subsequent ECU parts removed or added functions as necessary. This lead to an explosion of part numbers, but it must have been cost-effective enough to justify it. There are even more ECU part numbers in the Bosch parts list, but this overview is focused only on the known ones from the original ECU part numbers table from the Bentley Manual and the The Samba forum thread (see reference section).
• The different ECU capabilites were translated from some other Bosch document.
• It is unclear to me what the anti-bucking (Anti-Ruckeln in German) circuitry does, but it seems to have been an internal circuit in the ECU that acted upon particular changes of air flow and engine speed.
• A couple of ECUs have acceleration enrichment capability, but it is also unclear to me how acceleration was sensed without a throttle position sensor (instead of the WOT switch), such as the D-Jetronic one. The ECU could have sensed the rate of speed change, but there were limits to what a few discrete components could do. This was not a microcontroller-based ECU.

Reference

• Volkswagen Service Manual Type 2, Robert Bentley publishers, Section 10 (Fuel injection), pp 23.
• Differences in ECUs between the years, post on The Samba forum.
• Fuel injection (Bosch parts interchange) at type2.com
• Fuel injection swaps, ECU parts numbers by Richard Atwell

A quick visual reference to the pinout (wiring) of the Electronic Control Unit (ECU) used in the 1975-1979 VW Type 2 bus, from the harness side. The ECU was at the heart of the fuel injection system Volkswagen chose to reduce emissions and for better fuel economy: Bosch's L-Jetronic.

Connector pinout

Harness schematic

... for Type 1 and Type 4 engines (1968-1979)

Service and specifications

Installation tips

• 👉 Do the removal and installation with a cold engine. Metal expands when it's hot. The aluminium heads expand more than the spark plug when heated up, with the consequence that the spark plug can be fastened in place.
• 👉 Clean up the spark plug holes before installation. You don't want to torque against dirt or debris.
• 👉 Finger tight the sparkplugs first, then use a torque wrench set at the specified torque.
• 💡 Your fingers won't reach to the spark plug sockets where the engine tin covering them is higher. You can use a small section of fuel hose plugged into the spark plug's terminal end snuggly. Then do the installation from the rubber hose with longer reach.
• 💡 Consider removing only one spark plug wire at a time. That way there is no chance of possibly getting the spark plug wires mixed up.
• 💡 The firing order for all air-cooled Volkswagen engines is 1-4-3-2, so make sure the plug wires go around the distributor cap clockwise in that order.
• 💡 You should use a torque wrench for installation. Consider investing in one if you don't have it, or borrowing it from someone. If you are intending to continue servicing your bus yourself, you'll be using it for many other maintenance procedures. In a pinch, if you are in a situation where a torque wrench is not available, you can install the spark plugs finger tight and then add another quarter turn (90°) with a hex wrench.

FAQ

Should I use a spark plug with RFI suppression resistor?

ℹ️ VW busses with stock ignition do not need spark plugs with integrated resistor.

This is because the ignition circuit already contains the additional resistance: the rotor cap has 5 kΩ resistor, and each spark plug connector has an integral 1 kΩ resistor. That said, resistor spark plugs can be used (see more below).

Early fuel injector systems were designed to use resistor spark plugs to reduce radio frequency interference (RFI) that could affect sensitive electronic systems, such as ECUs. The effect of the resistor is to attenuate the spark's energy and thus minimize RFI. Not all vehicles featured the integral resistance in the ignition wiring as the Type 2 bus.

At the time of writing (2017), spark plug manufacturers have phased out non-resistor spark plugs and only the resistor type is available. Non-resistor spark plugs are still available from old stock, though. So if you want to be close to stock, the non-resistor option is still that option available.

That said, the additional resistance added by the spark plug resistor (generally 5 kΩ) is probably not critical. The extra resistance on the circuit affects the spark line section of the secondary circuit's voltage waveform by varying its slope.

According to this source, as long as a value of 20 kΩ is not exceeded:

Spark or burn section (spark duration): the amount by which this line slopes away from the horizontal is directly related to resistance in the plug and coil HT leads (ignition supression). [...] The total resistance between the centre terminal of the coil and the centre electrode of the plug should not exceed about 20 kΩ [...] Actual resistance is not critical but anything more than 30 kΩ may cause problems.

Testing your ignition with an oscilloscope, Electronics Today International, February 1977

Richard Atwell performed and documented some tests with varying rotor cap resistance as well. As it's a series circuit, the results can be extrapolated to be similar for varying spark plug resistance. The result was that with an increased resistance (up to the stock 5 kΩ of the rotor cap) a more optimal waveform was generated.

Understanding the ignition system, Richard Atwell (scroll down to the "Which rotor" section for the results)

Finally, to conclude that resistor plugs can be used nevertheless in VW busses, here is a quote from Bosch's spark plug documentation:

Resistor spark plugs can be fitted in non-resistor applications. Bosch recommends fitting a resistor type spark plug for every application.

Can I reuse the crushable gasket when reinstalling my spark plugs?

ℹ️ Plug manufacturers recommend the use of a new gasket any time a plug is reinstalled after inspection or cleaning.

You can purchase inexpensive 14 mm gaskets such as these:

That said, many bus owners have reinstalled spark plugs without replacing the gaskets over the years, so in a pinch, you can also do the same.

