Cool Mac Gear

That is the question. I haven't had time to do a whole lot with my new iBook yet. Last week was a bit hectic with MacWorld Expo, and I'm still waiting for the 512 MB RAM upgrade to arrive, but I'm pleased to note that the Finder response in OS X, even with the standard 128 megabytes of RAM, is much more lively than it is on my 500 MHz Pismo PowerBook. I'm also very impressed with how bright and clear the little 12.1-inch display is.

One of the setup issues I've been pondering he is whether to partition the hard drive or not. I've always partitioned my drives, and my inclination has been to create at least two partitions -- a relatively small one for booting directly from OS 9, and a larger main one for OS X. If I'm going to do it, now is certainly the time, before I transfer all of my archive and application files to the new machine.

However, the iBook can boot into OS 9 from the stock single partition, and I've been using OS X so much lately, even on the Pismo, that I'm doubtful that I will be running in OS 9 much on the faster iBook anyway. The principal advantage of creating a separate OS 9 partition is for maintenance and troubleshooting purposes. Most of my disk diagnostic and repair software is OS 9 native, and a separate bootable partition facilitates running it without the hassle of booting from CDs.

Having a second partition also makes it possible to keep a separate, fully-configured copy of OS 9 for booting from, and a stripped-down copy for Classic Mode in OS X, but a disadvantage to that is maintaining continuity of preference settings and the like.

Randy Singer's very helpful MacAttorney OS X site makes a strong case against partitioning, noting that OS X likes to grab a significant amount of free hard drive space for caches, virtual memory, scratch space, etc. OS 9 (Classic) likes to do this to some extent also. And not only do they like free space, but they like contiguous free space. (That is, the free space must be in one unbroken chunk, not strewn about in pieces on your drive.)...

"Installing OS 9 and OS X on an unpartitioned drive gives both OS's the maximum available amount of free contiguous space to work with that you can give them... Yes, there are some minor advantages that you give up if your drive isn't partitioned (the biggest one being the ability to do some maintenance and disk recovery on one partition while booted from another) but gaining system reliability is more important that these things, at least it is to this author."

I don't entirely agree with regard to OS 9. You can always assign OS 9 Vritual Memory to another drive partition if it's own gets cramped. However, I'll concede the point about OS X.

In the same article, the author makes some interesting observatioins about disk defragmentation in OS X, advising that "it isn't necessary to have and use a defragmentation utility routinely if you are running OS X," because:

"While UNIX-based OS's are subject to a lot of DISK fragmentation (i.e. there are unused spaces between areas filled with data on the drive), they automatically try to make sure that there is no FILE fragmentation (breaking up of a program into pieces strewn about your drive) if at all possible.

"Disk fragmentation does not slow a drive's performance. File fragmentation will slow down your drive's performance, but it only occurs under OS X if your drive is just about full."

So, I'm wavering. I'm still partial to partitioned drives, but it might be time to try the alternative.

Kickass Site

From Jima

Charles Moore,

I have read your stuff off and on for a long time.

I surfed in from MacSurfer, my default opener for Mac.

I must say, I am spending a lot of time, awake in the middle of the night, and really do appreciate your work. You are bookmarked, and passed on to a couple of friends. Thanks for what you write, really like the thread OSX Oddyssey 239.

I am now figuring out the system on Applelinks, and it looks fine.

(What I was really hanging out and looking for was someone who might make an estimate of the bottom end of the Mac food chain where OSX will run okay.

I am right now on a G4 400Mhz (AGP Graphics) 512RAM, and am not sure if I am wasting my time putting Safari on this machine. Then I noticed you are using OSX on a G3 500 Pismo. Any opinion on this Graphite 400 box and Jaguar? I have an eMac where I really enjoy using OSX. I was reluctant to accept it at first, and now, it is just fine)

Thanks again.

Jim Corti
LA/Bangkok

Hi Jim;

Thanks for the endorsement of Applalinks. We try.

I don;t have any experience with Jaguar on a 400 MHz (or any) G4 PowerMac, but I did have a G4 Cube for a while (450 MHz) and OS X was pretty lively on it. I would think that you should be able to get as good or better performance than I am with this Pismo.

