Chrome OS for web mapping delivery

27 Oct 12 Chromebook usage intensifies — blogging suffers


It’s surprising to note how quiet my writing has been since August.  The Cr-48 did a fine job on the road, receiving an update and returning to functional use of Verizon 3G along the way somewhere in southwestern Colorado. A few days later, the Cr-48 took in a beautiful sunrise from the Cal-Neva Inn on Lake Tahoe, and then came home.  There, things got intense.

But it wasn’t for work or practical use!  It was the diversion of reading free Manga from a variety of web sites.  Where chapters might have page counts between 12 and 30, from the time the Cr-48 got home until now it’s served up 640 chapters of one, 340 chapters of another, and 250 chapters of another (which averaged 32 pages per chapter).  In short, the Cr-48 has had more hours of use as a content delivery system than ever before in the past three months!  Yes, the hinge is a little wobbly on the right-hand side, but the machine gets taken everywhere but in the shower, so it gets set on its side at times.  It’s also gotten regular treatment with sanitizing wipes.  ;^)

Fact is, although the usage is sort of monotonous (it’s mostly been serving up content through free websites) the physical treatment has been intensive.

Cool apps have appeared like Camera, and the upgrades to system have been very regular.  I continue to use dev-Channel for Chrome OS.  As of today’s update, I’m up to

Google Chrome 24.0.1305.3 (Official Build 163672) dev
Platform 3083.1.0 (Official Build) dev-channel x86-mario
WebKit 537.16 (@132013)
JavaScript V8 3.14.5

Holy cow, that is soooo much easier to share than it has been for the past 22 months!  I’ve been quietly appreciating and not blogging about the UI updates that have become a steady stream as Chromebooks become more properly commercialized.
Where in the past it’s been a copy-and-paste exercise for each feature, my favorite status items to share have been stood up on a new
chrome://version/ page and the information there can be copied readily.

Other new stuff that I really like:  On the login screen, a pressure-touch on the touchpad will focus on user for password input, rather than the full button click.  The concentration of system status icons in the lower-right is evolving, and I look there with the same frequency I do in Ubuntu’s Unity interface (in the upper right) on my main workstation at home.

The collection of favorites along the lower left reminds me of system tray items that I look at in Windows 7, and they are really useful to the extent that they’re my most-favorite destinations.   The nine-square icon down there to launch an apps pop-up reminds me of the apps button in Android—and it even has scrolling windows that mimic a graphically flattened version of what I see in Jellybean.

It’s not purely Chrome OS, but the performance of Google Drive has ramped up noticeably in the past three months as well.  I am using G Spreadsheets to handle much of the logging tasks in my regular work that involves spatial data engineering.  In that, multiple tasks are set up and run, and progress or failure noted for each workflow thread.  At capacity, I make use of 15 threads on a (Windows 2008 Server) server through RDP and 12 threads on a (Windows 7 Pro) workstation with a dual-head display.  The problems that need to be solved are vague and mostly related to capacity issues of the ESRI Desktop software that I’m using, so I never really know what will work when I start out, and then I break things up in a quadtree to make them small enough to fit through all the limitations.

Tracking input data, output data size as items make it through a work flow, and logging completion is quick and easy.  I’ve been a major fan of Excel keyboard shortcuts since the mid 1990s, when I had to learn them to work spreadsheets on laptops during airline flights; to this day they continue to be so fast that its actually painful for me to watch people mousing their way through a spreadsheet.  Of course, having all my favorite Excel-style keyboard shortcuts working and functional within G Spreadsheets just melted my heart—Spreadsheets are my favorite Google Drive item at this time.  (Presentations will return to the fore when I’m back teaching next semester)

Did I mention that on my 12-thread Windows 7 workstation, I can launch Chrome Canary browser, then open Drive, and then open my G Spreadsheet and make a new entry within it faster than it takes Excel 2010 to launch on my workstation?  Well, I can. YMMV.

Details aside, I’m using G Spreadsheets for more than 80% of all my spreadsheet uses now, and that’s because I started using them with Chrome OS.  I’m using G Documents to contain text that needs to get pasted into web apps, so that I avoid the heartbreak of lost writing and also to keep a record.  I expect to return to use of G Presentations to set up class slides when I get back to teaching next semester.  And finally, I’ve make use of G Forms for something completely different: the Household Technology Grant Program (HTGP).