With terminal nut?

ℹ️ Use spark plugs with a threaded terminal stud. They are generally sold with a removable terminal nut that can be detached before installation.

The terminal is connected to the ignition lead connector via a small clip that pushes against the terminal stud's thread.

More on spark plug terminal types

Regular, extended or recessed center electrode tip?

ℹ️ Spark plugs for the Type 2 use a regular center electrode tip, raising 1 mm from the thread.

More on spark plug center electrode designs

Which heat range?

ℹ️ Spark plugs for the Type 2 are on the hotter heat range end (e.g. heat range code 8 for Bosch, or code 5 for NGK).

• More on spark plug heat range, from sparkplugs.com
• More on spark plug heat range, from Bosch

Should I use anti-seize to prevent galling of the threads?

Sparingly apply anti-seize compound to the upper two thirds of the threads to lubricate them. Don't get any anti-seize on the plug electrode or it will foul

(Tom Wilson on How to Rebuild Your Volkswagen Air-Cooled Engine, pp138)

Plug manufacturers do not recommend the use of anti-seize, though, as it modifies the torque specifications. If anti-seize is used, they suggest reducing the torque values by 20 - 40 %.

Appendix: historical plug type cross-reference

Fuel injection engine

From type2.com's spark plug reference charts:

From "Reparaturleitfaden" Aug. '81 (V.A.G):

ℹ️ The original Bosch part number was W 145 M2 / W8C0. The newer part number was W8CC0. It was reportedly a long life version of the W8CC, with reinforced electrodes. Insted of having ~2.5mm electrodes, the W8CC0 had ticker (ca. 3.0 mm) electrodes, so that the gap erosion was slower.

ℹ️ Bentley shows Bosch W 145 T2 (0 241 229 548). The latest designation by Bosch is W8CC (0 241 229 579). Stock gap is 0.024 inch, 0.60 mm.

Dual carburetor engine

All engines

From Reparaturleitfaden, Aug. '81

Further reading

• Spark plug terms
• Spark plug replacement
• Spark plug torque
• Spark plug heat ranges

⚠️ This article is several years old by now. If you're planning to self-host a wiki, you might want to look at more modern, containerized, and better maintained alternatives than Gollum.

Gollum is a simple wiki that uses git for revision control and markdown syntax for editing. It is also the engine that powers Github's and GitLab's wikis. It can be easily set up for either a collaborative or personal wiki.

After following this guide you will be able to get Gollum running on an Ubuntu or Debian server with secure OAuth authentication and a commenting system with Disqus.

💡 You can also run this setup for a personal wiki on your laptop or desktop instead of a server on the net, in which case you might want to skip the Omnigollum and Disqus sections. To make it even simpler, you can also skip the Nginx configuration section.

Installing Gollum

We're assuming an Ubuntu 16.04 or later system. Debian should work too.
We're not installing from source, but rather via Ruby gems

sudo apt install ruby ruby-dev make zlib1g-dev libicu-dev build-essential git
sudo gem install gollum

Updating Gollum

When a new version of gollum is released, you can easily upgrade your installation with the gem command:

sudo gem update gollum
sudo gem update gollum-lib

ℹ️ Other useful commands are gem list and gem outdated, to list installed and outdated gems, respectively.

ℹ️ See the Troubleshooting section if you've got issues upgrading on your VPS provider.

Basic setup

After the initial installation, Gollum requires a git repository to be pointed at to work. We'll create that and add a dedicated user to access that repository.

Setting up the repository

# Add the 'gollum' user to the system
sudo adduser --shell /bin/bash --gecos 'Gollum application' gollum

# Switch user in the shell to gollum
sudo su - gollum

# You can optionally set your personal name and e-mail to
# assign all commits to yourself. Particularly useful if you
# need to do manual commits as 'gollum' to revert pages,
# which cannot be done in the UI
git config --global user.name "David Planella"
git config --global user.email "david.planella@example.com"

# Create a folder to contain the repository to be used for the wiki
# This will be at /home/gollum/wiki
mkdir wiki
cd wiki
git init .
exit

Configuring Gollum

You can pass a set of command line parameters to Gollum whenever you start it to modify and adapt its behaviour to your needs. You can also define these parameters in a configuration file that Gollum reads, to the same result.

I prefer the second approach, as it's self-documenting and it allows you to keep the configuration file under source control. The file itself is called config.rb by convention, and it's a regular Ruby source file. The Ruby configuration variables you can specify as Gollum parameters map one to one with the startup command line parameters.

Let's start by creating the file and putting it in a standard location:

sudo mkdir -p /usr/local/etc/gollum
sudo vim /usr/local/etc/gollum/config.rb

File contents:

# Define the wiki options
wiki_options = {
  :h1_title => true,
  :user_icons => 'gravatar',
  :live_preview => false,
  :allow_uploads => true,
  :per_page_uploads => true,
  :allow_editing => true,
  :css => false,
  :js => false,
  :mathjax => true,
  :emoji => true,
  :show_all => true
}

# Send the wiki options to the Gollum app
Precious::App.set(:wiki_options, wiki_options)

As you can see, all we're doing is to assigning the values of the configuration parameters to variables in Ruby syntax, and then passing the array where they are stored to the Gollum app programmatically.

