Welcome to another “random” post – a collation of bits and pieces of random thought which don’t quite deserve their own posting. Last week, I didn’t get around to making one, so this week’s another “bumper” crop of randomness.
I can’t claim to hold any level of fame, nor can I claim to be inspirational at any level. But people have looked upon my work at this blog and wondered exactly why I even bother.
This week, I came across a web-comic drawing which serves as inspiration and one of the best ways of explaining just why I bother. My words could never do it justice, so I will let the comic speak for itself.
Go on, read it … I’ll wait.
Hopefully now you realize that the “gift” of producing the blog is the blog itself, but is also the hours of enjoyment and self-education that comes with exploring, destroying, reverse-engineering, creating and defining concepts and technologies which surround us everyday. I suppose I am an engineer at heart – although I can’t say it’s not nice to have money, fame, attention or affection. Sometimes I might wish for that too, but I remember that it’s better to do something rewarding in the interim rather than be hung up on wishes.
A Week of Failure
The last week has been a week filled with failures. One thing I didn’t expect was two power failures with a total downtime of about 40 minutes. It just so happened to disrupt a 40 hour x264 encode process which I had running to subjectively evaluate encoding quality and size at different CRFs (Constant Rate-Factors). Part of this was motivated by the new TV and a desire to determine just what bitrates and rate-factors correspond to acceptable quality to me, while minimising storage and allowing for stable wireless streaming (single-band, single-stream, backwards-compatible 802.11n doesn’t give you much). That was annoying.
The second failure was Dropbox. Last Saturday, Dropbox was not behaving at all. I tried to export an image from Procreate into Dropbox only to see it hanging at trying to list my Dropbox contents. I thought it was an issue with the app-linkage, so step 1 of troubleshooting is to unlink and re-link.
I unlinked only to find I couldn’t re-link at all – with different error messages!
What the hell!? My internet connection was good too. I opened Dropbox app, things seemed okay initially but things wouldn’t open. I did something silly – I un-linked the Dropbox app and tried to re-link it.
It twirled for ages, and then, that failed miserably as well! Now I had no access to any Dropbox files on the iPad. Cool.
It was not until I tried to visit Dropbox via the web-browser that they told me that they had failed. Instead of seeing Dropbox’s normal home-page, I was bounced to status.dropbox.com. Yep. Dropbox is not infallible – and it failed just as I wanted to use it. How convenient.
Interestingly, that persisted into Sunday, and while linking and file access was enabled, uploading from Procreate would stick for hours at 100% and never complete! At least it’s working again now – in case of problems in the future, give status.dropbox.com a check. You’ll hopefully see something different …
Enough about Dropbox though … even Google has its failures. This week, while using Google Analytics, I came across this rather “nice” explanatory message that a “dependency” had failed.
If that’s not enough, Yahoo Mail gave me another reason to sigh – just when I needed to send an e-mail, it went belly up too.
Maybe it’s the season, or maybe it’s just me. It all just goes to show, even with “cloud services”, it’s not wise to put all your eggs in one basket.
CeBIT Registration Time
Folks, it’s that time of year again – time to register for your free CeBIT Australia exhibition show floor tickets. The exhibition will be held 5-7th May this year, but this year at Sydney Olympic Park instead of the Sydney Exhibition Centre as in prior years.
If you’ve got an interest in technology, are involved in business and want to see what’s the latest and greatest from distributors, suppliers and factories direct – it’s a great place to just take a look around. It’s free – so what have you got to lose? One thing to note is that it’s not quite like the CeBIT of Hannover, Germany, or CES, or Computex – it’s a bit more low key, but the focus is more on the technology if that’s your thing.
When signing up, I overlooked a few must fill fields in my personal information and I got this rather unhelpful message – so if you’re getting “The change you wanted was rejected.”, just ensure you have filled in all fields with asterisks.
Updates – a land of opportunity?
Doesn’t it annoy you to see you have updates for Java or Flash almost every few weeks? While they’re often legitimate releases rushed to patch vulnerabilities and improve stability or functionality, they often serve another purpose – a chance to make some money.