The HTGP is how I’m dealing with the chatter around the house by certain individuals of diminutive stature who now desire Chromebooks of their own.  Using some questions adapted from a college technology grant form found on the web, we simply ask the interested parties to fill out the form and justify their desired technology.  Without specifying how much support they might be getting for any acquisitions, in the attractive and simple-to-format G Form they spell out the Hows and Whys of their need for funds to augment our household computer herd.

It’s been a blast watching the Chromebook advertising campaign as we keep an eye on baseball’s World Series progress.  For as long as I’ve been pounding away on the Cr-48 it’s seemed as though $250 would be a catalyzing price point for a browser-centric system.  But in the past three months, I’m excited about Chrome OS’s interface evolution more than just the price.

The assemblage of UI styles from Unity, Windows 7, and Android reminds me of that long-ago time when I saw an amalgamation of desktop shortcuts and UI phrasing in NeXTSTEP.  I really liked it then, seeing how shortcuts from the Mac and early Windows and X desktops were all there on the NeXT screen at the same time.  (And in a first for Steve Jobs, there were two mouse buttons!)  Sure, Chrome OS is not as complete an OS as NeXTSTEP, but my point is that there was strength in that amalgamation.  Anyone who doubts that need only examine OS X and start counting the NeXTSTEP features that have persisted for 20 years.  I think that there’s something very solid about Chrome OS’s UI changes that have taken place in the past couple of months.

So I’m getting the sense that Chrome OS is taking a polish that suggests its ready for the larger world, and I’m seeing devices for sale that hit the sweet spot where even those who (probably wrongly) imagine that they use browsers only 30% of the time would still see the economics work versus a $1000 MacBook.  Perhaps?  If not, then how about versus a $1600 MacBook?  ;^)

09 Jan 11 Four Weeks With a Cr-48 Chrome OS Notebook

Today finishes the fourth week that I’ve had a household Cr-48 notebook.  It arrived on December 9th the very afternoon after I’d filled out a late-night web form describing my interest in exploring Chrome OS as a medium for delivering web mapping applications.  Apparently that, and perhaps being the first in my Zip code to ask for one, was a good enough reason for Google to send an early Winter Solstice present.  I hadn’t even mentioned to anyone else in the house that I filled out the form, and so I didn’t hear until Friday morning that a box had arrived for me the previous day.  Oh, the magic of overnight delivery!

As a mapping person, I’ve been a big consumer of custom workstations with industrial operating systems and rather generous memory, network, display and storage resources.  When it came to mobile devices, for 15 years I have been disdainful of the value of laptops relative to the same money spent on a tower-sized workstation—at least when it came time to perform data development, analysis, or bulk processing.

The last laptop that I bought for myself was one that I’d waited years to see: the glorious Toshiba 100 MHz Pentium with 24 MB of memory and CD drive, circa 1997.  Sure, I’d used, enjoyed and occasionally admired the various corporate devices I was loaned.  My Toshiba logged quite a few trips to Europe, and went to the field around the US including Hawaii.  For several years, it was the last device I had with a serial port that let me download GPS tracks from my Garmin Vista.  But after 10 years of use, it finally went up on the bookshelf and hasn’t been down since. In the past couple of years, I’ve seen the value gap between laptops and desktops shrinking.   From 1990–2005, it seemed that a very decent workstation could be had for about US$3000 in any given year.  For much of that time, a higher-end laptop of the same price would have fairly diminished processor power, storage, or battery life relative to its tower-sized cousins.

But a couple of years ago the price of all components seemed to start dropping at once, and processor power had become so saturated once the ~3GHz clock speed was reached, that things just started getting cheaper.  So now I see very decent workstations for about $1000, and modest laptops at half that price.  This price drop actually turns attention away from hardware, and where should it turn to but the OS and the actual use case served by that interface. In the past five years, mobile devices have become recognizable computers too.  At the end of the dot-com boom, I experienced the heartbreak of being laid off; that very day I got myself a mobile phone of my own to replace the corporate one I’d been relieved of.  I went from an iDen team-player special to a more basic Motorola feature phone on Verizon so that any potential employer could reach me.