In the next section, we'll be telling the system how to start gollum and pass it the location of our config.rb file.

ℹ️ While there isn't a canonical list of configuration variables, you can infer their names and what they do from the Gollum configuration parameters list.

ℹ️ Another reason for using a configuration file is so that you can modify the supported markdown formats and extenstions. As a brief example, you can limit the list of supported syntax flavors on the editor to only Markdown:

# If present, undefine the :FORMAT_NAMES constant to avoid the 
# "already initialized constant FORMAT_NAMES" warning
Gollum::Page.send :remove_const, :FORMAT_NAMES if defined? Gollum::Page::FORMAT_NAMES
# Redefine the :FORMAT_NAMES constant to limit the available
# formats on the editor to only markdown
Gollum::Page::FORMAT_NAMES = { :markdown  => "Markdown" }

Starting Gollum

We'll be setting up Gollum to start as a service when the server starts, using systemd. This will allow us to control its start/stop/restart actions when we need to shut it down (e.g. on upgrades).

sudo vim /lib/systemd/system/gollum.service

File contents:

[Unit]
Description=Gollum wiki server
After=network.target

[Service]
Type=simple
User=gollum
Group=gollum
WorkingDirectory=/home/gollum/wiki
ExecStart=/usr/local/bin/gollum --config "/usr/local/etc/gollum/config.rb" --host localhost
Restart=on-abort

[Install]
WantedBy=multi-user.target

ℹ️ Note that we're passing the path to the configuration file (config.rb) upon launch. All required configuration parameters are then contained in that file.

ℹ️ Note we also do not need to pass the path to the git repository being served. If not specified, gollum defaults to using the WorkingDirectory we've defined on the gollum.service file.

ℹ️ Note that we're passing the --host parameter for Gollum to accept connections from the same host (localhost). By default Gollum listens to connections from every available network interface (0.0.0.0).

Exposing Gollum

Next we want to expose Gollum to the web. This describes a basic web server set up to give you some guidance and get you up and running. For critical sites, you'll probably want to have a more robust and scalable setup, which is outside the scope of this guide.

Essentially, we'll be configuring a web server that acts as a front-end and more efficiently deals with client requests than the Gollum app's web server (WEBrick).

The web server will be set up as a reverse proxy that forwards the requests to Gollum's own web server. We're chosing Nginx as one of the most popular, efficient and easy to configure web servers.

It's assumed you're familiar with Nginx and it's already installed on the server. Delving into Nginx's config and installation is outside the scope of this guide, but for the most basic setup, you'll at least need to install it from the Ubuntu archive:

sudo apt install nginx

We will start by creating an Nginx virtual host that listens to port 80, and sets up the reverse proxy that forwards the client requests to Gollum's web server. Let's define it in the usual Nginx way:

sudo vim /etc/nginx/sites-available/example.com.conf

File contents:

upstream gollum_app_webrick_server {
    server localhost:4567;
}

server {
    listen 80;

    server_name $YOUR_SERVER_NAME_OR_IP;

    location / {
        proxy_set_header Host $host;
        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
        proxy_set_header X-Forwarded-Proto $scheme;
        proxy_set_header X-Scheme $scheme;
        proxy_redirect off;

        proxy_pass http://gollum_app_webrick_server;
    }
}

Finally, we simply need to enable the configured site. Again, in the usual Nginx way:

sudo ln -s /etc/nginx/sites-enabled/example.com /etc/nginx/sites-available/example.com.conf

It's recommended to test and validate your configuration files:

sudo nginx -t
nginx: the configuration file /etc/nginx/nginx.conf syntax is ok
nginx: configuration file /etc/nginx/nginx.conf test is successful

And at this point, you might just restart (or reload) Nginx for it to reload its configuration files with your new site.

sudo service nginx restart

Adding Disqus comments

Disqus relies on a snippet of JavaScript code that is loaded on every page that needs comments. That code takes care of dynamically generating the comments feed. It is called the Disqus embed code.

Gollum relies on Mustache templates to dynamically render HTML for each page upon request. To modify the content that will be shown on every page (i.e. to add comment support), we will modify the template generates each page to inject the Disqus embed code. It's called page.mustache.

ℹ️ In these steps, we'll do a simple modification to the Gollum templates to add Disqus comments. If you want to broaden the scope and learn more about extending the stock templates, you can jump to the Customizing views section.

To keep things modular and simple, we'll be adding the code on a separate template file (comments.mustache) that will be loaded by page.mustache. In Mustache terminology, the page template will be loading the partial comments template.

To recap, we'll be:

• adding a mustache partial template named comments.mustache that contains the Disqus embed code
• modifying the page.mustache template to reference (and include) the partial template

Step by step:

1. Copy all stock templates to a new location where you can edit them without modifying the originals. We'll be using /usr/local/share/gollum/ as a standard FHS location.

sudo mkdir -p /usr/local/share/gollum/
cp -R /var/lib/gems/2.3.0/gems/gollum-`gollum --version | awk '{print $2}'`/lib/gollum/templates \
  /usr/local/share/gollum/templates
sudo chown -R $USER:$USER /usr/local/share/gollum/templates

1. Create a new comments.mustache file in the new location and add the contents from the Disqus universal embed code snippet.
2. Edit the page.mustache file in the new location to add the {{>comments}} line right after the end of the body div:

    <div class="markdown-body">
      {{{content}}}
    </div>
  </div>
  {{>comments}}

1. Point gollum to the location of your new templates directory via the template-dir option: add :template_dir => 'YOUR_TEMPLATES_LOCATION' to the wiki_options array in your config.rb file.

