Why Apple is betting on HTML 5: a web history - Page 4

Now you're just being silly.

I don't dispute that at all. The point was that adding Flash-embedding to PDF seemed like a stupid addition with very little benefit, but opened up vulnerabilities that otherwise would be completely moot.

One approach of Apple's that I truly admire is their streamlining of features and wisdom of cutting out little-used legacy features over time. Certainly it inconveniences a small segment of users and can be overdone, but paring down OS X not only keeps it leaner but also easier to maintain. In contrast, Adobe and Microsoft simply add new features (many unwanted by users but desired by marketers) and support legacy features forever (I think Win7 still uses a picture of a floppy on buttons for saving to 'disk', which pretty much sums up their clinging to the past). It's like they never met a feature they couldn't wedge into the toolbar somewhere. Sure, their feature lists look impressive, but at the expense of user simplicity and code maintainability. For Adobe to add Flash capabilities to PDF is just such an example and guess what - it came back and bit them in terms of security. Was it exaggerated? Perhaps, but why was it added in the first place?

And a little more on-topic, the feature set of Flash itself keeps growing and growing, while people wonder why it hogs up so many resources...

There's nothing wrong with adding features. No one would ever upgrade their software if features weren't added, and then companies would go out of business, because upgrades are a bigger income stream than new programs.

Shortly after Tiger came out well ahead of schedule, and proved to be buggier that many expected, I raised the issue of companies not releasing software until all the known bugs were fixed, and the software was as reliable as it could be. This is what's done in the mainframe and minicomputer markets, so it can be done. That rose a lot of irate responses. People want their upgrades, bugs or not. they then want to grouse about them.

Sometimes those bugs are security related. Bugs are bugs. A company doesn't think to fix some bugs over others if they DON't know why a certain bug is a bigger problem. And then others aren't found until later.

Apple has plenty of bugs in its stuff. Possibly not as many as MS who admitted that either XP or Vista, I don't remember now, had 68,000 bugs when released. Or look at this, to make your heart stop:

http://www.tomsguide.com/us/Blizzard...news-4685.html

Of course, there's x264 . . .

You seem to be forgetting something...



followed by this...

Sales for Snow Leopard Far Exceed Prior Launches

Of course the "0 New Features" is overly simplistic, but Apple's message is virtually unprecedented in the software industry. I personally like John Siracusa's take on possible Mac OS development:

"As for the future, it's tempting to view Snow Leopard as the "tick" in a new Intel-style "tick-tock" release strategy for Mac OS X: radical new features in version 10.7 followed by more Snow-Leopard-style refinements in 10.8, and so on, alternating between "feature" and "refinement" releases. Apple has not even hinted that they're considering this type of plan, but I think there's a lot to recommend it."

WHile you may consider it a stretch, Gore did start a bill that ended up funding NCSA where the graphical browser (NCSA Mosaic) was developed (1993). Gore did work on bills that ended up funding a lot of the connectivity/computing that was necessary for growing the "internet".

Ross Perot - have no idea of his connection.

The documents for the actual proposals for http /html were indeed developed by Tim Berners-Lee at Cern (Physics) lab in Switzerland. The initial purpose of the world wide web was to allow researchers / scientists to transfer information across the globe in a fast and timely manner. Collaboration was the desire.

My first browser was text only and was called "Lynx". Since it did text only it was super fast even on a 14400 baud modem.

Co-sponsorship is not at all the same thing, nor as notable, as what Gore provided to the efforts. To earn a co-sponsor notation, one simply can agree to lend their name on the legislation, even if at the hour before publishing. I state this for those unaware. In the spirit by which I believe you posted that tidbit, I think you are sufficiently aware, but just needling. If not, I'd be quite agreeable to provide you with the aforementioned's legislative records that should make the distinction quite clearly.

