Foo to you, hardy heron!

Using Linux is like being a third-class citizen in a troubled autocracy. Windows and Mac users can click on things and expect them to work, Linux users have to tread lightly and do their best not to upset the OSS or alsa daemons, or go too far in hoping a Flash plug in might actually WORK!!

I’ve been using Linux since 1992 or ’93. I’ve always had a Linux box around. It’s been an experiment that’s kind of lasted 15 years. Given my predilection for technology experiments, I suppose I’ll always maintain a Linux box on the side. But every now and then, I succumb to The Urge. I take one of my main computers, this time my home office PC, and install a Linux distro as the primary OS. Only when you do this do you really start to understand where Linux stands and what holes need to be filled. And when I say holes, I really mean gaping cavities that would put geological marvels like the mighty river Indus to shame. But let’s not get caught up in the semantics.

I’ve used tons of distros; Slackware, debian, Puppy, Mandrake, Damn Small Linux (DSL), Suse and Ubuntu. My latest adventure involves Ubuntu Hardy Heron (8.04) with a 2.6.24-19 Kernel. I’m running this on a 3 Ghz Pentium 4 rig with 2 GB of RAM and a couple of 80 GB disks. Not by any means, a slow machine.

So, what are my findings? I’ve been running this thing for 3 months now and have been using it daily. The good news is that the install was relatively painless, my NIC was recognized and so was my display. I have a widescreen LCD and the resolutions were picked up just fine – all goodness, and a far cry from the xf86config days when you had to manually select GPU clocks and the like.

Now, the bad news. In the course of usual business, you keep running into irritating limitations mainly around Linux being unsupported by so much of the new stuff out there (hardware, applications, websites.. you name it). Also, I keep running into system oddities that take a LONG time to debug and figure out. Combine this with the fact that Linux has been robbed of its stability and performance, and you’re in for an unpleasant ride. I’ll give you some examples:

1) The sound config on my machine is really wierd. At times it works ok, and then sound just stops. I think it’s the alsa daemon that keeps core dumping. When I manually restart it, things return to normal, and then it’ll go away again. Bloody irritating.

2) I had to go to lengths unrequired by Windows in order to configure Sun’s VirtualBox to run audio, but once I did, it worked. Indeed, when you run a Windows VM it outputs audio just fine. The problem, of course, is that the host Linux OS no longer does! I can go in and try to debug this and figure out if it’s a permissions issue or if VirtualBox is trying to talk to hardware directly, bypassing alsa or whatever other evil dynamics exist. But I will gain very little at the end of all this and I really can’t afford to waste so much time debugging everything manually in order to get little things to work that have been working on Windows since the good old days of 3.1. This is no longer fun. It’s miserable.

It’s not just hardware interaction stuff that causes you to tear your hair out. Commonly used apps like Flash are casualties too. If I want to upload an image to this wordpress blog when I’m using my Ubuntu box, and I naively attempt to use the default Flash uploader, it never works. The content area in the browser window is grayed out indefinitely. You have to use the HTTP upload functionality and God forbid you forget! Using Linux is like being a third-class citizen in a troubled autocracy. Windows and Mac users can click on things and expect them to work, Linux users have to tread lightly and do their best not to upset the OSS or alsa daemons, or go too far in hoping a Flash plug in might actually WORK!!

3) New apps, even when they’re released by the Champions of Free, Google, are unsupported on Linux. So what that tells you is that Linux is and probably will forever remain second or third class. Anytime a new app comes out, you’ve either got to wait for a low-quality, feature-poor knock off in the OSS domain, or wait on the developer to finally come out with the Linux version, if ever. In the case of Google Chrome, here is what I see when I try to download the new Chrome browser for Linux:

Google's Chrome development team prefers Windows

And oh by the way, yes, I do know some guy on the web got Chrome to work with WINE. Whatever. WINE is about as stable as a drunk trying to walk a straight line. I’ve unfortunately been using that for years also. I currently have the latest released rev which continually bombs and pegs the CPU to boot. I am not going to run Chrome under WINE (sort of sounds like Chrome running ‘under the influence’) and I doubt very many others will either.

Eventually Google will come out with Chrome on Linux, which won’t be as stable as Chrome on Windows. Just like Firefox on Linux is not as stable as Firefox on Windows. And there’ll be sound integration problems, and plugin problems and all sorts of other misery… The upshot is that I waited several extra months to get my hands on what Windows users had for a long time ago. And when I got it, it was uglier, slower, fatter and generally the sort of thing that ruins your weekend.

4) Video playback on Linux is, let’s face it, still a nightmare. Having tried three different video players (VLC included), I’ve found that the behaviour of codecs on Linux is much less stable than on any other platform I’ve used. This is still the case, as of September 2008. I’ve tried a dozen codecs. I truly, really wanted this darn thing to work. I get playback but with an FPS rate Windows 98 readily produces on a Pentium Pro!

