It would be interesting to try and pinpoint the reasons for this sudden surge.

And it would also be interesting to compare this with trends in Unicode use in e-mail communications. Sadly, my own experience suggests that things are probably not as improved in e-mail as they are on the web. I don’t have statistics about how many e-mail messages in my Inbox are encoded using Unicode, but I doubt very much that it represents more than a small minority.

And I am still occasionally encountering situations where messages (often forwarded messages) lose their proper encoding and are rendered improperly, which makes them unreadable and unusable if they contain any non-ASCII characters.

Of course it does not help that Apple’s own Mail application still does not seem to be able to use Unicode by default. The default text encoding when composing a new message is “Automatic” and, in my experience, that means that the message is encoded in ISO-Latin1. I have to manually select the Unicode option (in the “Message” menu) each and every time I compose a message.

Needless to say, I don’t do that, and I doubt very much that many Mac users do. If Entourage’s own interface is any indication, the situation isn’t any better on the Microsoft side.

It’s undoubtedly harder to force IT people to implement proper Unicode support and the use of Unicode as the default encoding in e-mail communications than it is for web pages, where the author of the web page is more in control and there are fewer potential pitfalls between author and reader.

But I am still hoping that, in the not-too-distant future, we’ll all be able to use the full range of Unicode characters in our e-mail communications without having to worry about whether they’ll be readable—and without being forced to switch to HTML e-mail (gasp!) either.

Thankfully, the Mac OS X 10.5.2 update addressed the most glaring issues, with the addition of options that enable you to bring back a certain level of sanity.

Even though these “sane” options are still not the default options—which means that there are still hundreds of thousands of Mac users out there enduring the idiotic Mac OS X 10.5.0 behaviours because they do not know that they can change them—at least by simply right-clicking (or control-clicking) on a “stack” in the Dock you can choose options to display the “stack” as a regular folder icon (as opposed to the utterly useless pseudo-3D display of a pile of multiple icons) and you can bring back the standard hierarchical menu of the folder’s contents that used to be the default option in previous versions of Mac OS X.

However, in spite of these (optional) improvements, I find it mind-boggling that, in Mac OS X 10.5, it is still impossible to open the corresponding folder in a Finder window in a simple way. You can open the corresponding folder in the Finder, but it is not simple. You have to click on the icon in the Dock, and then you have to manually select the “Open in Finder” command.

Worse still, if you are using the option to display the folder’s content in a hierarchical menu when clicking on it in the Dock, by default the “Open in Finder” command is actually hidden at the bottom of the menu:

As you can see in the picture above, the only thing that is visible by default at the bottom of the menu is the top 2 or 3 pixels of the “Options” command. And the “Open in Finder” is below that, so it’s totally invisible by default. You have to move your mouse pointer over the triangle at the bottom of the menu to cause it to scroll down the menu and make the hidden commands visible:

And then you can finally select the “Open in Finder” command to open the folder in question in a Finder window.

This problem occurs regardless of the actual length of the menu. Normally, the triangle is used when a menu is too long to fit on the screen. But that is not the case here. Even if you have tons of room above your Dock to display the menu in full, including its bottom two commands, Mac OS X still hides them by default and requires you to move your mouse pointer over the triangle to reveal them.

Compare this to the behaviour in previous versions of Mac OS X, where you could open the folder in a Finder window through a simple single click on the folder icon!

Mac OS X 10.5 changes the meaning of the single click, of course, which now brings up the contents of the folder in a fan, a grid or a menu. And that’s not necessarily a bad thing in itself.

But what is truly bad is that what used to be the default behaviour before Leopard has now become so complicated to achieve.

I’ve tried combining various modifier keys with a click on the icon in the Dock. The only thing that I have found is that a command-click on the Dock icon opens a Finder window showing the contents of the folder’s enclosing (parent) folder. But that’s not what I am looking for. What I am looking for is a truly simple way to open the folder itself in a Finder window (not its parent folder), which apparently only the “Open in Finder” command can do (with the above-described problems).

(To be more accurate, command-click shows the contents of the parent folder when the window is in list view. When the window is in column view, both the contents of the parent folder and the contents of the folder are visible, because the folder itself is selected in the parent folder’s column, and the next column shows its contents. What I would like is a simple shortcut that shows me the contents of the actual folder in list view, not of the parent folder.)

The only other thing that I have found is the “Show in Finder” command that appears at the bottom of the menu when you control-click (right-click) on the Dock icon:

This command is worded differently, because it does not do the same thing as the “Open in Finder” command in the other menu. Unlike “Open in Finder,” “Show in Finder” works the same way as the command-click shortcut, i.e. it reveals the folder in its parent folder. (In other words, it opens the parent folder, and then highlights the folder in question by selecting it.) This command is always visible by default when you control-click on the Dock icon, i.e. you don’t have to manually reveal it by moving your mouse pointer over a triangle. When using the list view mode in the Finder, however, it still opens the enclosing folder, which makes it less useful than the proper “Open in Finder” command.

I simply don’t understand why the “Open in Finder” command has to be hidden by default below the bottom of the menu when bringing up the regular hierarchical menu.

