Category Archives: Mac

giving up on Safari

I’ve been using Safari since its introduction in 2003.  Upon the death of Internet Explorer for Mac, I switched to Safari exclusively.  When other web browsers have come out, I’ve given them all a go, but I’ve always returned to Safari.  Until now: Safari 5.1 has forced me to switch to Chrome as my browser of choice.

Safari 5.1 now behaves more like Safari on iOS.  When Safari decides that you haven’t interacted with a page recently enough, it unloads that page to save memory.  This doesn’t match up with my usage of Safari in any way.  I often have several tabs open.  Those tabs represent a to-do list of sorts.  Some of the open are items that I simply want to read.  Others represent an action that I need to take: fill out a form, write a new blog post, write my weekly status report.

Forced reloading breaks every single one of these to-dos.  In the best-case scenario, the webpage that I’m reading hasn’t changed between when I started reading it and when Safari forces a reload of its content, so I haven’t necessarily lost anything other than my place on the page.  Even so, I lose the context of what I was reading, and I also lose the time necessary for the page to reload.  Occasionally, I lose the content of the page, if I’m offline when I’m trying to read the page but a forced reload has occurred.

In the cast of an action to take, the forced reload is even more irritating.  I lose my work: the partially-filled-out form, the incomplete blog post, the status report that I forgot to commit to the wiki.  At minimum, I lose the time that I invested in my half-finished work.  Recreating that work is always a losing proposition.

I tried to live with Safari 5.1 for a few weeks.  Slowly, I found myself trying out other browsers again.  I tried Firefox again, but its inability to respect my system proxy settings1 and its incorrect handling of keyboard shortcuts like option-arrow2 have made me move to Chrome.  I’m not sure if I really like Chrome yet, but it doesn’t break my workflows, and I don’t have the constant concern of losing my to-dos.

If a future update to Safari changes this behavior, I might try it again.  But Safari has really broken my trust with 5.1, and I don’t think that I’ll come rushing back.

  1. My wired access in my office is via proxy, but wifi has no proxy.  I switch between the two several times during the day: wired when I’m working in my office, wifi when I’m in a conference room.
  2. When editing text, option-arrow moves you to the beginning or end of the line.  Except in Firefox, where option-leftarrow is “back one page”.

LTFS wins NAB Pick Hit award

IBM Linear Tape File System (LTFS), which my husband Michael Richmond architected, won a Pick Hit award at the National Association of Broadcasters conference last week.

In short, LTFS puts a file system on top of magnetic tape.  This overcomes many of the disadvantages of tape, such as the long access time due to the linear nature of tape.  It also keeps the advantages of tape, such as the inexpensive cost per terabyte and its longevity.  It’s a big deal for media and entertainment companies who create and consume many terabytes of data per day, and who want to store it less expensively and more reliably than they can on hard drive.

I know that this isn’t quite one of my usual topics of user experience, VMware, or Macs, but I think it’s worthwhile.  Of course, LTFS does support Macs (obviously important for media and entertainment companies!), so maybe it’s not too far outside my usual topics.

iTunes wishlist

I think it’s safe to say that I put iTunes through its paces more than most people do.  As of this writing, I have more than 27,000 tracks in iTunes.  That’s not a complete representation of my music collection.  I’ve got another 40 or so CDs that I’ve acquired in the past month which will get ripped, and I’m in the process of re-ripping my existing CD collection to ensure that everything’s in ALAC.  I mostly like iTunes, but there are some things that I wish would change.

Over-the-air podcasts

Once I have listened to an episode of a podcast on my iPhone, I’d like to be able to update that podcast to get the next podcast in the queue.  I feel this pain most acutely on the weekends, when I’ve run out of podcasts on my iPhone but I can’t update since I sync my iPhone with my work Mac.  I know that there are apps like Podcaster, but they require that I only listen to podcasts on my iPhone.

Device-independent podcasts

Related to the previous point, the only reason that I sync my iPhone to my work Mac is podcasts.  I have some podcasts that I like listening to on my Mac while I’m working, such as the All Songs Considered live concert podcast.  I shouldn’t have to think about where I’m going to listen to a podcast and decide which of my Macs/iPhones/iPods will be with me when I am going to be there.

