Betalogue

Take the following excerpt from a simple word processor document:

Note where the insertion point is located (at the beginning of the second paragraph).

Now what do you think will happen when I start typing from that insertion point?

More specifically, do you expect the inserted text to be in bold or not?

Well, here is what happens:

The text is in bold. Why? The reason is simple. The previous paragraph ended with text in bold, and I only changed back to regular (not bold) at the beginning of the next paragraph before typing “Not bold.”

In such a situation, Pages, like other common word processors, “remembers” the exact location where the change from bold to regular took place, even though that location is now invisible to the user. And when the user places the insertion point in that same location (even though the user cannot tell that this is the location where the change from bold to regular was made) and starts typing, the resulting inserted text is… in bold.

On the other hand, if, when typing the first paragraph, I had changed back from bold to regular before inserting the paragraph mark (i.e. before typing Return), this problem would not be occurring now.

Yet, from a purely visual standpoint, the two different behaviours produce the exact same result. the first paragraph ends with something in bold (including the closing period) and the second paragraph begins with something not in bold.

Now my question is simple: Why can’t we get a word processor that is smart enough to understand that, in 99% of all cases, when the user goes to insert something at the beginning of the second paragraph, he expects that something to be in the character formatting of that paragraph (i.e. here the regular font) and not in the character formatting of the end of the previous paragraph (i.e. here the bold font)?

Is that really too much to ask?

Yes, I know that, from the perspective of the word processor program itself, the behaviour above makes sense. The change from bold to regular was made after inserting the paragraph mark, so to the word processor it makes sense to begin the second paragraph in bold. But it does not make sense from a purely visual point of view, which is all that the user can refer to once the typing has been done.

Now, of course, since I am more than familiar with this problem myself, I am usually careful to change my character formatting back before inserting a return character at the end of my paragraph. So for documents that I create myself, I don’t usually encounter the issue described above (except when unwanted text selection problems make it inevitable that I end up applying my formatting to unwanted trailing characters such as spaces and invisible return characters).

My real problem is that, in my job, I spend a lot of time editing other people’s documents. And the number of documents exhibiting the issue described above that I have to deal with is frustratingly high. There are probably various reasons for this, but the main one is of course the lack of intuitiveness of word processor interfaces in general and the lack of proper training of most word processor users.

It is for that same reason that there are no simple solutions to the problem. In the example above, it is merely a matter of switching back from bold to regular, which is not too painful. But far too often the character formatting changes are numerous: different font, different weight, different size, different colour, etc. It is simply too frustrating to have to change all these manually.

And reverting back to the paragraph’s underlying default character formatting by stripping all the inappropriate manual formatting with a “Remove Formatting” type of command usually is not an option either. More often than not, those word processor users who only change formatting options after inserting the paragraph mark are the same users who do not know how to use paragraph styles, and so apply document-wide formatting options manually as well. This means that the underlying default character formatting of the paragraph (i.e. the character formatting defined in the default paragraph style) does not match the actual default character formatting used in the document. Stripping the unwanted manual formatting also strips the manual formatting that needs to be preserved.

There is no easy solution here. The only tolerable approach that I have found when this happens to me is the following. Instead of putting my insertion point at the beginning of the paragraph, where I get the wrong character formatting when I start typing, I put it one or two characters to the right, inside the first word of the paragraph. I start typing what I want, then I go back to the beginning of the paragraph and remove the extra letters that I no longer want, and then I jump back after the first few letters that I have typed and resume typing.

It is utterly inelegant, but it is the only realistic solution that I have found.

Of course, as an experienced and demanding Mac user, I am profoundly bothered by inelegant solutions, even when I have no choice but to use them. Each time I have to use them, it adds a little bit to the low-level frustration that I experience on a continuous basis in my daily work due to the limitations and lack of intelligence of the tools that I am forced to use. I simply cannot resign myself and get used to this situation. I dream of smarter tools. I strongly believe that they are more than possible with today’s hardware and software. It is just a matter of “political” will. But the priorities of the market and of our current crop of “innovators” are obviously elsewhere.

In an ideal world, innovation would take place on a continuous basis on multiple fronts, including less “flashy” and less glamorous areas such as word processing and text editing. But that ideal world is far from today’s reality, where word processors are more or less just as dumb as they were 20 years ago (and not even faster!).

Try the following.

Create a test folder called “Test.”

Inside that folder, create another folder called “Enclosed.”

Inside the “Test” folder (next to the “Enclosed” folder), put two files of any kind. Let’s say the files are called “text1.txt” and “text2.txt”:

Select “text1.txt” and bring up the information for the file. Check the “Locked” option under “General”:

Now the file icon will have a small padlock icon in the bottom-left corner:

Now select both “text1.txt” and “text2.txt” in the window and try to drag them onto the “Enclosed” folder icon, in order to move them inside that folder.

Does the resulting behaviour strike you as normal?

Of course the “text1.txt” file is locked, so the Finder can’t move it.

But is that a good reason to:

1. put copies of both “text1.txt” and “text2.txt” inside the “Enclosed” folder
2. not remove either “text1.txt” or “text2.txt” from inside the parent folder
3. not give the user any feedback about what has just happened?

In my view, this is totally unacceptable. Compare that to what happens when you select both “text1.txt” and “text2.txt” and try to put them in the trash.

The Finder correctly detects that one of the files you are trying to trash is locked and asks you what you want to do:

Compared to this, the behaviour when attempting to move the two files to another folder is totally inconsistent and totally unacceptable. At the very least, there should be some sound feedback indicating that the files have been copied.

But what should really happen is that the Finder automatically moves the unlocked file and asks you what you want to do with the locked file.

Instead, we get… audiovisual silence, and a non-sensical behaviour.

I don’t care if my Finder is Carbon or Cocoa. This is unacceptable in any architecture.

UPDATE: A couple of readers have already written to point out that there is some visual feedback, in the form of a change of the cursor icon from the regular black arrow to the black arrow with the green blob containing the “+” sign that signals an impending copy.

While this is true, I had not noticed it when I wrote the article. The possible explanation is that the cursor changes only when your mouse pointer is over the destination folder. Before that, it remains the regular arrow. Since moving a couple of files to a folder is such a common task, I tend to do it very quickly, which explains that I didn’t notice the green blob, probably because I released the mouse button as soon as I was over the folder icon. So to me that’s not enough feedback.

In addition, in other circumstances where the green blob appears and the Finder performs a copy operation, it plays the “copy completed” sound once the copying is done. No such sound here, so if you don’t catch the green blob during the tiny fraction of a second that it appears, you have no other audiovisual feedback about what has just happened.

Finally, the fact remains that this behaviour is inconsistent with what happens when you try to move the files to the trash, and inconsistent with other comparable operations where some files in a selection of files have specific requirements. (See what happens when you try to copy a bunch of files to a destination that already contains a file with the same name as one in the selection.)

I still think it needs to be fixed.

Address Book has a useful command called “Go to My Card” (in the “Card” menu). Why is it useful? For a variety of reasons. For example, you might want to copy your own mailing address from your own card in order to paste it somewhere else.

Or, as often happens to me, you might need to refer to your own mailing address as it appears on your credit card statement, because such and such a web site requires it, and your mailing address as it appears on your statements is slightly different from the way you would normally write it, and you can never remember the exact differences, so you have it as a separate mailing address entry in your own card, and it’s handy to be able to double-check without having to dig out a recent credit card statement.

Unfortunately, Address Book suffers from one annoying flaw when it comes to using this “Go to My Card” command: It does not work what you have something in the search field.

When you type a text string in the search field, Address Book narrows down the list of addresses currently displayed in the “Name” column to those that contain the text string somewhere in their fields.

The problem is that selecting the “Go to My Card” command does not cause Address Book to automatically exit the search mode and jump to your own card. Instead, it just does… nothing. You are expected to understand that this means that you have to clear the search field and exit the search mode first, and then select the “Go to My Card” command again.

I don’t know about other Mac OS X users, but I often use the search field to search for something and then promptly forget to clear the search field before I move on to something else.

And so when I come back to Address Book and select “Go to My Card” command, I get frustrated because it does not automatically exit the search mode for me.

Granted, it’s a minor annoyance, but it’s representative of the lack of polish that the Address Book application still suffers from after all these years, even for those features that it does include. (It is also in dire need of some new features.)

