Category: Uncategorized

Acer C720 Chromebook: Initial Thoughts

Posted on by lhl

Last year I ordered a Samsung ARM Chromebook to test out as a developer device. The appeal would be a portable low/minimal maintenance low-power device I could use mainly for terminal/browsing work (I tried that unsuccessfully a couple years ago with a Toshiba AC100).

Unfortunately, it turned out the Samsung had poor battery life and a worse screen. Making it useful (a shell with vim and git) required dev mode (CTRL-D every time booting) and either installing chrubuntu, which had power/input issues, or a then very immature crouton (it’s now much better). The Chromebook didn’t do anything better than my MBA and had some significant drawbacks for development. The screen/performance also was a hinderance for general browsing/reading (my iPad was much better for that). A coworker eventually took the Chromebook home and found it useful/pleasant enough for that purpose though, so YMMV.

Given that experience, why buy another Chromebook? Well, it turns out one of the more annoying features, the secure booting, is actually fantastic when you’re worried about infosec when traveling. For an upcoming trip, I wanted a bulletproof system to be able to do basic browsing/web tasks and for copying (and maybe editing?) photos without worry.

There are a few new Chromebooks coming out, and I was looking at the HP Chromebook 11, which has a slick design, a great IPS screen and neat features like micro-USB charging, but I ended up going w/ the C720 due to the much better processor (Haswell Celeron w/ 4GB RAM vs Exynos 5250 w/ 2GB RAM), SD card reader, USB 3.0, and most significantly, the much better battery life (9h vs 5h).

It’s a shame that these ARM Chromebooks have such poor battery life when when compared to their tablet brethren using the same SoCs, but what can you do.

It’s still early going, but here’s what I can report:

  • The C720 is definitely much peppier, and browsing (scrolling, multiple tabs, YouTube) is smooth where w/ the Samsung ARM Chromebook, it was not, so that was a good call
  • Battery-life seems to be living up to its promise. I haven’t used it enough to run it down, even while installing software. It sleeps and wakes up well
  • The screen, an 11.6″ 1366×768 matte TN screen isn’t terrible (it’s brighter and a bit better to my eyes than the Samsung ARM Chromebook), but the blacks/contrast are mediocre at best and using the default black terminal and white web pages isn’t very comfortable, although w/ some tweaking I’m sure it’ll be workable
  • There’s now a file browser which means that I can copy photos from SD card to a USB 3.0 external drive, but I haven’t explored actually doing photo editing much. I doubt there’s anything with built in RAW support though, so I believe I’ll end up having to run in developer mode to install LibRaw, digiKam, or RawTherapee.
  • As mentioned, crouton is much improved, although I still wouldn’t pick a Chromebook as my first choice for development. Working in a Secure Shell window, I’m still missing basic stuff like having a decent clipboard, although the fix may be to succumb to running a full X/WM setup.
  • Although less secure in dev mode, you can still get some boot protection w/ verified boot with shell access. Just make sure that you have some good passwords for sudoing/your chroot. You might find reading the Chromium OS Security Overview to be worth reading. (In general, the design docs actually make for pretty interesting reading)
  • I’ve yet to setup my proxying/pac files, but you can see what I’m planning on my security plan page. (better safe than sorry)

Reference:

iPhone 5s LTE Bands

Posted on by lhl

As someone who does a fair amount of international travel, looking at the iPhone 5s’ LTE support of the different models is pretty interesting. The thing that popped out as possibly the biggest choice is the A1533 (T-Mobile unlocked version) vs the A1453 (Sprint). The A1533 supports Band 1-8, and 13, 17, 19, 20, and 25. The 1453 supports all of those and adds 18 and 26.

Based on the fact that you simply get MOAR bands and that its an option that’s available for me I initially was tending towards picking the 1453, however, looking at Wikipedia’s list of LTE networks, it turns out those additional bands are not so useful. 26 is only used by Sprint and KT in South Korea (KT also supports band 3 and 8), and 18 by au/KDDI in Japan.

Basically, not very compelling additions, so may not be worth the trouble. By comparison, the Nexus 5 will support band 2, 4, 5, 17, 25, 26, 41.

2013 MBA Battery Life

Posted on by lhl

I’ve been using an older (2011) 11″ MBA with very poor battery life (~2hrs?) and have been very much looking forward to upgrading to a Haswell MBA for quite a while.

TLDR: battery-life is much improved, however falls far short of the claimed 9 hours for my 11″ MBA. I get about 6hrs max (6hr 12min) and in regular usage, less.

