HDMI or "hit-me" as I like to call it, is one step forward and two steps backwards because:

You cannot hotplug HDMI

Initialisation of connected HDMI only happens on (cold) boot. So if you switch or connect HDMI devices it probably won't correctly negotiate the mode.

That sucks.

UPDATE: This just might be the case for MY hardware, not others for which it might "just work"

Displays cannot detect HDMI sources

Maybe it's our Samsung UE46C6530 TV, though if you search the interwebs many other devices seem to suffer with this auto switching "problem".

You end up having to manually choose the input device. Awkward.

Furthermore

It's a bit annoying that on my X220 Thinkpad DisplayPort to HDMI, that it doesn't supply the audio channel along with the signal. Probably should blame Lenovo, but why doesn't it work? Incompatibilities?

Confusingly:

> DisplayPort 1.2 is the future interface for PC monitors, along with HDMI 1.4a for TV connectivity.

I've been a Dreamhost customer since July 2005 and I've been a happy camper for the most part.

However I don't know if it's because I'm more reliant on them, or because they are beginning to suck, though there has been a few unresolved issues with their services that pissing me off, hence the rant.

DNS control fail

Bitfolk renumbered my VPS IP and I had to update the IP on a number of domains, managed by Dreamhost.

Dreamhost's panel is has always been slow and when you need update umpteen domains it's unworkable. There is no bulk edit or search & replace. Result? Down time on several sites, as I slowly worked out which sites were wrong.

Perhaps there is a trick to get the whole zone file, so I can better manage my handful of domains but I don't know it.

Site statistics

My username kaihendry has different passwords on different domains. There is no way to reset them all. This totally sucks.

No 1-click unsubscribe on Dreamhost Announcement Mailing List

I run an announce list from Dreamhost since 2007 with a couple of hundred users. The publishing UI sucks and worst of all, the is no 1-click unsubscribe which is almost "de rigueur" in this marketing medium.

Right now unsubscribe bizarrely works by asking subscribers to go back to Website where they subscribed from. This is too taxing for folks. Hence I end up fielding from time to time, depressing "how do I get off this list?" type emails.

After a long painful trawl through duplicate suggestions and a long support back & forth with Brian, Tyler and Heckman to end with:

> Be good if there was an open bug to point to here. Otherwise I feel it
> will just get buried.

I will try to get that task associated with your account once I'm able to
locate it. Keep in mind, however, that what my colleague said was not
simply a dismissal. Still, even though I believe he was being completely
honest, I'm going to still make sure you get that reassurance, since it
may seem the higher priority issues are pushing this one down in the pile
(which is, frankly, not small)

When reaching the summit of Mount Kinabalu in the early morning after spending almost 2 months in Kuala Lumpur without a time update to my Casio GW-M5600-1ER wrist watch, I was surprised to see it time syncing.

This made my think that perhaps Radio clock works better, the higher you are. 4095.2m in that case.

After moving to Singapore, again my clock failed to sync during the night until I (like the directions suggest) took off my watch one night and left it on the window sill. It synced at 2012-02-22T03:04+08:00 in the Mount Sophia area of Singapore.

So yes you can pick up low frequency time signals in Singapore!

However it must be patchy since I visited a Casio G-shock store in Harbour Front and enquired about their Radio clock range. They don't sell any Casio Wave Ceptor products!

So I'm not impressed that Archlinux has gone down the GPG route.

It's no secret, I don't like GPG and I long for a lightweight solution. However engaging in a security debate with "security experts" always seems to get heated. Armed with an asbestos suit and repeated calls for me to "Go take a security course", lets try understand this nonsense.

If you update Archlinux, GPG signing is turned on default. You need to:

sudo cp /etc/pacman.conf.pacnew /etc/pacman.conf

Copy in the "default config" to disable it. A little confusing, I know.

GPG stinks

Firstly GPG based package signing and verification is complex.

