Previous Issue | Search TidBITS | TidBITS Home Page | Next Issue
Last week Apple updated Mac OS X 10.4 Tiger to version 10.4.9 and provided a security update for Mac OS X 10.3.9 Panther. The security update is incorporated into the Tiger update, and could have been labeled "Fixes for the Month of Apple Bugs," a project we have written about before (see "MoAB Is My Washpot," 2007-02-19).
Security Update 2007-003 and the related code in Mac OS X 10.4.9 fix dozens of problems reported in the Month of Apple Bugs, including what was the most serious remaining problem, a way to exploit a flaw in Software Update by "enticing a user to download and open a Software Update Catalog file." We haven't seen reports of this - or any of the rest of the bugs - in the wild. Most of the non-MoAB exploits fixed by the security update require local users with access to an account and software that isn't enabled by default in Mac OS X.
There's no simple way to summarize 10.4.9's general enhancements. Like the last few updates to Tiger, this one is a grab bag of fixes for numerous individual problems, and it's likely the last big hurrah for Tiger, as Mac OS X 10.5 Leopard's ostensible ship date moves ever closer. Although Apple could release a 10.4.10, history shows us they prefer the numerical purity of single digits. (Jaguar ended its run with 10.2.8 and Panther with 10.3.9.)
Notable among the general changes are improvements to .Mac synchronization. As a regular .Mac sync user, I have seen lots of inconsistent behavior and long delays. I'm hoping 10.4.9 eliminates these problems. Another fix related to USB modems I have to call out as "I fax in your general direction": the note says that the update improves reliability in faxing in France or Belgium when using the Apple USB Modem.
Apple has made available separate incremental and combo updates for PowerPC and Intel systems running both Mac OS X and Mac OS X Server; you can use Software Update to download the best updater for your system or view all eight updates from the Apple downloads page. The combo updates work for 10.4.0 and later; the incremental releases work for 10.4.8. Set aside some download time, since the size of the updates runs from 72 MB (PowerPC incremental) to 350 MB (Mac OS X Server Intel combo). Panther's Security Update 2007-003 is also available both via Software Update and as standalone downloads for both Mac OS X (36 MB) and Mac OS X Server (49.5 MB).
As always, if you experience any unusual problems after updating, particularly with applications not launching, download and install the combo updater for your Mac, since it can provide a cleaner installation.
Apple also released iPhoto 6.0.6, which "addresses issues with EXIF data compatibility and photocasting." The photocasting fix is in response to another Month of Apple Bugs report. It's also available via Software Update or as an 8 MB standalone download.
A few months ago, Glenn Fleishman and I wrote a feature article for Macworld about using smartphones with the Mac (see "Get in Sync" in the January 2007 issue). When we tested BlackBerry devices, PocketMac for BlackBerry, which TidBITS also covered a year ago in "Putting BlackBerries in Your PocketMac" (2006-02-06), was the only method of synchronizing data between RIM's smartphones and the Mac.
Recently, a new contender has emerged. Mark/Space is currently offering a fully functional public preview of The Missing Sync for BlackBerry as a 9.2 MB download. Like other versions of The Missing Sync, this one uses Mac OS X's Sync Services to synchronize data between the BlackBerry and Address Book, iCal, or other programs that also use Sync Services. It also features photo and music transfers from iPhoto and iTunes on the BlackBerry Pearl. Remember that this is still beta software; Mark/Space has posted a list of known issues. It requires Mac OS X 10.4.8 or later, and a BlackBerry device running version 4.0 or later of the operating system. The completed version of the program is expected to ship by the end of this quarter; pricing has not been announced.
Apple has released iTunes 7.1.1, an update that "addresses a stability issue and minor compatibility problems in iTunes 7.1," according to Apple. The update is available via Software Update or as a 28 MB stand-alone download. A 36.1 MB download for Windows is also available.
The company also released iPod Reset Utility 1.0 for Mac (3.4 MB) and Windows (2.2 MB), which is designed to reset either generation of iPod shuffle if iTunes is unable to reset it.
After all my puzzling over how to create permanent links to articles in the New York Times (see "Easier New York Times Linking," 2007-02-26), I was amused to hear from a friend that the New York Times is now making TimesSelect free to any student or faculty member with a valid college or university email address. TimesSelect includes access to articles from the New York Times Op-Ed and news columnists in both text and podcast forms, along with up to 100 articles per month from the full New York Times Archive, which contains content back to 1851.