The specific model I have is the top-of-the line MacBook Air “11-inch, Mid 2013” with a 1.7GHz Core i7 processor, 8GB RAM, and 512GB SSD. Here’s the iFixit teardown. Anandtech did a recent test comparing Core i5 to Core i7 battery life (the Core i7 is very slightly better under low workload, and is 13-18% worse under medium-heavy workload).

The battery is spec’d by Apple at 7.6V and 5100mAh – 38.75Wh, 8.08oz. The battery had a Cycle Count of 6, a Max Capacity of 5326mAh (4.4% above spec) and an initial voltage of 8.545V – that’s to say it’s in better-than-spec condition.

Display was set to 75% for testing with the keyboard backlight off, wifi on, screen sharing off, SSH on.

Background apps include: Dropbox, iStat Menu, Evernote Helper, MiniBatteryLogger, and memcached (64MB, not being used).

The 6hrs was spent almost entirely with light web browsing w/ relatively low CPU/disk/network usage. It’s too bad there isn’t anything like Android’s battery/activity tools built into OS X (I suppose you could script something to sample/aggregate activity but that’s sort of beyond the time I really want to spend on this since I can’t really do much w/ that data).

I used MiniBatteryLogger 1.8.5 for getting battery stats and for graphing. Also here’s some clips at the end from iStat Menu:

For those interested, here’s the raw battery data export.

UPDATE: MacWorld also did a comparison of the i5 vs i7. Macrumors has a very long 2013 11″ Battery Life Thread (real world)

UPDATE2: The Verge published an 11″ review where they claimed to get 9-10hrs+ of battery life (cycling through popular websites at 65% brightness). gnoring any critiques on the battery testing methodology, I really wish I could find the disparity – getting that sort of battery life would be really nice… Things that could possibly be responsible for the difference:

  • Actual browsing uses a lot more network than refreshing a series of mostly cached sites
  • FileVault chews up a lot more power
  • Wifi signal/AP distance/type has significant difference in power consumption
  • Some long-running process, either ClipMenu, iStat Menu, Dropbox, SSH is causing my laptop to function way less efficiently than it should.

Basis Watch Report

Posted on by lhl

Next in my gadget backlog series…

Basis B1 WatchI had preordered a Basis B1 Watch a long while ago and promptly forgotten about it, so it was a bit of a surprise to find one sitting on my desk after SXSW (which has turned into this amazing construct of pure marketing – good for business though, I guess).

Since then, I’ve worn it almost every day (about a month-and-a-half). I figured I’d give a report after living with it for a while since Basis is now apparently starting to sell them for real (after being back-ordered for a while).

While I have had reservations on the watch form-factor after getting a MetaWatch and realizing in the intervening decade since I regularly wore one that I now couldn’t stand wearing watches, I’ve mostly re-adjusted, and the data the Basis collects has been worth overcoming the annoyance factor.

There have been a couple reviews (the Verge one isn’t a bad summary), but here’s the basic rundown of what you need to know:

  • Price is $199. It comes exquisitely packaged in a die/laser-cut box and overall all the industrial and product design is quite nice (package, device, site/app). The watch is less bulky than the typical smart watch and is pretty inconspicuous. The capacitive touch buttons work well and the display is simple but more than adequate.
  • Unlike almost all devices currently on the market (besides the BodyMedia devices), the Basis is a multi-modal sensor device that’s more than just a glorified pedometer. Sensors include:
    • Heart-rate via green LED optical flow-based sensor (I assume using something similar to the TI AFE44xx but sadly, there’s no current support for SpO2 or glucose measurements); An important note is unlike the similarly kitted MIO Alpha, Basis specifically notes that heart rate measurements are not suitable for exercise monitoring
    • Skin Temperature (accurate to the tenth of a degree) – mine seems to be about 92-94 degrees fahrenheit when I’m up and a few degrees lower when I’m asleep
    • Perspiration specified by μS/cm, so it looks like it’s using galvanic response to calculate? The Basis itself seems to be fairly well waterproofed, and I’ve used it in the shower w/o issues (although usually not, because that’s weird, right?)
    • Pedometer and sleep tracking via a 3-axis accelerometer
    • Calories burned are derived from BMR via Oxford equations multiplied against a “physical activity level value” that’s presumably derived from the sensor data collected. Note, if you have a better BMR value (ie, I have an RMR estimate from my lean body mass from my Bod Pod measurements, you can presumably use Table 5.2 to reverse calculate what weight to enter to get more accurate estimates)
  • Battery life is good for a few days (3-5?) and you need to use a custom charger to recharge. This is a bit of a pain actually, especially if you’re traveling – the charger is a plastic band that hasn’t broken yet, but is just begging to be and will soon enough, I’m sure.
  • This charger also syncs data from your device. This is a really important point. While the Basis has Bluetooth (2.1 eww) built-in, it doesn’t work yet and the USB sync is the only way to get your data. Also, there are mobile apps (Android first, then iOS?) on the way, but again, these aren’t available yet. Personally, I wouldn’t recommend this to most end-users until this gets sorted out.