ℹ️ Be aware of Gollum's issue 1221 and issue 1222.

Adding secure authentication

Unfortunately Gollum does not offer any means of access control or authentication, meaning that anyone will be able to edit your wiki in its most basic setup.

Omnigollum is a third-party library that builds on top of the Omniauth multi-provider authentication library to provide OAuth access control to Gollum.

With Omnigollum you'll be able to restrict access to your wiki via login through a given OAuth provider from the extensive set of providers supported by Omniauth.

:warning: Although Omnigollum seems to be the only way to provide secure access control to Gollum, a few words of caution:

• Omnigollum does not seem to be regularly maintained
• Omnigollum offers a very basic set of permissions
• The login/logout functionality is also very basic -i.e. you can log into your wiki via a 3rd-party OAuth provider, but you can neither see your status after login or actually log out

Enabling SSL

In order to be able to communicate with the SSL protocol for secure authentication on your server, the first step will be to install a set of certificates, that will be tied to the domain(s) Gollum is being served from.

For this, we'll be using Let's Encrypt a free and open Certificate Authority. The process of fetching, generating and installing the certificates is automated by the Certbot client.

The next sections describe how to install and operate the Certbot client to generate and install the required certificates.

Installing Certbot

The Certbot team maintains a PPA (Personal Package Archive) that contains the latest versions of the software. There it is updated more often than in the Ubuntu archives.

The first step is to install it:

$ sudo apt-get update
$ sudo apt-get install software-properties-common
$ sudo add-apt-repository ppa:certbot/certbot
$ sudo apt-get update
$ sudo apt-get install python-certbot-nginx 

Getting an SSL certificate

Before obtaining the certificate, we'll need to do a small change to the Nginx configuration.

Certbot will create some files under $root/.well-known and will try to access them from the Internet, so we'll have to make sure that we've set the root variable and allowed access to that location.

sudo vim /etc/nginx/sites-available/example.com.conf

Add this bit to the Nginx config file:

root /srv/www

location ~ /.well-known {
        allow all;
}

Then check if the Nginx configuration file still looks good, and if so restart Nginx

sudo nginx -t
sudo service nginx restart

You can test it by placing a text file there and trying to access the file's URL via your browser.

Finally obtain the certificate, specifying the domain it is for. You can specify multiple domains with the -d option, but note that they must at least be defined in a virtual host in your Nginx configuration files:

sudo certbot --nginx -d example.com -d www.example.com

The certificates will expire after a set time. To renew them:

sudo certbot renew

You can also set this command to be executed on a cron job. You can learn more about it on:

• Certbot documentation
• Letsencrypt tutorial on Digitalocean.com

Installing Omnigollum

Omnigollum requires an Omniauth provider to work. In our case, we'll use Github to handle authentication in our wiki, via the omniauth-github package:

sudo gem install omnigollum
sudo gem install omniauth-github

Configuring Omnigollum

These are the bits you'll need to add to Gollum's config.rb configuration file:

# [...]

## Omni Auth
require 'omnigollum'
require 'omniauth/strategies/github'

# [...]

options = {
  # OmniAuth::Builder block is passed as a proc
  :providers => Proc.new do
    # Found https://github.com/settings/applications/
    provider :github, 'Client ID', 'Client Secret'
  end,
  :dummy_auth => false,
  # If you want to make pages private:
  #:protected_routes => ['/private*'],

  # Specify committer name as just the user name
  :author_format => Proc.new { |user| user.name },
  # Specify committer e-mail as just the user e-mail
  :author_email => Proc.new { |user| user.email },

  # Authorized users
  :authorized_users => ["your_email@example.com"],
}


## :omnigollum options *must* be set before the Omnigollum extension is registered
Precious::App.set(:omnigollum, options)
Precious::App.register Omnigollum::Sinatra

Setting up Github as the OAuth provider

We'll be using Github as the OAuth provider to control access to the wiki. This is a very condensed set of instructions from Github's Basics of authentication.

Register your app at Github

1. Register your app at Github to get an OAuth Client ID and Client Secret
2. Fill in the details, and set the Authorization callback URL to your site.
3. Click Register application. You're nearly done.

Configure user authorization on the server

1. Go to your registered OAuth applications page at Github and simply click on your newly registered Gollum app
2. Copy the Client ID and Client Secret from there
3. Add github to the providers option in Gollum's config.rb file and paste the Client ID and Client Secret as shown:

options = {
:providers => Proc.new do
    # Found https://github.com/settings/applications/
    provider :github, 'your-client-id-goes-here', 'your-client-secret-goes-here'
  end,

# [...]

Ensure gollum is running (sudo service gollum status or sudo service gollum start) before the next step.

Authorize the Gollum app at Github

1. Navigate to your site in your browser (e.g. https://wiki.davidplanella.org)
2. Authorize the application in the Github prompt that appears on the browser.