Someone brought this up in relation to Al Gore being mentioned.


http://en.wikipedia.org/wiki/Ross_Perot

That is true to a large extent! But, maimframes and minis had a lot fewer: installations; 3rd-party software; OS versions/updates; and applications. This made the job quite a bit more manageable.

I worked for IBM 1963-1980 in the Data Processing Division (maimframes). The most successful application package ever (circa 1978) was a package called CICS (Customer Information and Control System). At its height, CICS/DOS had about 3,300 installations at about $3,000 month rental for the software only (you do the math). IBM was then, about a $9Bn corporation.

As an aside, CICS had its share of bugs (many known)-- to the point that the internal name was "Consistently Ignore Customer Satisfaction".

IBM's premiere database system in those days was a system called IMS-DL/I (Information Management System Data Language I). It was an Hierarchical DBMS -- Relational DBMS were still in the IBM San Jose Lab. The DL/I component (the database) could also be used with the aforementioned CICS. DL/I had been written [mostly] in the 1960s, by IBM System Engineers (non-programmers). It was the poster child for "Spagetti Code". There was one module in particular, the "Retrieve" module that was a nightmare. Retrieve was used by all the other modules and performed: I/O; buffering; resolution of relationships and intersection data; buffer flush...

Retrieve was so delicate that everything they did to it (bug fixes, maintenance, rewrite) introduced more new problems than were being addressed. It eventually got to the point that they left Retrieve alone and resolved bugs by trapping them and implementing the fixes in separate add-on modules. The system really broke when used for on-line multitasking.

There was one major IBM customer who managed their entire business with CICS-DL/I. They were so gun-shy that they implemented a shadow database for online processing (the changes were logged and physical updates were deferred to overnight [single task] batch processing. On-line processing would read yesterday's [batch] db, read today's logs and combine them on a per-SKU basis.

I think that DL/I eventually died of its own weight!

So, the maimframe business wasn't all that pure.

*

So did Canon, but at this point it's getting somewhat tangential to Apple's involvement with HTML5...

(Even more tangential - I still have one of the original BusinessLand/NeXT T-shirts in one of my drawers somewhere.)

I'm not forgetting it. Have you seen the price of the upgrade?

And what about all the articles on all the PC and Mac sites out there explaining all of the not new new features?

You're also forgetting that the almost total rewrite of the OS which includes such "features" as Open CL, Grand Central, Exchange support etc. is itself a very big feature.

...lazy Sunday, feeling argumentative-- what's the harm.

I have a Lisa button, an Apple I manual, and an IBM PC/jr "Chicklet" kb.

Seriously, I used to live in the valley, and only the coolest people had some of the more obscure T-Shirts. BL/Next makes you mondo-cool!

You're not Dave Norman posting under an alias, are you?

*

To remove "all known bugs" before release is, sadly, quite unmanageable. The fact is that bugs crop up anywhere and everywhere, and if they aren't "major issues" then it may be more helpful to just release the app, and send a patch later.

Killing all known bugs is a pretty HARD job to live up to - every time you get rid of one, another 2 pop up. Software developers almost always know about one or two minor bugs in their code, and do their best to get rid of them, but should they stop a release from going live? Unless they are very serious bugs, it may be unmanageable and financially not viable to hold back a release for 5 nasty but minor bugs, for example.

Perhaps we should take a less black-and-white approach, and say "hold back a release if there are a considerable amount of known bugs, and give some time to finding them.

Generally, as a software developer, I work on fixing bugs via several methods:

1. I build the app and try to fix any bugs as I go.

2. I'll 'beta' my software till my bug count is as low (it will rarely get to zero but I'll live with that)

3. I'll use the dog-food-principle. I'll try it myself. Run the app for a week as I plan it to be used. If I find any bugs that stand out in any general usage (be it even just an "annoying but unimportant bug") it will stop release. The point here is simple: bug "seriousness" is invariably tied to how often you run into it. If you run into it in everyday use, even if its not a "big technical bug" then its important to fix it before release. No options.