The web interface and the graphs that it gives you is actually quite nice. For me, the actual (almost) dealbreaker wasn’t the lack of wireless/mobile syncing (wireless power and data sync is definitely very high on my list for want-to-haves, personally), but rather that while there’s been the promise of an API/export for quite a while, there’s actually nothing (and no roadmap!) yet.

Paul Miller wrote a great piece on this earlier in the year about this, which I vigorously agree with and I won’t buy or support any self-instrumentation device that doesn’t give me full access to my own fucking biometric data.

Now, if you’ve read the last two paragraphs and are scratching your head on how/why I still have my Basis watch, well, that’s simple: there is an API, it’s just not publicly documented. If you’re an end-user looking for export, yes you are SOL and should send your Basis back and demand a refund if it’s important to you (and it should be!).

If you are a developer, simply pop up your Javascript console and look at the XHR calls to /api/v1 for nicely formatted JSON. Note: these calls are actually totally unauthenticated. That’s bad. However, since the tradeoff is easy access to my data w/ low likelihood of accidental leakage (data URLs have hashed unique identifiers), I’m ok with that for now.

Overall I like my Basis although I have an Amiigo on the way that, if it functions as promised, should be better than the Basis in pretty much every single way (sans telling time/wrist display – but that’s not so necessary anyway w/ a proper mobile app). It’s also half the price. For developers and data geeks, the Amiigo also promises to have a full SDK and actual developer support (although I hope they have their data as nicely formatted as the Basis data – it’s totally sweet).

However, the Basis is well designed, and if you’re absolutely buying something today, the Basis is by far and away the best device for those looking for serious self-instrumentation (ie not Fuel Points). The habit system is also on the right track, as well, in terms of general usefulness. However, as outlined, the B1 is also a seriously unfinished product at the moment and anyone expecting basic fitness-tracking functionality (like a mobile app/wireless syncing) will be disappointed. Early adopter caveats definitely apply.

It’s a pretty interesting time for fitness/activity trackers – I think we’ll soon be reaching a point where the sensor-suite will be more than good enough (as mentioned earlier, photometric glucose measurements and SpO2 should be possible w/ the existing sensor packages, and having wireless charging would be the last really nice bit), and the real value will be integrating with other data to create an integrated picture and assisting in behavioral modification/self-improvement. It’s also interesting seing some of the vertical applications as well. The LIT for example looks pretty neat.

UPDATE: Looks like a QS guy also reviewed the Basis and also spotted the JSON feed, and better yet, published some instructions and some code on Github for people.

UPDATE 2: I have a bunch of notes on wearables and smart watches on a hackpad

UPDATE 3: Current biggest discovered annoyance is that it’s impossible to change the Time Zone w/o going online to set it then syncing. WTF. Does no one at Basis ever fly?

Summary Mode

Posted on by lhl

While there are public tools like Storify that do an OK job for tweets, and I personally end up using Evernote or for most of my (private) tracking of topics, it’d be useful if there was a good tool for collecting/collating and snapshotting primary sources and making notes/comments on any particular topic (and groups of topics). Useful things would include seen-on and publish dates (a la Zotero etc). And being able to contextualize/parse interesting pieces (Evernote and Clipmarks do a pretty good job with this)

For example, for the controversy of the past few days on the whole PyCon fiasco:

* Summary: How “dongle” jokes got two people fired—and led to DDoS attacks

A series of discussions on Hacker News related to the topic:
* 2013-03-17 Inappropriate comments at pycon 2013 called out
* 2013-03-20 The PyCon Incident
* 2013-03-21 PyCon Code of Conduct changed to avoid public shaming
* 2013-03-21 SendGrid Fires Company Evangelist After Twitter Fracas
* 2013-03-21 Adria Richards, PyCon, and How We All Lost
* 2013-03-21 A Difficult Situation

RIP Aaron Swartz

Posted on by lhl

I was first introduced to Aaron (impossibly young), over a decade ago at a tech conference (OSCON?). And, while we were never close, we often floated in the same circles (tech, activism, civic and political tech) and over the years our paths crossed many times, in emails, projects, at conferences or meetups. The last time I saw him was in Boston, June 2010. We met up outside a food court in Cambridge and caught up on the projects we were starting/wrapping up and swapped some thoughts on civic and campaign tech.