And I also believe that we should be able to use a simple shortcut to reveal the contents of the folder in a Finder window in list view without having to navigate the parent folder. The command-click shortcut should probably be kept for “Show in Finder” because that’s what the Command modifier key is used for in a variety of contexts (to reveal a Spotlight search result in its enclosing folder, for example). But maybe command-shift-click could be used here.

The bottom line here is that, in their apparent obsession with this new fancy “stacks” feature that is actually far from innovative, Apple’s engineers have failed to preserve the most basic level of usability for Mac users who happen to find the new fan or grid display modes far from useful. And it is quite careless of them indeed.

The reason is pretty simple. There are two ways to initiate a Spotlight search in Mac OS X 10.5: you can either open a new Finder window (or use an existing Finder window) and type a search request in its “Search” field in the toolbar:

or you can use the global Spotlight menu in the top-right corner of the menu bar:

The Spotlight menu itself, however, only shows you a subset of the total list of search results, so if you want to view the full results of your search, you need to use the “Show All” option in the menu:

which opens a new Finder window showing the full list of results.

In both cases, you end up with a Finder window showing a complete list of the results of your search. Sadly, this Finder window looks like a regular Finder window, but it does not behave like a regular Finder window. It has no view options (so you cannot choose which columns you want to see) and suffers from some enduring bugs and flaws that don’t affect other Finder windows. But it’s a Finder window.

The option you chose to initiate the search, however, has a significant impact on the way that search results are presented in the resulting Finder window.

If you choose to start your search from a Finder window, then Mac OS X will use the current width of that Finder window as the basis to calculate the default width of the three columns for the display of the search results (“Name,” “Kind,” and “Last Opened”), which means that, the wider your original Finder window was, the wider the three columns in the results window will be.

If you choose to start your search from the global Spotlight menu, on the other hand, and then use the “Show All” option to view the full list of results in a Finder window, Mac OS X will always show the full list of results in a new Finder window with the same default size, and this size will always be… 750 x 460 pixels.

Because of the limited width of this default window size and the need to display four columns of stuff within that width (the sidebar on the left, and then the three columns for the list of results), you always end up with a “Name” column that is far too narrow to display most file names in full. So you end up with a list of mostly truncated file names:

Unfortunately, resizing the Finder window in question does not help. It increases the overall width of the window, but not the width of the columns. You just end up with useless empty space on the right hand side. You then have to manually adjust the width of the “Name” column in order to be able to see most names in full—and you have to do this each and every time you start a new Spotlight search through the global Spotlight menu and then choose the “Show All” option to view all the results in a Finder window. Because the control for resizing the “Name” column is pretty small (it’s the small line separator between the “Name” column and the “Kind” column), this is an adjustment that takes a non-negligible amount of effort, and having to repeat it again and again throughout the day is no fun at all.

My main screen is a 30″ monitor. This means that I have more than enough screen real estate to use a default window size for the Spotlight search results window that is twice as big as this 750 x 460 window. But there is no way to adjust Mac OS X’s default behaviour. Apple’s engineers have chosen this default 750 x 460 size, and there’s nothing you or I can do about it.

They undoubtedly chose this particular window size as some kind of compromise but, as with all compromises, the risk is that it is a compromise that attempts to please everyone and ends up pleasing no one. It certainly does not please me.

There are other aspects of the Finder interface where Apple’s engineers have made at least some semblance of an effort to adjust to the user’s preferences by making the Finder “remember” the last window size used. For example, if you create a new window in the Finder, then resize that window and close it, the next new window that you will create will have the size that you’ve just defined.

No such luck with the Spotlight results window in the Finder, however. All the effort put into resizing the window and adjusting its columns’ widths is forgotten as soon as you close the window, and you have to start all over again each time you use the global Spotlight menu to bring up a Finder window with a list of Spotlight results.

Of course, the “workaround” here is to avoid the global Spotlight menu altogether and always start a Spotlight search using an existing Finder window in the Finder. But that is not a user-friendly approach, and the command-Space shortcut is just too easy not to use. In addition, when you start a new Spotlight search, you don’t always know in advance whether you’ll need to use the “Show All” option or not. Sometimes the subset of search results that appears in the Spotlight menu is good enough, and so you never have to show the full results in a Finder window, so the global Spotlight menu fulfills its role without too much inconvenience.

I don’t know enough about the underpinnings of Spotlight in Mac OS X to offer a definitive solution here. But it seems to me that some kind of behaviour where Mac OS X would “remember” the last window size and column widths used for a Finder window spawned by the global Spotlight menu would help alleviate this particular problem and enable each Mac OS X user to customize the behaviour of the global Spotlight menu without adding complexity to the Spotlight interface.

(Of course, I would also like Apple to give us more view options for the search results window, but that’s another story.)

But of course at this stage of the season, you cannot really draw too many conclusions. The Gunners obviously only took the game half seriously, and it is hard to blame them. What really does boggle the mind is why Paul Jewell actually elected to take the management job at Derby. I mean, based on this season’s evidence, they will struggle not to be relegated next year too. What on earth attracted Paul Jewell here? It really is quite mysterious.