Freeverse has released Sound Studio 3.5, which adds many new features of interest to all users of this excellent audio editor and recorder. Sound Studio has long been my favorite tool for editing audio files, whether I'm producing a podcast, trimming a file in my iTunes library, or recording my son's funny snore one night (and later removing the laptop fan noise from the recording - it's important to have high quality embarrassment material for when he's a teenager).
Among Sound Studio 3.5's many new features are the addition of new audio formats for opening and saving, including Apple Lossless, ADTS AAC, NeXT/Sun Audio, and Ogg Vorbis. In addition, Sound Studio now supports 8-, 16-, and 24-Kbps bit rates when saving MP3 format files.
For podcasters, Sound Studio has added the capability to manage all iTunes-supported tags, including the podcast bit that determines whether a file appears in the Podcasts section of iTunes or in the regular Music section. This is a major boon to me, since I would like to move podcasts I want to keep into my music collection and let all others automatically delete themselves after listening, something that's been difficult to accomplish so far. Now I can open the file in Sound Studio, toggle the appropriate checkbox and re-import the file into iTunes where it appears in my Music section rather than in the Podcasts section. The reverse approach enables you to move spoken audio files from the Music section into the Podcasts section, if you so desire.
Podcasters will also appreciate the fact that markers set by Sound Studio within audio files are now automatically saved as chapters in podcasts. In other words, when playing back in iTunes, the marker titles are listed within the Chapters menu and enable one to jump directly to that spot in the playback. (For more information on how to use Sound Studio for podcasting, see my ebook, "Take Control of Podcasting on the Mac.")
Other enhancements in Sound Studio 3.5 include new preferences for how scrolling occurs during playback, the capability to loop sound in filter previews (rather than requiring you to play the sound over and over as you make adjustments), and new AppleScript support for referencing individual tracks and changing the pan and volume of each track.
Finally, Sound Studio now provides an innovative way to set the beats per minute (BPM) of a track. BPM is a piece of metadata that appears in the iTunes tag editor and as part of the display grid of the main Sound Studio window; it can be useful for generating smart playlists of slow or fast music in iTunes. All you have to do to set the beats per minute is click a button in time with the rhythm of the music.
Sound Studio 3.5 costs $80, and upgrades are free to registered owners of 3.0 or higher. Special upgrade pricing for users of earlier versions is available as well, as is a free demo (a 10.2 MB download) for anyone who hasn't yet tried Sound Studio.
[Andy J. Williams Affleck built Dartmouth College's first Web site in 1993, created the original Web site for the sitcom Friends, and started a virtual community that's still around a decade later. When he's not producing his Podcrumbs podcast or working on "Take Control of Podcasting on the Mac," he's a senior project manager and accessible Web design expert.]
At last month's 5th USENIX Conference on File and Storage Technologies, two academic papers - one from Bianca Schroeder and Garth A. Gibson of Carnegie Mellon University (CMU) and the other by Eduardo Pinheiro, Wolf-Dietrich Weber, and Luiz AndrĪ Barroso of Google - looked at the reliability of hard drives in large-scale installations. Among other conclusions, the CMU team found that real-world replacement rates were much higher than would have been expected from vendor-provided mean time to failure (MTTF) estimates, and Google's researchers concluded that there was little correlation between failure and either elevated temperature or activity levels. The papers weren't written for the lay audience and aren't easy reading, but they are worth a look if you're interested in when and why hard disk mechanisms fail.
Also interesting is the paper by James Cipar, Mark D. Corner, and Emery D. Berger of the University of Massachusetts Amherst on the Transparent File System (TFS). The goal of TFS is to create a contributory storage system in which multiple people could contribute unused disk space to a shared pool, much as the SETI@home project enables users to contribute unused CPU cycles to the shared task of analyzing radio telescope data. (And yes, there is still an active TidBITS team for SETI@home.) Apparently, TFS can contribute all of the unused space on a disk while imposing only a negligible performance drag on the contributor. Prototype source code is available; I'll be curious to see if anyone cleans it up and ports it to MacFUSE (see "MacFUSE Explodes Options for Mac File Systems," 2007-01-29).
The file-distribution system run by Pando has opened itself up to developers and podcasters (for more information about Pando and related services, see "Secure Transfer Using Civil Netizen and Pando," 2006-08-21). Instead of managing your own bandwidth and dealing with your service provider's limits or their overage charges on busy months, you can now employ Pando's combination of centralized and distributed downloading at no cost at all for files of up to 1 GB in size. Larger files can be distributed with Pando's paid levels of service.
