Internet-Standard

Grep-Envy

Grep is one of those programs that's been around forever as a part of *NIX. Its function is straightforward. Grep:

searches one or more input files for lines containing a match to a specified pattern. By default, grep prints the matching lines.

In other words, grep is a super-utility for finding information inside any file on your computer. The fact that grep uses regular expressions makes it even more useful. "Regex" is like wildcards on steroids: for example, let's say you're looking for all telephone numbers in the (212) area code stored anywhere on your computer. With wildcard searching, you could specify a search that might get most of those instances, but with regular expressions (and a little practice) you can probably catch every instance of the numbers "212" linked with a sequence of seven other digits in a telephone number, while ruling out instances of "212" occuring in some other sequence.

Anyway, a lot of that function can now be had with Google Desktop. But even so, full grep function has been missing from Windows from the beginning--and people have tried to implement grep for Windows (Google search) over and over, over the years.

There's GNU Grep for Windows, part of the GnuWin32  package of ports of key GNU utilities for Windows (up through Vista).

There's the shareware/trialware Windows Grep that you can try out for 30 days/buy for $30.

There's even a super-duper GUI-front-end PowerGREP, for the bargain price of $149/single user license.

And did you check that Google search? With over two million hits, I'm pretty sure there are other implementations you can buy. Or, you can just install Linux--grep is in there, for free.

Open Standards: Another Viewpoint

If you have any questions about whether open standards are better than proprietary standards, just check the past 25 years of computer networking: if you stuck with proprietary networking protocols, you were toast. Novell is no longer in the networking business because they ignored TCP/IP for too long; Microsoft is still in the networking business because, eventually, they acknowledged that they needed to support TCP/IP in addition to their own proprietary protocols.

So I'm always happy to see others writing about the advantages and benefits of open standards. For example, Robert Strohmeyer's recent article, Open Up to Open Formats at Maximum PC.

The reason we still see businesses selling proprietary standards/protocols is that they are still seen as achieving a benefit--for the vendor. If you buy into Software Vendor X's proprietary file formats (for word processing, accounting, audio or video), you're locked in to that vendor. If that vendor goes belly up, or decides to stop selling your favorite software, or just decides to switch to a new proprietary standard, you're stuck. You've got to find a way to migrate your data, yet again, to the standard du jour.

Open Source Drives Web 2.0

I'm totally not surprised that, according to this article, Web 2.0 is built on Open Source, the biggest/most visible Web 2.0 companies are built and hosted on open source software platforms.

Think about it: if you've got some crazy idea for a new web-based business, you have the choice of going out and spending kilobucks (to start) for commercial proprietary software--or doing your preliminary development/prototyping on no-cost open source software. If you choose the open source option, you're out nothing but time if your idea flops, so it's pretty easy to make that decision.

Let's say your idea starts to take off, though: with proprietary software, you've got to keep going back to the vendor to buy more and more licenses as you grow. If you'd started out with open source, though, all you need to scale is more hardware (which you need with the proprietary solutions as well). Again, it's pretty obvious that you do better with open source software, especially if you can't drive a revenue stream through your website until you've scaled up.

European politicians seem pretty smart...

Last year, it was French parliament switching to Linux. This year, the French picked Ubuntu as their Linux distro.

And just this week, we discovered that the Italian parliament will migrate to Linux as well. Which distribution? According to the Inquirer, Italian parliament bets house on SuSE Linux.

In both cases, money was a key factor: Linux was determined to cost less, despite startup costs associated with the migration and with training.

1): Open Source; 2): ???; 3): Profit!!!

Open source is the stealth fad: it's all over the place, and has been for years, but it seems to surprise so many people even now, in 2007. One of the problems is that unlike almost every new technology ever introduced, people have a very hard time understanding how you can use open source to generate profit--because how can anything be "happening" and "new" if it isn't going to make someone some money.

Though I've written about how to generate profits with open source software (see Open Source and the Profit Motive, for example), I'm neither the first to do so, nor will I be the last. In fact, I just read Top Five Open Source Business Models You Never Heard Of, and it got me thinking about open source and profits, again.