There really is not much else to say. Adebayor got a deserved hat trick, which brings his goal tally for the season to 30. It’s too bad he suffered from a lull at a crucial time in the season. But then, the whole team did.

The questions about the Arsenal defence remain unanswered, and it is quite obvious that Arsène Wenger will need to make at least one purchase in that department in the summer, especially if, as suspected, Senderos is allowed to leave.

Third place is guaranteed. The goal now is to finish with 83 points. Whether that will be enough to move ahead of one of the two top teams is highly questionable. We will obviously know more about this by kick-off time on Sunday. Until then, we are all spectators, and of course most Arsenal fans will probably be Barcelona supporters today.

As a die-hard Arsenal fan, I cannot really claim to be a “neutral.” But the truth is that I dislike both Manchester United and Chelsea almost equally, and much for the same reasons. I can tolerate Manchester United a bit more sometimes, because they do play attractive football at times, but I cannot stand the attitude of players like Rooney and Ronaldo (and van Nistelrooy before them, and Scholes, etc.).

Ronaldo is just so full of himself that it is painful to watch (except of course when he screws up gloriously, as in the missed penalty in mid-week at Barcelona), and Rooney is just a foul-mouthed, foul-tempered twit who happens to have good feet.

That said, in terms of being dislikable, Chelsea probably win the contest, as today’s game proved once again. And nobody does dislikable quite like Didier Drogba. Today’s game could easily have been titled “The Didier Drogba Drama Queen Show.”

It started with Vidic’s injury. Yes, it was probably accidental, and yes, Drogba did display some concern about the well-being of the player after the collision. But did you watch that slow motion replay carefully? Did you see how, immediately after colliding with Vidic’s head, Drogba’s immediate, instinctive reaction was to crumble to the ground in seemingly excruciating pain? It is only once he realized that what had hit his own precious knee was actually Vidic’s face that he corrected his reaction and decided that, after all, his knee didn’t hurt all that much and probably was not injured!

I don’t believe that, in the whole history of football worldwide since the very invention of the game, there can have been a single incident where any player got his knee injured through a collision with another player’s head. I think it is physically quite impossible, that even the toughest jaw or skull is simply no match for an elbow, let alone a knee. Yet Drogba’s instinct was quite obviously to try and make everyone believe that it was possible—until he realized how ridiculous it was. It just says so much about his attitude. As soon as anyone touches any part of his precious body ever so slightly, it has to be a foul, and he has to act like he’s seriously injured. It’s sad, just sad.

Then of course there was the whole Ballack/Drogba subplot. How embarrassing can it be to everyone involved with the club and to the team’s supporters to have the whole wide world witness such a spectacle on the pitch? There are of course big egos in English football these days, and there is a history of ugly confrontations between players from the same team in recent years. I distinctly remember the scene between Bowyer and Dyer at Newcastle a couple of seasons ago, and of course I am not forgetting the confrontation between Adebayor and Bendtner at Tottenham earlier this year.

But this one surely is the most high profile and the most embarrassing of the whole lot. And who else but Didier Drogba to be at the centre of it? How can any neutral spectator have any empathy for the “big family” at Chelsea after witnessing such scenes?

The bottom-line, however, is that they did win, and are now level on points with Manchester United. The Red Devils are still guaranteed to win the title if they win their two remaining games, because of their superior goal difference, but they cannot afford any kind of slip-up. It certainly puts more pressure on them, which they did not need ahead of the return leg of the Champions’ League.

It also means that mathematically, if Arsenal win at Derby on Monday, they still won’t be out of the race. There is of course, no realistic chance of both Chelsea and Manchester United slipping up in a major way in their remaining fixtures, but there still has to be some motivation to finish with the highest possible number of points, if only to prove that the Gunners really went pretty close this year.

Certainly, on the evidence of today’s game at Stamford Bridge, there is no gulf in ability between Arsenal and Chelsea and Manchester United. The football played today—to the extent that any football was played, which is debatable—was really not impressive. It’s just highly unfortunate that the Gunners went through that bad patch after the Birmingham disaster and didn’t recover from it quickly enough. It was a very fine line between success and defeat (or, in that particular case, a succession of draws leading to defeat), and the Gunners ended up on the wrong side of it this year.

But I didn’t see anything today that convinces me that these other two teams were indeed better and “deserved” it more. They are just more adept at turning a footballing contest into big drama, with a stomach-churning blend of bullying, cynicism, theatrics, and referee abuse. If that’s what it takes to be champions this year, then the Gunners should have no regrets, and they should only hope that they can redress the situation by winning with stylish and quality football in the next contest.

This will insert the current date in a default (long date) format:

It looks like regular text, but it actually cannot be edited. If you click anywhere on it, Pages selects the entire thing, and if you double-click on it, you get access to a small floating thing that provides further options:

In this little floating thing, you can use the “Choose Date Format” menu to select other date formats:

The list of available date format options, however, is still limited. What if, for example, you’d like an automatic date in the universal “yyyy-mm-dd” format (as in “2008-04-25”)?