But in a system as large as Mac OS X, Mel, you need to be realistic. I agree that Mac OS X should have a "dog food period" - where Apple's developers use their own creations for a week, purely to find the "in your face" bugs. I also agree Apple should delay their releases by a month simply to get the software "really polished". But "no bugs" - its a really tough ask. Any thoughts?

I've been using flash for quite a while now (When it used to be called Futuresplash). I was most interested in cel/vector based animation. Flash 3, 4, 5, MX, had its progressive feature set.

Around 2000, I was astounded at sites that used flash to the fullest. It was clean, and experimental. Websites looked amazing and dynamic, audio and everything. From 2advanced.com to Happy Tree Friends. It made me move from HTML to Flash. Just a note, Flash ran fine on my old 166Mhz.

Around christmas of 2005, Adobe had acquired Macromedia. Since then everything went downhill.
Flash player 7 was the best version out there for legacy computers. Flash player 9 and 10 is no longer exciting as it should be even if they introduced hardware acceleration and 3D like papervision(more on this in a sec). HTML, CSS, AJAX, JQuery is here and replaces Flash for its dynamic content and even animation.

As of now I don't use Flash for websites. One is its rendering performance and I hated Adobe for doing this. I had clients who said the images were laggy(I can't blame his computer.), while it works fine in some, so I get the blamed as lousy developer! Secondly, I rarely see websites that use Flash fully like back in 2000, so I just followed the standards. (HTML, CSS, etc). I only use Flash as vector based program and animations. OK I give Adobe a thumbs up on this one esp. when they added the bones feature.

Side story:
A friend of mine has worked and learned parallel with me ever since I discovered Flash years ago. Today he has interest on Flash game development and I "had" gained some interests as well. Actually he is more focused on the Papervision 3D project. We developed some samples using the api. I had bought a mobile phone when I was on vacation 6 months ago which is yet to be released to North America. It featured Flash Lite. I was very interested on making apps for it which I managed so in Flash. When we got hooked on papervision, I thought that it might not work on the mobile phone because its already taking 80% of the CPU. So I tried running it on the mobile phone, and of course it won't work it needs Flash 9. Even if my phone had 3D capabilities, I'm still no better. So, I sold my Sony Ericsson and bought the iPhone 3GS. I also ditched the Papervision3D project which I confessed to my friend that OpenGL is better.

Flash is still here as of today because of video playback. (And also because of flash based games but that will change). But right until HTML 5 and the introduction of WebGL, is like a polar shift waiting to happen. Predominantly, I have interest in WebGL. No longer I have to tell users to download O3D plugins which they refuse to install. I can have one site that renders my content on many platforms including iPhone (which already has canvas support and CSS 3D which works well). Goodbye Flash. Hello WebGL!

I point this out, because as I said, it's done for mainframe and mini systems. It can be done.

In a couple of articles written a few years ago, it was pointed out that personal computer users could have it one of two ways, get releases every five or more years apart that are well tested and have only very important new features, and essentially bug free, the way it's done on big systems, or get them much more frequently, with hundreds of important, and non important features, the way it's done for personal computers, and get inundated with bugs, the way it is for us.

PC, Mac, and Linux users want the latter, so that's what we get.

I try to aim somewhere in the middle. I'd say that's where Apple tries to aim with Mac OS X at the current stage. In the past, it seems Apple rushed out a lot of developer features, and justified them with a lot of user changes too, to try and "tempt" users to upgrade. Apple now has a far more mature system, and is moving further towards the "slow but important updates" from the "quick flashy" ones. But this is early days. We'll see where Apple is going with it, I suppose.

As for me? I aim for apps to be as free of bugs as possible. Its just at some point, holding back a release due to a few minor but aggravatingly stubborn bugs starts to become a complete waste of time, money, and resources.