When I moved across the Bay, the best coverage in the Marin countryside had been inherited by ATT, so I switched carriers and in 2006 bought an unlocked Palm Treo 650.  With that GSM/EDGE phone, I was able to enjoy Google Maps when visiting Hawaii and in places like Netherlands and Romania I could grab inexpensive pre-paid SIM cards and get a useful local phone number while visiting. When I changed jobs two years ago, I continued to be happy with the Treo 650 until one fateful day its use case caught up with it.  In August 2009, an air conditioner in the office started pouring condensate out of our drop ceiling, and I moved some furniture around to mitigate some of the damage.  In the process my shoe soles were unusually wet, and I strode out into our tile-floor lobby to contact the building manager and slipped—Treo in hand.  Well, the device gave its all, breaking my fall, but it was fatally damaged by having its charge plug broken out.  The phone was fully functional, but could never be charged without repair.  It was at this time that I admitted to myself that I really wanted to try out an Android phone. Ah, it was a simpler time, 16 months ago… For there was but one Android phone to have, the G1, and it was good.

T-Mobile wasn’t my first choice of carrier, but by having a GSM network, I was willing to give them a try.  Like I had a choice!  At first, they were a bit less coverage than ATT here in Marin, but now they have become much closer to comparable.  So suddenly I had GPS in the phone, and I no longer needed to fix the Garmin Vista that had been battered on bicycle handlebars a bit too much.  When I learned how easily I could share my travels with the MyTracks application, the whole business of downloading GPS files, processing them into GIS form, and doing stuff with them seemed so barbaric—relative to uploading and sharing the link to Google My Maps.  To save monthly costs, I eschewed any SMS plan; instead, with Google Talk I could still chat with most people who I wanted to reach that way.  A bit later, I installed Google Voice and very quickly got used to having voice messages transcribed and e-mailed to the phone.  Add in a WiFi sniffer app, and I quickly got the house set up with a wireless router so I could watch YouTube videos from any room.  Lots of nerdy fun.

So its been quite an experience for me to watch the expansion of Android adoption since then.  I sure knew that much of all the fun stuff I wanted to do on my phone involved free apps and wireless service that was $55/month.   Seems like I wasn’t quite as cool as some slick folks with an iPhone, but I had more than enough to play with, and it wasn’t sucking quite so many dollars out of my wallet each month as Apple folks get soaked for.  I still recall a conversation with a Linden Lab employee who noted that in their company, “all the engineers have Android phones, and all the marketing people use iPhones.”  It still rings true today, only the divide is edging closer to (and leaving less room around) the Apple camp.  Months ago I noted that Android new activations in the US had passed up Apple.  Now I note that total installed US base of Android phones has exceeded Apple.  I still like the feel of the Apple phone interface, but I view it with the same skepticism that kept me from purchasing a laptop in the past 13 years: I see the iPhone as poor value relative to competing choices. And so, with long deeply personal description of the background I bring to this tale, I find myself today blogging from wherever I feel like (except when I’m driving) thanks to a new device.

It’s too big to be a phone, and don’t call it a netbook, but it is not nearly as big as the laptops I see most people around me using.  The Cr-48 has a 31cm diagonal screen, and a keyboard that is a worthy facsimile of a 13-inch MacBook Air.  The machine weighs in at 1650 grams, and has a fairly beefy hinge on its screen.  When I’ve picked it up or handed it over to people to check out, I often pinch it with my thumb in the apex between keyboard and screen, and three fingers behind it, passing it around with one hand.

It’s no tablet, but it’s not that heavy.  I’ve gotten more than seven hours of use on its battery today, and the icon is just starting to turn red.  I can blog till the cows come home, and type just as fast as on a full-size workstation.  The keyboard is a very clever design, the best new design I’ve seen since the NeXT machine in 1988.  All the keys I want (with the exception of a numeric keypad option) are accessible with combinations of Ctrl- and Alt- and those two keys are available in oversized form near my left thumb.

I’ve taken the Cr-48 to the AGU Fall Meeting in San Francisco and happily used it among thousands of other laptop-wielding earth scientists.  I saw scores of netbook users, many of them grimacing and squinting, hunched over too-small keyboards while watching a too-small screen.  I saw scores of USS Enterprise-sized 19-inch laptop users who were generally clustered around power outlets.  I saw a few happy Mac users with 15-inch screens who were likewise looking for power outlets, and all the while I felt very balanced, knowing that I found my keyboard big enough, my touch pad kind of large with its fledgling multi-touch senses starting to develop, and an ability to graze WiFi or not sweat it and use Verizon EVDO 3G.