As you can see from the picture above, this particular date format is not part of the list of options. But this is where tight integration with Mac OS X’s underlying standards becomes important. See, the date format options provided in the floating thing above are actually based on the date format options defined in the “International” preference pane in System Preferences:

And the good news is that these date format options can be customized there. The “Customize…” button gives you access to a dialog sheet where you customize four different default values for the date format: “Short,” “Medium,” “Long,” and “Full”:

In the example below, I have customized the “Short” value and replaced it with the universal “yyyy-mm-dd” format:

(As this particular screen shot indicates, in order to get the number of the day with two digits, I just inserted the “Day of Month” control and then used the contextual menu to change from the default single-digit value to the two-digit value. You can do the same for the “Month” control in order to get the number of the month with two digits instead of the name of the month.)

Once this customization is done, I can return to Pages and double-click again on my automatic date, and now the options in the “Choose Date Format” menu have changed:

Et voilà! I can now insert an automatic date with the universal “yyyy-mm-dd” format in my Pages documents.

Sadly, Pages’s automatic date feature suffers from some frustrating limitations nonetheless. The major one for me is the fact that it is a unilingual feature, even though Pages itself is multilingual and lets you compose documents in a variety of languages.

This means that, regardless of which language your Pages document is in, the date inserted by Pages’s “Insert > Date & Time” command will be in the language-specific format selected in the “International” preference pane.

Since the majority of my documents are in Canadian French, I use the “French (Canada)” region in the “International” preference pane. Here are the four default values for the “Full,” “Long,”, “Medium,” and “Short” format options for the “French (Canada)” region:

As you can see, the “Medium” and “Short” values are exactly the same, which is pretty useless. This means that I can customize the “Medium” one to use the universal “yyyy-mm-dd” format mentioned above:

And I very much need this, because when I compose documents in Pages in English, the “yyyy-mm-dd” date format is pretty much the only acceptable one in English. Otherwise, I have no choice but to enter my dates manually in English, which means that there is no automatic updating—a particularly frustrating limitation when using dates in document templates.

Of course, in order to circumvent this limitation, Pages would have to break free of its total dependence on the “International” preference pane settings. It would have to provide its own preference settings for language-specific date formats, and that would undoubtedly add to the complexity of the feature. But I am afraid that, in a multilingual world, there is no other solution, and if Pages claims to be a multilingual word processor, then it needs to go all the way and provide multilingual options for dates as well.

The current situation might be acceptable to the majority of users who only ever use a single set of date, time and number formats—the one for their mother tongue. But it penalizes users who regularly switch from one language to another and, while these users might not be the majority, there are still quite a few of them, and not just in Canada.

(The same restrictions, of course, apply to automatic number formats in tables in Pages, Numbers and Keynote. Here again, it is impossible to use the applications’ automatic number formatting features in any language other than the one selected in the “International” preference pane.)

In conclusion and as a comparison, I was going to mention Microsoft Word’s own multilingual features. It used to be that Word was smart enough to change the date format used depending on the language setting for the underlying text. And indeed if I open old templates that I had defined in previous versions of Word, I can still find automatic dates in an English format in those templates that are in English. But when I try to insert a new automatic date in Word 2008, I now get the same limitations as in Pages, i.e. only it inserts the automatic date with the format options defined in the “International” preference pane, regardless of what the current language setting for the text is in the Word document.

So it looks like Word 2008 has been crippled in that respect compared to previous versions. Since I profoundly loathe Microsoft Word and use it as little as possible, I am not going to bother to try experiment to see if I can reproduce the language-specific date formats that I used to be able to achieve in previous versions. But if Word 2008 is indeed crippled, it is yet another reason to give up on Microsoft and switch to Pages. Word 2008 still has more flexibility than Pages (for example, it has separate field codes for print dates, which are dates that are only updated when the document is printed), but for all we know Microsoft’s MacBU is just as likely to further cripple the next version of Word and remove that functionality as well.

The real conclusion here is that Pages, while it is still somewhat limited in what it can do in terms of automatic dates, does provide a number of options, with further customization possible through the “International” preference pane in System Preferences.

Yet apparently that is still enough to cause Microsoft Word 2008 to misbehave. Here’s the proof:

Here’s another sample, at 150%:

And exactly what horrible abuse of computing machinery did I commit in order to achieve such catastrophic text rendering?

I scrolled up and down the document.

Wow.

Good job, Microsoft. Once again, you have exceeded all expectations. This is really fantastic engineering. Remind me again… How much did I pay for this crap?

An absolute bargain.

When in Print Layout View, as soon as you place your insertion point anywhere inside a table, Word displays this small proxy icon in the top-left corner of the table:

This little control is meant to be used for selecting the table and moving it around on the page.

The trouble is that, as soon as you start typing text inside a table cell, this control starts constantly flickering on and off. It’s a purely visual glitch and has no impact on the usability of the Print Layout View or the control itself (it stops flickering when you stop typing), but I find this flickering highly irritating and distracting.

And it serves as a constant reminder that Microsoft is unable to produce quality software with enough polish to qualify as real Mac OS X software.

Sure, there are many more important bugs to fix first. But the trouble with Microsoft is that there are also hundreds of such less important bugs, and Microsoft’s engineers simply never get around to fixing them—which means that we have to live with them for years, and cannot help but feel that Microsoft will never be able to achieve high enough levels of quality in their products.

