UltraEdit for Linux

You may already know that I'm a big fan of UltraEdit, the text editor for Windows. I have long bemoaned the fact that UltraEdit is available only for Windows. Mac users have the equally robust BBEdit (and the free yet surprisingly robust Text Wrangler) while Linux users have a variety of open-source editors. But not UltraEdit. Not yet, anyway. But soon. The Linux version currently is in beta and it is available for just a few Linux distributions. Fortunately, Ubuntu is one of them. And UltraEdit will soon be available for the Mac, too.

I haven't signed a non-disclosure agreement (NDA) with the developers, IDM Software, but I'm not including any screen shots of the new application. I can tell you, though, that it looks a lot like the Windows version.

There's still a lot of work to be done. I'm preparing this report on a Linux machine with UltraEdit. Instead of saving at the end of each paragraph (Enter, Enter, Ctrl-S), I save at the end of each sentence (Period, space, Ctrl-S). It's not that the beta code is prone to crashing because I haven't seen a crash yet. But it is prone to badly scrambling the text.

A double click, which should select a word seems randomly to select part of a word (from the space before the word to the cursor's current position) or parts of multiple words (from the middle of the previous word to the cursor's position in the current word). Sometimes just a letter or two.

Copying and pasting text is another interesting challenge. Sometimes all of the copied text is pasted. Sometimes part of the copied text is pasted. Sometimes it's pasted to the current cursor location and sometimes it wipes out parts of a previous sentence or paragraph. The display sometimes jumps forward in the document for no apparent reason and occasionally the cursor disappears. These kinds of anomalies are to be expected in early beta releases.

If you've never been involved in software development, you might find this list of "known bugs" informational. Developers and testers identify problems, analysts and developers then review the code to learn what's causing the problem and establish ways to resolve the problems, and finally the code is written to fix the problem. Once that is done, regression testing begins to determine if fixing one bug has broken something else. It's not an easy process.

As of the end of June, this is approximately half of the list of known problems that UltraEdit developers were working to fix:

  1. Doing a Find and pressing Count All causes UEX to crash.
  2. Often double-clicking on Find in Files results in the output window causes edit window to position to wrong line (one line below matched line).
  3. Using F3 or CTRL+F3 for strings that exist multiple times in file returns reference to correct string, but indicates that it doesn't exist in the file.
  4. "Include subdirectories" option doesn't work - only the specified directory is searched.
  5. When using Find in Files in "All Open Files" when there is no match (for example with "Match case" or Match whole word") no output is passed to output window.
  6. "Highlight All Items Found" doesn't highlight all occurrences (seems to skip first occurrence).
  7. When doing a replace, and the last item is matched and the "Search string 'x' not found" dialog is presented, the last replaced string replaces the search string when the prompt dialog is dismissed.
  8. A Replace using the Preserve Case option will skip the first match for the search string.
  9. Replace All skips the first occurrence of the find string.
  10. Pressing TAB in "Find what" field in Replace dialog should move cursor to "Replace with" field but doesn't currently.
  11. Painting issues with Replace when last line matched/modified by replace is not terminated by hard return.
  12. Search -> Incremental Search has no effect, and typing after selecting this option inserts text into active file.
  13. When a Replace occurs, focus does not automatically move to the next valid matching string.
  14. Find/Replace strings are not selected during TAB navigation.
  15. "Subdirectories to ignore" does not seem to be honored (FIF/RIF).
  16. "File names/extensions to ignore" does not seem to be honored (FIF/RIF).
  17. The actual search options in Find/Replace dialogs are never touched by tab order, only the Options heading.

Most of the configuration options (UltraEdit is highly configurable) are currently disconnected because they don't yet work properly. By "disconnected", I mean that the options are present in the configuration screen, but they don't yet work.

Still, I've seen enough to know that UltraEdit for Linux (and probably UltraEdit for OS X) will be welcome additions to the applications on my computers.

The Trouble Was Not in Your Set

On Sunday, July 5, you may have noticed that the TechByter Worldwide website was unavailable for a while and because the interruption was so brief, I delayed an account of the problem until now. This was the result of some maintenance work that, while not an emergency, needed to be taken care of promptly. Maybe "high-priority semi-planned maintenance" would describe it. We had to destroy the site in order to save it, so Sunday morning, I gave BlueHost approval to "nuke the domain."

A spammer caused this. A small amount of spam had been leaking through the domain. It wasn't a flood; just a few messages per day. BlueHost hadn't reported it to me, although they confirmed the problem when I reported it to them.

Your Address Was Not at Risk

That's the first thing you should be concerned about. Fortunately, I do not store subscribers' e-mail addresses on the Web server. They are maintained on the server that handles the weekly e-mail transmission.

Because the volume of spam was so small and no information was at risk, I proceeded slowly to see if I could find the problem. Eventually, I found what I believe to be the problem: My younger daughter, whose graphic design website is hosted here, had installed a contact form using some insecure PHP code.

Instead of securing the PHP code, I suggested that she use a Perl CGI script that is generally considered "secure". I used the scare quotes because it will be secure only until somebody finds a way to exploit a flaw and then the programmers will need to understand the flaw and fix it. But for now it's as good a script as you'll find.

Planning the Destruction and Renewal

Killing a website and reinstalling everything isn't particularly difficult, but several tasks have to be performed quickly if the goal is to minimize downtime.

Because I would need to reload everything, I thought that this would be a good time to change the primary domain from "techbyter.com" to "techbyter.com". I no longer have any association with WTVN, but I'll maintain the domain name for a few more months. If you point your browser at techbyter.com or technology-corner.com, you'll be directed to techbyter.com.

Once those associations were in place, I needed to create the other domains that are hosted on the server; then I could create the 70 or so e-mail accounts. This is much easier than it sounds because I could upload a configuration file to the server. Total time to create the accounts: About 3 minutes.

Simultaneously, the most critical files (about 4000 of them) were restoring to the TechByter Worldwide website: The 2009 program files, images, podcasts, cascading style sheets, and RSS files. Once that was complete, I could restore the root directory and TechByter was back in operation (at least as long as nobody wanted to see any program files other than those from 2009).

So the site was out of service for about 90 minutes and had limited resources for the next several hours. Anyone who visited during the first few minutes would have seen a "no site here" message. After that, there was an "under construction" message from BlueHost until I restored the root directory. I saved that for the second step to avoid having people get the home page but find that every link was broken.

It's Time to Dig Out the Old Microhoo Jokes Again

Microsoft and Yahoo have finally made it official. Microsoft bid $45 billion for Yahoo. Yahoo's resident geniuses decided that wasn't enough and rejected the offer only to see the value of the company shrink to that of a pack of gum. Now Microsoft and Yahoo have a a 10-year deal that moves Microsoft from a distant third to a solid second in the search engine category. Google is still in first place by a considerable margin. And by announcing its own operating system, Google has put itself squarely in the cross-hairs of Microsoft's big guns.

What's more than a little surprising about the deal is that Yahoo is abandoning its own search engine in favor of Microsoft's recently released Bing search engine. So Microsoft gets the name and the URL while Yahoo survives as a name, but not much else. Ads in search results will come not from Yahoo but from Microsoft. Essentially, Yahoo has thrown in the towel, but Microsoft has structured the deal to allow the company to save face. If nothing else.

Yahoo will still sell ads that appear on both Yahoo sites and on Bing. Beyond that, Microsoft lets Yahoo keep nearly all of the money from ads that appear on Yahoo sites. Microsoft obtains the information about users' on-line activities. Yahoo previously snatched defeat from the jaws of victory and now appears to be in a death march to the Internet scrap heap.

Shortly after Bing was released, it beat Yahoo, at least for a day. By combining Bing and Yahoo, Microsoft immediately moves from a weak third place to a strong second place. Google still has 70% of the search market, but the "Microhoo" or "Yasoft" combo now checks in with about a 30% market share.

The feds will still have to approve the deal because of the antitrust implications. Completing the deal is something analysts consider likely to happen because the 2 companies together have only 30% of the market. Expect Google to whine about it, but expect the deal to go through regardless.

Short Circuits

Amazingly Cheap Memory

I remember the first 8MB thumb drives that sold for $50. This week, buy.com offered a drive that was 500 times larger than the original for 20% of the price. If auto manufacturers could do that, I would have paid about $2 for my new Honda Fit.

Cheap Memory

Whether it's hard drives or memory such as the deals shown above, I'm almost always amazed by today's prices. An 8GB thumb drive for less than $17, shipping included. Amazing!

And speaking of cars, I did just buy a new Honda Fit to replace an aging (15 years old) Ford Explorer. The federal government's Cash for Clunkers program valued the Explorer at $4500, far more than I could ever have received in trade for it. So now I'm driving a car that gets double the mileage of the Explorer.

If you're thinking of using the program, you may have to wait a bit. On Thursday, I wrote "the time to act is now" but then the federal government announced that it was suspending the program because the money was already gone. The maximum number of sales under the program would be about 250,000 cars ($1B divided by an average of $4K). The program was scheduled to continue until the end of September or the funding is exhausted. On Thursday, I wrote "There's no way this program will last until September," but I didn't expect the money to run out quite that fast.

Spatio-temporal Network-level Automatic Reputation Engine kills Spam

SNARE (Spatio-temporal Network-level Automatic Reputation Engine), developed by the Georgia Institute of Technology is the latest (probably vain) hope to kill spam. SNARE scores each incoming e-mail on criteria that can be determined by examining just one data packet. If this works in the real world, BRAVO!

Analyzing spam is a challenge. We want our e-mail to arrive immediately, but analyzing takes time. Whatever defensive techniques you put in place will delay delivery, but some analysts say that spam accounts for more than 90% of all messages sent.

The GIT team examined 25 million e-mails from TrustedSource.org and concluded that good messages come from computers that have a lot of ports open. On the other hand, spam is delivered by automated bots that usually keep the SMTP port (25) open and close other ports. Hmmm.

Another test involves the distance between the sender and the receiver. Researchers determined that the distance between the originating IP and the receiving IP is important. Spam tends to come from further away than legitimate messages and spammers often have IP addresses that are numerically and positionally close to those of other spammers.

This kind of analysis is something that spammers can easily program to defeat, but it's another attempt to shut down the crap (I would prefer use another word that is more appropriate) that fills our mailboxes. For the effort, I can say only "I hope this works!"

Details: Here.

Totally Unrelated to Technology

A week ago, my wife and I and our younger daughter (age 25 in October) visited Amish country in Eastern Ohio and toured an Amish farm. Along the way, we were able to view 2 Amish homes (no longer occupied) and take a ride in an Amish buggy (the horse was not happy with my weight). Amish families do not use electricity delivered by power line or gasoline engines. This makes the homes quiet and peaceful. They do, however, use batteries that are recharged by solar panels to run power inverters so some electronic devices are available. I took a high-tech camera with me.

If you'd like to see the pictures, they're here.

What the area lacks in high tech it makes up in good taste. Some of the best cheese and bread you'll ever find can be had in this area. You'll see this in the pictures.

It would be strange to live without many of today's modern conveniences and it's hard to imagine an area of the country that's less hospitable to manual transportation, but maybe the slower, lower, quieter life style is worth it.