Mick West

Just downloaded the iPhone wordpress app. Seems to work quite well.

Is my computer now redundant? I think not, but it’s nice for blogging on the run.

So, my latest project is to actually wire the condo, to get around the slow and unreliable wireless connection and allow faster streaming to the XBox360 and PS3. The last wiring I did (for this same purpose) was in the Oregon house, where I laid Cat5e cables via the roof space. This was relatively straightforward, as I could get up there. But being in a condo means very limited access to the space above the ceiling, and one of my ceilings is twelve feet high. It seemed much simpler to use existing wiring.

There are actually four type of existing wires I could use. There’s the power lines, the phone lines, the cable TV lines and the TV antenna lines. The last two are basically the same thing, and luckily this condo is wired (more of less), with two sets of coaxial cables going to each spot where there’s a cable TV outlet. One is the CATV, and the other connects (presumably) to the rooftop antenna, shared by the building. That means it’s basically unused, since we just use cable TV.

Power line and phone line networking were discounted because of the slow speeds (10 Mbps, or so I thought, actually 200 is claimed now) and the unreliable nature of unshielded cables being dual-wielded, this leaves the coaxial (128 Mbps), with the obvious choice being the spare set of coaxial cables that can be dedicated to this new network.

There are a few competing technologies for this kind of thing, most notably: MoCA, and HomePNA, which comes in two flavors: phone line networking, and coaxial networking, which is sometimes called HCNA. The entire thing seems surprisingly fringe, and I had quite some trouble finding where to buy the equipment. I found a few options, but very few actually for sale.

• The Dynamix HP-30 Coax. (Unknown price)
• SendTek’s CET-330 (Unknown Price)
• Coaxsys TVNET/C PRO Target (”normally $199″, but price unknown)
• ReadyLinks IPCoax 602 (Unknown price)

Then some that you could apparently buy:

• The Highwire Connect ($179)
• The Alpha Telecom HPCE-322M ($83)
• Scientific Atlanta’s DPH548 ($128) (see pic, right)

The Highwire seemed more readily available, but is too expensive, and seems like it works in pairs, rather than a broadcast over shared wire. Of the cheaper two that I could actually order online, the HPCE-322M and the DPH548 looked fairly similar. Indeed I would not be surprised if they had the same chip inside. They seem to offer the same set of ports: cable in and out, two ethernet ports, power, a two-position switch (host/client) and some blinkenlights.

I decided on the DPH548, swayed by the “In Stock Now”, the nice diagrams of wired house, and the fact that it’s made by a company owned by Cisco. They have a nice little brochure on the adapter. It’s a bit more expensive, but I like to have a modicum of reassurance. I’m slightly concerned that there seems to be only this one obscure store on the internet selling them. It seems like they are restricting sales to installers? I don’t know. Perhaps this technology is just going out of fashion, since more people use wireless.

So I’ve ordered three of them. One will sit by the main router in my office, which is connected to the DSL line (hopefully soon to be some kind of FiOS). That one will act as the host. One will sit behind the TV, and will be connected via a switch to the Wii, the PS3 and the 360. The last one will be connected to Holly’s computer. Those will both be configured as clients.
Speeds are promised as a sustained 80Mbps. We shall see. That’s actually slower than the best wireless out there (802.11n), but I’m hoping that in addition to being much more reliable, the real-world throughput will work out higher (Wikipedia lists actual 11n throughput as 74Mbps).

Now, the diagram on the right here shows the adapter sharing the wiring with the Cable TV system. Apparently this should work just fine, seeing as the DPH548 uses the frequencies in the range of 12-28Mhz, whereas “existing RF Video” is apparently 54-860Mhz). However, I don’t trust this at all, and since I have these spare cable runs, I’m going to put it all on it’s own dedicated cables, disconnected from anything else. I think this will give me a cleaner signal, and prevent any interference with the cable TV. Plus it should give me room to change over to some meatier cable network later. I mean, coaxial cable can theoretically support several Gbps - like with all those cable TV shows in HD.