In Word 2008, Microsoft has reshuffled and renamed its various view modes (which are still far too numerous), and the venerable Page Layout view mode is now called Print Layout View.

Whether this has anything to do with the new bugs it introduces, I do not know. But here’s a fairly simple example of an utterly annoying problem that, as far as I can remember, did not occur in previous versions of Word.

I often have to work with Word documents containing large tables, and sometimes these tables have rows that too big to fit on a single page. If the table has the option to “allow rows to break across pages,” then such rows can be broken in two by an automatic page break, and you get something that looks like this:

No problems here. But now consider what happens when I try to copy the selected text in the picture above and paste it into another column in the same table.

Here’s the second column of the table before I paste the additional text:

Again, no problems here. But now see what happens when I paste the copied text at the insertion point, which is near the bottom of the page:

Great. Word inserts the copied text in the table cell on the first page, but completely fails to refresh the display of the continuation of the same table row on the second page. This makes it look as if I only pasted part of the text, and the rest of the clipboard has disappeared into the ether, along with the last portion of the text before the Paste command at the bottom of the cell on the first page!

In fact the Paste operation has worked just fine. It is just that Microsoft’s engineers have failed to ensure that the display of the continuing table row on the second page would be automatically refreshed.

And getting Word to refresh its display is not just a matter of scrolling up and down the document either. I tried this again and again, and still Word would not refresh the display of the second page.

I inserted my cursor at the bottom of the first page and used the cursor keys to navigate to the “invisible” part of the pasted text, and then I started typing something else “in the dark.” And then finally Word woke up and remembered that it actually had to refresh the display of the table on the second page and I finally was able to see my pasted text in full:

But really, this kind of shoddy work is completely unacceptable. It is yet another illustration of the utter lack of professionalism at the MacBU, which releases products that are clearly full of bugs and gleefully charges users hundreds of dollars for them.

I know, I know, it’s nothing new. But sometimes a visual illustration of just how bad Microsoft’s products are can be a useful reminder of the level of incompetence and carelessness of Redmond’s Mac engineers.

And here’s the exact same document when I opened it with Apple’s Pages 3:

Pretty sad, isn’t it? Microsoft’s own software engineers are unable to ensure a sufficient level of cross-platform compatibility, and Apple’s engineers, on their own, probably with no help from Microsoft whatsoever, are able to provide better compatibility with Word for Windows documents than Microsoft Word 2008 does.

Then again, when I try to open the same document with Word XP (Word 2002) in Windows XP under Parallels, I systematically get an application crash, so I am not even able to open it and edit it there. So clearly Microsoft’s engineers are just as careless about cross-version compatibility on the Windows side. I suppose Microsoft expects me to purchase Vista and Office 2007 just to be able to maintain compatibility…

Now, of course, this is just one particular file. I don’t have a lab of full-time technicians to test thousands of Word for Windows files in order to determine their relative compatibility with Word 2008 and with Pages 3 respectively on the Mac.

And it is also true that, for some types of Word documents, Pages has many compatibility problems—especially Word documents with “fancy” layouts making heavy use of text frame objects or whatever these things are called in the Word interface. These text frame objects have been prettified in Word 2008 compared to previous versions of Word for the Mac, but they are still extremely painful to use and atrociously designed and handled, so I certainly don’t blame Apple’s engineers for not even attempting to preserve “fancy” layouts in Word documents that make heavy use of text frames. It’s a nightmare just to use them in Word, so I cannot even imagine how difficult it must be to support them properly when converting Word documents to the Pages format. But this limitation in Pages’s compatibility with Word documents unfortunately means that I still have to use Word from time to time, for certain documents.

Still, my experience with Word for Windows documents is that, apart from documents with text frame objects, when I open these documents in Pages, I get results that are just as good, and often better, than what I get when opening these same documents in Microsoft’s own Word 2008. The above example illustrates this quite clearly. And it reflects the sad reality that Word 2008 is, well, yet another piece of Microsoft crap.

Out of curiosity, I tried to determine what type of image file it was that was causing the above problem in Word 2008. This is not easy to do, however, as neither Word 2008 nor Pages 3 provides any tools for identifying the type of picture file that was originally used to insert a picture in the document.

It is also impossible to just drag the picture from the document to the desktop to get the original file, like you can do with most pictures on web pages.

And when I try to copy the picture to the Clipboard and then paste it in another application, I get varying results. When I switch to Preview and create a new file, it automatically creates a new file based on the Clipboard, and that file is a PDF containing the logo in question, but I am not sure that this means anything about the actual format of the original picture file.

When I paste the Clipboard into a new Photoshop file, I get a layer called “Vector Smart Object,” with a low-res preview of the picture in black and white used as a some sort of background with an “X” across it, so that doesn’t really tell me anything either.