Charles

Reluctant Upgraders

From Ken Cavaliere-Klick

I just wanted to add to Nick Daisley's and Mike Etun's comments ( Moore's MailBag Friday January 10) about being counted as X upgraders. I am in that same count I have X (10.1.3 with the 10.1.5 upgrade) and I don't use it either. In fact, I didn't want it either. It was the only way I could find a retail copy of 9.2. I knew my old Bondi iMac would not run it well despite Apple's claims it would run it. I also knew many of my devices drivers were not available. To be fair, I did try it for a month, then reformatted my hard drive for 9.2.2 to get rid of it. My old iMac has 256mb of memory, 6gb video and a 30gb hard drive, hardly original for a Bondi 233. X was more 1.1.5 than 10.1.5. Slow, buggy, problematic. Of course, I also got caught in the "no upgrade" debacle making yet another expensive upgrade very unattractive and very suspicious.

If Apple wants more people to switch AND be satisfied they have got to come up with an alternate and appropriate installation for older iMac and Mac users. A lighter, simpler Aqua and Quartz engine is much needed, if only for older Macs. Of course, that won't sell new Macs. Before the knee jerk "can't be done" starts, it can. PPC Linux supports the older Mac's far more completely (and dual boots and runs 9 within Linux). Personally, I am at a cross roads here. I put off my new Mac purchase indefinitely since my existing system is just fine. But Linux is looking mighty attractive to me and I have lots of hard drive space to play with. The question becomes, once I try Linux and (if) I like it (most likely), what happens when the Bondi finally melts down? A Linux computer is considerably less expensive than a Mac.

Hi Ken;

Good point about Linux. The various Linux PPC distros prove that Unix can be made to work efficiently and speedily on older Macs.

Personally, I'm not inclined to forsake the myriad collateral joys of Mac computing, but I share your desire for an "Aqua Lite" as it were.

Charles

Switching to OS X

From Ben Dittman

I agree with your assessment that many people don't switch to OS X because many applications won't work with it. However, there is another reason that should be considered. There are a significant number of Mac users who own machines that don't support OS X. The G3 upgraded Power Mac 7100 that I am writing this message from cannot support OS X and probably never will. The same goes for 6100 and 8100 owners. I've also got a G3 upgraded 6400 that will never support OS X, and my brother has a 6500 with the same issues.

Many of us who own unsupported machines cannot afford machines that do support OS X. I can certainly count myself among that number. Even if I could afford to buy a machine that supported Apple's newest OS, I could not afford to switch many of my applications (in which I've built a significant investment over a period of time) to OS X native versions. This is just another point to consider when looking at reasons people don't switch to OS X. That's my 2 cents worth. Thank you for your time.

Sincerely,
Ben Dittman

Thanks Ben;

Good point about the cost of upgrading applications, which for some folks could more than double the outlay necessary for upgrading to a computer that supports OS X.

I am a fan of gradualism ans incremental change (as readers of this column are well aware ;-) ) Happily, Classic Mode works pretty well with many older applications.

Charles

5 million?

From: David Meyer

Charles,

The comparison of folks on OSX vs OS9 - 5 million vs 20 million? - must be viewed in with the understanding that people tend NOT to upgrade. You've heard "If it ain't broke, don't 'fix' it."? A lot of people feel that way about software. Worse, a lot of people will SAY that when they mean "Upgrading software scary, and the OS is the scariest!". Now pile on the systems that are simply too old or otherwise limited to run OSX.

• My son-in-law will not upgrade from OS9 because he's concerned about replacing software.
• My wife will allow ME to upgrade her G4 iMac - or it won't get done.
• My Mom's blue iBook won't get an upgrade - not enough memory and she is happy as is.
• The Quadra my Mom gave away can't even be upgraded to accept the new OS - but it works fine.
• My son and I (Yosemite and Quicksilver) are often at the head of the upgrade line.

I think five million is not a bad number. More will transition as they change systems - in a year, or a decade.

I have not booted to OS9 in months. There are a few games I run in Classic - shareware, only one has been converted to Cocoa or Carbon. Safari is working for me, but I have broadband. several iApps are doing things for me I've never really bothered to do before. (make movies, listen to music via my computer, keep a calendar, etc) This is a good time to be a Mac user, and the times show every sign they will be getting better.