Installation

While waiting for them to arrive, I tidied up the wiring I was going to use. There are eight antenna coaxial cables which converge in the wiring closet (actually just a closet) along with the CATV cables. I found the three of them that went to the locations I needed (by plugging in a small TV, and then unplugging cables until the signal went out). I then removed these three cables from the antenna splitter, and connected them all together with a single 3-way splitter.

The image above shows the closet. Originally there were two sets of cables: the “C” cables (CATV), which go into the two splitters on the left (for five TVs) and the “A” cables (Antenna), which used to go to the eight-way splitter at the bottom. Here you can see the new splitter on the right, and the three “A” cables I’ve combined into a new network with the 2-Way splitter acting as a 3-way combiner.

The adapters were shipped very promptly by NTI. The boxes they are in are labeled “WebSTAR”, which is Scientific Atlanta’s cable modem brand. There’s a sticker on the box that says ETH-COAX, which is what NTI call it, and a handwritten serial number (112). Peeling this off reveals the actual Scientific Atlanta sticker, which looks a bit more reassuring. Made in ‘06, I guess this is something they don’t make any more, and is surplus stock?

Setup is very simple. I just basically plug everything in. Connect the cable to the “COAX NETWORK” connector, plug the ethernet into the “PC/LAN” port, and connect the power. There’s a switch for host/client, which you don’t have to adjust as it will auto-configure. But I set the one near my router to HOST and the other two to CLIENT.

Results

It works! The boxes need about ten seconds to connect after you power them up, then it seems like I have a very solid 70Mbps (about 8MB/sec) Ethernet connection. I timed the speed by copying large files and observing the percentage used in the Process Monitor’s network window. There seem to be no problems at all.

While this is comparable to the reported peak performance of Wireless-N, it’s vastly faster than Wireless-G, which is what the PS3 uses. I did the same tests with Wireless-G (supposedly 54Mbp/s), and only got 18Mbps (This would not cut it if I were to get 30Mbps FiOS, which Verizon are offering here soon). Coax also does not degraded with distance or interference, so I’m getting the same solid 70 Mbps to the far corners of the house. Wireless seem inherently unreliable, for example, I just plugged in a D-Link wireless-N adapter to my laptop, and the speed DROPPED to 30% of the Wireless-G speed on the same laptop (should have been twice as fast)

I added a switch behind the TV to connect the Xbox 360, PS3, and the Wii, and that seems fine (I still need to get an Ethernet adapter for the Wii).

Overall I’m very happy. I’ve replaced a very unreliable slow wireless network with a fast wired network. I’ll still keep the wireless connected to use with the iPhones and laptops. But for the fixed devices, it’s a vast improvement.

Powerline 200Mbps?

Of course, now I’ve actually got this set up and working, I discover that powerline ethernet adapters actually go up to 200Mbps now, with 400 planned soon. Grrr! Still, reviews are mixed, and it’s not clear if you’d actually get that 200Mbps (and it’s probably 100Mbps on-way, which is the figure that counts), what with the power lines being unshielded, and having 110 Volts running through them. So I’m still pretty happy with what I’ve got.

It works great!

My wife and I both like to read. I recently got a Kindle, and Holly just had to have one too. One downside of e-books is that you can’t share a book with a friend. However, if two people can share one Amazon account (at least, just for Kindle purchases) , then they can share all their books between their two Kindle’s.

It works really simply, and really well. When you get the second Kindle, you just register it to the same account as the first one. Then when you buy a book, or download a sample, you get a drop-down box that lets you choose which Kindle you want the book sent to. In the image to the right I’ve selected “Mick’s Kindle”.

So it gets sent to whichever Kindle you like. There’s no option to send it to both Kindles, but once you’ve bought it, then it’s in your “Media Library” on Amazon, and from there you can send it again to either Kindle.

