[asxless]

Originally posted by pobodysnerfect:
<STRONG>ok, I am using SP60 and I STILL cannot log into my bank of america online account.
&lt;snip&gt;
I understand this is still a sneakypeek but others have said they can log into their BOA accounts. I am not sure why I can't... </STRONG>

I am able to log into my Bank of Ammerica online account with SP 61. FWIWoOne annoying characteristic of BOA is that they do not parse spaces out of the account number which can lead to a login failure.

BTW I have always been able to log in to my BOA account with OW but... the buttons for updating the screen display (e.g. after selecting a new report period etc.) do not work correctly. They either do the wrong thing (e.g. the action of a different button) or don't do anthing at all.

asxless in iLand

Ok...here's a few problems that have never, ever been fixed since the first days of OW. I was hoping it would get fixed along the way without having to say a thing but it didn't.

1. I know this was mentioned in one of the OW threads...the 'ForumJump' menu on MacNN's and AppleInsider's main page (and many other pages) works the first time around...but then if you hit the 'Back' button from the forum to go back to the main page and try using the 'ForumJump' menu, it won't work.

2. ArsTechnica forums...(this could very well simply be bad HTML code so I won't get too angry if it doesn't get fixed)...tables in the ArsTechnica forum cut off the 'thread state' icon (folder icon, locked folder icon).
ArsTechnica Mac Achaia Forum
Like I said though...I think it's just terrible HTML coding and I haven't seen this problem on any other sites so I don't expect Omni to fix this if it's bad-HTML-coding related.

3. Hotmail is broken, broken, broken!!! I don't care if people tell me it's fine...it's not!!! Java is broken (clicking on checkboxes is slow and will on really frequent occasions throw OW into some kind of infinite loop? Spinning rainbow wheel and all...force-quit is the only fix.) Search field is overlapped by the 'Go' magnifying glass icon. The Home, Inbox, Compose etc. tabs rarely work. When in the 'Home' tab, there are some missing graphics at the top of the page.

Broken I tellz ya...don't make me believe the opposite.


That's all for now.

BTW, this is in sp3.

Bye.

[Rickster]

Now my concern is that they will be forced to move onto 5.x because it is too difficult (or too time comsuming) to fix the javascript issues in 4.1.

"JavaScript compatibility" is a very general blanket statement. There are some features of dynamic websites that depend on how the JavaScript runtime interfaces to the rendering engine's document object model, and some of those issues will be impossible to fix in the 4.x architecture. There are a lot of JavaScript-related issues that we can fix, though, which is a big part of why we still haven't finalized 4.1 yet. However, there will come a point at which we have to start accepting that some bugs will have to be put off for a 4.1.x or future 4.x or 5.x release, or we'll never get 4.1 out the door at all.

the 'ForumJump' menu on MacNN's and AppleInsider's main page (and many other pages) works the first time around...but then if you hit the 'Back' button from the forum to go back to the main page and try using the 'ForumJump' menu, it won't work.

This is an unfortunate effect of an architecture tradeoff we deliberately made; it has to do with how we achieve "instant" results from the Back/Forward buttons. If we executed onLoad scripts when going Back or Forward, we'd have to re-render the page and it'd be quite slow. This tradeoff should be no longer necessary when we move to the 5.0 architecture, but in the meantime you'll have to manually Reload some pages after going Back or Forward to them in order for certain JavaScript-based features to work.

[Guy Incognito]

Originally posted by Rickster:
<STRONG>

This is an unfortunate effect of an architecture tradeoff we deliberately made; it has to do with how we achieve "instant" results from the Back/Forward buttons. If we executed onLoad scripts when going Back or Forward, we'd have to re-render the page and it'd be quite slow. This tradeoff should be no longer necessary when we move to the 5.0 architecture, but in the meantime you'll have to manually Reload some pages after going Back or Forward to them in order for certain JavaScript-based features to work.</STRONG>

Good enough reason.

btw, why not release 4.1 now and start working on 5.0? I'm scared that all that work you guys are putting in 4.1 will eventually have to be rewritten anyways for 5.0...or am I wrong?

And before there's any wise-cracks about my previous post...the 'BTW, this is sp3' post...I meant 'sp63'.

[ 03-26-2002: Message edited by: Guy Incognito ]

[groundwork]

my biggest problems with OmniWeb have been related to the IngramMicro and FedEx web sites. Can't do anything useful (like order or ship) on either one of them. That's pretty much the only time I have to revert back to IE 5.1.