More than a friend, Aaron Swartz was a fellow traveler. He was one of us. In many ways, the best of us. It was a punch in the gut when I read the headline last night. He dedicated much of his life and his many talents in fighting injustice and trying to make a difference.

And beyond the sense of loss, there’s a bitter taste that injustice and indifference has won the day.

Rest in peace Aaron Swartz.

Raw Nerve – some of Aaron’s best writing.

F2C2012: Aaron Swartz keynote – “How we stopped SOPA”

Fujifilm X100 First Impression

Posted on by lhl

I’ve been looking forward to the X100 for months, and have been in and out of a couple X100 forums, so it’s been interesting to finally have one in my hands now. (Actually, due to the way preorders overlapped with shipping, I’ll have two, but obviously one is going back, unless someone in LA wants to pick one up).

It’s still early days – I’ve only gone out shooting once (I’ll be heading out shooting tonight as well) and I’m still getting used to how it handles. I know it’s capable of capturing fantastic images, but coming from shooting tens of thousands of frames on Canon DSLRs over the past decade, adapting to the X100 has been a bit of a challenge so far.

The lack of rangefinder patch, focus screen or some other equivalent is killing me right now (the EVF is not my idea of a good time), but maybe that’ll pass. That being said, I’m actually looking forward to shooting the heck out of this thing. I’m sure it’ll make me a better photographer in terms of thinking about what I’m shooting, estimating distances, and thinking about DOF. It’s certainly not a P&S, and I’ll no doubt miss a lot of shots, but I plan on trying to carry it everywhere, so we’ll see how it goes. (I’ve ordered a Luma LoopIt strap.)

For those considering the X100, I’d probably have to say that this isn’t the camera you want to take with you to evening social events, not unless you really want to work for your photos. While the bright lens (F2.0) and amazing High ISO performance (totally usable at 3200) and quiet leaf shutter would seem to make it great for low-light candids, the AF seems more finicky than average, and sadly, the fly-by-wire focus ring is pretty much useless right now.

That being said, I plan on working primarily in MF mode and zoning, so we’ll see how that works.

Happy Accident

Also, there are a lot of niggles in the firmware. On the bright side, these could all be fixed and there seems to be some noises from Fuji that they will work on the firmware. On the negative, that historically hasn’t been the case. I’ll be keeping a running list of things I’d liked fixed here:

General

  • Startup Times: I’ve managed to get it starting up fairly quickly (about 1s?) but out of the box it took so long that I kept thinking it was broken.
  • File writing: file writes seem to block operations? Is this 2011?
  • ISO settings don’t persist across modes
  • Even in full manual mode, the aperture doesn’t seem to be physically set, so there’s shutter lag. More of an annoyance than a dealbreaker
  • Yep, the battery goes from “low” to shutdown in < 5min; in general power usage/management seems to be rather poor. There must be a better compromise for responsiveness/power usage. Canon DSLR powerpacks are about the same capacity and last forever.

AF

  • My current #1 focusing issue, at the cause of most of my misfocused shots is the OVF AF parallax. The FW fix that would help this greatly would be to recompose the AF point with the focal plane (just move it along with the framelines). More discussion here.
  • Low light AF; is this the best it gets?
  • The focus-frame size can only be adjusted in AF-S; I’ve find it enormously useful if it could be adjusted in MF

Focusing

  • The biggest problem is the focus ring doesn’t work at all for covering distances; it’s not helped by it being incredibly laggy as well. This should be an easy fix with a better acceleration curve, and allowing that to be user adjustable/distance-dependent
  • In MF, since the command lever is used to enter EVF zoom mode, wouldn’t it make sense to allow the lever to be used for focusing as well? Actually would be a lot less finicky than the ring; you could also use one or the other for rough/fine focusing

Buttons

  • It’d be nice to reassign the RAW button (I’m always shooting JPG+RAW, I’d much rather be able to assign that to movie mode or the ND filter)
  • Macro should just be a toggle instead of popping up a menu – it’s either on or off! Same w/ ND – these are probably my biggest usability quibbles after the first week.
  • Auto ISO should be a toggle in ISO menu (so you can easily flip it on/off in the OVF fn mode); actually, for night shooting, being able to toggle Auto ISO off even easier (long-press?) is even more important; it really screws with you when the metering if off by more than the exposure compensation…

Video

  • Manual focusing while shooting. The AF is especially clueless when shooting video
  • It’d be nice to be able to set the fn button (or the RAW button) to go directly into recording

OVF

  • The histogram doesn’t seem to be right? It doesn’t reflect aperture or shutter speed changes?
  • The range scale is good, but would be nice to show the hyperfocal distance as a marking (or as others have suggested, a mode/way to jump to the hyperfocal distance like the GRD)
  • In the post-capture EVF review, it’d be nice to have an option to show brights/histogram; being able to have a 1s or even 0.5s preview would be nice as well

I’ll be trying to post more regularly to my flickr account.