No, just an old NeXT user who was so disgusted with Apple in the 90s (after working for them as a contractor during the tail end of the Sculley era) with their mismanagement of Mac OS that I actually considered giving up on computers altogether. At the time, using NeXTSTEP, which was essentially what OS X is now, alongside System 7 (and Windows 3.1!) made me wonder why the rest of the world simply didn't get it. Obviously there was a lot wrong with NeXT, Steve Jobs, and the state of 'commodity computing' at the time that precluded NeXT's success, but imagine driving a 2001 BMW when everyone else was still in 1985 Volkswagons and Pintos and you'll have some idea how frustrating it was to see the platform fail in the marketplace. Fortunately, Steve Jobs gave a more compelling pitch to Gil Amelio than Jean-Louis Gassée at Apple's critical juncture, swallowed the company from the inside, and now I can wax nostalgic over my leftover NeXT schwag whenever someone brings up Ross Perot.

(I still have a couple of NeXT machines in storage, but, alas, no free space to set them up. I wonder if the motherboard batteries still work...)

Well, that's why I wrote, "Of course the "0 New Features" is overly simplistic...", but Apple was the one who made the slide, not me.

I'm hoping that this rewrite, will give us less bugs than before because of simplification, bringing code up to date, etc. but the fact that we've already had problems so close to release makes me wonder if we'll gain as much advantage in that area as we (and Apple) would like. Hopefully after they squash some leftover problems, we'll see fewer bug fix releases with this upgrade. Only time will tell.

I had been buying Macs for my own company since 1988, but didn't begin to use them at home for my own work until 1992.

Right before I did buy my own Mac, a 950, I went to the NEXT/Mac dealership on 23rd st between 5th and 6th aves here in Manhattan NYC.

It was set up well. On one side of the dark room were Macs, and on the other, NEXT machines. All machines were in booths.

The NEXT machines were set up as an 8 bit Greyscale machine, then a 16 bit color model, and then the 24 bit color top of the line.

The greyscale was $9,000, the 16 bit $16,000, and the 24 bit $24,000. I'm sure those last two prices were coincidental.

I looked at each, starting with the greyscale.

Afterwards, I asked what the upgrade policies were to go from one to the other. I was told that you got what you bought, and couldn't upgrade.

This was one reason they failed first with their hardware. I could afford the $9,000. I could even, manage the $16,000, but the last was too much. If I could have upgraded later, I likely would have bought a NEXT instead of my first Mac, which ended up costing me about $16,000 when all was said and done.

Computers were expensive in those days!

Wow, you can't buy a phone from another vendor?

EDIT: Never mind - Solipsism beat me to it...

Yes! And the removal of legacy support of PPC is good (though I still have a few PPC Macs).

Because of its smaller install base [of legacy systems] and continued support for multiple OS X versions, Apple has the luxury to move forward without dragging baggage along. That's smart strategic positioning, that Microsoft just cannot match!

That said, Snow Leopard does have its issues. I have it installed on 5 machines. While mostly good, there are flaws. Most seem to be app-related (especially iMovie 09). The pro apps appear OK. The F-Finder sometimes hangs and dialogs are not always displayed topmost.

As the BeachBall spins: Safari still hangs, Flash still hogs the CPU(s), apps still abort...

... but happily, SNL is a bold move forward and we shall reap the rewards for years to come!

*

Things are snappier. But my third party screensavers don't work, and I've had to upgrade all of my utilities.

I had problems with two printers, which I fixed.

So far, that's to be expected. But otherwise, things seem to be moving along pretty well.

I haven't had a single Safari crash from Flash. but I have problems with a couple of other utilities that haven't been upgraded yet.

And Norton isn't updated yet. I know how the kidding goes on about them, but I've been using it, and it's successors for many years.

Well, what software would you have run on the NeXT? There were a few apps that were, to quote Jobs, 'Best Of Breed', like the Postscript-native Illustrator and Lotus Improv--which had features that even Numbers still lacks--but the big killers, as always, were the lack of Photoshop and MS Office. The latter was more of a checklist item for business types, but not having Photoshop meant that all that advanced (and expensive) hardware never saw its full potential.