You can also buy books on the Kindle itself, and with that it works just as you would expect - the book goes to the Kindle you ordered it on, and to the Media Library, so it can be downloaded to either Kindle at a later time.

Finally you can also get a copy of the book on the other Kindle without using the computer. Just go to the “Content Manager” on the Kindle’s main menu. In the Content Manager, some books are labeled “Kindle”, meaning they are in your Kindle, and some are labeled “Amazon”, meaning they are just in your Media Store. If Holly buys a book, it will automatically show up here.

So, to download a book Holly just bought on her Kindle, I just select it in the Content Manager, and then select “Move to Kindle Memory” from the Menu. The book will transfer, and twenty seconds later you can start reading.

This all brings me to an unexpected advantage of the Kindle. We can read the same book at the same time. Normally you’d read a book and then lend it to someone. But since we have two Kindles, with two copies of the book (for less than the price of one paper book), we can both be reading it at the same time. We are currently both reading What is the What by Dave Eggers, and it’s a novel experience to be able to discuss the book while were are both still reading it.

I love Amazon’s Kindle.

I’ve had mine for about month now, and I’ve read four books on it so far:

1984 by George Orwell. (314 pages)
Water for Elephants by Sara Gruen (350 pages)
Sabriel by Garth Nix (496 pages)
Pillars of the Earth by Ken Follett (983 pages)

As a replacement for the paper versions of these books it is nearly flawless, and provides many advantages. The most obvious one being that the weight and space of multiple books is reduced to one slender 10.3 ounce Kindle. On the face of it this might not seem such a great improvement, but really it conveys two huge benefits.

Firstly it means no more shelf space is taken up by books you are never going to read again. We’ve got lots of bookcases in our house, full of books that we read once, perhaps as much as 20 years ago. It’s hard to get rid of books you have an emotional attachment to, but they are just taking up space - a lot of space. Look at how much nine books takes up.

You could of course get rid of your books after you read them, sell them, give them away, or only read books from the library. But one thing you can’t do with a book is make it smaller. Laying in bed reading Ken Follett’s Pillars of the Earth in hardback, is not a comfortable experience. It’s nearly a thousand pages, and weighs 2.2 pounds. Yet on the Kindle it weighs just the same as any other book. (See World Without End in the image above, same sized book)

Less obvious advantages abound. I read some Luddite suggest the Kindle would not be perfect until it had two screens to simulate the two pages of an open book. That’s just stupid. Two screens would weigh twice as much, would be greatly more prone to breaking at the hinge, and you’d have to shift your reading between the pages. The beauty of the Kindle is that you can read in one position, holding the Kindle in either hand (or both, or neither), and turn pages with an imperceptible press of a button.

People who scoff at this improvement over turning pages manually would probably have scoffed at the invention of the television remote control. “How hard is it to get up and change the channel”, they would laugh. This is a natural response, and it’s only when you’ve actually tried it for a while that you appreciate actually not having to turn pages yourself, and not having to heft a heavy book one way or the other so you can read each page. You appreciate being able to lay in whatever position you like, the Kindle conforming to your comfort, rather than your position being dictated by the weight and form of a traditional book.

Having a built in dictionary might not seem so great either, but I actually used it when reading Pillars of the Earth to look up “unctuous” (location 14409), which I originally took to be something like “uncouth”, and might have misunderstood the tone of the whole section had I not had the dictionary a few clicks away. It’s just very convenient.

The search function is nice as well. You can search all your books in seconds for a particular word or phrase. That’s simply impossible with traditional books.

The books are cheaper. Especially if you look at hardbacks. Look at Follett’s World Without End, which is $35 in the shops, $21+shipping online, and just $10 on the Kindle. It’s not available in paperback yet, but Pillars of the Earth is, and costs $14.97 for the paperback, and just $6.39 for the Kindle version.