So basically I still don’t know whether this particular logo was a JPEG, a GIF, a BMP, a TIFF, a PNG, a PDF, or something else. But whatever it was, it was something that Pages was able to handle properly, whereas Microsoft’s own Word 2008 was not. And I also have no way of checking what happens to the picture after I edit the Word file in Word 2008 and return it to its author, because I cannot open it in Word XP. I just hope that somehow the picture is preserved and still looks OK in the returned file. But this means, as usual, that I am penalized for being a Mac user, and running the risk of delivering damaged documents to my PC-using colleagues, which can impact me negatively in my work and how it is perceived by my colleagues. All thanks to Microsoft.

Manchester United might have dropped two points, but it’s a two-horse race between them and Chelsea now. The big clash at Stamford Bridge on Saturday will be crucial, but Arsenal fans will be watching it as neutrals, because the result will have no impact on the outcome of their own season. One can always dream of Manchester United losing their remaining three games and Chelsea losing their other two games, but it’s about as likely as lightning hitting Wayne Rooney and turning him into a beautiful frog.

It’s really too bad, because three more victories is definitely a possibility for the Gunners, which would mean finishing with a grand total of 83 points—a huge improvement over the last two seasons, but one that won’t count for anything except for a psychological boost at the beginning of the next season.

Still, even this comfortable 2-0 victory against relegation-threatened Reading couldn’t mask the fact that this Arsenal team still has trouble finishing off the opposition. What we really need is a string of more emphatic victories that reassert the Gunners’ superiority in an unequivocal fashion. And of course we need all the current players to stay at the club.

I was going to add “except for Éboué,” except that, shockingly enough, when Éboué came on on Saturday for the injured Touré at right-back, he actually had a pretty good game. Of course, it doesn’t mean much in a game like this, but it really made it look like this “Éboué as a winger” thing—which has been a bit of a nightmare, to be honest—had actually never happened in the first place. Does this mean we could use Éboué against as a right-back ahead of Sagna? No way. But maybe we could keep him as a capable substitute…

OK, maybe not.

This means that you cannot do a Replace All operation that applies to a specific section of the document only, without first copying that section of the document to another blank document, and then copying the section back to the document once the Replace All operation is complete. This is utterly clumsy and not worthy of a quality word processor.

I have also written posts about the many difficulties with punctuation characters in the Find/Replace dialog box, including smart quotes and non-breaking spaces.

Sadly, that is not the end of it. Pages’s Find/Replace dialog box suffers from yet more problems which can make it very dangerous to use it for batch Find/Replace operations in a document.

Here is an easily reproducible problem. First download this sample file called “ParagraphStyles.pages.zip”:

ParagraphStyles.pages.zip

Now decompress the Zip file and open the resulting Pages document titled “ParagraphStyles.pages” in Pages 3.0.x.

Once you have opened the document in Pages, invoke the Find/Replace dialog box, switch to the “Advanced” tab and enter the following Find/Replace request. In the “Find” field, enter:

[same as English]¶

where “¶” is a paragraph mark. (In order to enter a paragraph mark on the Find/Replace dialog box, use option-Return.)

In the “Style” pop-up menu for the “Find” field, choose the style called “Translator’s Note.”

Then in the “Replace” field enter:

[same as English]¶

where “¶” is, again, a paragraph mark. In the “Style” pop-up menu for the “Replace” field, choose the style called “List 1.”

These settings are obviously for the following purpose: to change the paragraph style of all occurrences of “[same as English]¶” (with paragraph mark) from “Translator’s Note” to “List 1.”

Now click on either “Replace All” or “Next” followed by “Replace & Find.”

Notice anything? Pages does indeed change the paragraph style of all occurrences of “[same as English]¶” (with paragraph mark) from “Translator’s Note” to “List 1,” but it also does something that is completely unwanted and highly undesirable: it changes the style of the paragraph following “[same as English]¶” from its current style (“Work Heading 3” in this case) to… the “Translator’s Note” style!

This is totally unacceptable. Yes, the scenario described above is a bit complicated, but the bug itself is not really all that complicated or difficult to reproduce. And it is a pretty serious bug when it comes to paragraph styles.

I know that I can easily select all occurrences of a given style and change them to another style, without even using the Find/Replace dialog box, but this is not what I am trying to do here. I am trying to find only specific occurrences of a given style, i.e. occurrences with a specific text string and a paragraph mark at the end.

The existence of such a bug—and of the other bugs and flaws described in previous posts—essentially means that Pages cannot be trusted when it comes to doing slightly complex Find/Replace operations. And that is a significant flaw for a word processor.

Since Pages’s main competitor, Microsoft Word, can rarely be trusted for such operations either (I speak from personal experience here), this means that we effectively have no alternative. And since these bugs are not immediately obvious, there is also little hope of getting the developers to fix them.

By this, I mean that there are a number of applications and sites that do attempt to establish connections behind the user’s back, without the user’s prior permission. Once Little Snitch is installed and running, this becomes abundantly clear.

I don’t mind a Mac OS X application connecting to a server owned by the application’s developer via http on port 80. This is a common procedure used by application developers to enable their application to check for available updates.

But I do very much mind applications connecting to third-party servers to transmit information without my prior approval. And the recent story with Adobe and Omniture proves that users have reasons to be suspicious of the practices of application developers.