My screen requires me to use reading glasses, but I need those anyway these days for my workstation screens. The app experience has been fantastic.  Before the Cr-48 arrived, I made occasional use of Google Docs, but now they make sense in completely deeper ways.  In the past month, I’ve been regularly collaborating on documents and spreadsheet with a colleague at work who is a part-time hire and who does not have a fully personalized e-mail account. They do, however, have an Android phone and make regular use of Gmail.  After watching my Google Docs collaborators type their changes in real time on my screen, whether that be a workstation browser or my Cr-48, I find packaging an MS Word document as an attachment to an e-mail composed in MS Outlook and sending for review via an MS Exchange server to be an antediluvian approach to collaboration.  After all, I didn’t install any app to make my Google Doc collaboration happen.

New experience: I just ran the Cr-48 into the ground after 8.5 hours of use today (it is getting late).  I  had six tabs open in Chrome OS, and WordPress open on my site.  The battery icon was hollow and red for about the last 20 minutes.  No lost data. As much time spent going upstairs to get the charger (90 seconds) as it took to hook up the power and get something more to drink.  One minute later, the Cr-48 is powered up and complaining that Chrome didn’t shut down properly and would I like to Restore.  Why, yes I would.  Two seconds later all my tabs are open, all my content is refreshing, and by 15 seconds all tabs were back to operation.  I’m not sure that I could reboot my MS Windows workstation in that time frame, but if I did, I wouldn’t have any of my applications running, much less restored to their state at the time of power loss.

Still reading? By now you probably realize that I keep running Google News searches on both “Android” and “Chrome OS”, and in fact that’s part of how I knew within hours on 8 December that it would be prudent to fill out a Cr-48 request form.  As I follow the relentless march of the Android army up Highway 85 to the Gates of Cupertino, I’ve reflected on how a clever truce could help the two Silicon Valley behemoths Just Get Along better.  Part of the cool factor here with the Cr-48 is the hardware.  It’s black, as black as anything created by Steve Jobs in his NeXT days.  I’ve made some careful calculations, and by my estimate, if one were to take 12.5 Cr-48 notebooks and stack them atop each other, they would form a reasonable facsimile of the size and color of a 12-inch NeXT cube.

Wouldn’t it be cool if Apple, just on a whim, were to make a special black edition of the 13-inch MacBook Air chassis, keep the display, gut the Mac hardware and OS X and its storage, and use Apple engineer savoir-faire to craft a Chrome OS-capable device?  Chances are, you could take six Apple hardware engineers, one Weber grill, a dozen steaks, two cases of beer, and a pile of chips with napkins and Sharpies and get a brilliant design sketch in one weekend.  (I’m assuming here that some other approach was used for the design of the Cr-48 itself ;^)  There would probably be enough space left over in the Air chassis to increase the battery enough to put the Cr-48 operating time to shame.  Imagine, if you will, Apple being willing to sell a non-Mac on Apple hardware, but having that occupy the Luxo end of the Chrome OS notebook space, and doing so for $495.

Chances are, that could be the upper end of what Chrome OS notebooks are like.  For goodness’ sake, if you have anything more in the way of hardware, then use your operating system of choice and perhaps the Chrome browser, and have all the same apps, plus bells and whistles that you see fit.  But there is a slice of the future that really makes sense for Chrome OS, and I feel that ever more clearly after a month behind the keyboard of the Cr-48.

In news items, blog posts are quoted where there is a muddled distinction between Android and Chrome OS.  There might be some overlap, but so far, the difference is obvious to me.  I poke the Android screen with my fingers, but I use pointers and keyboard, plug in a mouse and plug in a VGA cable to a TV or projector with Chrome OS.  I could collaborate on some documents with Android users, but I’d feel sorry for them if they wanted to punch in spreadsheet formulas, make a complex selection to change formatting in a document, drag items around in a presentation, or even begin to free-hand draw something.  In my daily work with mapping, I learn how dependent we are on very precise pointers; thus far, fingers on screens do not precise pointers make.

When I see the sales folks at an Apple store walking around with an iPad attached to their wrist, I see future repetitive-strain lawsuits and I see Really Slow Typing right now.  It’s one thing to have a tablet and a stand and a cool roll-up full-sized Bluetooth keyboard with one’s iPad.  But by then, the Cr-48 starts to look sleeker.

In 16 months I have seen Android evolve from the G1 obscure niche device for techno-cognoscenti to Samsung Galaxy S on all five carriers, plus a 7-inch tablet, even with a Galaxy Player to besiege the iPod Touch market—in this same way I am picturing this Cr-48 being followed by many Chrome OS devices.  These could be cheap enough to get attention; think of them as One Laptop Per Human devices.  Like my speculation above, picture an Apple 13 inch ChromeBook Air for $495 on the top end, many players selling full-price devices in the $250 range, and mobile carriers offering them for free with $40/month data-only two-year contracts.