In addition, there are other situations where Address Book does automatically exit the search mode, without requiring the user to manually empty the search field. For example, if you do a search and then, while viewing the search results, you click on a group in the left-most column, Address Book automatically empties the search field, exists the search mode, and displays the contents of the selected group.

To me, if it’s capable of automatically exiting the search mode in such a situation, then it should also be able to do so when the user selects the “Go to My Card” command.

Several years ago, I wrote about the numerous problems with the Edit mode in Address Book. To Apple’s credit, many of the problems described in that post have since been fixed. (I cannot reproduce them in Address Book in Snow Leopard.) But as the example above demonstrates, there are still other areas where Address Book’s user interface can turn out to be rather frustratingly rigid and user-hostile.

I should also point out that, sadly, one of the “hidden features” in Address Book’s search feature that I described in another post and made the rigidity of the search interface a bit more tolerable has since been removed from the application. Today, in Snow Leopard, if you start a search while a specific group is selected and there are no matches in that group, if you want to extend the search to all groups you have to click on “All contacts,” which exits the search mode and forces you to retype your search request.

Previously, Address Book was smart enough to exit the search mode, but not before it had put a list of search matches from all contacts in the “Name” column in answer to your search request. As one of the commenters on my blog post at the time noted, the disappearance of this feature appears to indicate that this was probably an “accidental side effect” and not an intentional thing on Apple’s part.

The bottom line here is that Address Book’s user interface is too rigid to be user-friendly. If you want to start a search in all your contacts, you need to first make sure that “All contacts” is selected. And if you want to use the “Go to My Card” command, you need to first make sure that there is nothing in the search field.

User interfaces that force you to do things in a specific order and do not offer an alternative when you forget to do things in the right order are simply not good enough in this day and age. I am all for iPods, iPads, iPuds, and what not, but come on, Apple, some improvements in the “smartness” of our systems—whether they are of the flat and portable variety or of the more traditional desktop variety—would be appreciated too. We are in 2010, after all. It is about time our computers started demonstrating a semblance of intelligence.

As a lone blogger in southwest Nova Scotia, I obviously only have anecdotal evidence. But as a Mac tech support person, I am consulted and hired by a number of local Mac users to provide help and advice, notably on their purchase of new machines.

And here is what I have observed just in the past few weeks.

A local newspaper decided to replace their entire fleet of aging Mac computers with brand new machines. It is a small business, so we are talking about one Mac Pro with a 30″ screen, one Mac mini to act as a server, and five iMacs.

For the last category, they decided to go with the 27″ iMac model, which is, on paper, a pretty convincing and fetching offering. And then they ordered all the machines from a national reseller, and the machines started trickling in after the holidays.

Of the five iMacs, two arrived with a cracked screen. Now, this might not be a flaw in the hardware itself, but at the very least it looks like a flaw in the packaging, which does not look to be sturdy enough.

Of course, these machines are going to be replaced at no cost, but it’s an additional hassle, and it’s definitely troubling to have the incident happen to more than one machine in such a small order.

Only Apple knows how many machines end up with a cracked screen when they are delivered, but maybe the fact that there apparently is a new delay in the delivery of 27″ iMacs confirms that there is indeed a problem and that Apple is trying to address it.

The Mac Pro seems to be OK, but the graphic designer is already having problems with the Magic Mouse, some of which might be tracked down to a problem with Bluetooth reception, which would seem to be a hardware flaw in the Mac Pro itself. (I am still investigating.)

(I am somewhat interested in the Magic Mouse myself, after having tried it during the installation of these new machines, but I won’t buy one for my new Mac Pro until I have good indication that there won’t be Bluetooth reception problems. Generally speaking, I distrust wireless peripherals, simply because I have heard of and seen myself so many glitches, if not actual problems and bugs, with the Bluetooth connections. Plus there is the issue of having to use batteries…)

Then I have another friend-cum-client who’s just bought herself a new 27″ iMac to replace her aging (and kernel-panicking) PowerMac G5. When she got it, the screen was not cracked, but she noticed right away that the machine was making way too much fan noise while running. She called me in, and sure enough, I was able to confirm that the noise was far from normal. I tried the usual troubleshooting steps (zapping the PRAM, etc.) to no avail.

The fan noise would start as soon as the machine was on (even before booting) and gradually ramp up, regardless of what the machine was being used for. I decided to do the hardware test by booting with the D key down (and was shocked to see that, even in these brand new Intel-based machines, Apple still uses a classic Mac OS 8/9-era user interface for the hardware test application).

Much to my surprise (I was already fearing a protracted battle with AppleCare…), the basic hardware test did produce an error called “4SNS/1/40000000:TH00-9.000.” The error code is not much help, but to me a brand new machine on which the hardware test gives an error is a dead-on-arrival machine and I advised my friend to get it replaced. She didn’t buy it directly from Apple, but we did call AppleCare and the representative immediately said that she should and would get it replaced by the retailer. He still gave us a case number just in case, and she’s currently in discussions with the retailer about getting a replacement.

Out of curiosity, I also did a Google search for the error code and found only one result, but a very interesting one. It’s a blog post at OWC (a US retailer from which I have bought stuff in the past, especially when US prices, even with the exchange rate and importing fees, were much lower than Canadian prices):

“Proprietary cable can put the brakes on upgrading Late ‘09 iMacs”

The error code does not appear in the body of the blog post itself, but in the comments, and the gist of the problem appears to be that 27″ iMacs customized with a 2 TB hard drive (the default configuration is with a 1 TB hard drive) that happens to be a Seagate Barracuda drive seem to have problems with thermal sensors and therefore exhibit abnormal fan activity. I quote:

BEWARE SEAGATE replacement drives don’t work.

I have a new 27″ iMac and the 2.0TB Seagate Barracuda LP. Once booted the fan will begin running at ~ 2500 RPM. It will continue to increase in speed until it reached very near max speed….. not quiet.

Running hardware diagnostics generates error 4SNS/1/40000000:TH00-9.000

Putting the 1TB Seagate back in restored the fans to normal operation.

Sure enough, my friend did get a customized iMac with a 2 TB hard drive. So I told her that, if she’s going to get it replaced with another iMac with a 2 TB hard drive, she’d better check and make sure the replacement machine does not have the same problem.

We will have to see how it goes, but again, it’s a rather annoying hassle, especially for people like us, who live in a community that is a three-hour drive from major authorized retailers.

Finally, there is, of course, the CPU heating issue with my own 2009 Mac Pro, which I wrote about a couple of weeks ago. The problem has yet to be solved. While it’s not a deal-breaker and my machine’s temperatures still stay easily within range (at least as far as I can tell, and unlike what’s happening to other victims), I am still rather annoyed by this.

I did submit a bug report to Apple a couple of weeks ago, and they responded to indicate that the bug was a “known issue” (original Bug ID# 7375647). But of course there is no indication of if or when the issue will be fixed. It is hopefully a software issue that can be addressed in a system or firmware update, but given that it seems to affect all 2009 Mac Pros, it has already been a full year with no noticeable action on Apple’s part.

All this adds up to a rather unnerving picture. Are Apple’s quality standards slipping? It is of course impossible to tell if you are not Apple and do not have access to hard data on machine returns, failures, complaints, etc. And I have not forgotten the not-so-distant years when Apple was shipping machines nicknamed “wind-tunnel” because of horrible noise issues. Or mooing MacBooks.

I suppose that, whenever a company ships millions of units, there are always going to be sporadic problems. Overall, Apple still seems to have good customer satisfaction ratings. But do we put up with more because of our minority situation or are things really as rosy as the customer surveys seem to indicate? And how much does the success of the iPod and the iPhone skew the ratings in Apple’s favour?

My own experience in the past few weeks certainly seems to indicate that Apple is far from immune from substantial quality control issues. One key aspect of such a situation is how promptly the company moves to acknowledge and then address these issues.

The culture of secrecy and (occasional) denial at Apple does not help in that respect. If Apple is aware of an issue such as the CPU overheating in Mac Pros and working on a fix, then it would be nice to be told more than just that it is a “known issue,” with no indication of when or how it will be fixed. (I have a number of bug reports that have remained in “known issue” status for years. Is that what is going to happen with this one? Or is Apple really taking it seriously and working on a fix that will be delivered soon?)