Secondly I don't like the whole Web of trust bullshit. Do I want to trust Eric Belanger? Do I want to attend GPG signing parties to establish some sort of link to Eric? NO

I want to trust the team behind Archlinux ideally. An Archlinux security team, that somehow look out for integrity issues in packages.

I do not want to trust individuals and therefore trust their individual handling of their private keys.

Individuals rarely follow good key practices like rotating their keys. Revocation is a PITA. What happens if a compromised package gets uploaded? People focus on all these GPG key procedures, but what the FUCK happens when a rootkit is installed on a machine?

falconindy argues that lots of people having private keys distributes risk. That compromising one key doesn't compromise everything. In my binary world, either the machine is compromised or it isn't. Having >5 individuals with private keys and the ability to upload rootkit packages increases the attack surface area.

It's HARD to maintain one's own private (ssh) key in a safe manner. Oh no, you can't have it on a server. Oh, but can you leave your private laptop un-attended? So WHAT if it has a password protecting it?

Sucky GPG procedures raises the barrier to entry for core developers. Now you don't just need to be good at packaging, you have to learn the whole key handling rigmarole. PAIN.

What's the alternative solution?

Whilst criticizing GPG, a common retort is ask me what my alternative solution is on the spot. That's not the point. The point here is GPG stinks.

So, Archlinux probably needs a way of securely verifying the integrity of packages.

I've informally proposed using the secure transport of HTTPS to distribute hashes (which I don't think is that server straining) to simply see if something has gone awry. This is criticized as it gives one massive failure point on the server. Is someone manages to manipulate the hashes or rather sneaks in a rooted package at the server, you are fucked. But, I'd rather have one castle. At least it (HTTPS) protects man in the middle attack and one disciplined team proactively maintaining security is simple and accountable.

I asked what would happen if a compromised GPG upload took place and I had the reply:

04:37 <falconindy> hendry: _if_ a key is compromised, the attacker still needs
to get a package built and onto the servers, which requires getting through
more layer of security. ideally in that time period, its revoked and we have a
new key in place

Reading the Archlinux news item & package signing wiki, I'm still struggling to understand how these "layers" work.

I'm all for layers, in the sense PKGBUILDs are reviewed again and again, like the AUR, but I fear falconindy just says "layers" as a deflecting that a serious compromise can happen after a developer's GPG key is compromised.

I fear people will think, "oh it has a valid signature, it must be safe" and it corrupts preventative review culture. GPG sets up such obstacles, that even if you found a problem, only the signer can apply the patch. Debian style stagnation results.

Please convince me otherwise security folks. Please avoid the adhominem attacks and lets have a sensible debate.

Runlevels is one of those things that struck me when first using Linux as a bit weird, only later to discover in life that it's one of those over engineered elements from UNIX that one just ends up ignoring instead of doing anything about it.

Runlevels:
0    Halt
1(S)       Single-user
2    Not used
3    Multi-user
4    Not used
5    X11
6    Reboot

In /etc/inittab I actually bother to use runlevel 5, like so:

x:5:once:/bin/su - -- hendry -l -c '/usr/bin/startx </dev/null >/dev/null 2>&1'

To boot into runlevel 5, you need to append '5' to your Linux boot /proc/cmdline, like so in my /boot/syslinux/syslinux.cfg:

# (0) Arch Linux
LABEL arch
    MENU LABEL Arch Linux
    LINUX ../vmlinuz-linux
    APPEND root=/dev/sdb1 ro 5
    INITRD ../initramfs-linux.img

As you can see this is quite the nightmare to script if you wanted your new machine to be automatically configured with your config from git. So I am making a note of this, so I have a reference when I spend the customary day setting up my Archlinux box once purchasing a new machine.

If you run runlevel, you can confirm which runlevel you're in. You can run init 6 to reboot and init 0 to halt. init 1 is to boot into an odd mode without networking IIRC, still don't actually know what it's all about.

Airasia like it or not is the cheapest way to get around South East Asia. Unfortunately AirAsia's Website sucks. It's incredible to see so many amateur mistakes.