The thing to keep in mind with open source is that it's not a technology or product: it's a new way of doing business with software. The millions of different companies go about their business in millions of different ways, each of them using different tools, different processes, different people, different vendors and different customers.

As it happens, more and more of those businesses are incorporating open source software into the way they do business: some of them are replacing proprietary software on their production systems with open source software; some are replacing their own proprietary software products with open source (or publishing that proprietary software as open source); some are marketing products or services to open source software users; some are building new products or services on top of open source software.

The thing to remember is that even though pundits, advocates, and random members of the blogosphere may imagine a dozen or so different possible models for creating profit from open source software, the most interesting and compelling business models will undoubtedly come out of nowhere.

It's instructive to look to history. Before Yahoo!, there was no Internet search business; when everyone thought Yahoo! had the business tied up, up popped Google.

So, will the first/next super-huge open source success be a software vendor like Red Hat or MySQL? Or will it be a company that incorporates open source into its product line, like IBM? Or will it be a company that depends on open source software to run its infrastructure, like Google?

My money is bet on some little startup, maybe taking shape right now in some coffee shop or home office, that will turn the whole business upside down. And whatever it is, it will be new and different and completely unexpected--but completely delightful in its novelty and utility.

Counting Linux Users

Did I mention that I've been "on" the Internet since 1988? Because I have: that's when I went to work at Draper Laboratory, in Cambridge MA. Eventually, after bouncing around a bit, I wound up in the network support group where we handled all the networking issues. And at Draper, unlike most organizations at the time, "network" meant "IP network", as in, "Internet Protocol" network.

Why do I mention this? Only because whenever Linux makes news, it makes deja vu for me. This time, the article at Slashdot, Attempts to Count Linux Users Remain Pointless, reminded me of the mid-90s when everyone with something to hype about the Internet and web would go on endlessly about the seemingly endless annual doubling (or faster) rate of Internet growth.

The original article cited at Slashdot, Can Linux Adoption Ever be Accurately Gauged?, is mostly about why it's so tough to figure out the degree to which Linux is penetrating the "market". Most of the Slashdot discussion is about whether or not it's even worthwhile to count Linux users, and if so, how to actually do it with some degree of precision.

I'm a big believer in back-of-the-envelope calculation, and I also believe that, although it may be difficult, it is possible to estimate Linux market share. I'm interested in hearing your opinions on this matter--but I'm also going to get to work on an article about how you'd go about generating this information, as well as the results I get.

Open Source iPhone Killer?

Are you sick of the iPhone yet? I sure am. Anyway, I was delighted to hear about an open source iPhone killer, from the OpenMoko project. Slashdot reported an Open Source Linux Phone Goes on Sale.

The phone in question is the Neo 1973, and the price for a developer's preview is $300. The really good news is that the OpenMoko people are trying to set an example for how hardware manufacturers can go about using open source software in a way that will benefit the manufacturer/vendor as well as users and third party developers.

Right now it's not entirely clear when they'll be ready to sell to mainstream users; at the moment their website seems to be knocked down by all the Linux fans looking for more information, but if you can't get in now they should be back up and running soon.

Red/Blue vs Windows/Linux?

Open source vs proprietary software is more than a business issue: it's very much a political issue. Politicians who take campaign contributions from every other industry are just as happy to accept money from the software industry as well. Microsoft spends as heavily as any other major corporation.

But politicians also have to use software, the same as businesses and individuals. So which presidential candidates use Windows/IIS for their campaign web servers, and which run Linux/Apache? Check out Douglas Karr's summary of the candidates' OS, webserver and web hosting service choices.

Eclipse news

It's been some months since I checked in on the Eclipse open development platform but they seem to be making strides in usefulness. I was reminded of Eclipse by this post about the latest Eclipse release, Europa.

Searching the "dark" web

Sometimes I forget there are other ways to search the web beyond Google; this article offers 8 ways for searching the dark web beyond Google. I hadn't heard about Clusty but I tried it and liked it: it's nice to have the "clusters" of hits around specific keywords that you didn't specify to really narrow down the search. I'll have to (re)discover the others when I get a few minutes.