With Pando, you upload an individual file or set of files (a "package") via a free application that incorporates advertising into its display. When you upload the package, you can provide email notification, where up to 10 people per transmission (for free publisher accounts) initially receive a Pando link to download the package or file. You can also receive a special link that you can send separately and some HTML to post on a Web site for download. In essence, Pando's client application works like a combination of an email application and a file manager. It's available for Mac OS X 10.3.9 and later, along with Windows 2000 SP4, XP, and Vista.
The copy of Pando running on your computer is also a "peer," or a potential uploading server, in a peer-to-peer network that Pando uses to connect anyone who has downloaded or wants to download the package in question. For files you want to distribute as broadly as possible - like a podcast, software download, or white paper PDF - the more people running Pando with a downloaded copy of that item, the greater the distribution network.
This method is what has driven the popularity of BitTorrent, but the addition of central servers to prime the pump and ensure a minimum level of bandwidth gives Pando some advantages. Pando requires accounts, and with a premium subscription as a publisher, you can even control who can download and use files.
With the addition of a developer site and associated toolkits and recipes for distributing files, you can bypass using the Pando program to upload files as a publisher. Anyone wanting to download files must first install Pando, but if you have sufficiently compelling content or if Pando really takes off - especially if they can get their software bundled by computer makers - that will be no bar. For anyone whose audience has no interest in using a separate application to retrieve files, Pando isn't yet a solution. But as a measure of their current popularity, Pando says they push 60 terabytes of downloads a day across their network.
For instance, you can convert your existing podcast RSS feed to work via Pando, and any audio or video enclosures are automatically retrieved by Pando and converted into retrievable packages. Currently, you need to request that Pando (the company) enable this option for your feed. But you can also use the Pando program to create URLs for uploaded files that you can embed in a Web page and that work in RSS feeds. A recent upgrade to the Pando application lets the program subscribe to Pando-compatible RSS feeds; for instance, they have a sample high-definition video channel that lets you download specific videos directly within Pando from that channel.
A software developer could offload some of the distribution burden for new files by using Pando's application programming interface (API), the programmer's toolkit, to create a package for files uploaded to the developer's Web site. Adding peers increases the efficiency of peer-to-peer networks; thus, a popular new application or update would have an extremely efficient Pando profile.
Pando uses advertising to defray the costs of its free service. The company also recently added three tiers of paid service; if you pay, you can transfer larger files to larger initial distribution lists, and your files can remain on Pando's central servers for longer periods of time. Right now, the free service allows packages of up to 1 GB each to be transferred, and it keeps the files active on Pando's servers for either 7 days (for Pando-emailed or IM-distributed files) and 30 days (for Web posted files) after the most recent download or message forward through their system.
All three higher level services - Plus, Pro, and Publisher - allow password protection of packages, weekday 24-hour technical support, and no advertising. They also let you send messages from within Pando to 100 people at a time rather than 10.
The Plus ($5 per month) and Pro ($20 per month) service increase maximum package sizes to 3 GB and 5 GB, respectively, and both allow Pando-distributed and IM-distributed packages to remain for 30 days without any downloads or forwards, rather than 7. The Publisher package allows packages up to 50 GB, and unlimited persistence on Pando's servers.
Pando versus Other Services -- Let's compare these offerings with other typical bandwidth charges. Amazon's Simple Storage System (S3) costs 20 cents per GB transferred, plus 15 cents per GB stored each month. Many co-location firms charge $1 to $2 per GB transferred with some monthly included amount. And some ISPs offer truly insane transfer amounts, notably DreamHost, which includes 2 TB per month in its $10 per month accounts.
At those prices, delivering, say, 10,000 one-megabyte files each month, or 10 GB of data, costs me nothing with Pando, $10 with DreamHost, $0 to $10 with a co-location host (if I'm under or over my allotment, respectively), and $2 with Amazon. Scale that up to 1 TB a month of smaller-than-1 GB packages, and it's free with Pando, $10 with DreamHost (unless they decide my usage is abusive), at least $800 and as much as $1,800 with a co-location host (assuming 100 GB to 200 GB of included usage), and $200 with Amazon (plus a few dollars for storage).