Besides their annoyingly slow & frustrating process to purchase a ticket, Airasia also require you to excruciatingly "check in" online. If you don't check in online, you need to pay 10MYR on the counter.

"Checking in" online asks you generally for the same information again with some clunky forms that try trick you into buying their travel Insurance.

Once you have checked in online, you then need to print your board pass. Why it can't be completely paperless is beyond me. Now you have two equally buggy options via this horrible form:

Choosing Email will send you a PDF. Why PDF instead of HTML, is one of life's little mysteries. The PDF must be printed before you get to the Airport, showing this PDF from your phone will be useless at the baggage drop off counter. Why? The PDF doesn't contain the important Aztec code (see below) and for some bizaare reason Airasia include four useless confusing barcodes in the PDF.

Another crazy element to the page is that the checkin session can expire. Which idiot coded that? Weep and start again.

If you go for the SMS option, you will get a URL link like so:

Confusingly I couldn't find this SMS at the airport as it uses the same from address as my taxi provider. I don't know how dumbphones are supposed to handle the URL, I expected some sort of booking code to be delivered by SMS... sigh.

Hopefully you will manage to follow "barcode link" the WAP (yes, seriously, in 2012) http://m.airasia.com site link to a Aztec code (not a barcode), that looks like this:

Hopefully your mobile renders it without being clipped due to large silly header "Scan 2D barcode in exchange for Boarding Pass". Now you need to scan this with one of the (hopefully working) printing machines in the airport. From there you get a slip of paper allowing you to proceed to the baggage drop off counter.

Dear Airasia, there is a lot of a room for improvement here. Don't make your customers jump through hoops and so callously upsell Insurance and holidays (adds misery on a currently poor user experience).

How about hosting your site in Malaysia instead of Europe!? Hire someone who knows what they are doing, look at Fluentspace for local Malaysian talent. Also please consider deploying a public BTS (like Bugzilla), so customers can file bugs with your process and also see some followup.

I was to pleased to purchase a Powershot Canon S100 #308030000785 from the Canon Concept store in Kuala Lumpur's MidValley mall. It's 80GBP cheaper than the Amazon.co.uk listing and I got it before its general release. I was pleased.

However 2 days later during my holiday which I bought the camera for, it stopped working. Canon S100 was taking black photos and videos.

Tbh I wasn't that impressed with the GPS lock speed and the camera being faulty made me instinctively pack it up for a refund.

However the small print of the receipt (which is incidentally typed over and hard to read) says Goods are not refundable & exchangeable which sucks a lot. In England I'm spoilt with consumer rights and Amazon's 30 day refund policy. Now my consumer confidence in Malaysia has been lost.

I've since returned from my holiday yesterday and took the trouble to go back to the Midvalley Canon store. I did ask for a refund and the sales assistant Liew said the best he could do was send it in for a priority repair. Which can take upto 3-4 weeks.

I did also separately email helpdesk.admin@cmm.canon.com.my who took a few days back to me tell me I should send the device to:

Canon Marketing Malaysia,
Block D, Peremba Square,
Saujana Resort, Section U2,
40150 Shah Alam,
Selangor Darul Ehsan,
Malaysia.

They did not offer to collect or pay for postage and packaging, even after asking.

Maybe I'm spoilt in Europe. But after this experience and the Canon IXUS 1000 HS, I don't think I will ever buy Canon products again. I'm also thinking of testing the Malaysia small claims court.

Other companies like Apple can offer good world wide customer service end consumer rights. Why can't Canon?

Update, two weeks later, it has been repaired!

Come pick me up (PITA):

Replace optical unit of the S100:

I found http://www.bbc.co.uk/news/business-16288247 quite interesting. Firstly I wish I could annotate the page and embed in the blog instead of commenting of items out of context.

Haven't played with "Enterprise social networking tools" such as Chatter, Salesforce, Yammer or Jive. I like to think a company's heartbeat is Bugzilla or Pivotal tracker. I'm guessing these are dumbed down or perhaps better focused on sales (the bottom line). Combat my ignorance, someone please.