The books are cheaper in part because of the obvious low costs of duplication. But another reason is the restrictions that are placed on the digital version. You can’t re-sell the book, or lend it to anyone. Since you don’t get a physical object, the value of what you receive is diminished.

This is not an issue for me. I rarely lend or borrow books. The only person I share books with is my wife, Holly, and this is something we can still do, as we now have two Kindles (Holly’s arrived today).

With two Kindles, you can share books if both Kindles are registered to the same account. This is a minor inconvenience to Holly, as she would have to be logged in as me if she wanted to buy Kindle books on her computer. But buying directly off the Kindle itself is exactly the same.

I also really like being able to download a sample of a book. That way I can read the first chapter and see if I’m going to like it before I buy it. It’s a simple thing, but ultimately could greatly improve the average enjoyment I get out of books, and save a lot of wasted half-read books that I’m only grinding through because I paid for them without reading any of them.

One major advantage is the VAST amount of books that are totally free because they are out of copyright. All you have to do it copy them onto your Kindle, and they read just like any other book. Here’s a list of some of the best free books available, thousands of dollars worth, from some astonishingly good authors:

http://www.diesel-ebooks.com/cgi-bin/category/download-free-mobipocket-ebook-titlesearch

So what are the downsides here?

The most obvious one seems to be the initial cost. $400 is a lot of money, but then again, if I save $10 for every book I buy, then I just need to buy 40 books to have it pay for itself. Certainly within a year or so I’ll be making a profit. In addition, you get all the benefits mentioned above - worth paying extra for, and yet we actually end up saving money. Not to mention the bookcases we don’t have to buy, and all the free books you are getting.

The real problem with the Kindle is the lack of available books. Not that there are not a lot of books; there are currently 110,704 books available on the Kindle. But of those, only 38,357 are fiction, and within that, 2,465 are “Literary Fiction”, which seems to encompass much of what I like to read.

Of course that’s still a lot of books. The problem is there are books I want to read that are not on the Kindle. There’s no Iain M. Banks, and he’s my favorite author, with a new book out in hardback (Matter) that I really want to read. There are others as well, books I’ve seen recommended, but which are not on the Kindle. It’s very annoying to have this wonderful way of reading books, but not all books.

So what to do? Right now I feel like I’ll just read the books that are currently on the Kindle. Eventually the other books will be on there, and I can wait a year of so for that to happen. There is certainly enough reading matter there, and while it rankles somewhat, I can wait. If I really need to read a book, I can buy it in paper - but I don’t think it’s likely.

Other things are mostly minor. The screen is a little gray, and so needs slightly more light to comfortable read than a traditional book. You can’t easily just riffle back a bunch of pages (but it’s easy to go back one page). I sometimes turn pages by accident (but you stop that fairly quickly, and it’s no worse than loosing your page in a traditional book). It’s perhaps a little slow and cumbersome in going through your library, but then you can’t search your bookcases while laying in bed.

The version of 1984 I read had several typos in it. This is obviously because of OCR errors from scanning the book in. This is not going to be an issue with any modern book (none of the other books had similar problems). However, cheap old scanned books might continue to have some errors. This is really a problem with the individual publishers, and will diminish as e-books become more mainstream.

It’s not great for reference books. While searching is great, the small screen, the low resolution, the lack of color, and the slowness of page flipping make it unsuitable for many kinds of reference book. It’s best for the type of book that you read once, from start to end. Those books occupy many feet of my shelf space, and cost me hundreds of dollars.

So the Kindle is great, I love it. It’s only going to get better as well. In three or four years I’ll probably buy a Kindle 2.0, with a crisper, faster screen, maybe color. All my books will transfer over to it and I’ll continue to save space and money, while reading more, and reading better.

Very easy to do, you just need to install two free programs:

1) DVD43: http://www.dvd43.com
(Reboot after installing)

2) Handbrake: http://handbrake.fr/
- More specifically, install the Windows GUI version.