[asxless]

Originally posted by Rickster:
"JavaScript compatibility" is a very general blanket statement. There are some features of dynamic websites that depend on how the JavaScript runtime interfaces to the rendering engine's document object model, and some of those issues will be impossible to fix in the 4.x architecture. There are a lot of JavaScript-related issues that we can fix, though, which is a big part of why we still haven't finalized 4.1 yet. However, there will come a point at which we have to start accepting that some bugs will have to be put off for a 4.1.x or future 4.x or 5.x release, or we'll never get 4.1 out the door at all.

Rick -- Thanks for the clarification/confirmation about javascript issues and for recalibrating expectations for their resolution in 4.1 vs 4.1.x, 4.x or 5.x. FWIW I'm just hoping www.fidelity.com falls in the "we can fix" in 4.1 catagory because I like to go back to using OW as my main browser sooner rather than later

asxless in iLand

[KidRed]

OW still has a problem rendering the tables on ESPN- http://sports.espn.go.com/nba/clubhouse?team=mia

[oeyvind]

Rickster et al,

Have you guys able to look at the problem I stated (see page 18 in this thread), and also on this:

OmniWeb can't access this page at all, you see the page correctly in MSIE.

It's an e-commerce site for a Japanese company running FileMaker Pro.

Thanks

[dj247046]

OmniWeb sp63 is out now!

[cpac]

Originally posted by dj247046:
<STRONG>OmniWeb sp63 is out now!</STRONG>

no way!

(sorry -just that its been out for over a day)

[Since EBCDIC]

Rickster, et alia,

Thanks (again) for a fine product. It's a pleasure to use; moreso when one is forced back to MSIE for a particular web site. Thanks also for keeping us up-to-date with your reasoning and constraints: for me, as a programmer, it's easy to accept some limitiation when knowing the issues behind it (such as the JavaScript item recently mentioned). Thanks. Your company's open culture is an inspiration.

A feature request, of a sort: I tried to set my background to black and my foreground to white, to make late-night dark-room surfing a bit easier on the eyes. But these prefs don't seem to override table prefs, and since damn near everyone renders pages as tables, it didn't do the trick so well.

A comment: web pages with lots of *big* pictures - a rather rare condition - causes me to swear at OW. I think it's the all-in-RAM rendering; are you trying to keep all the JPEGs in RAM too? These pages make OW buckle to its knees. (This has happened to me for at least the sp5x and certainly to sp61. I can privately give you an URL should you wish to watch the horror unfold.)

A question: why does OW take so long to quit? Other apps quit more or less instantly, OW seems to hang for, oh say five seconds on my G3PB 400MHz.

Kudos: to your staff at MacWorld SF. I had the very brief pleasure of chatting with some of them about feature requests, bugs, etc. and I found them to be very much on the ball. As I've seen mentioned elsewhere, YGFR.

Thanks again. I am so pleased that I paid for OW just after it appeared. I've never been so pleased as how my money was spent :-)

[JKT]

Originally posted by Since EBCDIC:
<STRONG>A question: why does OW take so long to quit? Other apps quit more or less instantly, OW seems to hang for, oh say five seconds on my G3PB 400MHz.</STRONG>

At a rough guess, because it is clearing its cache? Any browser set to clear its cache at Quit takes many seconds to do so. However, I can't say that I've noticed it taking as long as 5secs but I hvean't paid much attention either...

[Gregory]

Speed of the system, hard drive, waiting for a drive to wake up, unlock the bookmarks backup, history file, update cookies etc, and whether you were using swap space or everything was in memory.

It also has to look in /tmp to clear out anything held there, remove items from download list... how many windows are open at the time.

SP 63 shows more tweaks, speed, bookmark features.


Java: sounds like the Sun JavaOne is providing Apple with a chance to highlight Mac OS X - not sure but I'll bet 10.2 will 'close' OmniWeb 4.1 and begin 5.0 testing; as well as enhancements to java. just as 10.1 killed 4.06.

And finally, Norton Systemworks 2 is shipping so native disk utilities (off topic but not really).

[sinebubble]

YES!

With sp63 I can now use check boxes in Netscape mail! Thank you, Omni! Now you can have my $29.95. As far as I'm concerned 4.1 is done. Stamp it, go have a beer and then start working on 5.0 tomorrow morning!