All this makes me a big uncomfortable both as a Mac user and as a Mac “expert” asked to give advice about hardware and software purchases. I am not about to switch to the “dark side” and recommend that people get crappy PCs instead of their Macs, of course, but at the same time I wish I could be more enthusiastic and wholehearted about my endorsement of Apple’s products. It certainly is not the case at this point in time.

Most of the time, when I have to work on a Word document sent to me by a client, I open it in Apple’s Pages, save it as a Pages document, work on it, and, in the end, I export the Pages document back in Word format and send that to my client (after checking it in Word itself and making minor adjustments if necessary).

Every once in a while, however, I have no choice but to work on the Word document with Word itself. Most of such cases are when documents contain a large number of tables (Pages’s table editing tools can be very frustrating) or Microsoft Office-specific graphics typically generated by Excel and inserted in the Word document or generated within Word itself.

These days, I am working on one such document. It contains a large number of tables, as well as numerous graphics such as this one:

My task, as a professional translator, is to translate the entire document, including such graphics. If the graphics are not editable, I have to advise my client that I will provide the translation in plain text underneath the graphic, and his graphic designer or whoever designed the document will have to update the graphic himself.

But if the graphic is editable, I have no excuse but to do the editing myself:

(Things might look a bit different for you depending on which version of Word you are using. I am using Word 2008 for Mac OS X. Microsoft’s developers keep changing the look of the frames and handles and other controls for editing graphics, without ever doing anything about the flaws in the actual behaviours of the controls.)

As you can see in this picture, the “labels” for the pie chart are editable text boxes that can be clicked on and “entered” in order to edit the text inside.

And that’s where the frustration starts.

See, in most computer systems, a double-click on a word is a shortcut for selecting the entire word. When you want to edit text, it is a pretty convenient shortcut. In the example above, my task is to replace “Strongly Agree” with “Tout à fait d’accord” on the first line and to add a non-breaking space between the number and the “%” symbol on the second line (which is required in French typography). My first instinct is to double-click on either “Strongly” or “Agree” and then expand the selection to the left or to the right (depending on where I have started) in order to select the entire line and type the French over the selection.

The problem is that some software developers have other ideas and think that double-click is a convenient shortcut for all kinds of other things. In Word 2008, for example, double-click on the text inside such a text box, instead of selecting the text, will bring up… a dialog box for changing the format of the text box:

Argh. This might make sense as a shortcut for accessing the “Format” dialog box for a graphic object that does not contain any text, and is therefore not editable to begin with. But unfortunately, Microsoft’s engineers have decided that it should apply to all objects, including text boxes.

So now, each time I want to edit the text inside a text box (and I have dozens of text boxes to go through), I have to click just once on the text inside the box to select the box first, and then pause for a short while, and then click on the text again to “enter” the field.

But even then, look at what happens:

Depending on the actual position of the mouse pointer, Word throws a tool tip in my face to tell me what the label is for, which I don’t give a flying hoot about, and that tool tip hides part of the text that I want to edit! And even if I start typing in the text box, the tool tip does not go away. The only solution is to move my hand back to the mouse and move the mouse pointer elsewhere, so that Word will show a tool tip for something else and remove this one from my sight.

Double argh.

At this point, you might think: “Why doesn’t Pierre just do a document-wide Find/Replace operation to replace all occurrences of ‘Strongly Agree’ with ‘Tout à fait d’accord,’ especially if he has dozens of those to do?”

Well, I would—except that Word’s Find/Replace command does not work with the text inside text boxes! It’s as if this text didn’t exist. Word can’t find it, and Word does not include it in its word count. (Another bonus for translators who rely on an accurate word count for billing purposes.)

So I actually have to do each and every graphic manually, one at a time.

Now that I have finally entered the text box and am able to edit the text (as long as the tool tip lets me see what I am doing and does not force me to type in the dark), I have to deal with Word’s esthetic brilliance.

See, while you are editing text inside a text box, Word is apparently unable to recalculate the size of the text box on the fly, and so it continues to display the current borders of the text box, even though they are not longer correct:

This image is supposed to represent text that is centered inside the box. And when I am finished typing I get more prettiness:

Now, somehow, Word has realized that it needs to widen the text box, but the centered text spills over the other edge of the box!

It’s only when I exit the text box altogether that Word finally adjusts its borders to accommodate the new content properly:

Except that—Darn! I’ve forgotten to add the non-breaking space before the “%” on the second line!

And so we start all over again… Double-click on the text box’s second line—oh no, here comes that “Format” dialog box again. Escape that. And try again, this time with a pause between the clicks. And so on. And so on. And so on.

Repeat the same scenario over dozens of similar text boxes, and you get an idea of how maddening this can get.

Oh, and by the way, even if you uncheck the “Show Paste Options buttons” option in Word’s preferences (under the “Edit” pane), Word 2008 still shows the “Paste Options” button each time you paste something inside a text box, thereby adding to the screen clutter and hiding even more of the text you are trying to edit.

And I am not even mentioning Word’s performance issues, which mean that, even on a 2009 Mac Pro with 12 GB of RAM, half of the time, when you try to expand the selection inside a text box, there is a slight but very noticeable delay before Word actually draws the proper selection highlighting. Or responds to your mouse click. Or to your keystroke. Which causes you to make more errors, and to have to repeat all these stupid actions even more.

And I am not mentioning document scrolling issues either. I have dozens of such graphics one after the other, so when I scroll down to edit the next one, it’s not always fully displayed in the visible area of the document window. Or sometimes, even when the graphic that I am editing is fully visible, editing it still causes Word to jump up or down while I am trying to edit a perfectly visible text box, and so I have to readjust my own positioning due to Word’s uncontrollable scrolling, and scroll back up or back down, which causes yet more headache-inducing jumping, because Word is not designed for the scroll ball on my Apple mouse and thinks it’s a scroll wheel and Word’s scrolling is way too sensitive, and there’s nothing that can be done to adjust that either!

Oh, and did I mention that the blinking I-beam cursor keeps disappearing while you are typing, making it even harder to keep track of what you are doing and avoid more unwanted errors and more repetition?

I could go on and on. Working with Word is just horrible, and it’s probably even worse in Mac OS X with Word 2008, although other versions of the software are probably affected by most of the flaws illustrated above as well.

It’s hard to believe that we are in 2010 and we still have to use such crappy, badly designed, horrendously flawed software. But we are, and we do. And we have Bill Gates and his clique to thank for it (although Adobe’s own software wizards are not far behind, and are getting closer and closer all the time).

It’s maddening, and it’s profoundly discouraging.

Following my recent posts about message filing issues in Mac OS X’s Mail and the feedback I have received by e-mail, I have a couple of additional comments to make.

One is about the third-party MsgFiler plug-in for Mail. In my previous post on this topic, I wrote:

Unlike LaunchBar, as far as I can tell, MsgFiler does not have a smart abbreviation engine that “learns” your preferred destinations based on abbreviated versions of their names. It just matches what you type to the actual names of the mailbox folders.

If, like me, you have lots of subfolders that start with the same word, then you need to type not just that whole word but the beginning of the next one (or hit the cursor keys multiple times to select the right item in the list of matches). For example, I have mailbox folders called “Betalogue – Admin,” “Betalogue – Mail,” “Betalogue – Feedback,” etc. In order to select one specific mailbox in that group, I have to type “Betalogue -” and then the first few letters of the next word. That’s a lot of typing.

The same issue applies to family members. I have a mailbox for each family member and, of course, many of them have the same last name, which I usually type first so that mailboxes are sorted alphabetically. Here again, in order to narrow things down to a specific mailbox, I have to type the whole last name and then the first few letters of the first name. Again, that’s a lot of typing.

A couple of readers wrote to indicate that you don’t necessarily have to type out the first word in the name of a mailbox. For example, to get to “Betalogue – Admin,” if I have multiple mailboxes starting with “Betalogue,” I could type “Admin” instead.

The problem with this, of course, is that I also have multiple mailboxes with the word “Admin” in their name. Similarly, when it comes to people’s names, if the last name is common to multiple mailboxes, I could type the first name. But here again, I have multiple acquaintances with the same last name, and I also often have multiple acquaintances with the same first name.

A more useful note with a view to reducing typing is that you don’t have to type out words in full. For example, to select “Betalogue – Admin” in MsgFiler, I just need to type out “bet ad.”