"Another thing to watch for is that Google may lose control of Android." is a comment I guess because Amazon has usurped the tablet side of things for pushing its kindle Fire et al. Firstly I think the Honeycomb release bombed (funnily it's not used on the "Fire" AFAICT) and I don't think Googlers are interested in the tablet form factor anymore. They would rather concentrate on mobile devices. Also I think the Android based Kindles suck compared to Amazon's own un-named OS.

Didn't realise RIM were in such a bad state. 7.44BN market cap isn't great for this cruel industry. I don't see what RIM can offer. It's too entrenched in their retarded enterprise model.

NFC is something I've thought a lot about and I'm still a bit dubious. Is "wireless payments" the one killer app? Sounds lame to me. I think a VISA chip&pin card is far more versatile.

"big data" sounds like a new word to describe an old problem like "data mining". Bollocks.

Ipad/Apple/Iphone ... what will that do next... I find this technology deferment to Apple fucking irritating to say the least. Sad thing is Apple's competitors are completely retarded, so... I guess I will have to defer to Apple too.

3D printing has a nice mention. Check out http://www.tinkercad.com/ if you haven't already.

I do find Zynga quite interesting. Especially since its titles like "Word for friends" suck compared to Wordfued. I guess Zynga has the platform which has value... even though it's a sharecropper to Facebook... Right? The quote "Zynga has probably 50-60% of the social gaming market. The biggest company in mobile gaming has a 1.5% share." sounds like bullshit. 1.5% of what platform exactly?

Bring your own device (BYOD) is something I've always done. There is a term for it? I feel sorry for people working under oppressive IT regimes. Get a job that doesn't suck.

I really don't believe in the Enterprise "secure the network" intranet bullshit.

"internet of things" from Gartner sounds like an old idea, regurgitated for today. What do you expect from analysts?

My predictions

Ok after taking the piss from the "experts", what do I have to bring to the table?

I would like to predict a year where Internet users start to shun services like Facebook & twitter and even Gmail. I.e. folks start running their own distributed (de-centralised) standardised services. Unfortunately I think we are still a few years off that, so I predict things will just get worse. FB, twitter & incumbents like Google will just get more and more annoying.

I had to laugh when I saw a newsletter titled Advanced, affordable DDoS protection in my inbox from Neustar.

I think it was only recently Neustar actually publicly advertised their basic package fees. Usually Neustar has the business model of evaluating how much you think DNS is worth and charging you as much as possible for it.

Lets assume you're on the basic DNS package of a million queries a month.

Now if you go over a million queries a month, you get charged. They don't tell you how much on their Web page, but I know from being on a customer site who was "attacked" (for the first time, for no motive what-so-ever), and who went millions over that quota were charged something like 10k USD. Which is frankly madness.

I called Neustar on behalf of the client asking:

• Why didn't you inform the customer that they were being attacked?
• Why can't you (Neustar) deal with it?
• Why didn't you have some sort of cap so the customer doesn't get charged, lets say 10000 USD unexpectedly?

On the day Neustar said there wasn't much they could do, because we weren't on the more expensive "DNS DDOS protection" plan. Cue frantic calls to the "account manager" and "high level" negotiations.

On the client side, thinking of what to do... we removed the DNS record (this record was for a domain the client didn't even use!!) and according to Neustar we created even more problems for a reason I wish I knew.

What a joke. This is DNS folks. It's not that hard!!

I would like to think since they are Neustar UltraDNS, they are probably more prone to DDOS attacks then any other provider. Hence I really can't recommend Neustar UltraDNS. Avoid.

Firstly I did greatly enjoy Webcamp KL last night. Though I think we need a bigger venue and food/drink next time Wu Han.

So Wordpress, Drupal & Joomla were struck off my list for being firstly bloated, probably requiring mysql and full of crappy 3rd party plugins / modules. Hell noes!