Then:

3) Insert DVD, cancel any auto-play

4) Run Handbrake, under Source, browse to the VIDEO_TS folder on your DVD drive

5) Under Destination, browse to any folder on your PC where the converted video will be stored, pick an appropriate name for the movie.

6) Under Presets, click on iPhone/iPod Touch, do not adjust any other settings.

7) Click Start, and wait 30 minutes to 2+ hours (depending on computer speed, and DVD length)

8) Drag the converted (.mp4 or .mpv) video file into Movies in your iTunes library

9) On the iPhone in iTunes, click the “Video” tab, then select the Movies you want to sync.

10) Sync, done. Movie will show up on your iPhone’s iPod “Videos”.

Tip: When playing, double tap to adjust the zoom in letterboxed movies.

I’d been using wireless networking at home, and generally this worked fairly well. But a couple of things were problematic. Firstly it would occasionally go out for a few minutes at random (which I ascribed to the neighbor’s wireless phone). Secondly it was rather slow using Media Center Extender for the XBox 360. It seemed to work for streaming regular definition video, but anyhing higher was dodgy, and navigating menus of content (like photos) was painfully slow.

So I decided I’d upgrade from the podunk 54Mb/s to a stunning 1000Mb/s fully switched network.

This meant I had to install six wall sockets. One for the internet connection (in the kitchen, for some odd reason), one for the new switch (in a cupboard, out of the way), and four outlets (3 in the bedrooms, one in the living room for the Xbox. That means I have to lay five Cat5e cables, each one going from the switch to an outlet.
Having never done this before, I was a little uncertain how to go about it. Luckily many have gone before me.

Cutting the holes in the wall was not something I was looking forward to, as patching drywall always seems to go badly for me. I thought at first I’d have to cut large chunks out so I could install junction boxes to the studs. But I eventually figured I could just cut small holes in the wall (see pic), and then mount the faceplate with self-drilling drywall anchors. This worked out really well. Just pre-drill 1/4″ holes for the anchors to avoid crumbling the edges of the hole. It’s pretty solid.

Up in the attic there’s a whole mess of wires. It took me a while to find out the correct spots. I mostly went by existing wiring, and hence my network outlets ended up next to my cable outlets. I drilled 1/2″ holes in the header beams where needed. 1/2″ can take four Cat5e cables (so I needed 2 for the switch).

It took a few hours to get all the cable in place. Then it was simply a case of wiring the sockets. These use a standard “keystone”, which is all color coded, and very easy to wire (just double check it). The sockets come with a little plastic tool to set the wires. You don’t need any special tools, just remove an inch of the blue outer insulation (from the cable, not the wires). Then keep the wires as twisted as possible, and push them into the color coded connectors, and push them down with the tool. (Same color configuration at each end - the keystones have an ‘A’ and ‘B’ config, just use ‘A’ for everything).

I used a six outlet plate at the switch (five outlets, one cover plate). If you have more you’d use a patch panel, but this looks very neat. One problem was that the hole had to be larger, so the large anchors did not work very well. I used plain 1″ drywall screws, but you could possibly use smaller wall plugs (or actually stick a box in there). When everything was wired up, I hooked up patch cables to all the devices, and the gigabit switch, and turned it on. Everything worked first time. Woo hoo, 80 wires set in place without mistake.

Of course it’s not actually gigabit, since all the devices are actually 100Mb, but the 5-port gigaBIT switch (a D-Link DGS-2205) was only$30 or so. If I get some gigabit devices in the future, then all will be well. The cable runs were pretty short as well (like 25 feet max) so it could probably do 10Gb later.

Results? Well, the internet now no longer goes out. Media center is a lot faster - actually very usable now for browsing through photos. Everything went according to plan. Very odd.