This is an all too familiar sight – bundled offers. Why can’t they understand that we don’t want this? There are many bundled offer programs which allow for companies to make some revenue by being the “vector” which installs a given toolbar, hijacks a search engine, or installs some rather “unpopular” or questionable software of limited utility. Many of them can be easily overlooked during installation, or require meticulous clicking to ensure you don’t have any components of them installed. Experienced users will be familiar, but novice users can accidentally click on Next just a bit too fast and end up with more unpopular software burdening their system.
I think declining the option on the first install is a clear indicator that we don’t want this particular software installed on our machines. We don’t need to be re-presented with the same software every time! Some smarter ones will present several different offers, but we’re probably not interested in them either.
Maybe the only way this practice will stop is when the money stops when software vendors realize that installing software through near-deception “piggybacking” techniques isn’t a valuable way to build a userbase which actually cares about using their programs.
To CloudFlare or not?
This week, I was investigating CDN solutions, and there was an offer almost too good to refuse – CloudFlare. By using their caching solution, they will cache static data from your website at their numerous distributed points-of-presence bringing your content closer to your users and ensuring better load times with lower loading on the origin server. Best of all, you’re not charged for bandwidth, basic service is free and you can upgrade to paid service to improve options for attack mitigation.
It all sounds great, and many similar paid alternatives exist. In fact, if I were ever to serve this website from a home-connection, I will probably need the performance boost from such a solution – we see above 3Gb a day of bandwidth use here!
But ultimately, I decided it’s not the best solution for me at the moment – even if Ziphosting has issues with capacity sometimes.
The reason is that I am trusting a new third party to essentially reverse-proxy my content to my end users. CloudFlare is in a position to intercept all traffic, modify traffic or deny access to content (in fact, this is part of their DDoS protection). If CloudFlare was to have trouble, that could render the website inaccessible. I don’t have inherent trust in third party companies – they could be “in bed” with the wrong people, and introduce additional complexity in troubleshooting and support (e.g. does the issue originate due to DNS problems in CloudFlare, or in your web host, or does it come from issues with rate-limiting connections from CloudFlare?) The fact that big names use it does not provide any comfort if something unusual were to happen.
Without the right access to the server (to install modifications), the server logs which you retain will only contain connections from CloudFlare themselves and not the end users – this data could be important or valuable in some cases.
So while CloudFlare would probably improve my site’s performance for overseas (to Australia) visitors, I don’t find the risk and additional upkeep to be worth my while at this exact moment.
Tagging: It’s awful, but we need it
I’ve always been a bit annoyed at the lack of ease of navigation – there’s a lot of content, but unless you know what you’re searching for, or Related Posts have shoved it in your face, you might not even realize it’s here.
One way around this is to tag your posts – but I’ve found that tagging has many issues of its own. One of which is the consistency and specificity of tagging.
So far, I’m finding it hard to ensure either – for example, parts of certain articles may be tagged with some tags which aren’t shared by other articles in the series, or general tag terms are used initially and more specific terms are used as we go along. Filtering by the tag cloud will leave you with the wrong impression that there was no other articles about a given topic.
I also don’t want to go overboard with tagging and pollute the pool of tag terms – but we have silly instances of plural vs non-plural as well. Regardless of the shortcomings, it’s still a valuable tool, although sub-optimal. I’m hoping that users will use the search some more, and look at related posts.
This week, I’m pleased to announce that I’ve made my first contribution to Lyngsat, a very often used transponder directory for direct-to-home satellite broadcasts. I’ve provided updates for Optus D2 and Intelsat-19 which was accepted. It took a while – I’m sure the webmaster receives many e-mails about the constant changes, but he’s provided a very valuable service which I’ve benefited from, so I’m ecstatic to have been a part of it and I hope it benefits the rest of the satellite TV community.
Small excerpts below:
Multimeter Firmware Upgrade
Just the other day, I was running something with the Keithley Model 2110 5.5 digit multimeter of mine, which I was awarded thanks to element14’s RoadTest program and while downloading updated software, I came across a firmware update to version 2.00. This is the second update offered for the multimeter and is mainly a bug-fix. It applied with no problems, and was very quick to update from version 1.02.
Apple iOS Safari Image Resizing – What Retina?