That's it!

Making a wiki private

Optionally, you can make your wiki or part of it private. You can achieve this by specifying the path of your wiki you want to make inaccessible in the :protected_routes variable in config.rb.

As an example, the following values will not allow unauthenticated visitors to:

• See any subpages you create under the private namespace in your wiki
• Revert, create, edit, rename, upload or delete pages
• See individual page history or latest changes globally
• They will still be able to access (i.e. view) the open parts of your wiki, that is, those outside of the private namespace.

They will still be able to access (i.e. view) the open parts of your wiki, that is, those outside of the private namespace.

:protected_routes => [
    '/private/*',
    '/private',
    '/revert/*',
    '/revert',
    '/create/*',
    '/create',
    '/edit/*',
    '/edit',
    '/history/*',
    '/history',
    '/latest_changes',
    '/rename/*',
    '/rename/',
    '/upload/*',
    '/upload/',
    '/delete/*',
    '/delete'],

:warning: Be aware of Omnigollum's issue #44 and issue #45 before you do this.

:warning: You might want to set the :authorized_users variable in the config.rb file to restrict which authenticated users have access.

Customizing Gollum wiki's appearance

Customizing styles

You can add your own styles to augment or modify the stock Gollum theme by creating a custom.css file and placing it in the root of your wiki's git repository. To enable its use, you'll need to either set :css => true in the wiki options of the config.rb file or start gollum with the --css command line option.

ℹ️ You'll need to have committed the custom.css file on your git repository for custom styles to be taken into account.

Code injection

You can also inject your own JavaScript code styles to augment or modify the stock Gollum theme by creating a custom.js file and placing it in the root of your wiki's git repository. To enable its use, you'll need to either set :js => true in the wiki options of the config.rb file or start gollum with the --js command line option.

ℹ️ You'll need to have committed the custom.js file on your git repository for your custom code to be used by Gollum.

Customizing views

If you've followed the Adding Disqus comments section, you will know that Gollum renders its wiki pages based on a set of Mustache templates. In plain terms, it is a set of files with the .mustache extension, with a combination of HTML markup and Mustache syntax that you can modify to your own needs. The file hierarchy looks like this:

/var/lib/gems/2.3.0/gems/gollum-4.1.2/lib/gollum/templates/
├── compare.mustache
├── create.mustache
├── edit.mustache
├── editor.mustache
├── error.mustache
├── file_view.mustache
├── history_authors
│   ├── gravatar.mustache
│   ├── identicon.mustache
│   └── none.mustache
├── history.mustache
├── latest_changes.mustache
├── layout.mustache
├── livepreview.mustache
├── page.mustache
├── pages.mustache
├── searchbar.mustache
└── search.mustache

The two perhaps most important files you'll want to modify are:

• layout.page: this is the template that defines the basic layout of all pages. As such, it defines the HTML5 doctype, which CSS files and JS files will be loaded and in which order. It also defines the <body> tag and the most basic CSS styling.
• page.mustache: most of the content on your wiki will be viewed as pages. This template defines the layout of each single page. It can be used, for instance, to add Disqus comments to each page.

This makes it a modular and versatile system for extension and modification. You can create your own views or modifying existing ones by simply adding (and including) new .mustache files, or modifying the stock files. In theory, you could also create your own full theme, but I have found that currently the core gollum JavaScript code and CSS classes are not decoupled enough from the view generation in order to be able to do this cleanly. It can be done, though.

Here's a suggestion on how to customize the Gollum views to your own needs:

1. If you haven't done it already, copy all stock templates to a new location where you can edit them without modifying the originals. We'll be using /usr/local/share/gollum/ as a standard FHS location.

sudo mkdir -p /usr/local/share/gollum/
cp -R /var/lib/gems/2.3.0/gems/gollum-`gollum --version | awk '{print $2}'`/lib/gollum/templates \
  /usr/local/share/gollum/templates
sudo chown -R $USER:$USER /usr/local/share/gollum/templates

1. Also, only if you haven't done it already, append the :template_dir => '/usr/local/share/gollum/templates' line in the wiki_options array inside your config.rb configuration file
2. If you've changed the config.rb file, restart gollum: sudo service gollum restart
3. Now open the template files you need to modify and hack away :)

Troubleshooting

Cannot allocate memory when updating Gollum gems on Digitalocean

ℹ️ In some cases some gems cannot be updated, returning a Cannot allocate memory message. This seems to happen on Digitalocean servers, where due to their SSD storage, swap is not enabled. A workaround is to temporarily create a swap file, do the update and then delete the swap. In a nutshell:

sudo fallocate -l 4G /swapfile
sudo chmod 600 /swapfile
sudo mkswap /swapfile
sudo swapon /swapfile
sudo gem update gollum
sudo swapoff -a
sudo rm /swapfile

For the '74-'79 fuel-injected Bay Window bus (Type 2)

Vacuum hose identification

Vacuum control hose identification

Vacuum control hose routing

The guide is geared towards a bus, but for the most part it should apply to other models and makes equipped with the same engine (VW Type 4, Porsche 912E, Porsche 914). For simplicity, and with the hobbyist with the basic toolset in mind, the bus is kept on the ground. A seasoned mechanic with access to heavier equipment would most probably lift the vehicle.