Yes, in order to work its magic, Little Snitch does need to install a kernel extension in Mac OS X’s “System” folder, and that does bother me a bit. But in truth, in several years of constant use of Little Snitch, I have yet to encounter a single bug, and I have yet to see a single report of Little Snitch having a negative impact on the stability or reliability of Mac OS X.

But my point today is not about Little Snitch’s main function, which is to intercept unauthorized outgoing connections. My point is that I recently discovered that Little Snitch has another feature that I actually find quite useful.

See, I still have to live with limited bandwidth, because my Internet connection is with a satellite Internet service provider. And because of this, I need to be able to monitor incoming and outgoing Internet traffic on my machine.

For many years, I have used MenuMeters for this purpose. MenuMeters resides in my menu bar (which is fairly long with tons of room, since I have a 30″ screen) and is visible at all times, unlike Mac OS X’s own Activity Monitor, which also enables you to monitor network traffic, but only in a window. Also, MenuMeters provides more flexibility and more options than Activity Monitor.

But MenuMeters, like Activity Monitor, suffers from one significant limitation: It can only report on overall incoming and outgoing Internet traffic. It cannot tell you where this traffic is coming from, i.e. which application is causing it. More often than I would like, I find myself in a situation where there is incoming Internet traffic, and I have so many applications open and running that I am not sure which application is causing the traffic.

That’s where Little Snitch comes in. In addition to its main feature, Little Snitch actually also includes a feature called “Network Monitor,” which can be accessed through Little Snitch’s menu extra in the menu bar:

And this Network Monitor feature does not just show overall incoming and outgoing network traffic. It can actually show a breakdown of the incoming traffic per application, in a small floating window that you can keep open in the foreground at all time:s

I find this feature particularly useful in my situation with limited bandwidth, where I really do need to know where the incoming traffic is coming from and what application is causing it. I suppose that, when I finally get a proper high-bandwidth connection, I will care less about those things, because background network traffic will have little impact on what I am doing with the foreground application.

But until then, Little Snitch’s Network Monitor is definitely a very useful addition to my panoply of monitoring tools.

Arsène Wenger took his final gamble of the season by replacing Senderos with Song in central defence and bringing Lehmann back in goal in lieu of Almunia.

The latter was a highly questionable move following Lehmann’s latest ugly rants earlier in the week. What signal was this supposed to send to Almunia? As it happens, Lehmann did have a couple of impressive (albeit slightly fortuitous) saves during the game. But he had his big chance to prove that he was better than Almunia in the big games by saving Ronaldo’s penalty, and he did not, so the conclusion here is pretty obvious.

I don’t think Almunia is a better keeper than Lehmann, but he does not seem to be worse to me, and at least he does not get himself booked for idiotic behaviours. Wenger will still need to find a world-class replacement sooner rather than later, but it’s not exactly fair to blame Almunia for our lack of silverware this season.

As for the rest of the game, yes, it was yet again a somewhat dodgy penalty decision, but handling the ball in the area at Old Trafford is always a pretty risky behaviour. It was unfortunately the highlight of Gallas’s game. And since Adebayor’s head probably got a helping hand as well for the Arsenal goal, it is hard to feel cheated here.

The truth is that the Gunners played very well during the first half, but had absolutely nothing to show for it at the half, which was unfortunately an all-too-familiar refrain. Given that Manchester United were missing their best central defender, it was simply unacceptable for the Gunners to have all this possession and not get ahead.

Then they did get ahead early in the second half and, once again, were unable to hold on to their lead for more than a few minutes. The rest of the game was a bit of a blurry mess, but the tide had turned, and Lehmann failed to stop that Hargreaves free-kick. (I could have said “Hargreaves scored a beautiful, unstoppable free-kick,” but then Lehmann clearly implied in his comments he could make the difference in a big game. This was another chance to do so. So much for turning your words into actions, Jens.)

I don’t understand what Wenger didn’t give a starting berth to Walcott ahead of Éboué, but then I am not a manager. Clearly Wenger sees something in Éboué that no one else can see. Seventy minutes later, I don’t think anyone can see it any better. And it was too late for Walcott to really have an impact.

For all of the Gunners’ dominance, there were not enough of the truly incisive runs that would have created unstoppable chances. Adebayor wasted a number of good chances and Hleb still refuses to shoot, which I find highly problematic in that the opposition’s defenders clearly know that they don’t have to bother trying to block his shots. Van Persie still hasn’t recovered his pre-injury form, and Fábregas still squanders too many opportunities himself.

The big question now is: How all of this going to change next season? Wenger does not seem to be willing to adjust his thinking, and continues to complain that his players “get slaughtered” by pundits even though “we fight until the end of the season with teams who have spent much more money - some teams spent over £100m.”

While it is perfectly true, of course, it’s hard to be surprised that the media don’t give credit to Arsenal for not winning anything. There’s no trophy for still being in the title race in April while having spent £100m less than the others.

Don’t get me wrong: I don’t want trophies at any cost. I don’t want a takeover financed by Russian oil money, and I don’t want a team of mercenaries. I also don’t want an idiot like Rooney in the Arsenal team, regardless of his actual talent. I don’t want to “win ugly,” at least not with constant ugliness (bullying, referee harassing, cynical fouling, etc.).