[Rickster]

No, we don't have a disk-based cache to clear when we quit. The most time-consuming task when OmniWeb shuts down is probably writing the HistoryIndex file. But that souldn't be taking long anymore... try deleting any HistoryIndex files that exist in your ~/Library/Application Support/OmniWeb folder; maybe one from an older sneakypeek is corrupted. On my iBook 500, OmniWeb takes a second or two to quit, but it's faster to disappear than iTunes, Mail, and iPhoto.

All-in-RAM rendering? As opposed to what? (Disk-based caches only make a difference in performance the second time you load a page, not the first. In any browser, rendering of pages/images tends not to involve the disk.) Since EBCDIC, could you send a URL for this page that's causing you trouble?

[Sharky K.]

check this site: http://www.cyberarmy.com/crash.shtml

(all browsers will crash, OmniWeb, Mozilla and IE on all platforms, so be careful.)

I downloaded the html source with Download Wizard so I could take a look at it and it is pretty nasty.

If you would like me to post the html source here, just ask me.

[unfaded]

Originally posted by Rickster:
<STRONG>On my iBook 500...</STRONG>

Rick uses the same computer as me! I feel like I've found a long lost brother...

or something :o)

[Don Pickett]

Got two unresolved issues with SP62:

1) After I have hidden OW it still takes cpu cycles, sometimes spiking to 50% before calming down after about fifteen to thirty seconds. Sometimes it never gets to 0, but hovers around 3%. OW is hidden and doing nothing.

2) It likes loads o' memory. This afternoon, with three windows open, it was taking ~400MB of swap space. Any way to cut this down?

[Amorph]

As of sp61, http://www.crfh.net/ now sports a great deal of whitespace that it never had before (and which doesn't appear in other browsers). Granted there's a great deal of ugly JavaScript code and the like, but it never munged the formatting in OW before, and it doesn't affect other Keenspot.com sites with the same JavaScript code.

[Sine]

SP63 still has the lame copy bug that cause it to freeze and not work. *sigh* I hate writing a paragraph post, go and try to copy a quote and have to force quit Omniweb and retype my post again. It's REAL lame.

[Gregory]

This site Chip Geek, uses "tabs" to switch between categories, but none of them work.
http://64.55.181.130/osgeek/mac/index.htm

[Gwenhiver]

Originally posted by Sharky K.:
<STRONG>check this site: http://www.cyberarmy.com/crash.shtml
(all browsers will crash, OmniWeb, Mozilla and IE on all platforms, so be careful.)</STRONG>

Huh ? This looks to me like some very stupid crash-test...
What's the point here ? Trying to see if a browser can display a form element with a size of 9.10^100 or so is just stupid.

The other tests are not very interesting either. Actually, some of them --I think the form test is of that kind-- don't even test the browser, but the ability of the OS underneath to deal with huge integers...

[Since EBCDIC]

No, we don't have a disk-based cache to clear when we quit

That's what I'd been led to understand. Good to have confirmation.

The most time-consuming task when OmniWeb shuts down is probably writing the HistoryIndex file. But that souldn't be taking long anymore... try deleting any HistoryIndex files that exist in your ~/Library/Application Support/OmniWeb folder; maybe one from an older sneakypeek is corrupted.

I did, and with sp63 it's still taking a bit of time. It's probably my slightly corrupted disk; Norton can't fix it and I haven't wiped it clean and started again. I've seen OW respond to disk integrity before...

All-in-RAM rendering? As opposed to what? (Disk-based caches only make a difference in performance the second time you load a page, not the first. In any browser, rendering of pages/images tends not to involve the disk.)

Because it seems to bog down on pages with lots of big images, and MSIE doesn't, I brilliantly concluded that it was your no-cache architecture to blame. I thought you downloaded images into RAM (and swap files :-) rather than into the cache and this was causing the problem. I'm not sure that I'm doing a good job of describing my intuition, but it's moot because...

Since EBCDIC, could you send a URL for this page that's causing you trouble?

It is done. Thanks for looking at it.

[Sharky K.]

Originally posted by Gwenhiver:
<STRONG>

Huh ? This looks to me like some very stupid crash-test...
What's the point here ? Trying to see if a browser can display a form element with a size of 9.10^100 or so is just stupid.