But the underlying OS... ... .. .

I truly believe that if the breaks went the other way, where NeXT became self-sustaining early on, and Jobs learned to control his impulses, we'd be 15 years more advanced in software development, and Microsoft would just be another third-party vendor making office apps and wind-up Ballmer action figures.

You know, I honestly don't remember anymore. Illustrator, yes, but I don't remember too much about what was available. Publishing programs were available, and graphics and publishing was mostly what I needed it for, though PS would have been nice, I had it at my company.

There were some custom apps. I sold my house in Saratoga, in 1989, through Alain Pinal Realtors. They were NeXT-only in the office.

I agree that NextStep and WebObjects could have technologically set the bar... but they were too far ahead of their time, and took too much of an investment to become the lingua franca of their day.

Oddly, the iPhone SDK is plowing that same ground, with great success, more than a decade later.

And, IMO, this will bring the developers to the Mac... what goes around...

*

I really do not get the big deal with Flash. I have been using it since MX 2004, but I see it declining steadily. It is a processor hog and terrible for SEO. I try to steer my clients away from using Flash and instead towards QT and Javascript when needed. If you can not play it on the iPhone, what's the sense. Not every company is Nike. Adobe seems to be going in too many directions with it, it's just getting way over complicated and confusing. I wish Macromedia didn't sell out. Adobe is really screwing things up.

I am looking forward to HTML 5. Hopefully, while Balmer is not looking, the smarter engineers at Microsoft will slip in a modern web browser with the latest OS.

Even more tangental, I have one of the original MacUser N E V R t-shirts around here somewhere

It had FrameMaker (albeit not as powerful as the Solaris version); a number of really cool but amateurish Stone Design apps for graphics; Pages, which was the first true WYSIWYG page-layout app (the Apple version is similar in recycled-name only); a number of really good diagraming and utility apps from LightHouse; and the still-surviving Omni Group apps. Every one of them was written in Objective-C and had the full power of UNIX, Display Postscript, TCP/IP (as opposed to Apple Talk or Novell's IPX), and other forward-carrying technologies to build upon. All initially implemented back in 1988!

Unfortunately, the initial enthusiasm and third-party investment waned when it couldn't compete financially with Apple, Microsoft, Sun, HP, IBM, and all the other established tech companies of the time. So NeXT retrenched over time, first shedding the hardware, then opening up to other platforms in limited enterprise-only deployments, but hung on long enough to see it's Chess.app eventually become one of the first 64-bit applications available on OS X. How's that for an accomplishment?

Holy crap, Linux and OS X users can't stomach Flash.

The fact that it's just a container for the avi, mp4 or other codec means nothing to Hulu and other content distributors.

Moving to an AJAX UI on the client to manage the VIDEO/AUDIO HTML 5 tags is not a problem for them to switch.

Well, I've been a regular reader of this site for awhile and I thought the irrelevant political interjections were an unnecessary distraction too. Especially the other recent post on using text to speech programs. It was interesting until it diverted off into a diatribe on health care politics that really had nothing to do with the technology being described.
He does make a fair point.

So what, a minority player in the Mobile devices world supports (cut-down, not proper) Flash. All mobile devices have Flash LITE support, and it isn't really that great anyway.

I’m not quite getting the significance of the slide you posted. That was all about showmanship. To get your attention with something unexpected. P. T. Barnum would have approved.

Note that after they stated “No new features” they went in and introduced new features, including an a new enduser accessible feature in Mail for Exchange support.

MOV, AVI, MKV, WMA/WMV are all containers for codecs that may or may not be open.

Quicktime/Windows Media Player/VLC are all players of which support playback of multiple containers supporting numerous codecs.



Of course, because Apple's proprietary DRM scheme (FairPlay) is only implemented by the Quicktime container.