Parts List:

• 150 feet of Cat5e Cable (I got two 100 foot rolls). Measure first.
• 10 RJ-45 keystone sockets
• 5 single outlet keystone wallplates
• 1 six outlet keystone wallplate
• 1 blank cover
• 12 Drywall anchors or drywall screws
• five short patch cables (24″) for the switch
• five long patch cables (7′ to 14′, as needed) for the outlets/devices
• 1 five port gigabit switch (D-Link DGS-2205)

Tools used

• Drywall saw
• Wire setting tool (comes with keystone sockets)
• Drill with long 1/2″ bit for drilling through headers
• Fish tape for pushing/pulling wire, especially through walls with insulation.
• String (sometimes easier to push string through with the fishtape, then pull cable)
• Screwdriver

Took about six hours of work.

The Panasonic HDC-SD1 is a very nice little video camera that records High Definition video. We bought it for our trip to Alaska earlier this year, but did not really use it very much. But a couple of days ago Holly used it in videoing the birth of a new cousin, and this gave us 55 minutes of high definition video to do something with.

The SD1 records on SD-HC (Secure Digital - High Capacity) cards. We were using a 4-gig card (that’s what the camera comes with), and the 55 minutes resulted in 3.38GB of files. Most of these are the actual video and audio that is stored in .M2TS files - this is part of a kind of super mpeg format used in BluRay DVDs called AVCHD. The .M2TS files are really all you need, but they are surrounded by a directory structure which starts with the AVCHDL folder, and the M2TS files are hidden away in the AVCHDL\BDMV\STREAM (or more precisely in HDWRITER\071109_1\PRIVATE\AVCHDL\BDMV\STREAM).

This is all starting to sound rather complicated. The problem here is that AVCHD is still a fairly new format, and the programs for dealing with it are not particulalry user friendly. A case in point is the software that comes with the SD1, called “HDWriter”, which is of limited functionality, and rather confusing to use. However, it does let you burn regular DVDs of various clips, with some simple editing. And given what I was trying to do, I think I should have just used that for my initial DVD burning.

But no, I decided I needed some proper video editing software, so after a little bit of research I downloaded the trial version of Ulead Video Studio Plus, and started to use that.

The first problem you encounter was how to get the video into Ulead. I tried dragging the M2TS files in, but no dice. I then tried connecting the camera, and importing from the camera, but still no luck. A bit more googling led me to the correct solution, which was to click on “Capture from Disk”, and treat the camera as a removable hard drive, navigating down until you find the AVCHD, at which point you can import everything. That was a rather tedious process, as you have to select each clip individually, and selecting a clip takes a few seconds. Bu eventually all was selected, and I started to import.

Things seemed to be going well, and I burnt a DVD fairly easily using the first clip (about 45 seconds) and it did not seem to take particularly long. So I though I’d make a DVD with all the clips. That’s 55 minutes in total. So I plonk them all on the timeline, chose one of the cheesy DVD menu templates, and start burning.

20 hours later, out pops a non-working disc.

Yup, 20 hours. It actually took 20 hours to make a DVD. I was expecting maybe 2-3 hours, given that it had nearly an hour of high definition video to decode and recode, but 20 hours? I guess I was just being overly optimistic. I was doing the encoding a 1.66Ghz Centrino duo laptop with 1GB of RAM. Not the meatiest beast in the world. It probably would have been 3 times as fast if I’d used my Desktop machine. At least with the laptop we could just set it going, and leave it going, all day, all night, and a bit of the next day.

Then it did not work. It turns out though that this was just a bad disk. Luckily I’d told it to save an image of the disk, and I was able to simply burn a copy of that and it all worked.

What does this fun tell me? High definition video editing is not quite ready for prime time. I seems like everyone just wants to share their video over the internet anyway, and HD support is being left for people with a bit of time on their hands. Simply recording a one hour DVD from HD source took 20 hours. A faster machine would cut that down to 6-7 hours, but still.