5) The mythological stability of the Linux OS is officially GONE now. Argue with me all you want, but Elvis, my friends, has left the building. I’ve used both platforms sufficiently to know the difference. Linux today, out of the box at least, with a mainstream distro like Ubuntu is both SLOWER and LESS STABLE than Windows. Flame away. It won’t change the truth. I have more inexplicable Firefox crashes on Linux, I have more instances when my screen grays out (GNOME does this when an app isn’t responding) leading to either the app disappearing and the screen ‘ungraying’ or me having to kill the process the old fashioned way. Good old ps -x, hunt for the PID, kill -9 it.

6) Process bloat. Those who complain about Windows running a lot of processes, blah blah, should do a ps -x | nl on the command line as root. I’m counting over 70 processes right now when I have TWO applications running… Firefox and Open Office’s Calc spreadsheet.

7) Bad resource management. I don’t mean this to be a lecture delivered in an Operating Systems class, but just take a look at the following image. I don’t have any Flash playing under Firefox, I don’t have any active apps or heavy Web 2.0 Javascript type apps running at the moment. And yet, Firefox usage is 56% of a modern, 3Ghz CPU. What the HELL!!??

Firefox consuming 56% with no heavy Web 2.0 apps, no Flash...

What is Linux good for? Maybe a minimal kernel that’s been thoroughly tested to run a single app in kiosk mode? Maybe as a way for embedded device makers to save some RTOS royalties by massaging a Linux kernel to run a couple of their point-applications? It certainly continues to have utility as a server platform that doesn’t need to worry about the peripherals an end user needs, or compatibility with the latest graphics stuff you would expect to pick up from a typical computer store, or any semblance of UI uniformity and ease-of-use etc. Also, netbooks and UMPCs like the MSI Wind, eeePC etc. are using Linux primarily to keep their costs low, but also, the limited apps you’re going to run on a platform like that can probably be caressed and cajoled into working when the hardware manufacturer is involved in building the specific Linux kernel, and hand picking the apps that go on to the device.

When it comes to the desktop, I think Linux is not ready and the question is, will it ever be? I don’t think so. I think we’ll sooner migrate to apps in the cloud where Chrome, IE and Firefox running fast JavaScript engines will allow us to do most of our stuff online. Ironically, that might be a simpler target for the Linux community to aim for. Let the browser be the “Window Manager”, cut the kernel down to the basics, remove some of the bloat – which should hopefully restore a bit of the speed and stability we used to have 10 years ago with Linux – and just try to make it a great platform to run Cloud apps.

Meanwhile, here in TechLahore Land, I’ll be scrubbing Ubuntu off my drives and moving to Vista-SP1 (been running fine for me post SP1, thank you). Maybe next winter I’ll come around, like I always do, and give the latest distro fare another try. Until then, adios penguin!

Share this:

Celio Corp's REDFLY

You’ve got to hand it to the brave and the few. Because that’s what the mavericks at Celio corp are. While you and I would try to learn from products that the world rejected and, oh, I don’t know, maybe NOT repeat the same mistakes, Celio has a different idea. They want to re-introduce the worst idea Palm ever had – the failed Foleo – to prove to the world that the world is just a random collection of idiots who don’t really know what they want. Celio will now teach us otherwise.

First, let’s get the concept out of the way. What is the REDFLY? It’s a “Smartphone Mobile Companion”, the same as Palm’s doomed Foleo. I suppose the marketing team at Celio didn’t feel they needed to recast the category which, thus far, has a population of one dead product. The basic idea is that you will pay $400 for the REDFLY, and connect it via bluetooth to your $300 smartphone to get a painfully slow computing experience, albeit on a larger display and more comfortable keyboard. Hey, if it’s going to be slow, it might as well be big too!

Celio claims that the REDFLY neither has a CPU, nor an OS, nor memory. Now the engineer in me rebels at definitions at like those, because it HAS to have an uber-controller, processor or FPGA because it hosts a full fledged bluetooth stack, not to mention a display interface etc. I guess their point is it doesn’t run Windows, Linux or an OTS RTOS… why not just say that then? But anyway, we’ll ask the engineer in me to shut-up for a bit so we can continue with the product description. So, where were we… yes, the REDFLY has an 8 inch screen, a commensurately sized keyboard, decent battery life at 8 hours (and about the same as a $399 MSI Wind, which, by the way, is a fully functional UMPC). Windows Mobile seems to be the only Smartphone OS supported thus far, but that’s fine, because it’s pretty popular. And I’m sure Celio can expand the list if enough people buy their product to allow them to go do that.