The other tests are not very interesting either. Actually, some of them --I think the form test is of that kind-- don't even test the browser, but the ability of the OS underneath to deal with huge integers...</STRONG>

OmniWeb should have some max limits I think.

[Rickster]

Because it seems to bog down on pages with lots of big images, and MSIE doesn't, I brilliantly concluded that it was your no-cache architecture to blame. I thought you downloaded images into RAM (and swap files :-) rather than into the cache and this was causing the problem.

Ah, I see. Well, I'll look at the URL you sent, but I can assure you that our lack of a disk-based cache would have nothing to do with this problem. All browsers do "RAM-based rendering"; in order for them to decode an image and render it to the screen, it has to be in RAM. What a browser's cache does is make it so that the second time you view a page or image, you don't have to download it from the network again. The difference in browsers with disk-based caches is that you can wait longer before viewing the page/image again -- even quit and relaunch the browser -- and still not have to fetch it from the network. If you're on a fast broadband connection and tend to leave your web browser running, there's little performance difference.

As for that deliberate-crasher page.... that looks like a good thing to put in our internal test suite.

[starfleetX]

Originally posted by Rickster:
<STRONG>As for that deliberate-crasher page.... that looks like a good thing to put in our internal test suite. </STRONG>

Yeah, I have had that CyberArmy page bookmarked for quite some time (I'm a proud member ) and at one time it loaded fine in OmniWeb. Though, I think they continuously add to and change the code to adapt to browser fixes.

[Sharky K.]

I still think it is wrong that OmniWeb creates weblocks instead of .url files...

OmniWeb has even a nice icon for the .url files and .url still works when you delete the resource (compressing it for example, .tgz or .zip)

[cpac]

Originally posted by Rickster:
<STRONG>

This is an unfortunate effect of an architecture tradeoff we deliberately made; it has to do with how we achieve "instant" results from the Back/Forward buttons. If we executed onLoad scripts when going Back or Forward, we'd have to re-render the page and it'd be quite slow. This tradeoff should be no longer necessary when we move to the 5.0 architecture, but in the meantime you'll have to manually Reload some pages after going Back or Forward to them in order for certain JavaScript-based features to work.</STRONG>

Rickster - is this what causes dynamic forms (like Adobe's product registration or apartmentguide.com) to fail in Omniweb? In those cases, choosing an item from one pop-up menu changes the content of one of the other menus, but after using the first one, the form doesn't work anymore. (e.g. selecting "photoshop" in the product pop-up in adobe registration, changes the contents of the "version" pop-up menu, but one can do nothing further with the page)

Might there be a work around for 4.x where we could allow the scripts to reload the page so forms like these will work, or are we stuck until 5.0?

OW rocks for most everything else I use it on.

cpac

[Mr. Blur]

...sp64 is up now....

[Sharky K.]

Omni should look in the new ad sizes like: 600 x 90 and these: http://www.iab.net/iab_banner_standa...nnersizes.html

[Diggory Laycock]

I still think it is wrong that OmniWeb creates weblocks instead of .url files...

I have to disagree with this strongly. Omniweb is a well written App that follows both the Aqua HI Guidelines and is a good mac citizen. This means that it uses APIs correctly and does what you would expect. Apple built URL handling into the OS for a reason. Internet Explorer's bastardised URLs are a perfect example of bad practice - The OS provides you with a standard behaviour, but you decide to implement your own - thus introducing inconsistency in the OS.

Following the proper methods means that your app not only works well in itself, but also plays well with other well written apps and most importantly works well with the OS.

Would you be happy if an image app generated gif images that could only be viewed by itself?

[ImpishLM]

2 completely unrelated issues:

1) I've discovered another page to test OW's Java capabilities.
www.eyeplaygames.com hosts a multi-player version of the board game RoboRally, called RoboRunner. Registration went smoothly, and the Java game board, chat lines, and game control panel are drawn with no problems. Chatting with other players also works.

Players have to order "chips" left to right for each turn. In MSIE, one just drags the chips around, and when ready, clicks a button to submit the "program."

In OW sp 63, however, clicking on a chip makes it appear all by itself in the center of a blank screen.

Too bad, as the game board is so much prettier in OW than it is in IE, but then, that's always been true, and on any site!

2) Has anyone mentioned to sign up for Billpoint/eBay payments using OW, including getting your checking account verified? Billpoint won't verify my account, but won't tell me why, and I'm trying to track down the root cause. Thx!