Dave Meyer

Hi Dave;

A good time indeed.

Charles

Re: Why I Don't Use OS X

From Tamas Jakab

Paul Delcour wrote:

"I do not use OSX because I have far too many OS9 only programs. Besides it makes my Mac G3 beige very sloooow. After installing I thoguh I'd switch 'tween the two. No way! My Mac refuses to restart with OS9, it immediately switches to OSX. I then deleted OSX: there's no way to uninstall but by deleting. That however left me with some files that made my Mac still start in OSX. I then erased the OSX partition. I then ended up with a kind of debug screen in which typing BOOT or the likes didn't work. I then reinstalled the 9.2.2 update and hey presto: I'm back with the living Macs. I find all this scary. It seems impossible to go back to 9 after installing X. It's still no go for me."

I ran into a similar problem on my Beige G3 after I upgraded the CPU to a G4. Apparently the G3's PRAM is not fully compatible with the G4. The "debug" screen he's ending up in is Open Firmware, because it doesn't know to look for the Mac OS 9 ROM file.

Anyway, the remedy for this is to use Apple's short-lived System Disk Control Panel (from the Mac OS X Server 1.x days) to set the startup disk in OS 9. It should also fix the problem switching between OS 9 and X. It can be found here.

Tamas Jakab

Scanning software for X

From Anonymous

Charles,

http://www.hamrick.com/ makes a very popular scanning software for Mac OS X.

There is also another by Silverscan.

This is in response to your article:

http://www.applelinks.com/articles/2003/01/20030110133456.shtml#92

Sincerely,
anonymous

Check over 400 Macintosh and 200 GIS/Mapping websites at
http://www.macmaps.com/

Thanks for the links.

Charles

Browser Tests

From Jon Smith

Charles,

Thought I would use your links and do some browser testing myself. I tested Mozilla, Chimera and Safari. (Opera has been off my radar for a long time) I am running OSX 10.2.3 on a G4500/DP with 1 GB RAM on a cable modem with attbi.com. Below are my times.

Applelinks.com

Mozilla 5 seconds
Chimera 3 seconds
Safari 3 seconds

Low End Mac

Mozilla 2 seconds
Chimera 2 seconds
Safari 2 seconds

National Post

Mozilla 11 seconds
Chimera 10 seconds
Safari 9.5 seconds

Environment Canada

Mozilla 5 seconds
Chimera 3 seconds
Safari 3 seconds

I added the times together and the results are,

Mozilla 23 seconds
Chimera 18 seconds
Safari 17.5 seconds

I found Safari the clear winner. What added to this was the "snap back" in Safari. Returning to the sites via the keyboard was instantaneous in Safari while it was as long, sometimes longer (Mozilla) returning to the original page.

For the first beta release of Safari, I am totally impressed! I have used all the of the above through different versions and it took many releases to get as fast as Safari is on it's first! I have a new browser! You go Apple!

Thanks for the columns and insight you provide through your columns

Jon

Hi Jon;

Your results underscore the assertion that Safari seems to be a better relative performer on broadband than it is on dailup.

Or is it always? Check out this comparison with Chimera:
http://www.maccritic.com/reviews/comments.php?id=143_0_13_0_C

Thanks for the report.

Charles

Safari speed

From: Edward Starkie

Dear Mr. Moore,

Tried opening the National Post site--it took approximately 17 seconds with Safari on my Pismo 400 running 10.2.3. Low End Mac took about seven seconds. Applelinks took about three seconds from the MacNewz link.

Chimera opens Applelinks in seven seconds with my connection compared to three with Safari, where National post took 10 seconds on Chimera and 17 with Safari. Seems hit or miss. By the way, I enjoy your column.

Edward Starkie

Hi Mr. Starkie;

Glad you enjoy the Odyssey. It seems that Safari hiccups a bit on that National Post site even with DSL.

Charles

Safari performance and other issues...

From Jaime Hancock

Charles,

I have been a fan of Applelinks for awhile and especially your articles. I don't always agree with you, but I sense that we share many views in common, not the least of which is a Judeo-Christian worldview and outlook on life.