My post itself was intended as an attention-getter. The context of the discussion was I lamented that Adobe had added too many features to Acrobat (such as the inclusion of being able to embed/read Flash within a PDF, which turned out to open up a [rare] vulnerability), to which Mel responded that adding features was a natural course of software development, and such inevitable bugs will eventually get fixed. His post included, "No one would ever upgrade their software if features weren't added, and then companies would go out of business because upgrades are a bigger income stream than new programs."

The slide was a smart-ass way to say that Apple itself was pushing the concept of releasing a software upgrade (namely 10.6) that ostensibly contained no new features, where refinement was the emphasis of the effort. We all know that Snow Leopard contains many new features, but unlike previous releases, these new additions were downplayed (at least to the developer audience) so that everyone could instead concentrate on tightening up their compliance with the latest sets of APIs and underlying technologies.

Years ago (well, decades, actually) Volkswagon made the same appeal with the Beetle, comparing its (mostly) unchanging appearance with the superficial styling changes of american cars of the time. They were saying, in essence, "We don't rely on trivial marketing changes to sell our product - it stands on its own engineering merits." It's a tricky message to pull off, because it depends on your own confidence, and thus the customer's as well, that the product has inherent value independent of the distracting marketing battles that pit one feature set against another. I personally think Adobe has lost sight of that message with their current business directions.

One more example.

That's frightening! I could see making web analytics available as ColdFusion system functions so web developers could easily incorporate them into web sites. Sadly, I don't think that's Adobe's goal as they seem to have abandoned ColdFusion.

Rather, it appears they are going to incorporate analytics into Flash (likely as another ugly layer of Flex). Ugh!

On second thought, they could use it to monitor when Flash redlines the CPU or to measure [Flash-induced] browser hang time and crashes. We can look forward to the Flash three Rs being replaced with the Flash four Rs:

Rich, Reach, Retch and Report.

*

No, Quicktime is one way to play H.264 video, it's not the only way. H.264 is a standard, companies are free to implement that standard any way they want to. Apple uses Quicktime, Microsoft uses something else. When Microsoft comes out with a new browser/os/platform, they don't need Apple to software to play H.264 video for their platform because there is no dependency or requirement for a quicktime runtime to play H.264 within a HML5 browser.

Take that same video and wrap it with flash for no other reason beyond the limitation of HTML and suddenly Microsoft, Apple and anyone else is dependent on a 3rd party to write code that allows them to play that video.

"A growing contingent of increasingly important players on the web, led by Apple, Mozilla and Opera, began looking at how web standards could be advanced to support open, vendor-independent, standards-based alternatives to the then existing landscape of the web, which was increasingly becoming dominated by proprietary Flash content that relied upon a separate plugin to render it, effectively turning the open web back into a proprietary system like AOL had been in the late 80s: closed, buggy and slow."

This is somewhat insulting & extremely inaccurate.

Statements like that make me question the author's accuracy across the entire article.

In fact, AOL's proprietary was extremely fast & in GM (gold master released software versions) very stable. Many people use AOL 5, 6 & 7 today without a hitch.

While truly closed and while presenting other issues for usability; this criticism has no merit, whatsoever.

Thanks guys for a most enjoyable thread - we've had it all; an excellent history lesson for Prince Mclean (I always like these articles, Al Gore invented the Internet!) - anti-Apple trolling and then the heavyweight contributions from Solipsism and Melgross - always good value.

Seriously its been great.

I hate Flash with a passion - in the Advertising industry in which I occasionally swim it had become an excuse to charge clients ridiculous fees for months of development for websites of dubious value and appalling useability. For some reason Flash based sites are de-rigeur in the luxury sector. I make a point of strongly advising against whenever possible.

Clearly HTML5 will win the day - it has strong backing and wins the 'moral' argument (in tech terms) hands down. Adobe's management have been moving in strange, and multiple, directions over the past few years.