I'm particularly interested in this subject because of the potential risk for average people as an increasing number of us host audio, video, and other huge files, and thus face the peril of popularity. Nearly four years ago, I ran into just this problem when I tried to give away an electronic version of "Real World Adobe GoLive," a book I co-authored with Jeff Carlson (for the full story see "Publish (Electronically) and Perish?" 2003-03-24 and "The Boy Who Cried Bandwidth," 2003-04-07). I wrote about how it all worked out in the end in the New York Times, but I could have been on the hook for up to $15,000. (In that case, an older method of sustained transfers had ridiculous tiered levels of fees: one toe over the line, and whammo!)
Most network service providers charge you for the bandwidth you use over your monthly allotment; these companies tend to run co-location facilities that house hundreds or thousands of servers. Many Internet service providers, along with Apple and its .Mac service, cut you off when you hit your limit; ISPs tend to provide bandwidth to residential and business customers, and they try to preserve bandwidth rather than serve hosting customers. It's annoying to be taken offline, but at least you know that you're just out of luck, but not out of pocket, when you hit the limit.
I like the notion that Pando is bearing some of the risk for popularity, but balancing that with fees for higher levels of use and support. I also like the notion that bandwidth is such a commodity that Pando can use advertising to offset the cost of most file delivery. Pando needs to hit a mass audience to make the mental cost of downloading its application approach zero - or get pre-installed deals with computer makers - but this latest addition makes it increasingly likely that "Pando me that file" could become a common phrase.
[Editor's Note: Some of the URLs in this article use the gopher:// scheme rather than the familiar http:// scheme. These gopher URLs can be viewed directly in Camino or Firefox, but if you are using Safari, which does not support the Gopher protocol, view these pages using the HTTP<->gopher proxy. For more information, check out this document describing Gopher support in most Web browsers.]
Back in the early 1990s, when I was an undergraduate at the University of California camping out in the beat-up, 24-hour-a-day, VT100 terminal lab under one of the lecture halls, the World Wide Web was, well, not very wide and certainly didn't encompass much of the world. Graphical interface to the Internet? Are you kidding? Most of what the Internet had to offer then could be viewed on those text screens. All my activity happened while logged in over a serial port to one of the campus Unix servers.
Still, that monochromatic interface was the gateway to an interconnected world of computers very much like the Web - a world accessible both to the people typing away on those ancient dumb terminals, and to the lucky folks on the spanking new Mac IIci computers in the Mac labs. It had weather, headline news, music, search engines, and even video clips (if I could use one of the Macs). This was Gopherspace, and it's still alive today.
Back in 1991, Gopher sprang out of a University of Minnesota campus information service project aimed at building a "friendly" method of accessing university documents and services. (The University of Minnesota's sports teams are the Golden Gophers.) In those days, most campuses and corporations maintained their own walled-garden services and access policies, and almost all of them operated in unique and sometimes wildly different manners.
In contrast, Gopher provided a unified, consistent hierarchical interface to access everything. The approach translated well to both text and graphical interfaces, and better still, it offered an easy way to connect a varied set of hosts using simple links. This beat the stuffing out of getting files via FTP, which usually required using a command line. Gopher's method was a large improvement over interacting with library and campus directory systems via Telnet and trying to remember how to compose searches from system to system. Thanks to Gopher, the public resources other servers offered weren't merely accessible - they were usable.
Within a year or two, many other campuses were using Gopher for their own local operations, along with some private users and corporations. Gopher servers and gateways pulled together many disparate Internet resources, such as local directories and white pages (using CSO), and access to FTP servers and WAIS (Wide Area Information Servers - WAIS was an early standardized way to search remote databases).
At the same time that Gopher's purview expanded, point-and-click Gopher clients appeared, including some for Macs - remember TurboGopher, anyone? An increasing amount of information started to pour into Gopherspace, including electronic books, email magazines, pictures, programs, software and more.
Sorting through that growing mass of content required yet another piece of software: a search engine called Veronica (its name was a play on Archie, the search engine for FTP). No accounting tells us exactly how many Gopher servers existing during Gopherspace's heyday, but I remember all seven Veronica servers being busy during the day. As the Web become more generally popular, Gopher links were still rampant on Web pages because a lot of data was still in Gopherspace.
By 1995, Gopherspace had largely evaporated, thanks to a combination of the University of Minnesota's restrictive and expensive licensing policies (they eventually released Gopher under the open-source GPL license, but years too late) and the wide availability of a better technology. The new technology had the same interconnectedness of resources, while offering a prettier interface and wider possibilities for creative and informational communication. Naturally, that was - and is - the Web.