As I write, I feel we may be approaching an echo of market situations that led to my earlier disdain for laptops as poor value against tower PCs.  Perhaps what will truly distinguish Chrome OS devices from smart-, super-, or ultra-phones will be value.  For where an unlocked no-contract high end Android device may well cost $750, a very respectable Chrome OS device might cost less than $300 and support much more productivity and business relevance than many tablet solutions.   Meanwhile, the tablet may have a larger screen than the phone, but its cost will probably stay closer to the phone—and both of these are Android devices.  With Chrome OS, one is likely to experience more performance, more laptop-like hardware features, and more business processes served with less disruption than with phone-only or tablet-only mobile solutions.  No doubt there will be places for both phones and tablets, but I suggest that these will shake out to be either personal communications (like phone) or specialized field work (meter reading, field condition recording, or filling out forms) for the tablet.

If one can have much of the basic functionality of the PC: e-mail, written documents, spreadsheets, presentations, and drawings, in an environment that is robust enough to have minimal loss of documents, lost productivity due to file management issues—then how awesome to do so on very low cost hardware with familiar keyboards, reasonable screens, a very green low-power footprint, and an almost throw-away commodity-like sense to the hardware.  Remember: while typing this I learned that my Cr-48 won’t quite make it to nine hours on a battery charge, but I lost less time than a Windows reboot and lost nothing from my in-process work. This is very good, and it seems likely to be very big.

It is also fully distinct from Android, at least at this time.  But the distinction is not exactly a continental divide.  Case in point: while working on the Cr-48 this afternoon, I’d left my Android phone in vibrate-only mode from a meeting, and I had set the phone over on a shelf nearby.  I missed a call, and the caller rolled over to Google Voice, which is not hooked up to my office phone.  They left a message, which was transcribed and mailed to me—showing up in my Gmail tab.  I got a sense of what was in the message from the transcription, but I wanted to hear the tone of voice that the speaker used: was she really stressed, or just sharing information?  So I played the message on my Cr-48, and I returned the call with a headset from my office phone.  Android passed the call off to Chrome OS, and tried to help me with the transcription, but ultimately I just played the message back.  My phone stayed on the shelf the whole time…

Enough for now.  Suffice it to say that I’m having a blast, and I’m feeling fully experienced enough now to take the random blog posts that make the news with a grain of salt.

Oh, one more thing.  It’s that Open Source thing.  The Cr-48 comes with an internal switch that allows one to boot Chrome OS images that are not signed by Google.  That means that folks like myself, who assemble (either by just having one or in my case, imaging an older version to-purpose) an x86_64 Ubuntu 10.4 Lucid Lynx workstation can compile their own Chrome OS image for the Cr-48.  Last month, as a matter of principle, I did this.  I ended up running an hours-old version of a nightly build and it was all possible using standard search tools and a modicum of patience.  The open source site provides a Developer’s Guide and lots of information.  The forums are typically timely, and the only issues I had involved certain steps of the build bot relying on network path information that works within Google, but needed a tweak to work for those of us in the outside world.  By the third evening of my attempts, I had a successful build, and was given key supportive advice that the things I was having happen really shouldn’t have been like that, so I felt ready to keep at it.  And things got fixed within a day so that us out here can actually make our own Chrome OS builds using Ubuntu 10.4 LTS  x86_64 as the standard development platform.

After all that development joy, I’ve also started running Chromium browser (it’s been version 10 for a couple of weeks on Ubuntu) and both Chromium and the Google Chrome Canary Build (just a version or two behind Chromium nightly) browser on Windows.  So now on Windows, I run three variants of the browser: Google Chrome Beta, Google Chrome Canary Build, and Chromium.  This gives me three icons: red-gold-green with blue dot for standard and Beta, all yellow for Canary, and all blue for Chromium.  So far, I don’t have much preference, although only the Google Chrome variants support Google Cloud Print, so that my Cr-48 can scavenge Windows printer use both at work and at home.  Sort of like Google Docs, or plugging in a wheel mouse to the Cr-48:  Cloud Print just works.

One of the Chrome OS posts that I enjoyed recently was titled The Cr-48 running Chrome OS is so dull as it captures an essence of how many aspects of Cr-48 use are complete non-issues, and that is not a bad thing at all!