This is good to know, except, of course, that, in this particular case, in my mailbox drawer, “Betalogue – Admin” is not just a mailbox itself, but also a folder containing multiple subfolders. So I still get multiple results when typing “bet ad,” although the “Betalogue – Admin” mailbox itself is the first result (i.e. the one selected by default) and the other mailbox subfolders enclosed within the “Betalogue – Admin” mailbox folder are listed afterwards and can be selected directly by typing portions of the words in their own name.

It all depends on the names used for one’s various mailboxes and mailbox folders, of course. I have a long-established system that I use and I have no desire to change it after all these years. It means that, in my case, MsgFiler might require more typing than it would for other people with other naming systems.

It can still be argued, however, that, even in my situation, MsgFiler is still more convenient and efficient than the default message filing options available in Mail. I will have to give it another try over a number of days and see if I can really get used to it and find it more efficient than Mail’s options and consider it worth the expense and the additional burden of having to manage yet another third-party plug-in. (I am not giving up Mail Act-On for the single keystroke shortcuts that I have defined for my most common mailbox destinations.)

Speaking of Mail Act-On, another reader wrote to point out that Mail Act-On has its own feature for filing messages in any mailbox, aside from its rule-based keyboard shortcuts that can be defined by the user and used for specific mailboxes.

Under “Menus” in the Mail Act-On preference pane, there is a shortcut for a command named “Move Messages.” The default shortcut is F3.

However, as I wrote in my reply to this reader, I have to report that, on my system, that command is unusable. On my system, the window that pops up for message filing is totally unresponsive, or responds to key strokes (or mouse actions) after a delay of many seconds. It is unacceptable.

It was already like this a couple of years back when I first tried the feature, and it is still like this on my 2009 Mac Pro.

The reason might be that I have a large number of mailboxes (nearly 2,000). The reader who wrote to me and who uses this feature all the time only has 40 mailboxes. He has tens of thousands of messages, but only a limited number of mailboxes, and mostly relies on Mail’s search feature to locate individual messages.

I obviously have a different approach. And it looks like Mail Act-On’s “Move Messages” command is useless when the number of mailboxes is large. (Other Mail Act-On features still work fine.)

In closing, I think it’s fair to say that, for people frustrated with Mail’s built-in options for moving messages around, there are third-party alternatives, but that these third-party alternatives are an extra expense and might not work for everyone. My wish is still for Apple to include more efficient built-in solutions, but until that happens, I will definitely give MsgFiler another try.

After reading my post on spring-loading and scrolling in Mail’s mailbox drawer, a couple of Betalogue readers wrote to suggest that I use a Mail plug-in called MsgFiler.

Unlike AppleScript scripts or Mail-Act On rules, MsgFiler can be used to move messages to any mailbox in your mailbox drawer. The idea is that you define an application-wide keyboard shortcut for the “Move with MsgFiler” command that gets added to the “Message” menu in Mail after you’ve installed the plug-in.

With this keyboard shortcut, you invoke a small window where you can start typing the name of the intended destination, i.e. of the mailbox folder to which you want to move the selected message(s). MsgFiler finds and lists all mailboxes matching your typing, much like a utility like LaunchBar, which I use all the time.

Once you have narrowed down the list in MsgFiler to your intended destination (either by typing more characters or by using the cursor keys to select the desired item in the list), you can just move the selection to the destination by pressing Return.

MsgFiler works as advertised and is reasonably priced. But I had tried it a while back and decided not to purchase it because, in my experience, it still required too many keystrokes to obtain the desired destination.

Unlike LaunchBar, as far as I can tell, MsgFiler does not have a smart abbreviation engine that “learns” your preferred destinations based on abbreviated versions of their names. It just matches what you type to the actual names of the mailbox folders.

If, like me, you have lots of subfolders that start with the same word, then you need to type not just that whole word but the beginning of the next one (or hit the cursor keys multiple times to select the right item in the list of matches). For example, I have mailbox folders called “Betalogue – Admin,” “Betalogue – Mail,” “Betalogue – Feedback,” etc. In order to select one specific mailbox in that group, I have to type “Betalogue -” and then the first few letters of the next word. That’s a lot of typing.

The same issue applies to family members. I have a mailbox for each family member and, of course, many of them have the same last name, which I usually type first so that mailboxes are sorted alphabetically. Here again, in order to narrow things down to a specific mailbox, I have to type the whole last name and then the first few letters of the first name. Again, that’s a lot of typing.

That said, I will definitely give MsgFiler another try and see if I can live with its limitations and still improve the overall experience of moving stuff around in Mail.

(I am also reluctant to spend additional money and of course increase the risk of incompatibilities and instability in Mail by adding yet another plug-in.)

There is something slightly annoying about the way the mailbox drawer in Mac OS X’s Mail responds to user actions when the user tries to drag a message or a selection of a messages to a specific mailbox folder nested somewhere within his mailbox folder hierarchy.

Here’s the situation as I encounter it on a daily basis. By default, I keep all my main mailbox folders in the “ON MY MAC” section of my mailbox drawer collapsed (i.e. closed) so that I can see them all. The last one in the list, as illustrated in the picture below, is a folder called “Work”:

As you can see, when all my folders are collapsed, the last one is still visible above the bottom edge of my mailbox drawer. But of course, if I expand (open) any of these main folders, the list of enclosed folders extends far beyond the bottom edge of the drawer and I have to scroll down to see them all:

Now let’s go back to the initial situation, i.e. when all the folders are collapsed:

If I now drag a selection of messages from my message list to the “Work” folder and wait for a fraction of a second, the automatic spring-loading mechanism causes Mail to expand the folder to reveal its contents:

So far so good. But let’s say the enclosed folder to which I want to drag my selection is lower down in the list. In order to reach it, I will have to scroll down the list of enclosed folder. But I have my index finger on my primary mouse button, because I am dragging stuff. So I cannot use the scroll wheel for this.

The expected standard behaviour in such a case is that, when I drag my selection near the bottom edge of the drawer and wait for a fraction of a second, Mail will start scrolling down the list to reveal the rest of the enclosed folders.

But that’s where there is a problem. The “near the bottom edge of the drawer” area that I am supposed to target with my dragging is not clearly defined and, more important, it overlaps with the area where one of the enclosed subfolders appears (in this case, the folder called “Graphic Designers”).

So what’s going to happen when I try to drag my selection near the bottom edge of the drawer and wait for a fraction of a second? Instead of simply scrolling down the list, Mail will also activate… the automatic spring-loading mechanism for the “Graphic Designers” folder:

But that’s not what I want! I do not want to drag my selection to a sub-subfolder inside the “Graphic Designers” folder! I want to drag my selection to a subfolder after the “Graphic Designers” folder.

Unfortunately, because of the overlap between the area that triggers scrolling and the area that triggers spring-loading, I cannot have one without the other. And the end result is that I have to scroll down through the list of sub-subfolders first before I can reach the rest of the list of subfolders.

Fortunately, Mail doesn’t continue this silly behaviour ad infinitum. Otherwise, it would be a nightmare. It only confuses scrolling and spring-loading once, presumably because after that, since the drawer is scrolling down, I am not longer lingering on any given position long enough to trigger spring-loading, until I actually choose to stop the scrolling by lifting my selection back up a bit.

And once I’ve actually dropped what I was dragging in the desired location and released the mouse button, of course Mail collapses everything back up and I am back to normal.

Still, it’s quite frustrating that, almost every time I want to archive something somewhere inside my hierarchy of mailbox folders, I am forced to deal with this confusion between scrolling and spring-loading and scroll through more things than I should have to.

Given that this is a problem that only affects people with a long list of folders in their mailbox drawer, and not the majority of users who never archive their mail outside their inbox and let it balloon up to thousands of messages, I wonder whether Apple is even aware that there is a usability issue here and, if so, whether they can be bothered to think about a way to fix it.

Admittedly, it’s a small problem, and one that I have learned to live with. But it still grates that I always have to do more scrolling than should be necessary and deal with unwanted spring-loading which, as far as I can tell, cannot be turned off (and one wouldn’t want to turn it off anyway, since it is necessary; the only option would be to turn it off only in this particular area of the drawer).