One of the things I have enjoyed is reading your experiences with your computers and with OS X, even if my own experiences have been somewhat different. Interestingly enough, part of the reason that I have enjoyed your articles as much as I have, is because they are not too objective. Pure objectivity is not possible by anyone, and I believe the best we can hope for is proper subjectivity which strives for a certain standard of objectivity.

This was clear to me in reading you articles about your mileage with Safari. While I personally have found Safari to be comparable in speed to Chimera, sometimes faster sometimes slower, I don't discount that your results are accurate. As you yourself said, there are several factors that may be influencing your performance that are necessarily different than mine. I'm using a 400Mhz "Pismo" PowerBook with 576 MB of RAM, running 10.2.3 and using a DSL connection. That setup is different than yours, and should be expected to produce different results. But the tests that I ran on my computer, while perhaps less rigorous than those performed by some others, are no less "real". I thought that you're articles were fair and stated quite clearly that they were your results, not proof of some vast benchmark conspiracy on Apple's part. Anyway, please keep writing your articles, it's refreshing to read well thought out opinions.

By the way, I notice you have mentioned the performance of OS X on your computer as being less than desirable. One thing I have found is that X runs much faster if there is no OS 9 on the computer, not even for Classic. I understand that you have some applications that require OS 9, but if you ever get to the point where all of your apps are on X and you can financially afford to have them all updated, remove OS 9 completely from your HD and you will see a noticeable speed increase.

Ever since Accordance come out as a Carbon app, I have been X only, and I have to say that I was amazed at the speed increase after I reformatted my HD and only put X on it. I have noticed the same thing on my iMac at work. Once I reformatted with only OS X on, it ran noticeably faster. The difference wasn't because I was actively running Classic on my computer, because I rarely used Classic and always quit it after I finished using Accordance, but simply having OS 9 on the same HD, including a different partition, caused my computer to run slower. Just some food for thought if you find yourself able to switch to OS X all the time in the near future.

Even today I can't understand the complaints about X's speed. It runs fine for me. Of course, once I go digital with my camera setup, I'm going to invest in a TiBook. I hope Apple updates them by July with the new Aluminum case, integrated Bluetooth, Airport extreme (with the better Antenna placement), and of course a speed bump. Then I'll be able to do photo, video and Bible teaching all from one computer. Of course I'm also thinking of some ways that I might be able to integrate Keynote into some of my lessons, at least the background/historical lessons that we do before we start a new Bible book. I could get my pictures from the Bible Lands Photoguide CD. Hmmmm. Or maybe I'm too much of a tech geek for my own good. ;)

Anyway, I look forward to reading some more your interesting articles.

Grace and Peace,
Jaime Hancock

Hi Jaime;

Thanks for the comments.

I'm very fond of some commentary on the "objectivity" topic from John Fraser's book “Saturday Night Lives: Selected Diaries” (MacLelland and Stewart). Fraser was editor of Canada's Saturday Night magazine for seven years in the '80s and early '90s. Prior to that, he was a reporter and foreign correspondent for the Toronto Globe & Mail for 17 years. In the book, Fraser delivers a withering critique of journalism as it is currently practiced, and is especially caustic about the "journalistic objectivity" cult -- calling it “one of the vainest goals a humble craft ever set itself.”

“There is no such thing as a strictly objective story,” declares Fraser. “It isn’t possible. Everything -- from the structure of an article to the choice of facts is filtered through a particular outlook and a prejudiced mind.... The most you can hope for... is relative honesty. And the very best (ie: the most honest) journalists always let their readers know their specific prejudices and the general nature of the intellectual equipment through which they distill their stories.”

Words to live by in this game.

I doubt that I will be ditching Classic anytime soon, even on the new iBook. Too many mission-critical (for me) applications still require Classic. However, your point about removing OS 9 speeding things up is intrigueing. I wonder why this is.

I do note that in the http://www.barefeats.com/pb17.html benchmarks run last week by BareFEATS Rob Art Morgan, pitting the new AlBook PowerBooks against a couple ot Tis and an 800 MHz iBook, the little G3 machine, which presumably had OS 9 aboard, actually smoked the new, OS X-only, 867 MHz G4 AlBook in the Quartz Graphics and the CPU Tests (but see some notes on this from Trans Tex Software's Tom Bender in Moore's MailBag today). :-)

All the best for your Keynote presentations.