After a couple weeks with the X100, I posted an updated list on Engadget.

SXSW Music 2011 Playlist

Posted on by lhl

I haven’t been very diligent listening to the SXSW torrents this year, but my schedule is already plenty packed: my current SCHED. (My SXSWi is much less structured; primarily planning on catching up with old friends).

Anyway, before things get officially crazy this week (I head into Austin on Thursday), I made a little share a quick and dirty playlist of some of the bands I’m most excited about catching live:

Bonus Track: Not playing at SXSW, but I just stumbled on a great track that Quantic spun at his set last year:
Nicolas Jaar – Mi Mujer (Wolf + Lamb 2010) by Clown and Sunset

Installing Windows 7 on a Mac Pro via USB Key

Posted on by lhl

First of all, holy shit this was a miserable experience. The tl;dr version is that on a 2008 Mac Pro (2 x 3.2GHz Quad-core) running a freshly installed OS X 10.6.6 with rEFIt 0.14 I couldn’t successfully install a Windows 7 x64 ISO from USB.

To successfully install Windows 7, I ended up driving out to buy some DVD-Rs and burning the ISO in a Win7 VM (VMWare Fusion 3.1.12) installed from the ISO via Boot Camp. If you are trying to install on a machine without a DVD drive, I suggest either simply using a VM or if you must Boot Camp, acquiring an external drive.

Additional caveats: once I was able to boot was that the Windows 7 installed did not recognize the BOOTCAMP NTFS partition, and the after formatting, threw an apparently common Error 0x80300024 on install (the solution for this was to shut down and pull out my 3 other drives out so the Windows installer didn’t get confused).

At the end of the ordeal, I ended up with a working Windows 7 install that used 28.4GB of space. o_O

Disabling components actually only increased usage. There is a tool called DISM (must be run as admin) that can manage packages, but appears to treat almost the entire Windows 7 install as one big package. The only way to strip Windows 7 down seems to be using a tool like RT Seven Lite to create an ISO that removes the packages entirely beforehand.

Note: if you need to zap the BOOTCAMP volume during the Windows 7 install, you’ll need to grab the Boot Camp drivers from the Snow Leopard CD or from BitTorrent (Apple doesn’t offer 3.0 drivers standalone, and 3.1 and 3.2 require 3.0 to be installed) and run those manually. (You may need to do this anyway as apparently Boot Camp 3.0 may fail by trying to installing the x86 instead of the x64 drivers).

In case you didn’t get it by now, this experience has been a pretty convincing argument for continuing to stay away from Windows as much as possible, certainly on Macs where the unholiness combination of EFI, Boot Camp, and Windows just makes a complete mess.

For your edification/amusement/horror, additional things that did not work on my Mac when trying to install from USB:

dd
The recommendations to use dd if=$iso of=/dev/$usbkey -bs=1m is a total fail as it’ll create a UDF image on the USB key, which apparently the Mac’s EFI won’t detect
Windows 7 USB/DVD Download Tool
I had an older Windows XP partition, so I was able to try Microsoft’s ISO tool. This created an image that was bootable in the sense that rEFIt could see and select it, but subsequently did not show up as a boot device (DOS prompt for inserting a bootable disk)
UNetbootin
Another try once I started down the WinXP path, I tried using UNetbootin to make the USB stick bootable, but to no more success. Same result as the Microsoft ISO tool.
WinToFlash
Next into my increasingly desperate descent into madness was to use an app that could write Vista/7 boot procedures using EFI. Unfortunately, these resulted in an “Error: Unsupported while loading bootmgr.efi” message when booting in rEFIt
Windows 7 in VMWare Fusion + DISKPART
Finally, I decided to install the ISO in VMWare Fusion on OS X, and then follow this method with DISKPART (you can’t do it in XP because it’s version of DISKPART doesn’t recognize removeable drives as disks). This also failed spectacularly (also, trying bootsect), at which point I threw in the towel on USB booting.

UPDATE: Here’s a method of getting BootCamp Assistant to create a bootable USB Windows installer that might work, although I haven’t tested it at all and it’s probably also dependent on how old your Mac is if it does work…

UPDATE 2: On a recent PC, I was able to make a bootable install disk via Linux which may be helpful, but may not be applicable w/ old Macs w/ 32-bit UEFI.