(And yes, I know that I can avoid all this by using right-click to browse my mailbox folder hierarchy via the “Move To” contextual menu command, but that comes with its own set of problems, because the menu font is bigger, and hierarchical menus require greater mouse movements with a greater risk of overshooting and having to start all over again.)

I have a few AppleScript scripts and keyboard-activated Mail Act-On rules for moving stuff to my most frequently used destinations, which helps alleviate the problem, but I cannot have a rule or script for every possible destination, so I still have to deal with this problem on a daily basis. And that, in my book, makes it more than a “minor” annoyance.

I first read about this a few days ago, over at MacInTouch. I was able to confirm this myself, so I thought I should report on it and add to the chorus of voices asking for a fix.

I bought a new Mac Pro (2 x 2.26 GHz Quad-Core Intel Xeon) with a few bells and whistles back in the fall of 2009. I am quite pleased with the machine and haven’t had any significant complaints so far.

But it looks like most, if not all 2009 Mac Pros are affected by a bug relating to audio playback in Mac OS X. There is a thread about it on the MacRumors forums, another one on Apple’s own Discussions forums, and even an article dated Dec. 16 about it at MacNN.

I personally tried the following on my own Mac Pro.

I left the machine running but doing only background tasks (checking for new e-mail, basically) for a couple of hours in the evening. iTunes was open, but not playing anything.

I came back to the machine and checked the temperatures with Temperature Monitor:

CPU A Heatsink: 30 °C / 86 °F
CPU A Temp Diode: 35 °C / 95 °F
CPU B Heatsink: 27 °C / 81 °F
CPU B Temp Diode: 30 °C / 86 °F

Then I simply started playing an MP3 file in iTunes and monitoring the temp values. Within a few minutes, all four values had increased significantly and eventually stabilized around the following values:

CPU A Heatsink: 45 °C / 113 °F
CPU A Temp Diode: 52 °C / 126 °F
CPU B Heatsink: 38 °C / 100 °F
CPU B Temp Diode: 44 °C / 111 °F

While the values are lower than those mentioned by other MacInTouch readers using early 2009 machines, they are still ridiculously high for just playing a song in iTunes. That’s an average increase of approximately 14 °C just for playing an MP3 track in iTunes. Imagine if the temperature of your iPod in your pocket increased by that much (on top of its normal operating temperature) each time you played music with it.

I was also able to verify the following:

A similar temperature increase occurs when I play stuff with audio in Safari (video clips, etc.). The temperatures drop as soon as I stop playing audio.

A similar temperature increase occurs as soon as I launch Logic Pro, even if I don’t open anything in the application and therefore do not use the application for anything at all.

And again a similar increase occurs as soon as I launch Elgato Video Capture, right at the first screen, before I even start configuring or recording anything.

I was also able to replicate the problem when playing a music track in iTunes under Windows XP in a virtual PC environment in VMWare.

The problem is therefore not just with iTunes and not strictly only with audio playback. There’s obviously something about both Logic Pro and Elgato that activates whichever OS component causes the problem as soon as the applications are launched, even before any audio is played by the user.

I did not try to play a music track in iTunes under Windows XP after rebooting under Boot Camp, because I do not have a Boot Camp volume on this Mac Pro. (I have one on another, older Mac Pro.) But other users report not seeing the problem on the same machine under Windows.

It therefore looks like it’s a software problem, albeit one that affects only 2009 Mac Pros, when running Mac OS X. People seem to suspect a faulty kernel extension. This would be good news in that it should be fixable via a software update.

The bad news is that, in spite of the number of reports mentioned above (and presumably numerous direct bug reports to Apple, to which I have added mine), after several months there is still no fix in sight.

To be honest, I hadn’t noticed anything about my machine’s temperatures until I read about this on MacInTouch. I hadn’t installed Temperature Monitor before last week, so I hadn’t noticed these significant temperature variations. As indicated by others, while abnormally high, the values are still within normal range, and I haven’t noticed any audio-related kernel panics in the several months I have been using this new Mac Pro.

I also use Logic Pro on a regular basis and haven’t noticed any issues with that either, even with several other processes running in the background.

But like other people, I am concerned about the durability of my machine and about wasteful power consumption. It just does not seem right that a simple task such as playing back an MP3 file (something that an iPod achieves on what I am sure is a tiny fraction of the power consumption reflected by the temperatures above) should require so much electrical power and generate such an increase in temperature. It definitely makes me feel uncomfortable about running audio-related tasks on my Mac.

I can understand it not being given the highest priority at Apple, but I sure hope that the problem can be addressed soon, especially if it’s been around for a while now.

It is one of those small details that help restore one’s faith in Apple’s on-going ability and willingness to fine-tune its user interface in the name of consistency and usability.

In the Finder, in column view, when a file preview is displayed in the last column on the right because a file is selected in the next column on the left, there is a “More Info…“ button underneath the file information:

This button is effectively the equivalent of the “Get Info” menu command or command-I keyboard shortcut. By default, it opens a new file information window for the selected file.

But the the “Get Info” menu command and command-I keyboard shortcut can be modified using the Option key. If you hold the Option key down while selecting the “Get Info” menu command or press command-option-I, instead of the “Get Info” command, you get the “Show/Hide Inspector” command, which, as the name implies, shows or hides the file inspector window that stays open in the foreground in the Finder and whose content is updated each time you select something else in the Finder, to display the file information details for the new selection.

Prior to Mac OS X 10.6 (Snow Leopard), holding the Option key down while clicking on the “More Info…” button in the preview column in column view did not have the same effect. The Finder effectively ignored the fact that you were holding the Option key and simply displayed the stand-alone file information window for the selection, like “Get Info.”

I am pleased to report that, in Snow Leopard, holding the Option key down while clicking on the “More Info…” button, while it does not modify the look of the button in any way, does change its function equivalent from “Get Info” to “Show/Hide Inspector.”

This makes the behaviour of the button consistent with the behaviour of the menu command and of the keyboard shortcut, and it is a much appreciated improvement. Prior to Mac OS X 10.6, even though I knew that it didn’t work, sometimes I just couldn’t help but instinctively try to option-click on the “More Info…” button to open the inspector. It just made sense, and it didn’t make any sense that the button’s behaviour was not consistent.

Yes, it’s a very small thing, but it’s not insignificant, and it demonstrates that at least some people at Apple do pay attention to bug reports and enhancement requests and try to eventually implement the suggestions if they are valid.

It used to be that Mac OS X’s Finder application would only save its current status—i.e. its currently open windows, along with each window’s size and position—when quit.

Since Mac OS X’s Finder does not have a “Quit Finder” command in its application window, the only way to quit it is to go through the Activity Monitor application (where you can select the “Finder” process and click on the “Quit Process” toolbar button) or to use a third-party utility such as TinkerTool to add a “Quit Finder” command to the “Finder” menu in the Finder.

But even the use of such a “Quit Finder” command did not solve the fundamental problem, i.e. that the Finder would only “remember” its currently open windows when quit. Unless you remembered to quit and relaunch the Finder on a regular basis, there was no way to force the Finder to save its status more often.

This wouldn’t have been a problem in an ideal world where computers never freeze and require a hard reset and the Finder application never crashes or freezes (requiring force-quitting, which, unlike regular quitting, does not save the current status, for obvious reasons). In the real world, however, the Finder does occasionally freeze or crash and even Mac OS X itself occasionally throws a fit, in the form of the dreaded multilingual kernel panic screen that requires a hard reset.

In all those situations, until recently, the current status of the Finder would be lost and, upon relaunching, it would reopen the windows that were open the last time it was quit “properly,” which could be several weeks before and might have no relation to what you were last busy doing in the Finder.

Well, something has obviously changed in this department, because, after several weeks of perfect system stability (I didn’t have to restart my Mac Pro once during the holiday season), my computer suddenly had a kernel panic this afternoon. (It might be related to my use of the Elgato Video Capture device, especially since the kernel panic report specifically mentioned an error related to the “com.apple.driver.AppleUSBAudio” extension and I also see all kinds of “IOAudioStream“-related errors in my kernel log.)

Much to my surprise, upon restarting the machine, I found that, when relaunching, the Finder reopened, not all the windows from the last time I properly quit the Finder, which was some time in December, but all the windows that were open today, just before the kernel panic.

In other words, the Finder now seems to “remember” its current status, even in the event of a system crash and hard reset.