Charles

Safari

From Frank L. Sinz

Hi Charles!

As a regular reader of your columns I feel free to address you by your first name. I truly enjoy your writing, although I have fully switched to Jaguar 10.2.3. Only on very rare occasions do I use Classic. But every time I read one of your columns I learn something.

What compels me to write to you is the posting of David Chilstrom in Odyssey 239. In your postings you always note, that you strictly report about your experience in your set of circumstances. The vitriolic remarks of Mr. Chilstrom are therefore uncalled for. One does not lie if one reports ones experience correctly with all the pertinent information about the hardware and software, especially version of the OS. Mr. Chilstrom did not allow for that in his posting but simply pitted you against Steve Jobs. Keep up the good work and don't let postings like these dishearten you!

My own experience with Safari is very good, it seems faster than Chimera 0.6. But I do run Jaguar 10.2.3 and Apple clearly states, that the browser is optimized for that Version. I run the browser on a Dual 500 Mhz G4 with almost a Gig of memory and broadband connection. My experience, therefore, is different from yours. I am sure you will have a different report once you have your new iBook up and running with sufficient memory and Jaguar 10.2.3. I am looking forward to your rave reviews of Jaguar once you have reached that point.

What I like about Jaguar the most is its stability and the multitasking. For many years I owned an Amiga with a true multitasking OS, which BTW was co-written by my son Michael Sinz. After the demise of that platform about 8 years ago I switched to the Mac. Almost immediately I was annoyed by the frequent - at least for my experience - crashes of the system. Starting out at OS Version 7 all the way to 9.1 I just could not get used to the fact, that I had a crash almost every day. But I knew, that a new OS was coming. I also had no alternative because my experience with Windows (at work) was much worse than the Mac OS.

Needless to say I was an early user of OS X and have not looked back since. I do not shut down my machine for months at a time and I have yet to experience a single system crash! That - to me - is the reason not to use OS 9xx and I can overlook the few shortcomings of Jaguar. I know, most of the features that made the "old" OS so efficient in many ways will be coming in future upgrades of Jaguar. I have switched from Quark to InDesign and even that was a blessing because Adobe's product is superior to QuarkExpress.

And yet I enjoy your musings about OS 9 versus Jaguar every day and I do learn a lot from them.

Happy New Year and cheers

Frank L. Sinz

Hi Frank;

Thanks for your thoughtful comments. David may have been having a bad day, but I do realize that my style of reportage does get under some people's skin -- not just in the computer genre. I'm not a techie or a scientific reductionist by nature or temperament, which I think sometimes tends to irritate those who are.

As I have said here a number of times, if I had been suffering the instability in OS 9 that some folks report, I would have switched to OS X long ago and not looked back. However, the OS 9s have been very satisfactorily stable for me, and with the extra speed......

That said, I'm working more and more in Jaguar, and I anticipate that when I get the iBook set up as my production machine, I'll probably rarely boot it into OS 9.

Charles

Safari

From: Steven Cades

Charles--

I downloaded Safari yesterday, and while it's not a fully-fleshed-out browser yet, it is--on a pretty fast college LAN--a very slick product.

But that's not what's exciting about it. Somewhere on the Web--I think in an Applelinks item--I read an excerpt from a message from someone in the Safari development team that indicated that Safari was built on an open-source base, and that it would remain in the open source domain.

That suggests a development strategy that's really subversive (in a good sense, I think). If Apple continues down this path, we will see a whole series of applications--maybe only in the "iLife"category, but maybe not--that will be freeware based on the work of all those wireheads who love making software that is smaller, faster, more elegant, and more bugproof than MS could ever produce. That is, a major American profit-making corporation--Apple--will leverage the work of amateurs (in the true sense of the word) to benefit not directly by selling their free product, but by selling hardware and operating systems made more desirable by their work.

Whadda'ya think?
--Steve

Hi Steve;

Yes, Safari is based on the same KHTML browser engine as the Linux KDE Konqueror browser. I'm delighted to see more convergence between OS X and Linux application development.

Don Melton of Apple the engineering manager of Safari sent a greeting communique to the KDE forum.