As a (hobby) photographer, I often like to upload and enjoy my images at high quality. No doubt, you will have realized the quality of the images featured on this blog have improved since the early days – this is all effort I put in to ensure it remains a high quality resource.
One thing I realized was an inconsistency in the way my iPad first generation running iOS 5.1.1 and iPad third generation running iOS 7.0.4 seem to present data about images. Despite the iPad third-generation featuring the Retina Display, high resolution images loaded in Safari looked decidedly smeared.
Loading my test image, taken with my D3200, measures exactly 6016 x 4000 pixels. Interesting to note is that loading the same image on the iPad first generation results in it displaying a resolution of just 1504 x 1000 pixels.
But this is the stinker – the image looks poor on both Apple devices, and is clearer on the desktop browser and Android 4.4.2 Chrome browser. It is the exact same file. Just see the comparison:
Desktop (approximately the same view, 100% zoom)
iPad 1st Generation (maximum zoom level)
iPad 3rd Generation (slightly more zoomed in, not maximum zoom level but higher zoom reveals no new detail)
Google Nexus 7 (approximately the same view, 100% zoom)
A quick investigation revealed that the resolution of 1504 x 1000 pixels is all the iPad will decode as it uses JPEG subsampling at 1/16th resolution for all images greater than 2 Megapixels. I think this is a very poor practice – while it may improve performance, it definitely lessens the appeal of the retina screen and possibly limits the ability of websites to present high resolution content correctly.
- The maximum decoded image size for JPEG is 32 megapixels using subsampling.
JPEG images can be up to 32 megapixels due to subsampling, which allows JPEG images to decode to a size that has one sixteenth the number of pixels. JPEG images larger than 2 megapixels are subsampled—that is, decoded to a reduced size. JPEG subsampling allows the user to view images from the latest digital cameras.
I’m quite disappointed to be honest, I was expecting to see the detail in the leaves …
Unfortunately, this week, the prognosis is not good. It was exactly two weeks to the day of my first steroid injection and the cyst has definitely recurred. As a result, I’m back to a no-walk state, at times, with dull pain around the clock that gets markedly worse on rotating or applying pressure to the ankle.
I’ll wait and see if it recovers on its own – and whether the other cyst recurs before contemplating a second (and possibly last injection) prior to organizing to see a specialist and schedule surgical intervention (which will necessitate even longer recovery times, without much guarantee of improved probability against recurrence).
It’s a slow road to recovery – I really miss being outside, talking to people, or even just watching things go by. It’s been literally months since I’ve had some time to myself outside … it gets a little lonely. That being said, keeping this blog going helps keep me going.
Micro SD Adapter and SIM Pictures
To round off a relatively “long” but non-exhaustive random post, here are some items which I’ve had on my desk for a while which deserves a picture or two. The first is a microSD to SD adapter.
As you can see, this is an entirely passive adapter. One of the pins is entirely unconnected. I’m sure there are many pictures of this sort lying around – but there’s my rather dusty one from my desk.
Finally, here are pictures of the actual substrate which is contained in mobile phone SIMs. As I had a few lying around that had expired, and the practice of “trimming” SIMs to fit micro or nanoSIM form factors is relatively common, it’s easy to show why this is the case. By carefully bending the plastic carrier away from the contacts (one got damaged), it is clear to see that the contacts and thin PCB substrate forms the smart chip itself.
The left one is a more modern 6-contact SIM, whereas the right one is an older 8-contact SIM which is based on the smart-card standards. The irony is that the two extra contacts aren’t used, however, due to the way the substrate is made, those 8-contact SIMs are not suitable for trimming to nanoSIM size. They can be trimmed to microSIM, but will cut very close to the edges.
When we turn it over, we can see the silicon die for the SIM itself as the tiny “square” in the middle, with bonding pads in a line in the middle (rather than at the edges, which I find strange). The chip is “gob topped” onto the contact substrate with clear adhesive which makes it possible to see the bond wires which are arranged radially. Just five of the six contacts are actually bonded.
Technically, the SIM *could* be made even smaller … but I’m not sure we’d want to.
That wraps up another week in Gough’s rather busy head. More posts forthcoming when time, motivation and energy permits!