This is a welcome improvement. In the real world, crashes do still happen (though thankfully much less than they used to do ten years ago!), and it is always frustrating to have to painstakingly rebuild one’s work environment the way it was before the crash. Now, as far as I can tell, there is one more thing that does not need to be rebuilt, namely one’s current window setup in the Finder.

I don’t know exactly when this improvement was introduced. I don’t remember noticing it before today, and I have had crashes in Snow Leopard before, so it cannot have been in Snow Leopard itself, i.e. in Mac OS X 10.6.0. It might have been introduced in Mac OS X 10.6.1 or Mac OS X 10.6.2. I also need to experience a few more crashes in order to confirm that it does indeed happen every time.

Out of curiosity, I just made a slight change in my window setup and then option-clicked on the Finder’s Dock icon and chose “Relaunch,” and the slight change that I had just made was not preserved when the Finder was relaunched. It returned to the previous state, i.e. the state that was correctly preserved when the kernel panic happened.

As far as I can tell, using this “Relaunch” command is the equivalent of force-quitting the Finder, which should mimic what happens in the event of a crash.

So it does not look like the Finder saves its current status all the time. Maybe it saves it at predefined intervals of a few minutes or something. It’s hard to establish for sure.

But it certainly looks like the Finder does save its status more often than it used to do, i.e. not just when at the time it is quit “properly,” when the user logs out or restarts his machine or when the “Quit Process” button is used in Activity Monitor.

Last year, I took advantage of a promotional offer at Mostly Digital to purchase a USB device called the Elgato Video Capture. I have a stash of old VHS tapes (mostly Prince-related stuff) that I have been meaning to convert to a digital format for years and it seemed that this particular device would be a decent solution for this particular purpose.

I finally got around to trying the device during the holiday season. The hardware setup was pretty simple, since I already had the required equipment nearby. I just needed to free a USB port on my Mac Pro.

The first snag came when I tried to install the required software. Of course, the version of the Elgato Video Capture application that came on a CD with the device was out of date, so I went to the Elgato support page for the latest application update.

As you can see if you visit this page yourself, there is no link to download the software there. The only solution, as far as I can tell, is to run the version that you currently have and use its own built-in software update feature, not just to check for an available update, but also to actually download it.

And of course, once the application has finished downloading the software update and installed it, the update is nowhere to be found on your hard drive. So the next time you need to install the software, you’ll have to go through the entire process all over again (unless you assume that the application is all that is needed and you can just archive a copy of the application itself).

I just hate it when software designers assume that a built-in software update mechanism is the best one for all their users and don’t bother to offer an alternative to those who might want to archive their software downloads.

At 30+ MB, the application is not a tiny download that only takes a fraction of a second.

Anyway, I proceeded with the installation of the update and relaunched the application. Everything went smoothly and there was no requirement to restart the computer, which was much appreciated. In that respect, it was pretty much just plug-and-play.

I then tried to digitize a VHS tape. I was a bit disappointed to see that the only interface offered by the application was a wizard/assistant type of thing, with step-by-step instructions divided into a series of “screens”: Get Started, Connect Video, Connect Audio, Record, Trim, Finish Up. There was no “manual” alternative for users who might want to do things in a different order.

But I figured that, as long as it worked properly, I didn’t mind too much.

So I went ahead and gave a name to my tape in “Get Started,” checked that I had a video signal with the right aspect ratio in “Connect Video,” checked that I had a good audio signal in “Connect Audio,” and then started to record the tape, with the option to stop the recording automatically after two hours.

I watched a bit of it, then checked the “Mute Sound” option on the “Recording” screen and left the thing to run in the background while I was doing something else on the Mac. (Thankfully, the application makes full use of Mac OS X’s multiprocessing capabilities and spreads the processing load across multiple cores, which guarantees smooth operation in the background.)

When the two hours were up, I switched back to the application and let it save the recording in MP4 format. (Unfortunately, the file saving process can make your entire system less responsive, but that’s probably due to the limitations of Mac OS X’s file system itself, because I have experienced the same thing with other hard-disk intensive tasks.) And then I tried to do another tape.

I went through all the steps again, but when I got to the “Connect Audio” screen, I discovered that I just couldn’t get any sound out of the device:

Normally, when you have sound, you should be able to hear it and you should see the animated level indicators in the corresponding location in the screen:

But in this case, while I could see the tape’s picture, I couldn’t hear any sound. I checked on my amplifier to make sure that there was indeed sound coming from the VCR, and there was. So the problem appeared to be with the Elgato Video Capture device itself.

I tried quitting and relaunching the application. I tried unplugging and replugging the USB device. I tried connecting the VCR’s audio-out directly to the audio-in connectors of the USB device. All to no avail.

Finally, I went to the Elgato web site looking for help. I checked the FAQs to make sure I wasn’t missing something obvious. Then I had to register with Elgato in order to be able to submit a support request with a form. They guaranteed an answer within 1-3 business days, but of course it was December 30, so I didn’t expect an answer until after New Year’s Day.

In the meantime, I decided to check the support forums. I didn’t find any reference to a similar problem in the forum for the Elgato Video Capture, so I decided to post a plea for help. Much to my frustration, I discovered that I had to register again, this time with the forums section of the Elgato site, and that my existing registration with the support section of the Elgato wasn’t recognized. Worse still, the registration required an administrator’s activation.

Fortunately, I did receive a message confirming the activation of my forum registration four hours later, which shows that at least one person was still working late in the afternoon at Elgato on Dec. 30.

The next day, I posted my plea for help, and then moved on to other things, namely enjoying my time off work during this holiday season (which also explains the dearth of posts lately).

I never did get any replies. On the other hand, I did eventually (yesterday) get an answer to my support request:

I’m sorry to hear about the audio problem you are having with the Elgato Video Capture.

It’s possible that if you selected to mute the audio playback during the recording of the previous tape, the software has remembered that setting which would mute the audio during the recording setup. Have you followed the steps all the way to the recording step where the mute checkbox is shown?

This would not affect the final export, so if you record a small portion of the new tape and play back the resulting file in QuickTime Player, does the recording have sound?

Thank you for providing the requested information. We will review your answers and get back to you as soon as we can.

That’s when I first started to suspect that the whole thing was not a bug or a defect in the device, but a pretty blatant user interface flaw. And indeed, after further testing, I was able to confirm the following.

If, using the Elgato Video Capture assistant/application, you activate at any point during the recording the “Mute sound” function, this function stays on even after you have finished your work with this particular project. Even if you quit and relaunch the Elgato Video Capture application, even if you unplug and replug the USB device, the muting remains active… until you uncheck that particular “Mute sound” checkbox in the “Recording” screen.

The absurdity of this user interface is threefold.

First of all, if you stop recording, the application proceeds to the next screen automatically, which gives you no chance to uncheck the “Mute sound” box. And the muting remains active, not just for this particular project, but for the next one too.

But the “Mute sound” control only appears on the “Recording” screen. So if you start a new project, and get to the “Connect Audio” screen (which comes before the “Recording” screen), and try to check if you are getting any sound, you can’t hear anything.

And worse still: when the sound is muted, the Elgato application does not animate the sound level indicators in the “Connect Audio” screen, even though the device is properly connected and is receiving the audio just fine.

In other words, you are expected to ignore what the “Connect Audio” screen seems to indicate, i.e. that the audio is not working, and move to the next screen, and uncheck the “Mute sound” box, and then go back to the previous screen to check the audio connection—although by then, of course, you can also hear the audio in the “Recording” screen, which confirms that it’s working, except that you don’t get the sound level indicators on that screen.

This is rather absurd. And of course, since I didn’t anticipate such a level of absurdity, I had to contact tech support and wait several days to find out that what I thought was a bug or a defect in the device was simply a bad UI problem.

As far as I am concerned, the whole situation just confirms my views on assistant/wizard-like interfaces in general. They are too restrictive, even when the process only involves a few simple steps, because they impose an order on things that the user should be able to do in whichever order he pleases—for instance, I should be able to check the audio connection first and then the video connection, if I wanted to—and because, if the software designer is not particular careful, they can easily introduce absurd behaviours such as the one described above.

If the sound muting feature is an application-wide setting that is even saved as a preference when you quit and relaunch the application, then this setting should be accessible in all screens—at the very least, in the “Connect Audio” screen—and not just in the final “Recording” screen.