But to me these three trophy-less seasons mean that something has to change. There needs to be a shift in attitudes. There needs to be a combination of youth and experience, of agility and physicality, etc. Wenger seems to have taken things too far in one direction, and it’s not paying off. Yet he still seems to believe that he is right.

Belief is of course laudable, and you need plenty of it to win things. But that belief needs to be backed by concrete facts, and there have been far too few of those in recent weeks, during the crucial part of the season. There is no point in building up expectations if you cannot deliver. Do we really want to go through the same thing all over again next year?

For a club of Arsenal’s stature, I cannot help but feel that three trophy-less seasons is about the longest barren spell that they can afford. If this continues, sooner or later the fans will turn against the players and the manager, or key players will become dissatisfied and leave, and so the cycle will continue.

The parsimonious approach was understandable and acceptable when money was indeed tight. But what is the excuse for continuing with it now? Will the focus on youth development really be lost with the addition of a few key players with experience and grit?

Arsène knows—supposedly. But sooner or later he’ll also have some explaining to do.

When editing text inside a table cell, instead of doing the expected, i.e. extending the selection to the beginning or the end of the current paragraph, the shortcuts are actually interpreted by Pages as commands for adding new rows above or below the current table row. Argh.

But the situation is actually worse than that. First of all, the problem also affects the keyboard shortcuts for extending the selection word by word to the left or to the right, i.e. shift-option-Left and shift-option-Right.

The additional quirk for these last two shortcuts is that, if you use them inside a paragraph of text, they work as expected (i.e. they extend the selection). But once you reach the very beginning (or end) of the text in the cell, then Pages hijacks them again and interprets them as commands to add new columns to the left or to the right of the current tablecolumn. Argh.

Here’s an example of a table cell where I have selected a word and extended the selection to the left (i.e. to the previous word, which is actually on the previous line because of the text wrapping) by pressing shift-option-Left once:

So far so good. Now I press shift-option-Left once more to extend the selection by one more word:

No problem here. But now I have reached the beginning of the text in the cell. Look at what happens if I use the shift-option-Left shortcut once more:

Argh. Pages has not only added a new column to the left of the current one, but also lost my selection in the process.

Why is this such a problem? Because when you are a regular user of keyboard shortcuts for text navigation and selection, such as myself, you use them all the time and often repeatedly, in quick succession, in order to select a bunch of words. And sometimes you overshoot by using the shortcut once too many. It’s perfectly normal. But in such situations you would expect Mac OS X to ignore the extra keystrokes.

Instead, Pages has a very destructive behaviour where, once you reach the beginning or the end of the cell, the meaning of the shortcut changes, the selection is lost, and unwanted columns get added to the table. It is not good, and I frequently encounter the behaviour, which effectively punishes me for using text selection shortcuts in Pages. (It looks like Apple’s own engineers never use these shortcuts themselves.)

It gets even worse. The problem affects not just tables in Pages, but also tables in Numbers. Both applications share a number of behaviours when it comes to tables. Unfortunately, they also share this dual use of the shift-option-Up, shift-option-Down, shift-option-Left, and shift-option-Right shortcuts both for text selection and for adding rows and columns.

Interestingly, however, the behaviour in Numbers is not exactly the same as it is in Pages. If you take the scenario described above, and try to use the shift-option-Left shortcut repeatedly inside a table cell in Numbers in order to extend the selection to the left, when you hit the edge of the cell (i.e. the beginning of the cell’s contents), Numbers actually does the right thing and simply ignores additional occurrences of the shift-option-Left shortcut. It does not hijack the shortcut by interpreting it as a shortcut for adding a new column before the current column.

Numbers only interprets shift-option-Left as a shortcut for adding a new column before the current column when the cell itself is selected, not the contents of the cell. (Like Pages, Numbers has two selection modes in tables: a cell selection mode, and a text selection mode when editing the contents of a cell.)

It is much better, because it means that you can overshoot with repeated shift-option-Left or shift-option-Right shortcuts inside a table cell in Numbers without running the risk of accidentally insert new columns. And Numbers is similarly smart about interpreting the shift-option-Up and shift-option-Down
properly when editing text inside a cell, i.e. it correctly extends the selection to the beginning or end of the paragraph and does not add new rows.

Numbers only interprets the shortcuts as commands for adding new rows or columns when the cell itself is selected, without being in editing mode inside the cell.

I still do not agree with the dual use of these shortcuts, even in Numbers, but at least Numbers does not let the two functions of these shortcuts interfere with each other while editing text inside a table cell.

Sadly, the third member of the iWork ‘08, Keynote 4.0, suffers from the same flaws as Pages 3.0, i.e. it does let the two functions of these shortcuts interfere with each other while editing text inside a table cell.

The bottom line here is that, if Apple’s engineers really insist on using the text selection shortcuts as shortcuts for adding rows and columns as well, at least they should implement the dual meaning properly, like they did in Numbers, in all iWork applications, including Pages and Keynote.

Until they do that, Pages and Keynote users who are regular users of these text selection shortcuts will continue to be punished on a daily basis when they attempt to use these shortcuts in tables and end accidentally adding rows and columns to their tables.