A friend Simon Waters comments on this 3rd party nightmare:

Some of what I see is module authors trying to poke too hard, when what they are doing might be better done by extending the system itself rather than as a module. Some of it no doubt is just incompetence. I don't seem to have problems with Drupal plugins in the same way as much, but that might be luck, or it might be that Drupal has so much basic functionality is implemented as modules that the handling of them is more mature?

MongoPress looked interesting, but I thought it looked too fresh to try. It's worringly 26k of PHP with a rewrite coming.

So I'm not your typical CMS user. I don't need a nice interface. I want my content sensibly in git and to generate static content. Much like I am doing already.

Jekyll introduced by Lim of Hacker Monthly fame took my fancy. The problem I have with ikiwiki or suckless's werc is they don't do pagination which I'm a fan of when it comes to navigation.

I'm wondering how easy it is to migrate to Jekyll from my ikiwiki markdown ikiwiki sources to a Jekyll produced site?

yaourt ruby-jekyll and off we go:

==> ruby-jekyll dependencies:
 - ruby (package found)
 - ruby-albino>=1.3.2 (building from AUR)
 - ruby-classifier>=1.3.1 (building from AUR)
 - ruby-directory_watcher>=1.1.1 (building from AUR)
 - ruby-kramdown-last>=0.13.2 (building from AUR)
 - ruby-liquid>=1.9.0 (building from AUR)
 - ruby-maruku>=0.5.9 (building from AUR)
 - ruby-redcloth>=4.2.1 (building from AUR)
 - rubygems (package found)

This takes ages to install on Archlinux. I'm never too sure if I should rely on Archlinux packages or gem... though I would seriously resent having to run gem update --system or whatever it is as well as Arch's yaourt -Syua. I blame MacOSX for having a "non-package" management system.

I do moan about Jekyll dependencies, but I guess ikiwiki's dependencies are just as many. It's just that I don't notice them much when I do a apt-get install ikiwiki.

My first run of Jekyll looks like:

x220:/tmp/foobar$ jekyll 
/usr/lib/ruby/1.9.1/rubygems/custom_require.rb:36:in `require': iconv will be deprecated in the future, use String#encode instead.
WARNING: Could not read configuration. Using defaults (and options).
        No such file or directory - /tmp/foobar/_config.yml
Building site: /tmp/foobar -> /tmp/foobar/_site
Successfully generated site: /tmp/foobar -> /tmp/foobar/_site
x220:/tmp/foobar$ ls _site
ls: cannot access _site: No such file or directory

Er, wtf? Usage docs aren't inspiring me. Need to use YML (urgh), rename my posts to a "YEAR-MONTH-DATE-title.MARKUP", ok... I can do that. Luckily I got unstuck by using some Jekyll site sources.

find ~/wikis/natalian-org.branchable.com/archives -name '*.mdwn' | grep -E '20[0-9][0-9]/[0-9][0-9]/[0-9][0-9]' | while read blog
do
    IFS=/
    set -- $blog
    postname=$(basename ${10} .mdwn)
    dest=./$7-$8-$9-$postname.md
    IFS=""
cat <<END > $dest
---
layout: post
title: $(echo $postname | tr '_' ' ')
---
END
    cat $blog >> $dest
done

Wow, Jekyll has forced me to correct my posture. It's so fast compared to ikiwiki (could be my new SSD though). Pagination works! I like the clean layout.

So far I am very impressed with the output at http://jekyll.natalian.org!

Outstanding issues:

• Related posts don't work. Oh well, a feature I don't need.
• Need to figure out how I should rewrite the "archives" bit to make sure my links don't break between for e.g. http://jekyll.natalian.org/2004/02/01/homecam.html & http://natalian.org/archives/2004/02/01/homecam/
• How to import the comments? $ find -name '*._comment' | wc -l says I have 1358 to worry about
• Does Jekyll do a calendar? gah, I don't care.
• I expect URLs like http://jekyll.natalian.org/2011/04/ (all post from April 2011) to look a little nicer, not sure how to solve this