And since the audio muting only applies to the sound played by the Elgato Video Capture application in the Mac environment and does not affect the recording at all, there is simply no way that the muting should affect the sound level indicators on the “Connect Audio” screen.

Of course, now that I know this flaw in the UI, I can work around it. But that’s not an excuse. The application should be properly designed in the first place. Application-wide settings should not be presented as screen-specific features. And you’d think that engineers responsible for an audio/video recording application would be particularly aware of the need to distinguish between the sound recorded by the application from the USB device and the sound played by the application in the Mac environment. (Rogue Amoeba’s Audio Hijack Pro does this just fine.)

More generally, the case of the Elgato Video Capture application confirms how difficult it can be to provide the right hardware/software combination. Apple has a lot of experience with this. Other companies, evidently much less. That a software engineer would think that such a UI is good enough to release as a 1.0.x product speaks volumes about the standards of software designers in general.

Don’t get me wrong: The Elgato Video Capture application is otherwise a decent looking Mac OS X application and it works pretty well, once you know that thing about muting. The application is fairly limited, but at least it’s not a horrible non-Mac-like thing like the type of stuff you usually get with other providers of consumer-level hardware/software combinations, such as printer suppliers (HP, I am looking at you) and scanner makers (HP, Canon, Epson, etc.).

And yes, I am quite grateful that we live in an age where some many tasks that used to be the reserved domain of the “pros” are now available to the rest of use thanks to the powers of computing technology. But that does not mean we users should have to put up with absurd design decisions in consumer-level products and accept to live with badly designed technology all around us, which is pretty much the situation today. (I am not just talking about computers here, but also PVRs, cars, kitchen appliances, etc.) On the contrary, it is precisely consumer-level products that should be particularly well-designed and user-friendly.

There is one other pretty obvious flaw with the Elgato Video Capture application, which is that, when you do not just move the application to the background, but actually hide it in your Mac environment while it is recording, the video portion of the signal ceases to be rendered properly. Instead of the moving frames of the video recording, you just get a frozen frame of the last picture that the application was displaying at the time you hid the application.

At that point, you might think that the recording is no longer working, but that is not the case. The sound is still playing properly (provided that you haven’t muted it) and the “Duration” display with the length of the recording and the file size is still being updated from time to time. But the video is frozen. You can actually force the Elgato application to update the video by hiding the application again and then bringing it back to the foreground. But all you’ll get is an updated frozen frame. As far as I can tell, there is no way to restore the proper video playback without stopping the recording altogether and starting again from scratch.

But the important thing is that, even if the playback is frozen, the recording is still working as expected, and you’ll get your fully playable audio/video recording once the process is complete. Again, it’s not a bug that I’d expect to find in a 1.0.x product, but it’s not a deal-breaker either.

To conclude, I wouldn’t discourage people from buying the Elgato Video Capture. It is reasonably priced (I paid $120 CDN + tax for it last year) and it does what it says on the box. Just be prepared to live with the “glitches” described above, which force you to try and remember to uncheck the “Mute sound” box when the recording is over and not to use the “Hide” function in Mac OS X if you don’t want to lose the ability to watch the video playback while the recording is taking place.

And there’s always hope that Elgato will fix these things in a software update… (I am sending a link to this post to the tech support person who wrote to me.)

(UPDATE: I just got another e-mail from Elgato—in answer to an earlier e-mail from me—saying that “[t]he persistence of the audio muting is a bug which should be corrected in a software update so that it resets when the software is relaunched.” So there is hope, although I believe it’s a design flaw and not a bug per se, and the interface itself needs to be rethought.)

“Sticky Patch” (MP3 file – 160 kbps – size: 6.5 MB – duration: 05:22)

“Knock Them Down” (MP3 file – 160 kbps – size: 4.8 MB – duration: 04:00)

More LATEXT music here.

Today, I got an e-mail from my dad in France about Mac OS X’s Mail. While he was a math teacher by training, my dad was there at the beginning of the personal computing revolution in the late 70s and early 80s and ended up teaching various computer science courses at the community college level in France. (That’s also how I caught the computing bug as a teenager.)

For a long period in the 80s and 90s, for his personal computing needs at home, he was a Windows guy, and that was the source of many passionate, yet light-hearted Mac vs. PC arguments between him and me (and my brother, another long-time Mac user).

Then, when Apple came up with Mac OS X and its Unix underpinnings, my dad made the switch and, as far as I know, he hasn’t regretted it one bit. He takes care of his own tech support needs. But from time to time, since I am the “Mac expert” in the family, I get an e-mail from him about something that puzzles him.

And so today I got this e-mail about Mail. The gist of it was that he wondered why, even though all his settings were correct, sometimes Mail didn’t check for new mail for certain accounts (he has several) when it was supposed to do so, i.e. during the automatic mail checking procedure that takes place at regular intervals. And he wondered whether that was linked to the fact that, “quite frequently” (his words), Mail would ask him for the password for this or that account… a dialog box that he usually got out of by clicking on the “Cancel” button.

I thought our e-mail exchange would be a good opportunity to revisit the long-standing and pretty serious problem with Mac OS X’s Mail and less-than-ideal network conditions.

Now, you need to know that, unlike me, my dad lives in an urban environment and is subscribed to a fast DSL Internet service with plenty of bandwidth. But of course, he uses a MacBook Pro laptop with its wireless connection to his router. That means that, even with more than enough bandwidth for his needs, he can still encounter occasional problems with his network connection, either because of a temporary failure in his wireless connection due to interference, physical obstacles or software or hardware flakiness beyond his control, or because of occasional problems with the provider itself.

This, in my experience providing tech support services over the years to a variety of clients with a variety of means to connect to the Internet, is pretty much the norm, and not the exception. (The exception is my own situation, i.e. a lousy satellite-based Internet connection that, in addition to all of the factors mentioned above, can also be affected by weather conditions and bandwidth throttling during peak hours.)

The single most scandalous thing about Mac OS X’s Mail is that its behaviour when checking for new mail appears to have been “designed” (if you can call it that) based on the assumption that network failures never occur. I don’t know what kind of weird bubble Apple’s Mail engineers live and work in, but one thing is for sure: it is completely disconnected from reality. I don’t know a single person who has never experienced network outages. Even with the best equipment, the most up-to-date software, and the most ideal technical environment, there will be occasional failures. It’s just a fact of life.

That means that the job of an Apple engineer working on Mail includes ensuring that the application’s behaviour when such a failure occurs will be reasonably user-friendly.

But that obviously is not part of the job description. Because Mail’s behaviour in such conditions is appallingly bad, and it has been that way for years now. How any self-respecting software engineer working on Mail at Apple can sleep at night knowing that thousands, if not tens or hundreds of thousands, of Mac users are currently experiencing the idiotic behaviour that I am about to describe once again is beyond me. These people should be ashamed of themselves.

If Mail tries to check for new Mail while a network/connection failure of some kind is occurring, here’s what happens.

If you are lucky, Mail takes your e-mail account off-line without telling you.

If you are not so lucky, Mail’s Dock icon starts bouncing endlessly, because the application is displaying a modal dialog box asking you to… re-enter your e-mail account’s password.

If you are wise enough to know better, you do like my dad and you click on the “Cancel” button in that modal dialog box. It gets rid of the dialog box, but it also takes the account off-line.

If you don’t know better, you actually follow Mail’s request and re-enter your password. But of course, like most people, you have set Mail to remember your password for you, so you are not sure you actually remember it correctly. “Is it the name of my dog or the name of my grandmother or something else entirely—maybe something weird that Pierre Igot asked me to use the last time he visited me because he was concerned about the safety of my e-mail account and didn’t want me to just use as a password the name of my dog or my grandmother, which dozens of people who know me directly or indirectly know very well themselves?”

“Oh well, let’s try the dog’s name.”

And then all hell breaks loose. Unless you are really lucky and you remembered the correct password, and then, if the network failure has not miraculously stopped during the interval (and even then, I am not sure that’s enough for Mail to recover at this point), the same thing will happen again and you’ll get the dialog box again. And this time you’ll hit “Cancel” and Mail will stop trying to check for new mail and take the account off-line.