While the whole process, including oil strainer cleaning and oil filter replacement is described, they are not necessary to do at each oil change. Check the service intervals on your owner's manual to decide when you need to do these.

Materials

Consumables

The most essential items you will need to do an oil change. Note that the picture is simply showing one brand and grade of engine oil for the sake of example, but not advocating for its use

1. ✅ Engine oil
2. ✅ Oil filter [021 115 351 A]
3. ✅ Oil drain plug washer [N 0138492]. ID: 14mm, OD: 20mm, thickness: 1.5mm
4. ✅ Oil strainer gasket set [021 198 031] (two gaskets and a crush washer)
5. ⭕ Oil strainer [021 115 175] (only necessary if you can't reuse your original one)

❓ Which oil should I use?

❓ Which oil filter should I use?

💡 Unlike other procedures in your car, and perhaps except for the oil itself, the parts required for doing an oil change are cheap. Do yourself a favour and order each set more than once. It can be a lifesaver if you lose or damage any part, and it will save you postage fees for the next oil change if you ordered them online.

💡 Unless your old strainer is damaged, a new one is optional. But it can save you the time to clean the old strainer. You can then keep the old strainer, clean it in your own time, and swap them on the next oil change. Also, oil strainers are made of aluminum and thus relatively fragile, so if you happen to crush one of them, you will then have a replacement.

Tools

Essential

• ✅ Socket wrench
• ✅ Torque wrench in·ft or N·m (e.g. 6 - 30 N·m)
• ✅ 19 mm socket
• ✅ 13 mm socket
• ✅ Drain pan (min. 4 l or 1 gallon)

Good to have

• ⭕ Disposable gloves: nitrile or latex, you don't want blackened nails when you go to dinner later
• ⭕ Large paper roll: to clean parts and any mess left behind
• ⭕ Bowl or container: for oily removed/replaced parts
• ⭕ Funnel: to do the oil refill
• ⭕ Old toothbrush: to clean oil strainer
• ⭕ Carb and choke cleaner or dish soap: to clean oil strainer and plate
• ⭕ An old mat: to lie on the floor while doing the oil change

⚠ Use the torque wrench for tightening only, you do not want to overstress the torque limiting mechanism while applying lots of force. Use the regular wrench for undoing nuts and bolts.

💡 If you use lbf·ft where you are, you will be better equipped with an inch pound torque wrench for low torques such as here. Torque wrenches with a pound foot scale tend to have a wider range and are less accurate on the lower torque band.

Procedure

Drainage and cleanup

1. Warm up the engine: optionally drive or run the car idle for a few of minutes to heat up the oil. It will make it less viscous to ease draining
2. Find a place on level ground: it will ensure effective draining and accurate oil level check
3. Chock the vehicle: use wheel chocks for safety and to prevent the vehicle from moving
4. Prepare the change pan: place it underneath the drain plug
5. Remove drain plug with a 19 mm socket
6. Let the oil drain for some minutes
7. Remove the oil strainer nut with a 13 mm socket
8. Clean or replace the oil strainer: an old toothbrush works wonders here. It's also a good opportunity to inspect the old oil for particle deposits: metal filings in your strainer could mean main bearings damage and are bad news.
9. Clean the strainer cover plate with dish soap
10. Remove the oil filter (by hand or with tool). Move the change pan underneath the filter first.

💡 As the filter is obviously filled with oil (half a liter/quart of it), a thick black mass will start quickly dripping out of the filter as soon as you remove it from the thread. Have your oil pan ready for it underneath. If your pan is not large enough to cover all the places where oil is dripping from now (oil strainer, oil drain, oil filter holes), put the filter in another container and then pour the oil back to your pan. Alternatively, assemble the oil strainer and drain plug first, and then undo the filter.

💡 When removing the oil strainer nut, you will notice the old crush washer will be firmly attached to it. Use a thin knife or metal plate to pry it out and then replace it with the new washer.

❓ What is the inspection plate?

❓ What is the oil pressure relief valve?

Assembly

1. Drain plug: use a new copper washer [3] and tighten the drain plug [4] with a 19 mm socket at 22 Nm (16 lbf·ft)
2. Oil strainer: use a new gasket set [1] and crush washer [2], tighten the nut [5] with a 13 mm socket at 10-13 N·m (7-9 lbf·ft)
3. Oil filter: coat its rubber seal with engine oil and install the oil filter hand tightening it.

⚠ Do not overtighten the oil strainer nut. This is probably the most important thing you need to watch out for while doing the oil change: you can break the engine case otherwise. Keep to the torque spec of max 13 N·m or 9 lbf·ft. Do yourself a favour and get a torque wrench. If you are however in the middle of nowhere only armed with a regular socket wrench and sticks and stones, get it tight and then a tiny bit more.

💡 Fitting the oil strainer and the two sandwiched gaskets while lying on the floor on a tight space and trying to keep it level is not everyone's idea of fun. An easier way is to attach the gaskets to the strainer so you are dealing with only one part. Glue them up with something that's oil and temperature resistant and that's easy to remove next time you need to replace the gaskets. I use Hylomar M (non-setting gasket compound) as it was what I happened to have at hand last time. I've heard of other folks using grease for instance. Your mileage might vary.