In the email, which is posted in its entirety here, Mr. Melton notes:

"

It's important that you know we're committed to open source and contributing our changes, now and in the future, back to you, the original developers. Hopefully this will begin a dialogue among ourselves for the benefit of both of our projects...

"The number one goal for developing Safari was to create the fastest web browser on Mac OS X. When we were evaluating technologies over a year ago, KHTML and KJS stood out. Not only were they the basis of an excellent modern and standards compliant web browser, they were also less than 140,000 lines of code. The size of your code and ease of development within that code made it a better choice for us than other open source projects. Your clean design was also a plus. And the small size of your code is a significant reason for our winning startup performance as you can see reflected in the data at http://www.apple.com/safari/"

More about the technologies underlying Safari:

WebCore

WebCore is a framework for Mac OS X that takes the cross-platform KHTML library (part of the KDE project) and combines it with an adapter library specific to WebCore called KWQ that makes it work with Mac OS X technologies. KHTML is written in C++ and KWQ is written in Objective C++, but WebCore presents an Objective C programming interface. WebCore requires the JavaScriptCore framework.

The current version of WebCore is based on the KHTML library from KDE 3.0.2. Changes that are specific to WebCore are marked with #if APPLE_CHANGES. Other changes to improve performance and web page compatibility are intended for integration into future versions of the KHTML library.

JavaScriptCore

JavaScriptCore is a framework for Mac OS X that takes the cross-platform KJS library (part of the KDE project), combines it with the PCRE regular expression library, and makes it work with Mac OS X technologies.

The current version of JavaScriptCore is based on the KJS library from KDE 3.0.2. The few changes that are specific to JavaScriptCore are marked with #if APPLE_CHANGES. Other changes to improve performance and web page compatibility are intended for integration into future versions of the KJS library.

KDE

Information about KDE, the K Desktop Environment, and both the KHTML and KJS libraries is available at http://www.kde.org.

PCRE

Information about the PCRE regular expression library is available at http://www.pcre.org.

For more information, visit:
http://developer.apple.com/darwin/projects/webcore/

Safari

From Stephen Green

I am running "bleeding edge" hardware and I can vouch for Steve Jobs statement and agree with you all at the same time. Whenever Apple releases new software whether it be an OS or application it is always tested and marketing statements are based on Apple's latest and greatest hardware. I am comparing Safari to IE and Netscape since whenever I have run Chimera/iCab/Opera I run into screen problems and the inability to log into my company remote email server. I haven't had any of these problems with Safari and it is faster than IE and trounces Netscape.

Stephen

Hi Stephen;

A balanced assessment. Glad to hear that Safari is working well for you.

Charles

Safari

From Lee Lamb

Charles,

Safari is a lot faster on my Ti 667 Powerbook with 1 gig of ram running from a cable modem through airport than anything else I've tried.

If they can make it remember my login names and passwords for sites I go to regularly it will be a improvement over IE, Netscape, Mozilla, iCab, and the like.

Later,
Lee

Hi Lee;

Good to hear.

Charles

Re: Safari speed

From Marc Reed

Charles -

I just tried your speed test using http://www.nationalpost.com/national/ and found it loads in about 7-8 seconds. This is a cable connection on a B/W G3 400 w/ 448 megs of ram running OS X 10.2.3.

I downloaded Safari as soon as the keynote was over and ran it immediately and I'm definitely not an early adopter of anything.

I have to say that from the first my impressions were that it ran so much faster than IE 5.2 and Netscape 7.01 that I made it the default browser. ( And it has nothing to do with it being an Apple product. ) My benchmark sites include several weather sites with animated radars and some Mac sites ( Macdirectory and, I'm sorry to say, Applelinks ) that are interminable when all I want is a little news.

I know that these " shootouts " can be subjective, but in this case I'm willing to give the hype it's due.

Marc Reed

Hi Marc;

No argument from me. If it works better than the others for you, cool! If it worked better for me I would use it. I'm certainly not biased against Apple products!

Charles

Safari load-out

From Joe Ballo

Well of course... ...

"One possible pertinent variable is that I am running Jaguar 10.2.1, which is supported by Safari, but the literature says that it is optimized for 10.2.3. Anyhow, I decided to run some timed tests to backup subjective evaluation with some objective numbers."