Better auto-complete

Since I have a massive iTunes library, there’s a lot for auto-complete to draw from.  But auto-complete also gets in the way.

This is most noticeable when capitalisation between what it’s trying to auto-complete and what I’m actually typing differ.  Once I get past the point where iTunes gives up on auto-complete because my new entry isn’t in its database, it keeps the capitalisation that it already had.  For example, I was entering the band name ALO today.  I typed A, and it came up with Al Franken.  Then I typed L, so it kept Mr Franken, but changed my upper-case L to a lower-case l.  And then I typed O, but it still had the lower-case l there.  In other words, I entered that information, don’t lose it.

Another potential improvement for auto-complete would be to match the composers field with the artist field.  If I’m in the composers field and start to type the first letter of the artist’s name, don’t give me the first entry that matches that alphabetically.  Give me that artist; if I keep on typing more, then maybe the song’s composer isn’t the artist.  But it usually is, so save me a little bit of typing.

Improved ability to update an album

As I mentioned earlier, I’m going through my CD collection.  Since I’ve been using digital music for years and years, my music is in a variety of formats.  As I re-rip my discs, I delete the old version.  I’d love to be able to select the old version and tell iTunes that I’m updating this album, and thus keep its tags and metadata.  I know that I can do this if I’ve kept all of its tags the same as what’s in the CDDB, but I correct the tags on my files when there’s incorrect or incomplete information in them.  I don’t want to lose the metadata because the CDDB is wrong.

Better handling of the “composers” field

Today, the composers field is just treated as a single text-entry box.  But modern music usually has multiple composers.  I’d like them to implement some kind of delineator (slash or semicolon would work well).  Then I could have auto-complete work on the individual composers of a song, instead of the whole list of composers of a song.  For example, Michael Stipe’s sister Lynda has writing credits on a song on Fables of the Reconstruction, so I’d love to be able to type “Bill Berry; Peter Buck; Mike Mills; Lynda Stipe; Michael Stipe” for that song and save myself a little bit of typing since everyone but Ms Stipe is a composer on another song in my library.

Carriage returns in comments

I love the comments field.  It makes my geeky little heart sing with joy.  I’d love to keep more information in the comments field, but it’s hard to do so because I can’t use a carriage return to provide visual separation.  I don’t want to write a novel in the comments field.  For example, I have a smart playlist for “covers”, which looks through the comments field for that string.  I also use the comment field for live songs, to note when and where they were recorded.  Without a carriage return, a song which has both of these pieces of information is messy.  It’s keeping me from using the comments field for other smart playlists, too.

What improvements would you like to see in iTunes?

the silent installation of Growl

I know plenty of other Mac geeks love Growl, but I’ve never liked it.  To be more accurate, I dislike any kind of notification system; it’s just that Growl is the most visible example of it.  I find notifications to be disruptive.  I don’t mind an Adium window appearing when I get a new instant message.  Beyond that, though, I prefer not to be interrupted from whatever I’m doing.  This isn’t to say that other people shouldn’t like Growl.  Everyone works differently.  If it works for you, I’m perfectly happy for you.  It’s just not welcome on my Macs.

Today, Macworld published an article about the mystery of spontaneously installed Growl.  This is one of the things that drives me crazy: other applications which install Growl without notifying me.  Adium gives me the option of installing it (and I thank them for that option), but Adobe CS5 doesn’t.  This is an especially frustrating user experience, given that I did a custom installation of CS5 and so have an expectation that it shouldn’t install anything other than what I selected during that custom installation.

I really appreciate that one of the developers for Growl said this in their interview with Macworld:

We hate it when people install software—any software, including ours—on other people’s systems without permission.

That’s pretty classy.  An official comment from Growl is a great thing, even though Growl has no way to enforce it.  It’s something that I hope that software developers take to heart.  If your software is going to install something else that’s not advertised as part of your software on my system, then you’ve got to both tell me about it and let me opt out of it.