💡 In a pinch, if you do not have a new copper washer for the drain plug, you can anneal the old one by heating it with a propane torch after removal. This will soften and expand the metal, preparing it for a fresh crush.

Refill

1. Refill: open the oil filler and refill with oil to the required capacity
2. Check level: check the oil level on the dipstick (no need to wait for it to settle, as you've only filled the sump). At this point it might seem high, but particularly if you've changed the filter, which will be filled with half a quart (liter) once you start the engine.
3. Don't be that guy: ⚠ Put the the oil filler cap back on before driving away.
4. Start: start the engine, and check that the oil pressure warning light goes out after a few seconds as expected.
5. Inspect: check for leaks and address them if necessary –e.g. retighten the filter.

Once you're done, remember to bring the old oil to the nearest recycling center for safe disposal.

⚠ If you forget to put the oil cap back on, you will be creating a vacuum leak in the crankcase. The oil filler is not part of the oil circuit, so the oil pressure will not be affected. But other than potentially adding contaminants to the sump (e.g. you've been driving offroad with the cap off), the main issue is the vacuum leak, which won't allow the engine to run well. Particularly in fuel-injected engines, you will be introducing unmetered air to the intake, the symptom of which is a poor idle, generally stalling the engine.

💡 While checking for leaks, you might want to put a cardboard on the floor underneath the engine. This will help you see the most obvious leaks, particularly if some oil was already spilled on the floor during the change.

ℹ There seems to be one school of thought that favours prefilling the filter before fitting it in, as opposed to fitting it dry. As described in Tom Wilson's "How to rebuild your Volkswagen air-cooled engine", pp148, the idea behind it is to speed priming at engine start. YMMV.

Reference

Service interval

TBD

Torques

Capacities

FAQ

Which oil should I use?

Ah, you've happened to reach a very hot topic, and outside of the scope of this guide.

Folks with more expertise than I have been discussing this at length, so I'll just point you to that conversation and some other resources to make your own decision. For instance, this website, despite the "welcome to the 90s" look goes into the concepts related to oil characteristics and usage in a very concise and clear way.

TL;DR:

1. Use a multi-grade oil with at least API Service Classification SE or higher (current is SN). Then,
2. decide on mineral vs. synthetic, then
3. decide on its grade (thin vs. thick) depending on the age of your engine and the area where you'll be driving it.

I myself ended up picking up the same grade the previous owner had been running the bus with for the last 10 years.

Which oil filter should I use?

Just get one of the OEM ones for VW's 021115351A:

• Mann W920/17 filter
• Mahle OC 28 filter
• Bosch 0 451 103 084 (or P3084) filter

From what I gather, some folks in the US also recommend the NAPA Gold 1521 or the WIX 51521.

ℹ Note that only Mann and Mahle seem to be doing their own filters. All of the rest outsource fabrication to one of a set of companies. Often the same filter manufacturer delivers filters to different companies. While not exactly for the same filter model, if you want to learn more about oil filter manufacturing, their quality and more, you can check out this great comparison of German (read for VWs) oil filters

What is the inspection plate?

While not used for an oil change, in case you are wondering about the inspection plate hole, its main purpose in a bus is to house an engine heater, used in cold climates to pre-heat the oil to ease engine start.

In Type 4 engines on other vehicles such as the Porsche 914, it used to house the oil temperature sensor, with a modified plate and cover often referred to as the taco plate for its supposed resemblance to a taco. As such, some bus owners use that location with either a modified aftermarket or an original Porsche taco plate to mount their oil temperature senders.

Regardless of its name, very little inspection, if at all, can be made from the inspection plate.

What is the oil pressure relief valve?

Also not used for an oil change, you might have noticed the big screw for the oil pressure relief valve. The valve effectively controls the flow of oil depending on its temperature, e.g. allowing it to bypass the cooler on startup.

To better understand this, I recommend reading the Lubrication notes from the IAC forum and the Dual Relief Oiling System post on the Samba forum.

Learn more

Here's a collection of great external resources for further reading or watching on the subjects of oil change and selection.

• Changing the engine oil, from the 1979 VW Type 2 Owner's Manual
• Oil change, by Sean Bartnik
• Oil selection, by Richard Atwell
• Engine oil service categories from the American Petroleum Institute (API)

I took some pictures while my bus was being lifted, for reference and to assess the status of everything under-floor.

The pictures were stitched together into one, so the perspective is not always right.

Ubuntu has been around for just over a decade. That's a long time for a project built around a field that evolves at such a rapid pace as computing. And not just any computing –software made for (and by) human beings, who have also inevitably grown and evolved with Ubuntu.

Over the years, Ubuntu has changed and has lead change to keep thriving in such a competitive space. The first years were particularly exciting: there was so much to do, countless possibilities and plenty of opportunities to contribute.

Everyone that has been around for a while has fond memories of the Ubuntu Developer Summit, UDS in short. An in-person event run every 6 months to plan the next version of the OS. Representatives of different areas of the community came together every half year, somewhere in the US or Europe, to discuss, design and lay out the next cycle, both in terms of community and technology.

It was in this setting where Ubuntu governance and leadership were discussed, the decisions of which default apps to include were made, the switch to Unity's new UX, and much more. It was a particularly intense event, as often discussions continued into the hallways and sometimes up to the bar late at night.