If you’re not really lucky, you’ve typed the wrong password, and now you’re really screwed. Because even if you hit “Cancel” after getting the same dialog box five times and entering five different variations on that password and giving up—and then wait for the network failure to stop and try to check for new mail manually, the mail checking will fail, of course, because the password that Mail now remembers and uses is wrong, and you’ll be caught in an endless loop of the dialog asking for your password and Mail taking your account off-line.

Time to call Pierre, and hope that he can remember your password. If you’re lucky, this has happened before, and he’s stored your password in his encrypted password database. Otherwise, you’ll have to call your ISP’s tech support line and wait for half an hour until you can speak to a live individual and then identify yourself with personal questions and confidential information and so on and then the ISP might give you your actual password or reset it for you. (By that time, one hopes that the network failure will be resolved and you’ll be able to verify that it is the right password indeed.)

And all this, because Apple’s engineers apparently have not noticed this idiotic behaviour and how aggravating it is for so many people using Mac OS X’s Mail in the real world, where network failures do occur.

Now, let’s analyse this behaviour for a second here. First of all, why does Mail ask for your password when your password is perfectly correct, simply because there is a network failure? Is it really impossible for Mail to tell the difference between a failed connection and a wrong password? Is it really reasonable to assume that, if the connection to the server fails, the most likely explanation is that the user went behind Mail’s back and changed his e-mail account’s password using his ISP’s web interface or by calling the ISP on the phone?

It is not. It is not just unreasonable. It’s completely stupid. How often does the average user change his e-mail account’s password or have it changed by someone else?

And then, there is that thing with taking accounts off-line. Who on earth got that idea that e-mail accounts should ever be taken off-line when the user does not explicitly ask Mail to take them off-line? Why on earth does cancelling this idiotic modal dialog box asking for a password that Mail already knows and is perfectly correct cause Mail to take the account off-line even without the user’s permission?

It’s bad enough that this dialog box appears in the first place. But on top of it, whichever way you choose to respond to the dialog box, you will probably screw things up. The least disastrous option is indeed to hit the “Cancel” button, as my dad correctly figured out by himself, possibly in part because he has 40 years of computing with various user-hostile systems behind him. But even that least disastrous option is still a major pain in the neck, because it means that you have to know and remember that you have to take your account back on-line manually, either by forcing Mail to manually check for new mail or by clicking on the warning icon that appears next to the account’s Inbox in the mailbox drawer.

Of course, you cannot take your account back on-line until the network failure is gone, so there’s no point in trying to do it right after you’ve clicked on that “Cancel” button. No, you have to actually wait until the network failure is resolved, and then remember to take the account back on-line manually.

And then, if you were lucky or wise enough not to screw up your account’s password, things will work properly again.

But even a seasoned computer user like my dad couldn’t bring himself to think that Apple’s engineers would be so clueless that they didn’t even design Mail to recover automatically once the network failure is resolved. So, while he correctly figured out that he had to click on that “Cancel” button and that this particular weirdness probably had something to do with the fact that Mail didn’t seem to reliably check for new mail in all his accounts automatically all the time, he didn’t go as far as to think that he actually had to manually bring these accounts back on-line after the end of the network failure, even though he had never taken them off-line himself in the first place.

And if my dad with his 40 years of experience couldn’t quite figure it out, you just know that there must be hundreds of thousands of Mac OS X users enduring that very same problem and not knowing how to deal with it “properly,” i.e. by doing manually what Mail is utterly unable to do by itself automatically.

Who knows how these users finally get out of the situation that Mail has put them in. Maybe quitting and relaunching Mail actually forces Mail to take the accounts back on-line. I don’t know. There are so many different situations in Mail that it is impossible to even tell when an account is actually in an off-line status that requires the user to manually take the account back on-line or just in a greyed-out status that will somehow resolve itself without user intervention.

I have seen greyed out Inboxes with the warning icon (exclamation mark inside triangle). I have seen greyed out Inboxes with the danger icon (lightning bolt). I have seen Inboxes not greyed, but still with the warning icon, itself greyed out or not. I have seen it all. I have never understood what the difference is, and how to tell whether Mail will actually try to check for new mail in the affected account automatically or not.

The whole thing is just so stupid, and so incredibly frustrating. And it’s not just me and my lousy Internet connection. I have talked about this problem before, and I have received feedback indicating that it happens to all kinds of people, in all kinds of situations, with all kinds of Internet connections.

There clearly is something very wrong here. It’s been like this for years, and it does not look like Apple is interested in doing anything about it. What do we need to do? Do we need to start a petition? A boycott? Do we need to sabotage the Internet connections at 1 Infinite Loop so that Apple’s own engineers get a taste of the action? What exactly will it take?

A while ago, I ranted about Pages ’09’s on-going apparent inability to do what I ask it to do when it comes to replacing straight apostrophes with curly ones or vice-versa.

This week-end, I was able to better pinpoint the exact source of the problem. In a nutshell, the problem is that the “Use smart quotes” option in Pages ’09’s preferences supersedes the choices you might make in the “Find/Replace” dialog box:

For example, if you uncheck the “Use smart quotes” option in Pages ’09’s preferences and then bring up the “Find/Replace” dialog box and enter a straight apostrophe in the “Find:” field and a curly apostrophe in the “Replace:” field and then hit “Replace All,” Pages ’09 replaces any occurrence of an apostrophe, whether it’s straight, curled to the left or curled to the right, with… a straight apostrophe.

Conversely, if you check the “Use smart quotes” option in Pages ’09’s preferences and then bring up the “Find/Replace” dialog box and enter a straight apostrophe in the “Find:” field and a curly apostrophe in the “Replace:” field and then hit “Replace All,” Pages ’09 replaces any occurrence of an apostrophe, whether it’s straight, curled to the left or curled to the right, with… an apostrophe curled either to the left or to the right, depending on what Pages ’09 think is appropriate in the context in which the apostrophe appears.

Why would you want to uncheck the “Use smart quotes” option in the first place? might you ask. Well, I happen to be a long-time Spell Catcher X user, and the software came with its own “smart quotes” feature long before Apple even came up with a word processor of its own. So I have become used to using Spell Catcher X’s smart quotes feature, and I don’t need the one in Pages ’09.

But that does not mean that I don’t want to be able to use Find/Replace in Pages ’09 to change existing quotes and apostrophes! I use Spell Catcher X, for the most part, as an interactive tool that acts on my text input as I type it. But when I receive texts from other people or copy text from various sources, sometimes I simply want to be able to use Find/Replace in Pages ’09 to effect batch changes.

Because the result of Find/Replace operations in Pages ’09 is tied to the “Use smart quotes” option, it makes it rather difficult and frustrating to try and control what happens to straight and curly quotes and apostrophes.

In addition, there is simply something nonsensical about Pages ’09 replacing curly apostrophes with straight ones when I ask it to do the exact opposite in the “Find/Replace” dialog box. It seems to me that, when I take pains to enter a straight apostrophe in the “Find:” field and a curly one (specifically curled in one direction) in the “Replace:” field, I indicate quite clearly what I want to achieve. But I don’t get the expected result at all.

I would understand the “Use smart quotes” option applying if all I entered in the “Find/Replace” dialog box was straight quotes and straight apostrophes. But that’s not what I do, and I expect Pages ’09 to be smarter about this than it currently is.

The situation in French is even worse, because the nonsensical behaviour extends to French quotation marks as well. If Pages ’09 were consistent, I would expect it to automatically change any occurrence of a straight English quotation mark (“) with the appropriate French quotation mark and non-breaking space (either “« ” before or “ »” after) depending on the context when I use the Find/Replace feature to replace a English quotation mark with an English quotation mark with “Use smart quotes” on. But that’s not what it does. It replaces the straight English quotes with… curly English quotes.

And if I turn the “Use smart quotes” feature off and try to replace occurrences of English curly quotes with French quotation marks (“ with « and ” with  »), that totally fails too, because Pages ’09 is unable to differentiate between a “ (English opening curly quote) and a ” (English closing curly quote). Argh.

Unfortunately, I don’t expect Apple to care much about all this, both because I am a Spell Catcher X user and because I am a bilingual user. I am obviously in a small minority and my bug reports will be ignored.

But it seems to me that Apple has really not put much thought into this at all, and that the default behaviours of the Find/Replace feature in Pages ’09 can easily become fairly nonsensical, whether the “Use smart quotes” feature is on or off.