The University of Minnesota tried to salvage Gopher, with neat tricks like merging Gopherspace with virtual reality via the GopherVR project, but the Web had already passed Gopher by. Fascinated as I was with the Gopher world I used to inhabit, I threw together my first bits of HTML and put up my own home page on the Web in 1994, and Gopher became history to me, too.
Or was it? In 1998, while working as a programmer for Point Loma Nazarene University in San Diego, I wondered what happened to the old world down the Gopher hole. I brought up my own Gopher server software on the Apple Network Server sitting in the office early in 1999 and told it to go find the other Gopher servers out there. Surprisingly, a few answers came back.
The University of Minnesota's Gopher pages still worked, and they still had most of their links to former Gopher peers. Many of those hosts had turned into Web sites, and some had utterly disappeared, but a few were not only still operating but also still maintaining their content. I started compiling a list and trying to index their content, and eventually I put my database up for searching and browsing as gopher.ptloma.edu (with the IT department's blessing), the host that was the forerunner of Floodgap's Gopher server.
Other people had been wondering what happened to Gopher, too, and had erected their own servers independently. One day I got an email message from a fellow named John Goerzen, who had also written his own new Gopher software to run a service he whimsically called quux. Better still, along with his new content, he had managed to preserve a fair number of the archives of old Gopher sites that I thought had disappeared without a trace.
John was only the first of many people I would hear from who remembered the quick simplicity of Gopherspace. It got to the point where I started tracking all the new hobby and user servers that were cropping up. I even received a letter from Mark McCahill, one of Gopher's original architects, after he noticed the new Veronica clone that I had thrown together out of the data the Gopher crawler had acquired.
Gopherspace had never disappeared after all; it had just gone underground. Even after the University of Minnesota finally turned off gopher.tc.umn.edu a few years later, Gopher hobbyists live on, writing new features (like the Gopher "phlog"), creating clients and software, and adding new content to their own little worlds. Plus, most of the old Gopherspace archives now have new homes, meaning most of their content is still available today.
Nevertheless, Gopher remains more than just a living fossil. In a world where flash (and sometimes Flash) is often more important than substance, Gopher replaces all the trappings with a clean, sterile, and consistent interface of folders and files. The Gopher sites that people visit have real content and real function, so there's nothing but a menu between you and gigabytes and gigabytes of data. You can still access Gopherspace with a dumb terminal just as well as you can with a Mac Pro. It loads quickly over a dial-up link, and it's instantaneous over a broadband connection. You can still get weather reports in Gopherspace, you can still read mailing lists and headline news, there are still lots of files for downloading, and heck, you can even read TidBITS! (Thanks to Adam Engst for granting permission.)
More people are discovering that there's an alternative to the World Wide Web for many functions, and better still, an alternative that can co-exist seamlessly with the Web - all the Mozilla-based Web browsers work fine as Gopher clients too. Maybe it's for that reason that the Power Mac 7300+G3 that runs gopher.floodgap.com today still gets a few thousand hits daily.
Yes, there are far fewer Gopher hosts than there used to be (86 hosts and 740,000 unique resources, as I glance at the robot statistics file while I perform maintenance on the Veronica-2 index). But the world down the Gopher hole is still alive more than 15 years from its inception. If the Web seems to be a heavy or fluffy distraction as you wait for your browser to grind through another Flash animation and a pile of ads, perhaps it's time you took a trip back underground for a glance at the simpler and cleaner world that the Internet used to be.
[Cameron Kaiser is a recovering database administrator and programmer who unwisely got an MD instead and now works as a county health physician in Southern California. He drives old United States highways, maintains old Commodore and Apple computers, and relentlessly implements old information technologies on his "$50 Wal-Mart server rack" in his rapidly disappearing spare time. He has used Macs since 1987 except for a brief stint we shall not talk about.]
Nike+iPod Only for Fitness Runners -- Adam's article about the running device elicits discussion of stride lengths and the variation in the Nike+iPod's readings. (14 messages)
TidBITS 2007 Reader Survey Results: Who Are You? Readers respond to the first results of our reader survey, including readers under the age of 21 and a question of how the data breaks down. (3 messages)
I lived through Daylight Saving 2007 and survived! Disaster averted, readers note some of the side effects of this year's adjusted Daylight Saving Time. (3 messages)
Mac OS X update weirdness? Mac OS X 10.4.9 causes some sporadic problems, which seem to be solved by using the combo updater on Intel-based Macs. Also, a new delay in ejecting discs appears to be a feature, not a problem. (11 messages)
Previous Issue | Search TidBITS | TidBITS Home Page | Next Issue