What did you expect. The Safari beta is explicitly stated to be optimized for 10.2.3? So you run it on 10.2.1 and - surprise of surprises - it doesn't run as fast as what the designers say it will!! But these are not objective numbers - they are numbers that appear to be purposefully fudged by not using the platform that the application is best designed to run under. Please get with the program, keep your system up-to-date, and then start testing software and systems in a meaningful way. I have an old IIfx in the closet . Maybe I'll test Safari on that with system 7.1. Oh - I forgot -it'll probably not run at all!!!

Dr Joe

Invalid results

From Dave

In your recent speed test of Safari vs. the other guys, you begin by speculating that the reason why other users seem to have more luck with Safari speed than you is that they are probably running 10.2.3, while you are still using 10.2.1. Safari is published as "optimized" for 10.2.3, so that logic seems to make sense.

Then you proceed to run a series of tests on several browsers, and observe that Safari is the slowest of the bunch. You also mention that you are still running 10.2.1.

One important question your article did not answer: why?

10.2.3 is a free update, and not even a very big one. Why on earth you would run these tests on anything other than the latest revision of 10.2 simply baffles me, especially when you came right out of the gate indicating that using 10.2.1 might result in poorer performance from Safari.

Also, running the tests on an ultra-slow modem is a very bad way to score a browser's rendering speed, especially when you load each site only once for each browser. The test becomes skewed even more when you reboot and dial out again before testing each browser. What speed did your modem handshake at for each phase of the test? No way to tell here. Also, what time of the day was it for each test, and how bad was network traffic at your ISP, the sites you visited, etc. Issues of traffic and line quality introduce a huge fudge factor under these circumstances. So much so, that I would insist that all of your results fell within the margin of error for the test, and therefore your tests were completely inconclusive.

The ideal tests would be to run them on 4 identical systems, all on the same network, all at the same time. Since that's not practical for most people, a good alternative would be to run the tests 20 or 30 times and publish average results.

Either way, I think you owe your readers an explanation why you did not bother to upgrade to 10.2.3 before running this test, especially when you believed that your choice to use an out-of-date revision of 10.2 was crippling Safari's performance in this shoot-out.

Hi Dr. Joe and Dave;

Apple's Safari page says:

"The Safari Public Beta works on all Macintosh computers running Mac OS X v10.2 or later, but runs best when using Mac OS X v10.2.3."

No purposeful fudge at all. It's officially supported by 10.2.1 (which is only what -- three months old? -- and what came preinstalled on my brand new iBook that arrived last week), so Dr. Joe's IIFX/System 7 analogy is gratuitous, and calling 10.1.2 "outdated" is a bit much. In the real world. Aside from a small minority of enthusiasts, people don't as a rule scramble to grab the latest system upgrade the moment it comes out. For example, I sold a WallStreet PowerBook back in 1998 to a woman who uses it to run her contracting business. We did get her upgraded to OS 8.6 when she brought the machine in for service a year later, but that's it. My friend with a 2001 Ruby iMac is still running OS 9.something, and My local newspaper is still working with the OS 10.1.x that shipped on their iBook 500 MHz.

Trying out Safari on a supported, three month old system version is fair ball. I expect it works best on broadband with a dual processor 1.25 GB Power Mac as well, but most people don't have those. I clearly stated what equipment I'm using.

As for "optimized," IBM ViaVoice X, for instance is "optimized" for the G4. Does that mean I have no business evaluating and reviewing it on a supported G3 machine (which I have done)? Puh-leez!

I've had the 10.2.3 updater file downloaded since the day after it came out, but there have been so many negative reports on the forums that I haven't been inclined to run it yet on my mission critical Pismo. Now that I have the iBook, I'll have a bit more flexibility, but I make my living with these computers on tight deadlines, and I'll be updating the iBook to 10.2.3 forthwith, but I can't afford to risk downtime for troubleshooting or other hassles just to stay at the bleeding edge with my number one machine.