Right now I’m recoding everything as 720×480 DVD format mpegs, seeing as that’s my target format anyway, hopefully it will be a bit easier to work with. It’s nearly finished converting now, been going about six hours. This is just using the batch convert in ULead VideoStudio, there’s probably some much faster method. I’ll have a look into it.

[UPDATE] We ended up using Adobe Premier Elements 4.0 with the converted mpegs to put together the DVD.  It turned out though that the MPEGs were not really that much smaller than the m2ts on the disk.  Of course they are easier to work with as they use less processor power.   The mastering time for a ten minute DVD was practically real time, like 10-15 minutes, which I was very happy with.  I’ve no idea how it would be with m2ts files.   It did crash a few times, but just about bearable.

I just upgraded my server software to cPanel. Previously it used some crappy free thing called “Basic Control Panel”, which kind of worked, but really did very little in terms of helping me manage my sites. I’d avoided changing because: A) it cost money, and B) it required a “reprovisioning” of my server. That meant the server would be wiped clean, and I’d have to back up everything and restore it.

Still, the move was precipitated when GoDady sent me 46 emails reminding me that I was on the old Fedora Core 4 and I really should be using Fedora Code 6, for security and stuff. So I decided to bite the bullet.

First I had to back everything up. All my sites are basically the same, based on wordpress, with a bunch of files. I just FTPed everything down, and then dumped the databases to SQL files. This worried me a bit, as I’d had experiences where very large SQL files caused problems (and my largest was 16MB or so). So I also decided to try and copy my databases onto another MySQL server on my backup hosting account. This proved to be the first of my many difficulties.

First I tried the freeware HeidiSQL, but it just crashed when connecting to my backup hosting account. This account is just a regular web hosting account at GoDaddy, with 100GB and 25 MySQL databases for $6.95 a month I just keep it for FTP backups, and for this kind of thing. Anyway, databases, I downloaded this nice tool NaviCat (on trial, so free for the duration), which is a professional MySQL wrangling tool, and it connected to my live databases just fine, but was unable to connect to the backup databases.

I eventually figured out (via much googling) that GoDaddy’s regular hosting does not allow external accesses to their databases. So I needed to install this little “HTTP Tunnel”, which was pretty straightforward - just a little php script which you upload, and then set it up in NaviCat’s HTTP Settings. once this was in place I was able to do a “data transfer”, which copied the entire database from the live server to the backup server.

So, everything was (hopefully) backed up. At 12:35 I tell GoDaddy to upgrade to fedora core 6. My sites all go down, and about 45 minutes later my new server rises from the ashes, empty. Unfortunately I now have to do it again to install cPanel, the server goes down again, this time for about twice as long. Eventually things spring back to life.

I then begin the task of restoring everything. Somewhat tedious. NaviCat worked very well in transferring the databases back (once I added my home IP address to the new database). I had a very annoying problem with Wordpress not being able to connect to the database. This apparently happens a lot, and the error message covers a multitude of evils. You can actually get a much more useful error message if you write a little PHP file that tries to connect, and then echos the error message on failure. But what was actually happening for me was that the installation of MySQL uses a longer hash for password authentication than is supported by the version of PHP that is installed. The solution was to downgrade the passwords by running an SQL query as root for each user/password:

SET PASSWORD FOR mysite_user@localhost = OLD_PASSWORD(’password’);

Not trivial. It took me an hour to figure this out. Again, Google was very useful, as other people had had the same problem. See here:

http://dev.mysql.com/doc/refman/5.0/en/old-client.html

I later tried to install phpBBS, but had the same problem. At this point I discovered I could set MySQL to use old style passwords all the time in WHM->Server Configuration->Tweak Settings->My SQL->USe old Style passwords. Much better.

