Betalogue

About six months ago, our local newspaper decided to replace its entire fleet of aging Macintosh computers with brand new machines and the director asked me to advise them and help them out with the installation process.

We chose a Mac Pro with a 30″ screen, a few 27″ iMacs, and a Mac mini to act as a local file server where they would store most of their files, with a couple of external FireWire hard drives for backups. On the software side, the main purchases were six licenses for Adobe CS4 to replace their older versions of Adobe’s applications, and a few other things.

This brand new Mac mini was replacing a older Mac mini that had been used for the same purpose for many years. I was pleased to be able to show them that, with Snow Leopard, they would be able to use screen sharing to control the Mac mini remotely without having to physically connect a display to the server machine. (Their old Mac mini had an old version of Timbuktu installed on it for that same purpose, but things were no longer working right, they didn’t have the serial number for the Timbuktu software, and they obviously no longer needed the software with the new machines anyway.)

On the whole, things went fairly smoothly. But the transition introduced one new problem that they had not experienced with their previous setup. Quite often (but not always), when they were in the process of browsing the file/folder hierarchy of the Mac mini server remotely on their own machine (whether it was one of the iMacs or the Mac Pro), things would be “jumping around” uncontrollably. They tried to describe the problem to me over the phone, and at first I thought it was some kind of BlueTooth issue that was causing their computers to lose track of their mouse pointer position on the screen.

But they were also able to reproduce the problem even when BlueTooth was completely turned off and they were only using a wired mouse, so obviously the source of the problem was somewhere else.

Of course, as is often the case with such intermittent issues, at first they were not able to show me the problem, because it would invariably fail to happen whenever I was there on site. And then one day I did witness it myself. And I saw quite clearly that it was not at all a mouse pointer problem, but that something was causing random refresh issues in Finder windows and Open/Save dialog boxes when browsing the file/folder hierarchy of the Mac mini server. (The problem didn’t happen when they were browsing their own local file/folder hierarchy, only with remote browsing of the Mac mini’s hard drive.)

All of a sudden, while they were in the process of scrolling through the contents of a specific column or selecting a file or folder in a Finder window or Open/Save dialog box, Mac OS X would jump back up several levels in the file/folder hierarchy and select something higher up instead.

Needless to say, it was quite maddening and made using the server rather frustrating.

I suggested a number of basic troubleshooting steps, such as ejecting and remounting the remote volume on their machine, closing the Finder window and opening a new one, etc. But these steps didn’t do anything to solve the problem. Sooner or later, it would happen again. Sometimes they went hours without seeing it, sometimes it was happening all the time. It seemed completely random.

I said I would have to think about it and do some research, but it looked like a pretty obscure bug to me and one of the worst kind too, i.e. a bug that is impossible to reproduce reliably and involves a network setup where the source of the problem could be a number of different things.

Then the other day I was there again to wrap up the installation of the last couple of machines and answer questions that they might have. They raised the issue of the “jumping around” again. They said it was still happening regularly, and they were still hoping that I could find a solution. I said that I would try to take some time to seriously research the problem, but that we might have to try and talk to Apple about this if the problem was persisting.

Of course, the problem was not happening while I was there. But I had yet another look at things just the same, just in case I would notice anything that might explain the problem or at least give us some clue as to what was going on.

And that’s when I had a sudden small flash of inspiration. I saw that, in the file/folder hierarchy on the Mac mini server, they had several folders that had accented characters in their names. (We are talking about a French newspaper here, so accented characters are quite common everywhere.) In particular, I noticed that one folder at the root level of the hierarchy had an accented “é” in its name, and it was precisely when browsing the contents of that specific folder that, according to them, the problem was occurring with a particularly high frequency.

I then suggested that we try changing the name of the folder to something “simpler” without any accented characters in its name, just plain ASCII letters and spaces.

Unfortunately, the change in the name of that folder would break all kinds of links in their InDesign publications, since link file paths are stored by InDesign in its publications when placing external items and any change to the name of a file or folder in any path causes the link to break and requires that the user manually restore the link in InDesign after opening the publication.

But they were willing to put up with that if it meant getting rid of the “jumping around,” so we decided it was worth a try.

Today, more than a week later, I just got a “thank you” note from them telling me that, since we made the change, they haven’t experienced any more “jumping around” while browsing the contents of the Mac mini hard drive remotely.

Now, I would have to do more testing and experimenting in order to establish with certainty that it was indeed the presence of these accented characters in folder names that was causing this particular bug in Mac OS X’s file sharing feature. And I don’t really have time for this, so I guess there is no point in trying to report the bug to Apple.

But I still find it rather disappointing that even today, in 2010, we still have to deal with such a bug that is apparently caused by something as simple as accented characters in file names. It just seems to indicate that Apple’s engineers, like most other computer engineers I suspect, still do most of their testing with plain vanilla ASCII characters and neglect to focus more of their testing on issues that would primarily affect international customers that speak a language other than English.

And of course, by not reporting this bug because of time constraints, I am also playing a small part in perpetuating such problems. It’s not that I feel guilty about it. It’s just that, over the past three decades of daily involvement with personal computers, I have been able to observe again and again this big “Anglocentric inertia” in computer science and engineering that makes it inevitable that international computer users who work in languages other than English will always experience more problems, bugs, and annoyances than their English-speaking counterparts.

Don’t get me wrong. Things are improved a lot over these three decades. But even today, I still encounter, from time to time, French-language e-mails that are not encoded properly and have become illegible garbage by the time they reach me, or badly encoded web sites where the French text only renders properly on Windows machines, or other similar annoyances—to say nothing of badly translated user interfaces and controls which, as a professional translator, I find so difficult to live with in my daily activities that most of time, I end up choosing to use an English-language user interface even when I work in French. (And it’s also no coincidence that I am writing this blog mostly in English even though my first language is French.)

I am afraid that, at this rate, we will never have the expected and desired quality or reliability in international computer products. While some bugs are reported and eventually get fixed, others stay unaddressed for years simply because foreign users are not proficient enough to address them properly or do not have the incentive to do so. As a bilingual Mac user, I am in an ideal position to report on such bugs and help companies like Apple fix them, but I don’t have an infinite amount of free time to devote to such things without compensation, and I also often feel that it is a losing battle, that even if I report some bugs and they get fixed, other similar bugs will inevitably crop up, simply because most of our computer products continue to be designed by English-speaking engineers and developers for a primarily English-speaking audience.

I only have so much time to devote to repeating the same things over and over again in the vague hope that someone will get the message and implement testing processes and procedures that ensure that such problems are caught and fixed before products are released on the market. I am just glad that I was able to fix this particular problem for this particular client, and I guess that, by documenting it on this blog, I also hope that my observations can be of some use to other people in similar situations.

I am afraid that, at this point in time, that’s pretty much the best we can hope for.

This entry was posted by Pierre Igot on Friday, April 30th, 2010 at 4:27 pm and is filed under Macintosh. You can follow any responses to this entry through the RSS 2.0 feed. Responses are currently closed, but you can trackback from your own site.