In a traditionally distributed community, where discussions and planning happen online and across timezones, getting physically together in one place helped us more effectively resolve complex issues, bring new ideas, and often agree to disagree in a respectful environment.

Change takes courage, it takes effort in thinking outside the box and going all the way through, but it is not always popular. I personally believe, though, that without disruptive changes we wouldn't be where we are today: millions of devices shipped with Ubuntu pre-installed, leadership in the cloud space, Ubuntu phones shipped worldwide, the convergence story, Ubuntu on drones, IoT... and a strong, welcoming and thriving community.

At some point, UDS morphed into UOS, an online-only event, which despite its own merits and success, it does admittedly lack the more personal component. This is where we are now, and this is not a write-up to hark back to the good old days, or to claim that all decisions we've made were optimal –acknowledging those lead by Canonical.

Ubuntu has evolved, we've solved many of the technological issues we were facing in the early days, and in many areas Ubuntu as a platform "just works". Where we were seeing interest in contributing to the plumbing of the OS in the past, today we see a trend where communities emerge to contribute taking advantage of a platform to build upon.

Yet Ubuntu is just as exciting as it was in those days. Think about carrying your computer running Ubuntu in your pocket and connecting it to your monitor at home for the full experience, think about a fresh and vibrant app developer community, think about an Open Source OS powering the next generation of connected devices and drones. The areas of opportunity to get involved are much more diverse than they have ever been.

And while we have adapted to technological and social change in the project over the years, what hasn't changed is one of the fundamental values of Ubuntu: its people.

To me personally, when I put aside open source and exciting technical challenges, I am proud to be part of this community because its open, welcoming, it's driven by collaboration, I keep meeting and learning from remarkable individuals, I've made friendships that have lasted years... and I could go on forever. We are essentially people who share a mission: that of bringing access to computing to everyone, via Free Software and open collaboration.

And while over the years we have learnt to work productively in a remote environment, the need to socialize is still there and as important as ever to reaffirm this bonding that keep us together.

Enter UbuCons.

The rise of the UbuCons

UbuCons are in-person conferences around the world, fully driven by teams of volunteers who are passionate about Ubuntu and about community. They are also a remarkable achievement, showing an exceptional commitment and organizational effort from Ubuntu advocates to make them happen.

Unlike other big Ubuntu events such as release parties -celebrating new releases every six months- UbuCons happen generally once a year. They vary in size, going from tens to hundreds to thousands, include talks by Ubuntu community members and cross-collaboration with other Open Source communities. Most importantly, they are always events to remember.

A few months back, at the Ubuntu Community Team we started thinking of how we could bring the community together in a similar way we used to do with a big central event, but also in a way that was sustainable and community-driven.

The existing network of UbuCons came as the natural vehicle for this, and in this time we've been working closely with UbuCon organizers to take UbuCons up a notch. It has been from this team work where initiatives such as the UbuContest leading to UbuCon DE in Berlin were made possible. And more support for worldwide UbuCons general: in terms of speakers and community donations to cover some of the organizational cost for instance, or most recently the UbuCon site.

It has been particularly rewarding for us to have played even a small part on this, where the full credit goes to the international teams of UbuCon organizers. Today, six UbuCons are running worldwide, with future plans for more.

And enter the Summit

But we were not content yet. With UbuCons covering a particular geographical area, we still felt a bigger, more centralized event was needed for the community to rally around.

The idea of expanding to a bigger summit had already been brainstormed with members of the Ubuntu California LoCo in the months coming to the last UbuCon @ SCALE in LA. Building up on the initial concept, the vision for the Summit was penciled in at the Community Leadership Summit (CLS) 2015 together with representatives from the Ubuntu Community Council.

An UbuCon Summit is a super-UbuCon, if you will: with some of the most influential members of the wider Ubuntu community, with first-class talks content, and with a space for discussions to help shape the future of particular areas of Ubuntu. It's the evolution of an UbuCon.

As a side note, I'm particularly happy to see that the US Summit organization has already set the wheels in motion for another summit in Europe next year. A couple of months ago I had the privilege to take part in one of the most reinvigorating online sessions I've been in recent times, where a highly enthusiastic and highly capable team of organizers started laying out the plans for UbuCon Europe in Germany next year! But back to the topic...

Today, the first UbuCon Summit in the US is brought to you by a passionate team of organizers in the Ubuntu California LoCo, the Ubuntu Community Team at Canonical and SCALE, who hope you enjoy it and contribute to the event as much as we are planning to :-)

Jono Bacon, who we have to thank for participating in and facilitating the initial CLS discussions, wrote an excellent blog post on why you should go to UbuCon in LA in January, which I highly recommend you read.

In a nutshell, here's what to expect at the UbuCon Summit:

• A two-day, two-track conference
• User and developer talks by the best experts in the Ubuntu community
• An environment to propose topics, participate and influence Ubuntu
• Social events to network and get together with those who make Ubuntu
• 100% Free registration, although we encourage participants to also consider registering for the full 4 days of SCALE 14x, who are the host to the UbuCon

I'm really looking forward to meeting everyone there, to seeing old and new faces and getting together to keep the big Ubuntu wheels turning.