A few other problems arose, like my mail not working. This turned out to be a couple of problems, one of which was solved by simply restarting the various mail services (easy to do, but rather obscure). The other was a little more insidious. I tried setting up a mail server for one domain (hollymick.com). Then to test it, I sent mail from my gmail account at mickwest.com to and account I’d created called test@hollymick.com

The mail got bounced with:

Delivery to the following recipient failed permanently:

test@hollymick.com

Technical details of permanent failure:
PERM_FAILURE: SMTP Error (state 13): 550-Verification failed for <mick@mickwest.com>
550-No Such User Here
550 Sender verify failed

This is a rather confusing error, as it looks like it’s saying that the user “test@hollymick.com” does not exist - which actually seems reasonably, since I’m attempting to set it up. So initially I go off and try to figure out why it does not exist. But if you look more closely, it’s actually saying mick@mickwest.com does not exist, which is rather odd, since that’s what I’m sending the mail from, and I know it’s there.

Again, other people had had this problem. What’s going on is that mickwest.com is hosted at the same server as hollywest.com. Hollywest.com has the mail server here as well, and it’s configured by default to check to see if the sender’s email address is real, so it checks with the sending mail server to see if mick@mickwest.com is real. Now this SHOULD check with Google, but because mickwest.com is hosted here, it first checks the local zone file for mickwest.com, and the default setting for the MX record is mickwest.com, meaning it’s asking the LOCAL mail server if mick@mickwest.com is good, and of course it’s not, as mick@mickwest.com lives at Google.

Solution? In WHM, under DNS Functions, Edit MX Records for mickwest.com, and set them to the Google mail servers (aspmx.l.google.com, etc).

That was about it as far as problems go. I’m really liking cPanel so far. It’s very easy to make changes, and a lot of things are automated. Like I was missing the GEDCOM perl module for my family history stuff, so I just clicked on “Install Perl Module”, typed in “gedcom” and ten seconds later it was installed.

It’s also set up to repair and update itself every night, and it emails me a nice list of things that have changed. The backup seemed to work fine as well, backing up everything including MySql databases via FTP to my backup server.

There’s also a shedload of software that can automatically be set up on a site, under “cPAddons”. This actually includes, I now notice, Wordpress, which might have saved some time. There are also a bunch of CMS, which I might have to take a look at.

One more problem that cropped up overnight was the FTP backup failed at one point with:

Can’t call method “login” on an undefined value at /scripts/cpbackup

The solution, apparently, is to enable “Passive Mode” under “Configure Backup” in WHM.

[[Summary: C:\Windows\system32\pvmjpg21.dll can cause Vista to crash in various ways including "COM Surrogate" errors. Deleting or renaming it can fix the problem, but may cause your camera to stop working. Seek updated software. Techniques are shown for tracking down similar problems]] Read the rest of this entry »

The Playstation 3 is a very powerful machine. So powerful that it needs a lot of electricity to run, and that means it produces lots of heat. It has a very efficient built-in fan to keep it from melting down, but when it gets too hot, the fan starts to run very fast, making a noise something like a small jet taking off.

My PS3 is in a cabinet with a glass door, and I use mostly for watching DVDs. If the door is closed, then the PS3 will overheat in about 30 minutes which makes it hard to hear, but if I leave the door open, then all is cool, except you can now hear the normal fan noise from the PS3. I did not like this state of affairs - for one thing I would forget, and half-way through watching a DVD I would hear this horrible high-pitched roar coming from the cabinet, and I’d have to either ignore it and hope it did not explode, or open the door and listen to it at full volume until it cooled down.

I could move the PS3, but there is really nowhere for it to go in a way that my wife would find aesthetically pleasing, so it’s kind of stuck there. Adding an aftermarket PS3 fan such as the “intercooler” is pointless, since the problem is the air inside the cabinet getting too hot, and having nowhere to go. The PS3 is quite capable of cooling itself, so long as it is “well ventilated”. So I decided what I needed to do was mod the cabinet by adding a cooling fan that would blow in cold air.
Read the rest of this entry »