If the pictures don’t do it for you, you can watch a Celio demo here: http://www.celiocorp.com/nd/o/vDemo.asp

So why do we think this is a really-bad-idea? Well, here’s all the reasons why:

1) High Price. Most people are NOT going to pay $400 to get a device exactly the same size as a UMPC, with the same battery life as many UMPCs, but with none of the capabilities of a UMPC. And let’s not forget, I still need a $300+ smartphone with enough juice to run the kinds of apps the Celio REDFLY would be needed for.

Me thinks you'll be better off with the MSI Wind

2) Poor Experience. The video redirection over a bluetooth connection, while technically interesting, is not going to result in a good user experience. The Celio video demonstration at the link above is ample proof of this. And it’s not just about the bluetooth connection, there’s other problems. For one, Windows Mobile apps don’t adapt well to resolutions larger than 320×240. You can drive a lot of Smartphone displays up to 640×480 and when you do, you notice that more than half the apps break. So while you have a large screen, you’re either stuck using the same resolution, or having a significant percentage of your apps croak. Also, the Celio has an 8 inch screen whereas a similarly sized UMPC typically has an 8.9′ display. The Celio provides an 800×480 resolution, whereas most UMPCs deliver a 1024×600 resolution. That translates into 40% more pixels!! Big difference.

3) Battery issues. The REDFLY battery may last for 8 hours, but my Smartphone battery probably won’t. Especially when the Bluetooth connection is constantly active. The 8 hour battery on the REDFLY, in fact, does me very little good. And by the way, if I had a UMPC I wouldn’t be using my Smartphone battery which I want to preserve to make calls. (Yes, calls. Remember that’s what we typically use phones for.)

4) No ergonomic advantage. Will the Celio device and it’s charger really take any less space than a fully functional MSI Wind, eeePC, HP mininote or Dell mini? NO. Will the Celio and a smartphone laid out on a typical airline fold-out table take any less space than a UMPC? NO. In fact, with the smartphone, it will take more.

5) Application Limitations. The Celio demo shows you a Powerpoint but neglects to mention that the Windows Mobile edition of Powerpoint doesn’t let you edit anything. You can only view previously produced decks. Fat lot of good that does. In situations where you have to whip the Celio out, you would expect you’re about to perform an activity not do-able on the smartphone. You can see a Powerpoint deck just fine on most smartphones. You’d sooner use the zoom feature on a phone than pull out the REDFLY from your backpack, set it up and synch it to your phone via Bluetooth. This isn’t just an issue with Powerpoint, many Windows Mobile applications, since they were intended to be used on a Smartphone, have similar limitations. None of them expect to be attached to a large screen and keyboard so they limit input activities, typically. This is not Celio’s fault, but not factoring this reality into their product design most certainly is. Now, you can pooh-pooh these limitations by posting links to apps that DO let you edit a powerpoint on your Windows Mobile device (Yes, I DO know there are several out there), but will I really pay $15-$30 for apps like these to enable my Celio with functionality I get for free on a UMPC? Don’t know about you, but I certainly won’t.

Palm's failed Treo

6) Poor entertainment capabilities. Why pay $400 for a small device with slow video driven off a bluetooth link, that’ll consume Gigs of space on your Smartphone storage card for a single movie, and still won’t be able to play 720p full screen video while you’re cruising over the Atlantic, when you can pay $400 for a small device with great video and 80GB of space to store your videos.

Let’s not make this a much longer list now. I think the point has been made. I am completely astounded that a company with an experienced executive team of 3 C-level executives and 5 Vice Presidents – yes, count them, a startup with an exec team of eight Powerful-And-Experienced-Executives – would go do something like this. I wish all eight of them the very best of luck, but my (unsolicited) advice would be for them to please not trade salary for equity.

In the meanwhile, please feel free to flame me for expressing my views on the product concept. Just keep it within the realm of civility.

Share this:

This is possibly the largest WiMax network currently operational, in the world. After much poking and prodding, it seems Wateen has listened to us and delivered the goods. While I don’t yet have a personal experience to report to our readers, suffice to say the Wateen website is proclaiming general availability. Moreover, this piece of news has received quite a bit of coverage in the international media – especially with the uber cool set at sites like engadget.

The extent to which this network can revolutionize business in the country can hardly be overstated. With mobilink soon to follow suit with a competitive WiMax offering, and the PTA encouraging other Telcos to launch 3G and data networks based on a slew of alternate technologies, the time is now ripe to launch services and applications that leverage the high speed, anywhere, anytime connectivity. We’ve previously complained about Telenor’s poorly executed TV over GPRS service for instance, and once mobile phones and UMPCs like Samsung’s SPH-9000 become widely available in the local markets, WiMax based streaming will probably sound the death knell for pretenders such as telenor TV.

Share this: