Much ado about scripting, Linux & Eclipse: card subject to change

2007-07-28

EMF: Crumbelievable?

+1: Argo2ecore releases v3.0.0, updated for Eclipse 3.3. Argo2ecore allows users to create graphical UML models using ArgoUML and then use the EMF codegenerator to generate java code.

-1: Kraft repurposes EMF's classic song as "Crumbelievable". Cheez.

2007-07-27

Windows: A Survival Guide, Part 3

To complete the trilogy of 'Windows for Linux geeks' posts, I wanted to close by talking about WinTabber, which was suggested to me in Part 2. WinTabber is fantastic for grouping Cygwin windows (a la Konsole), or even Windows Explorer windows (a la Konqueror), though there's no way to split windows into left-right panes. For that, it's back to FileAnt or Beyond Compare. Problem solved.

However, grouping other application windows is less than helpful -- for example, grouping Eclipse windows causes the top menu to disappear:

But that's no problem, really. One snag I did run into is that a couple times the ext driver I'm using to read/write my linux partition from Windows left the partition mounted when I shut down so on reboot I had to do a full fsck to recover the journal. Of course having written that, I can't reproduce the effect with either a read or a write. Go figure.

2007-07-26

Help me help y'all

Hey there, cowpokes! Do you spend long days and longer nights wondering how to get your code to work? Eclipse got you down? Feel like there's no one to turn to for help?

Well, friends, look no further!

Y'see friends, one fine day I was hanging out in #eclipse and was amazed at the sheer volume of quickly answerable questions coming in from Eclipse fans like y'selves -- questions already well documented! Now, I'm not talking about questions about CTRL-3 and the several hundred blogs singing its praises -- no sir! I'm talking about real -- and real easy -- questions like 'How do I start Eclipse?' or 'What's the deal with gcj?' or 'How can I solve my memory problems?'

Now, I know it's sometimes hard to know where to turn in a moment of crisis, and the internet can be a mighty scary and complex place for a greenhorn at times. We've all been there.

And so, without further ado, I give ya'll: my Top N List of places to go to for help...

  1. Read The FAQs: IRC FAQ, Eclipse FAQs, More FAQs
  2. Learn about the tools/projects you use from their websites
  3. Ask a question in the newsgroups
  4. Submit a question / discuss an issue / keep up to date in the mailing lists
  5. Browse source code in CVS to find an existing solution
  6. Read the online ISV help docs and APIs at help.eclipse.org
  7. Read articles, blogs, and other resources
  8. Watch videos of demos, walkthrus, and howtos
  9. Find, report, patch, triage, fix, or verify a bug or feature request
  10. Search with Eclipse or Google
  11. Seek legal advice about 3rd party code and IP
  12. Ask a question or discuss an issue in real time in #eclipse on freenode

But wait! That's not all! As if by coincidence, most of the items in the above list are also ways to contribute to Eclipse and the community. Once you find your answer, pay it forward by blogging about it, writing or expanding a wiki entry, contributing to a FAQ, answering a questing in a newsgroup or hanging out in #eclipse to help others.

Hell, ya'll could volunteer to be testers and take a few minutes to report a detailed bug -- with steps to reproduce, platform/config details, screen shots, logs, JUnits, and/or patches (as appropriate). Or, if you prefer to squash bugs than to unearth them, participate in BugDay. Want to start your own project? Check out next year's Google's Summer of Code and start contributing now!

Now I expect you're asking yourself, "Self, here comes the sales pitch where they ask for all mah hard earned greenbacks." But thanks to the magic of Open Source... everything you see before ya'll is free. Free as in 'free beer' -- open, transparent, and ultimately funded by your time and effort. So, please, give generously. ;-)

2007-07-22

phpeclipse vs. PDT

I've always enjoyed the phpeclipse developer experience, but I've noticed lately that the project appears to have slowed its release schedule, though they have started publishing N builds to an update site. Meanwhile, PDT, as of v1.0.0.v20070712 is IMHO, finally usable -- and also installable by update site (but don't forget to select the Europa site too in order to get all the prereqs).

I'm also impressed that PDT uses WTP's Web Standard Tools (WST), which in turn depends on EMF. So, it's a bigger install than phpeclipse, but it's full of modeling goodness, so it's worth it. Why?

With modeling comes interconnected views such as the PHP Project Outline view (a list of all defined functions (with signature), constants and classes, linked with editors), the PHP Explorer view (like the Package Explorer for Java, but unfortunately using its own set of Working Sets, rather than the shared ones available in other views), and the PHP Functions view.

However, there's a few features in phpeclipse I'll miss, such as:

  • select an include() or require() statement's argument and hit F3 to navigate into that file
  • warning of unused/undefined variables in files and functions
  • search results highlighting (other than in the right margin)

One major inconvenience I ran into, however, is that installing PDT into an existing Eclipse install with phpeclipse 1.1.9 or 1.2.0 causes a whole raft of NullPointerExceptions when attempting to open PDT editors (phpeclipse editors still work fine). I wonder if this is a namespace conflict between two "PHP Editor" editors. Anyway, removing phpeclipse and restarting seems to fix this problem, on Windows XP, anyway. I wasn't so lucky with Linux, but it never hurts to try again sometimes.

Maybe some day these two projects will be able to share resources and features instead of competing, both technically and politically. I'd love to see the great aspects of phpeclipse merged into PDT, and have those developers become committers for PDT, if they're not already. I have to agree with the comments posted on this blog about pooling resources to solve new problems instead re-solving the same ones.

Reacting Badly

I live not far from Pickering Nuclear -- in fact, about 5 blocks from where the nuclear fallout from an explosion would apparently reach. Yes, the cloud is supposed to stop at Morningside Ave. Riiight. Luckily, the reactors which had a bad rep in the 90s were cleaned up and are back in service, though Greenpeace has its doubts.

Anyway, with the summer in full swing and my A/C working hard to keep my house habitable, I'm reminded of the past few years of outages, brownouts, and full-blown blackouts we've had here over the years. This morning I did a little reading into where things are at in the world of nuclear energy, and I have to say I'm disappointed.

Back in the 40s, the West led fission technology and it unfortunately led to a rather effective (if deadly) end of WWII. There's been much speculation about the reasons Germany never built the bomb before the US did, but I digress. Since then, we've enjoyed nuclear energy in a number of forms, but they've all mostly been water-cooled systems which are now, in retrospect, highly inefficient and expensive. In Canada, we're pretty much stuck with CANDU-style reactors because they're of a Canadian design, and at least are safer than the infamous design used at Chernobyl. I'm all for 'support locally-developed tech' but when there's the potential for a better, cheaper, and safer solution, it's time to look past our own backyard and toward what the neighbours are doing. (Why can't the AECL license the new tech? They'll happily build CANDU 6 reactors for other countries -- why not these new ones?) In the world of nuclear energy, we're not keeping up with the Joneses.

So what's the alternative? Well, there are several, but there's one that I've been tracking for a while now, since I was clued into it via this WIRED article, "Let a Thousand Reactors Bloom". China is moving along nicely in their plan to surpass Canada with their use and consumption of nuclear energy, but they're not doing it with the old, expensive designs. They're pioneering a new design, developed in South America by PBMR, which instead of fuel rods and water cooling, uses graphite-covered uranium "pebbles" and some inert gas (eg., helium).

Waste
The design of the of PBMR fuel makes it easy to store the spent fuel, because the silicon carbide coating on the fuel spheres will keep the radioactive decay particles isolated for approximately a million years, which is longer that the activity even of plutonium. [T]he PBMR fuel can be stored on site for at least 80 years, [and] has a greater "burn-up" [...] which makes it less valuable to recycle. [...] The spent coated particle fuel can be disposed of in a deep under-ground repository. (Coated particle fuel will maintain its integrity for up to ~ 1 million years in a repository, ensuring that spent fuel radionuclides are contained for extremely long periods of time. The plutonium will have decayed away completely in 250 000 years).

Safety
The PBMR is walk-away safe. [T]he reactor will cool down naturally on its own in a very short time. This is because the increase in temperature makes the chain reaction less efficient and it therefore ceases to generate power. The size of the core is such that it has a high surface area to volume ratio. This means that the heat it loses through its surface (via the same process that allows a standing cup of tea to cool down) is more than the heat generated by the decay fission products in the core. Hence the reactor can never (due to its thermal inertia) reach the temperature at which a meltdown would occur. The plant can never be hot enough for long enough to cause damage to the fuel.

Radiation Leakage
The helium itself, which is used to cool the reaction, is chemically and radiologically inert: it cannot combine with other chemicals, it is non-combustible, and non-radioactive. Because oxygen cannot penetrate the helium, oxygen in the air cannot get into the high temperature core to corrode the graphite used in the reaction or to start a fire. If, through some accident, the helium gas duct (inlet and outlet lines) is ruptured, it would take some nine hours for natural air to circulate through the core. Even if this could happen, it would only lead to less than 10-6 (one millionth) of the radioactivity in the core being released per day. That means that the amount of activity released in 24 hours under this very severe (and recoverable) situation would be some 10 000 times less than that requiring any off-site emergency actions. To avoid such a total failure of the main gas ducting it is designed to leak before it breaks, so that the depressurisation will be gradual and cannot lead to such a rupture. The helium pressure inside the closed cycle gas turbine is higher than the air pressure outside it, so nothing can get inside the nuclear circuit to contaminate it. [1]

As of 2004, China has one of these reactors in use, with another planned for 2010 and at least 30 more old-school water-cooled reactors coming online in the next 15 years. It seems they want to do more research on the new tech before leaping forward with it [2], but if all goes well the plan is to "spread this technology both at home and to the whole world" [3].

Meanwhile, back in North America... nothing much is happening. Well, that's not entirely true. A couple CANDUs in Ontario are being fixed up for re-use [4]. There's talk of building new CANDUs in Alberta to support the oil industry there. The US is watching what China is doing and predicting the world's nuclear outlook, suggesting China will be among the top five nuclear countries by 2025 (along with the US, France, Japan, and South Korea). Oddly, Canada, Germany, and Russia are predicted to decline in nuclear output. I wonder why, considering the AECL is developing a new CANDU design, the ACR-1000 for service by 2016. Maybe we'll be switching to wind and solar? Yeah, right.

2007-07-19

Real World Bug Reporting, Part 2

I've given up on Rogers' ability to think outside their carefully cloistered support pillars, but for comic relief here's their final reply in this thread, interspersed with a little froth:

Thank you for taking the time to write to us, we appreciate your use of online customer service.

Gotta love the opening remarks drivel "We appreciate your use of online customer service." Geez, that's fantastic.

In your recent email, you have informed us that you would like assistance so you can view your services online and be able to manage your services successfully.

We apologize that you are experiencing difficulties accessing our www.rogers.com website.

Holding breath, waiting for the other shoe to drop...

We are also sorry to hear that you have not been happy with the customer service you received. The expectation is that you receive world class customer service. If that is not what you experienced, we do apologize.

Silly me, I thought I'd made it clear I was dissatisifed. But by all means, couch your replies with "if", if it gets you through the day. Note also the distinct lack of "... and to make it better, we're going to actually do something other than simply sending this FOAD email to pacify you."

Please ensure that when you are navigating our website you are using Internet Explorer or Netscape as your browser as we do not support other browser types at this time. It is also suggested to clear all your cookies and your cache folder.

Kudos to Rogers for being forward thinking and choosing to mention Firefox here instead of Netscape. Oh, wait, I'm dreaming. Unfortunately, I can't argue the Internet Explorer point since according to some recent stats, they're still about 70% of the planet. I've never seen very flattering stats for Mac or Linux usage, so I guess it's reasonable to only support Windows and its de facto browser. You win this time, tech support. Oh, but wait -- this isn't a browser issue, it's a VBScript compilation error!

You are able to login as many times as you would like on our www.rogers.com website.

Actually, that was the reason I filed this complaint in the first place, and no amount of "just use IE or Netscape" will resolve the fact that this isn't a browser problem.

If you continue using our website and experiencing errors, we do suggest that you contact our eCare department at 1-877-343-5745. They are open to serve you Monday-Sunday 7am-3am.

They are truly our website experts and with further troubleshooting, they will be able to assist you so you can view your services online successfully.

Oh, let's be real. They'll probably tell me that they've never heard of Linux, that Konqueror is a guy on a horse with a big sword, and ask me what version of Windows I'm using repeatedly until I hang up in disgust. I doubt they'll acknowledge the possibility that the browser is not the problem or that it's their code that's causing the compilation error.

We understand that you would like to receive assistance via email, however some matters are more complex and better if we discuss via telephone.

Like the ones where I say, "Read this Microsoft KB article and fix your website accordingly." Yep, that requires discussion.

And true, it's way more efficient to read a screen to someone over the phone than to send him an email so he can read it himself. Besides, manually entering a URL as someone reads it to you is *much* more fun (because errors = comedy, right?) than just clicking a link.

We do appreciate your cooperation concerning this matter. You are a valued customer and we thank you for your business.

So highly valued that they can't communicate between their departments to resolve problems. Ah, well, you can lead a monopoly to water, but you can't make it think.

Anyway, despite this, I'm stuck with these folks for the following reasons: the phone lines in my neighbourhood suck and the best throughput I can get is about 2-3MB/s download. With Rogers over cable lines I can get 5MB, and yesterday I got some junk mail saying they're upgrading everyone to 7MB for free. (It was postal junk mail, not email, because they're not quite *there* enough to send me a free email when wasting postage is an available option.) Ah, Rogers. Always going that extra mile to justify jacking up their rates while giving us a little something extra so we won't complain too much.

So, what can you do? Monopoly, yes, but ultimately, it's all about the bandwidth.

'Till I was Blue In The Face

Went to see GWAR tonight w/ my friend and die-hard GWAR fan Ulysses. Caught the last two opening bands (Chimaira, I think, and Shadows Fall), then the Scumdogs took the stage. As a GWAR concert noob, I wasn't sure what to expect, but things got underway pretty fast... beheadings, dismemberment, puppets getting torn apart and even a bit of bestiality/necrophilia for, ahem, colour.

Speaking of colour, the highlight of the experience has to be the "blood" that's rained over the audience during the aforementioned comic atrocities. I started the evening in a white shirt and ended up looking like I'd torn apart a Smurf.

2007-07-16

EMF: One up, one down

+1: Legisway states they'll probably "use EMF for in memory [data] access."

-1: In other news, people with EMF 1.0 or less get kudos in Michigan, but I have to say I disagree: they need to have EMF 2.2 or 2.3 to really acheive cost savings. ;-)

This Just In...

The results of this year's JDJ Readers' Choice Awards are in... well, sort of.

According to the JDJ site:

Voting has now closed, stay tuned for the audited voting results to be announced at SOA World Conference & Expo 2007 East June 25-27, 2007 in New York City.

After a few weeks of intermittently looking for results (only to find past years' results), I finally managed to unearth the results.

Oddly enough, however, there is *NO* mention of any of the Eclipse technologies listed in the various 29 categories of the original survey, and only a few mentions of Eclipse member companies like Borland, IBM or Oracle. Categories, like Best Book, seem to have been renamed to make them more SOA-centric with all non-SOA contenders dropped from the pile. Some categories seem to have been merged, like the Best Java Application Server (Free) and Best Java Application Server (Commercial) categories. But perhaps the wierdest part is how votes seem to have just vanished from the registry. It's like the US 2000/2004 elections all over again!

Let's look at the first category listed in the results: Best App Server. Here's a copy of the survey data posted here and here, which shows:

ProductCompanyPoints
Sun Java Application Server 9 Sun Microsystems 21
Enhydra enhydra.org 37
JOnAS ObjectWeb 53
WebSphere Application Server Community Edition IBM 345
Geronimo Apache Software Foundation 508
Project GlassFish Sun Microsystems 680
JBoss Application Server Red Hat Middleware 2150
ProductCompanyPoints
Interstage Application Server FUJITSU 4
uCosminexus Application Server Hitachi, Ltd. 18
EAServer Sybase Inc. 27
None None 27
Pramati Server 5.0 Pramati Technologies 29
NetWeaver SAP 42
JRun Adobe Systems Incorporated 47
Resin Caucho Technology, Inc. 88
Borland Enterprise Server Borland Software 162
Oracle Application Server Oracle 454
Sun Java System Application Server Sun Microsystems, Inc. 511
WebSphere Application Server IBM 536
WebLogic BEA Systems, Inc. 2169

* Note: These results are un-audited numbers. All voting results will be tabulated, and will go through Java Developer's Journal's extensive auditing process at the conclusion of voting. The results you see here may alter slightly or significantly as a result of that auditing process. JDJ makes no guarantees, and gives no endorsement to the live results which you see here.

So the winners would appear to be BEA's WebLogic and RedHat's JBoss with their commanding leads of over 2000 votes each, followed by Glassfish, the combined Websphere, Geronimo, and the combined Sun App Server in the 500-900s. But after tabulation and the "extensive auditing process", here's the final results, in the order reported with un-audited numbers added:

Best App Server
Java System Application Server, Sun Microsystems, 511 + 29 = 540
Oracle Fusion Middlware, 454?
Pramati Application Server, Pramati, 29
Borland Enterprise Server, 162

I guess those 5000+ votes for WebLogic, JBoss, Geronimo, Websphere, and even Glassfish didn't count. Maybe the voters used Diebold browsers to cast their votes. ;-)

Another turnout I find surprising is how Eclipse lost the Best IDE / Best Java IDE Environment category, despite its huge lead (1726 votes) over the second-place contender, MyEclipse (1040+12), also conspicuously absent from the final results. Even the third-place contender, BEA Workshop (563) seems to have not managed to keep its lead over TIBCO (not listed in unaudited data!), JetBrains (375), Oracle (338), and Sun (320, 185).

Anyway, auditing discrepancies aside (could the process be more transparent next year?), I'm also surprised that several of the more interesting (to me, anyway) categories have vanished.

EMF was in third place for the Best Java Modeling Tool and second for the Best Java Component, but neither of these categories made JDJ/SOAWorld's 'worth a mention' list, so I've still no idea how we did after the audit.

Does anyone who attended SOAWorld know if the other 9 categories were mentioned anywhere, and if so, who won? What about the Most Innovative Java Product category?

2007-07-15

Windows: A Survival Guide, Part 2

Back in Windows today (to try out Joost thanks to an invite from Lucas Bigeardel, component lead for EMF Search) so I thought I'd blog a few tips for my continuing survival here.

Out of the box, Pidgin 2.0 uses ESC as the 'close tab' key, which is a pain when you're used to Gaim 1.5's CTRL-W binding. So, to fix that discrepancy, shut down Pidgin, edit the file C:\Documents and Settings\Administrator\Application Data\.purple\accels and change this line:

(gtk_accel_path "<main>/Conversation/Close" "<Control>w")

For more on Pidgin keybindings, see the wiki.

Still have no perfect solution for the absence of Konqueror, but I found two alternatives worth mentioning:

1. Explorer XP: Tabbed file manager; session memory (reload previous tabs on startup), keybindings for all menu items
2. FileAnt: Left-right paned file manager w/ tabs; runs in task tray; session memory; reports disk/memory usage & folder sizes

2007-07-14

CTRL-ALT-DELETED!

To celebrate Friday the 13th, I went to Ontario Place's Molson Amphitheatre tonight to see Billy Talent tonight w/ SWMNBN, our niece Skylar, and and her friend Katrina. It was Katrina's first time at a concert, and everyone else's second time seeing BT play the home town crowd. Gotta support the local, ahem, talent, after all.

This gave me a chance to try out the video recording on my phone, along with Blogger-in-draft's new video upload feature.

Here's a couple of examples of the a/v quality achievable from my phone, after converting the .3gp files to .mpeg using Miksoft's Mobile Media Converter. I have to say I'm very impressed with the converter (yes, there's a version for linux!) but not so much with the source data captured by the phone.


Surrender


Where Is The Line?

In addition to video, I tried to take some photos throughout the show (including the encore where Ben got everyone to light up cellphones and lighters), but was repeatedly thwarted by the manic fans in front of us. (I can't fault them for having fun, however, as that was the point of the evening after all.) Good times.

2007-07-12

You got wiki in my webpage!

Inspired by Karl's post about transclusion of wiki content in other webpages, I thought I'd post my solution to this issue, as used to include all the EMF-tagged wiki documents in our docs page.

Caveats:

  1. it's prone to breakage (if the wiki markup engine changes, as it did recently to drop the "/index.php/" prefix on URLs in category pages); and
  2. it only works for turning category pages into Phoenix-styled bulleted lists, but that's pretty cool IMHO.

Here's the code:

/* convert a wiki category page into a series of <li> items */
function wikiCategoryToListItems($category)
{
 $collecting = false;
 $wiki_contents = "";
 
 // insert wiki content
 $host = "wiki.eclipse.org";
 $url = "/Category:" . $category;
 $vars = "";

 $header = "Host: $host\r\n";
 $header .= "User-Agent: PHP Script\r\n";
 $header .= "Content-Type: application/x-www-form-urlencoded\r\n";
 $header .= "Content-Length: ".strlen($vars)."\r\n";
 $header .= "Connection: close\r\n\r\n";

 $fp = fsockopen($host, 80, $errno, $errstr, 30);
 if (!$fp) {
  $out .=  "<li><i>$errstr ($errno)</i></li>\n";
 } else {
  fputs($fp, "GET $url"."?"."$vars  HTTP/1.1\r\n");
  fputs($fp, $header.$vars);
  while (!feof($fp)) {
   $wiki_contents .= fgets($fp, 128);
  }
  fclose($fp);
  $wiki_contents = explode("\n",$wiki_contents);
 }
 if ($wiki_contents && is_array($wiki_contents))
 {
  foreach ($wiki_contents as $wline)
  {
   $matches = null;

   // find stop line
   if (false !== strpos($wline, "printfooter"))
   {
    $collecting = false;
    break;
   }
   
   // collect link(s)
   if ($collecting && 
    preg_match_all("#<a href=\"/([^\"]+)\" title=\"([^\"]+)\">([^\<\>]+)</a>#", 
     $wline, $matches, PREG_SET_ORDER))
   {
    if (is_array($matches) && sizeof($matches)>0)
    {
     foreach ($matches as $match)
     {
      $out .= "<li><a href=\"http://wiki.eclipse.org/" . 
       $match[1] . "\" title=\"" . $match[2] . "\">" . $match[3] . "</a></li>\n";
     }
    }
   }
   
   // find start line
   if (false !== strpos($wline, "Articles in category \"". $category ."\""))
   { 
    $collecting = true;
   }
  }
 }
 return $out;
}

Then, I use this in our docs/index.php thus:

echo "<div id=\"midcolumn\">\n";
$contents = file("docs.xml");
$matches = null;
foreach ($contents as $line) { 
 if (false !== strpos($line, "<!-- DO NOT REMOVE: placeholder for wiki content -->"))
 {
  print wikiCategoryToListItems("EMF");
 } 
 else
 {
  print $line;
 }
}
echo "</div>\n";

And the result is that as new docs are tagged with [[Category:EMF]] in the wiki, they automatically appear in two places: http://wiki.eclipse.org/Category:EMF and http://www.eclipse.org/modeling/emf/docs/#references.

Obviously, this is not a full solution for full-page transclusion, like Karl's idea of using iframes, but does solve the issue of not having to produce lists of links on the website, when the wiki will do it for you, automatically.

Real World Bug Reporting

I've seen some kvetching lately in #eclipse that people (Eclipse users/customers) open bugs which never get resolved. Also, last night I saw an article from earlier this year discussing whether or not Vista would be fit for purpose, which spawned further kvetching about Microsoft and IBM and how much/little support there is from large organizations with large customer bases.

That said, let me present a real life example I just went through of how much worse life would be in OSS if we, the developers, treated our customers (the world at large) as if we were a monopoly with nothing to gain from helping you out, with an insipid insistence in NOT working together as a community.

Without further ado, I give you ... reporting a bug to Rogers Communications (unabridged).

 Original Message Follows:
 ------------------------
 name: Nick Boldt
 emailId: nickboldt
 replyEmail: nickboldt@gmail.com
 issue: account
 subcategory: other
 problemdescription: Logging in using Firefox or Konqueror browser
 (which normally works), I periodically get this error:

 Microsoft VBScript compilation  error '800a03e9'

 Out of memory

 /web_auth/verifpwd.asp, line 0

 Given you're running a Microsoft server -- maybe it's in need of a
 reboot? ;-)
Dear Nick Boldt,

Thank you for your email.

We understand your concern with this issue.

Rogers Yahoo! Hi-Speed Internet technical support representatives are qualified to diagnose and troubleshoot issues relating directly to our client software. We will therefore not be able to assist you with this matter, as it is beyond the scope of our support parameters.

For assistance regarding peripheral devices and other software products, we recommend contacting the vendor of the software in question. We apologize for any inconvenience.

If you have any further questions or comments regarding our service, please fill out the online form on our Customer Support page listed below or contact us by phone at 1-888-288-4663.

Regards,
name removed

There's nothing more annoying that people who tell you to solve your problem by repeating what you've already done (namely, to contact the software vendor, who in this case is in fact Rogers).

Guys, this is either one of two problems:

a) your webserver has a memory leak

b) your website relies on activeX controls and therefore doesn't (consistently) work with anything but Internet Explorer

In either case, it has nothing to do with my "client software" -- it's a website/webserver problem.

Please escalate this to your webmaster for review.

Cheers,

Nick

Granted, I was wrong about what the problem here was, but that's why I was suggesting escalation. Besides, I don't do ASP anymore.

Thank you for taking the time to write to us, we appreciate your use of online customer service.

We would be more that happy to assist you with your concerns.

We appreciate hearing from our customers and thank you for your comments and feedback. However, please note that the Rogers Yahoo! Hi-Speed cable internet service, currently supports only Internet Explorer and the Rogers Yahoo! browser. We apologize for the inconvenience.

Please let us know if we may be of further assistance to you.

If you have any further questions or comments regarding our service, please fill out the online form on our Customer Support page listed below or contact us by phone at 1-888-288-4663.

Regards,
name removed

I think the point you're (once again) missing is that it's not the browser (client) that's the problem but the website/webserver (server). Since this problem is intermittent and happens with multiple different browsers, I'm reasonably certain it's not a browser issue.

Further, since VBScript compilation occurs ON THE SERVER, not in the client browser (we're not talking about Javascript or another client-side scripting language, but a server-side one), and since my computer (and thus browsers) have ample memory to load other websites (2G of RAM and a 2G swap) an "Out of memory" error must be caused by a problem ON THE SERVER.

A quick Google search confirms that this is a known problem in VBScript, with an available workaround provided by Microsoft:

Microsoft VBScript compilation error 800a03e9 Out of memory
http://support.microsoft.com/kb/174634

That said, if you don't feel this issue needs to be escalated to your web server support staff for review, or that it's perfectly reasonable to expect that at least once a day your website won't let customers sign in (it worked fine a moment ago but not when I first filed this bug report), then by all means, please send me another FOAD "we only support MSIE" note. After all, it's much more fun (and costly for you!) for me to call your support staff instead of using the website for self-service every time it won't let me log in.

Yours,

Nick

Dear Nick Boldt,

Thank you for your email.

We sympathize with your situation.

Please provide further information on the nature of the problem you are experiencing and we will be happy to assist you in any way we can. Information such as error messages or codes you receive may be helpful in determining the source of the problem.

Please include when you are recieiving the error message as well, including the full URL of the site and any other information you can provide about it, may be useful as well.

If you have any further questions or comments regarding our service, please fill out the online form on our Customer Support page listed below or contact us by phone at 1-888-288-4663.

Regards,
name removed

Fine, if you need me to do your job for you, I will.

Steps to reproduce:

1) Go to http://www.shoprogers.com/homeen.asp?CustomerType=Consumer&Language=En

2) Enter username / password; check box for 'Rember my username'; click go or hit enter

3) Marvel at the following error which tends to occur at random intervals but particularly the second or third time I try to log in using the same browser session:

Microsoft VBScript compilation error \'800a03e9\'

Out of memory

/web_auth/verifpwd.asp, line 0

4) Per advice in http://support.microsoft.com/kb/174634, fix this bug.

5) Send me a $50 account credit for helping you fix your website. (Hey, I can dream.)

Cheers,

Nick

Dear Nick Boldt,

Thank you for taking the time to write to us, we appreciate your use of online customer service.

We would be more that happy to assist you with your concerns.

Although we thank you for your input regarding the Corporate Website, for further issues related to the Customer Service feature, we recommend contacting the support number 1-877-343-5745. They may then escalate such matters to the network administrators, who manage the server.

Please note that the email support team, addresses issues related to the various Rogers Cable services (such as Hi-Speed cable internet, Home Phone, Digital Cable TV, etc.).

If you have any further questions or comments regarding our service, please fill out the online form on our Customer Support page listed below or contact us by phone at 1-888-288-4663.

Regards,
name removed

1. It's not the Corporate Website, it's the public customer website. I'm not a corporate user, I'm just one of your many regular customers. I would have thought that would be obious from the URL: http://www.shoprogers.com/homeen.asp?CustomerType=Consumer&Language=En -- note the "CustomerType=Consumer" part.

2. Why should I have to report this issue a second time to yet another unsympathetic ear when you can just forward this email chain to the appropriate people for resolution? Not only do you need me to document (twice) the problem, but you now want me to report it twice as well? Are your webmaster(s) so afraid of email that they need a buffer of phone-answering lackeys around them to protect them from valid, documented, and solvable bug reports?

3. If your staff "addresses issues related to the various Rogers Cable services", then why can't you deal with issues related to me checking my Rogers bill online -- namely, the fact that about half the time I try to log in, I get the aforementioned error? Shouldn't supporting the website on which my bills reside be part of supporting Rogers Cable services?

Thanks,

Nick

Dear Nick,

Thank you for contacting us again, we appreciate the support you are providing us.

I have read the history behind your e-mail requests along with your current e-mail. First I would like to mention that the Rogers corporate web site is www.rogers.com. I'm sorry if that information was confusing to you.

Second, my question to you is why are you logging in 2-3 times in the same session? (from e-mail sent 7/11) You only need to log on once to do all your maintenance, bill viewing and reviewing of your account. This may be the root cause of the problem.

Third, you need to call the staff in our Internet support office. We are not qualified to help you. We do not do any technical troubleshooting via e-mail. Please understand that we are not equipped with this type of knowledge, we answer a limited range of questions about billing, sales, general inquiries. Please call 1-888-288-4663, they are available 24 hours a day, 7 days a week, and would be very happy and able to help you.

Thank you for contacting us, we hope that this issue will be resolved by our technicians.

Sincerely,
name removed

So, the moral here is that even when you provide steps to reproduce, solve the already-documented problem (via Google and the Microsoft KB), and request multiple times that they simply forward the email chain to someone who can address it, the oh-so-diligent lackeys at Rogers' email support desk refuse to help while claiming they're "happy to assist." Yeah, right.

Contrast that with the 1000s of bugs closed this year for Europa, and I'm sure you'll agree Open Source is much better than dealing with the close-source, closed-process, closed-access monopolies. Not only do we have several documented way to ask for support (bugzilla, newsgroups, mailing lists, IRC), but we actually use them, and the projects & member companies share information and collaborate to solve issues. Oh, and all this without having to pay $100+/mo for "service."

Really, what more could you want?

And now, back to Linux

This afternoon I had my power off while SWMNBN'd friend and co-worker (and apprentice electrician) fixed his wiring problem from last weekend's 2-day rewiring adventure. Apparently he's miswired the panel for the clothes dryer so while the lights were on, nobody was home (no heat produced and therefore no actual drying).

Blue Screens Of Death (In Stereo Where Available)

Anyway, after the juice came back on, I docked my T60p back in its crap-o-matic port replicator and attempted to tell it to open both the laptop and monitor screens. It complied... but not before BSOD'ing in stereo.

2007-07-11

Windows: A Survival Guide

*ponderous sigh*

After several months of enduring painful multi-monitor support (like cursor rendering problems and the inability to have monitors at different resolutions) and even more painful support for LEAP (for wifi access at work) ... I've finally given up on Linux on the Laptop and gone back to Windows XP.

This comes at a time when I no longer have my own cube at the Lab, because as Wassim puts it, I'm now "working from exile." Ed and I have elected to go mobile to save Big Blue some real estate costs (and to *cough* help the environment by staying home more).

Along with our switch to mobile status, Marcelo and Dave have moved too. They were made to give up their windows in D3 and now sit in Upper Middle Cube-land over in B3. So, to preserve the balance in the universe, Marcelo and I have gone back to Windows. (For those keeping score, that's 2 down, 2 up.)

Today was my first day at IBM without my own cube, sweet cube. Because of the aforementioned LEAP problems and because I no longer have a second monitor keeping my desk from blowing away in the breeze, it was time to revisit my Windows partition. I spent a portion of the afternoon setting up cygwin for the first time, including apt-cyg (a commandline way to update cygwin a la apt-get from the Debian/Ubuntu world). Then, in order to faciliate migrating my desktop settings from my linux partition into cygwin, I discovered Ext2 Installable File System For Windows.

This kicks ass and almost makes life in Windows bearable. It supports both ext2 and ext3 (if properly unmounted), and allows me to read & write into my linux partition via a mounted drive letter (though I haven't dared to try writing to it yet!). With this I can browse naturally w/ Windows Explorer or even with cygwin.

Next up to install was Pidgin, aka Gaim 2.0, which now supports Sametime natively, and imported my old Gaim 1.5 settings & buddy lists automatically... except where it lost my IRC buddies/chats and forgot which plugins I had enabled. Meh, close enough. Interface is a bit cleaner than the old Gaim 1.5, but I don't like the new icons.

Anyway, all the apps I use daily are back, bar one:

  • Mail/IM: Gaim + Meanwhile -> Pidgin; Notes 7; GMail; Thunderbird
  • Web: Firefox 1.5 or 2
  • Development: Eclipse 3.3 w/ phpeclipse, shelled, eclipsecolorer, mylyn; Cirrus or Beyond Compare w/ WINE -> Beyond Compare 2; (g)vim/Kate -> EditPlus
  • Console: Konsole --> cygwin console
  • Filesystem: Konqueror --> ???

Does anyone know of a filesystem browser that works for Windows that allows pane splitting and supports tabs & ssh? Anything's better than Windows Explorer, but I really miss Konqueror!

Maybe I'll have to play with Cygwin/X to see if I can get Konq back that way...

2007-07-07

What I Did On My Summer Vacation

With Europa out of the way (well, mostly), it was time to take a week off (give or take a few bugs and some IRC lurking), catch up on my TV viewing, do some paddling with my new kayak over at Rouge Beach, cycle a few dozen kms on my bike, play with my dogs and take in a few shows, including Transformers (the new Michael Bay whatsit at the Docks drive-in, thankfully free courtesy of the folks at Chevrolet), Vida, We Will Rock You, and a second helping of Evil Dead: The Musical (because you just HAVE to see it more than once).

Groovy.