I wasn't using a slow modem; but rather the built-in 56k unit in my Pismo. The bottleneck is the 12 miles of ancient, poorly maintained, copper phone lines between me and the telco switching station. 24,000 or 26,400 are the best handshake speeds one ever sees here on the one available ISP. I did not reboot and redial before each browser download. I did reboot the machine and freshly dial up before running the tests back to back to back to back. The relative numbers for Mozilla, Chimera and iCab were remarakbly consistent with previous similar tests I've run.

As I said in the article, "This is just an observation of real world performance over a dialup connection at a particular point in time. Milage may vary," Safari was not last but second slowest overall under the clearly described parameters of the comparison. The results from similar comparos posted by readers here and others elsewhere running on broadband and different Macs indicates that browser performance is very ideosyncratic.

Run the tests 20 or 30 times? Uh.....there are certain time constraint realities that obtain.

The POINT of this exercise is to provide a frame of reference, not scientific lab test results, whih I am unequipped to offer. The performance I cited is the performance I'm experiencing under conditions and with equipment that a lot of readers can relate to. The vast majority of Internet users are on dialup and don't have the latest hardware. My original impression of Safari was that it was slower than Chimera and Mozilla *on my setup.* The timed downloads I ran confirmed this. When I get the iBook set up with 10.2.3 and the RAM upgrade I'm waiting for, I'll be interested to see if the results change.

Charles

From Marc X

Hello Charles,

The misleading point in Safari is its 'blue-ing" out of the address field while loading a page. Other browsers show progress much less visibly.

As in most browsers, the page is already readable and has loaded, while the browser is still loading stuff.

So Safari is still blue-ing, while the page has loaded to my satisfaction... Using a DSL 512 mbps connection, Safari loads in general fastest of all browsers (Mozilla, IE, iCab, Omniweb), and ties with Chimera, although even there generally quicker.

Chimera wins largely in loading big pictures though (applied SpeedChimera to Chimera for pipelining).

"Maybe dialup isn't it, then. I'm beginning to wonder whether the fact I'm running 10.2.1 rather than 10.2.3 may be a factor in the lackluster Safari performance I experienced."

Charles

I thinks it definitely is: even in Apple's software page it sais "for 10.2.2 users" (which sould probably say 10.2.3).

I recall reading somewhere it runs best in 10.2.3.

Thanks,
Marc

Hi Marc;

It's just that the URL field/progress bar in Safari is so prominent. It would be cool it it really did represent the actual progress of the page load proportionately.

I'm gradually getting my new iBook set up for production work, and will upgrade it to 10.2.3 (it shipped with 10.2.1). Will be interested to see how Safari performs when my RAM upgrade arrived (hopefully this week).

Charles

Browser Shoot Out

From Winsor Crosby

Seems like you might get more variation with a faster connection. But I could be wrong.

It seems like drawing the page is the bottle neck. I am amazed at how quickly a previous page redraws when I hit the back button on Safari. It is almost instant, unlike my experience with iCab. I cannot imagine that a tabbed browser would be any faster.

Winsor Crosby

Hi Winsor;

That snap bck feature is nice in Safari, but clicking on a tab to return to a page is pretty instantaneous too. I still think tabbed browsing rocks. Hope they add it to Safari.

Charles

Safari Update?

From John Dennis

From what I have read there is a new update for Safari that you may want to try out. I am not sure what the changes are, but it may be worth trying.

Hi John;

Sounds promising. I'll check it out soon.

Charles

Need help

From Bobby Pellegrini

Who can I contact on problem when downloading Safari thinking it would work in X1.5?

Hi Bobby;

I don't know. Perhaps our readers can help.

Charles

Bold

From: Eolake Stobblehouse

From your mailbag:

I am now using Safari. The text on your site is now all bold. I sent this bug to Apple.

I think I know why. I got the same thing on one of my pages!

In that case at least, it was caused by a <btag before a table. Other browsers do not take that to mean that text in the table should be bold, but Safari does.

Safari is gratifyingly faster on my widescreen iMac. Enough that I keep noticing it.

10.2.3.

Yours, Eolake

http://MacCreator.com
About the Mac and creative use of computers

Also writing about the culture and the Mac at: http://MacObserver.com

The OS X Odyssey archives may be accessed here:
http://www.applelinks.com/news/odyssey/

CM

Charles W. Moore

Cool Mac Gear