[Sharky K.]

Originally posted by Diggory Laycock:
<STRONG>

I have to disagree with this strongly. Omniweb is a well written App that follows both the Aqua HI Guidelines and is a good mac citizen. This means that it uses APIs correctly and does what you would expect. Apple built URL handling into the OS for a reason. Internet Explorer's bastardised URLs are a perfect example of bad practice - The OS provides you with a standard behaviour, but you decide to implement your own - thus introducing inconsistency in the OS.

Following the proper methods means that your app not only works well in itself, but also plays well with other well written apps and most importantly works well with the OS.

Would you be happy if an image app generated gif images that could only be viewed by itself?</STRONG>

I disagree with your example...
the .url file can be opened by all applications that need it. I also think weblock is an old thing from MacOS Classic and that is the only reason why it is still here.

.weblock is very wrong now with MacOS X. Apple wants to get rid of resources as much as possible. But .weblock depends on resources. So this is also not possible on other than HFS(+) hard disks or Unix's own compressions.

[starfleetX]

Originally posted by Sharky K.:
<STRONG>Omni should look in the new ad sizes like: 600 x 90 and these: http://www.iab.net/iab_banner_standa...nnersizes.html</STRONG>

Even better would be an addition to the Privacy preference pane to enter custom sizes like iCab and Mozilla's "Banner Blind" project do. I was actually surprised when I started using OmniWeb that this wasn't available.

[Sharky K.]

I notice that the latest SP version of OmniWeb crashes a lot at versiontracker but does not start the crashcatcher...

[davidb224]

Originally posted by Sharky K.:
<STRONG>I notice that the latest SP version of OmniWeb crashes a lot at versiontracker but does not start the crashcatcher...</STRONG>

I go to Versiontracker nearly every day and use my OW4.1 SP's to download stuff and its never crashed on me there.

[starfleetX]

Originally posted by Sharky K.:
<STRONG>I notice that the latest SP version of OmniWeb crashes a lot at versiontracker but does not start the crashcatcher...</STRONG>

For the last 5 to 10 sp's, OmniWeb seems to locks up or crash a lot without launching the crashcatcher.

Plus, Omni still doesn't follow anchors like on these forums. I'd gotten so used to accommodating this behavior that I'd forgotten how well the other browsers worked.

has anyone mentioned these?

1) www.playsite.com
java board games for the most part. i usually play scrabble or chess. in scrabble, when i get to the games window, i can only start a game and hope other players join because i'm unable to see the "games" window.

2) www.nba.com
watch game updates instantly with "courside live" by clicking on "java" updates. basically it opens up a java window that is updated constantly. well, after being open (this java window) for about 3 minutes, OW crashes and logs me out of the system.

i dont know if these problems exist with other browsers because i dont use other browsers...i hope these are fixed at some point before 4.1 is released.

[Rickster]

I still think it is wrong that OmniWeb creates weblocks [sic] instead of .url files...

It's not really something we can avoid, regardless of whether one thinks it's better or worse than .url files. The drag-and-drop mechanism for Cocoa applications doesn't allow developers as much control over what file gets created in a drag to the Finder as the Carbon mechanism does -- we were able to "hack" it for image-dragging, but that's not without its problems (note that dragging the same image twice results in an error from Finder). I don't believe we'd be able to do that for URLs anyway, because we still need to be putting URL types on the pasteboard for dragging to other apps that accept URLs.

Ironically, earlier releases of OS X (Public Beta) had a variation on the format for those "internet location" files that used XML in the data fork and was thus safe across filesystems, but it went away in 10.0. Whazzupwitdat?

Rickster - is this what causes dynamic forms (like Adobe's product registration or apartmentguide.com) to fail in Omniweb?

I doubt it. The problem I described was specific to cases in which you try to use a form or other JavaScripted feature of a page after visiting the page once, then leaving and going Back or Forward to see it again.

[Sharky K.]

is it possible to create .url files with an other app or an other way with OmniWeb? (maybe in the future)

[neutrino23]

Originally posted by starfleetX:
<STRONG>For the last 5 to 10 sp's, OmniWeb seems to locks up or crash a lot without launching the crashcatcher.

Plus, Omni still doesn't follow anchors like on these forums. I'd gotten so used to accommodating this behavior that I'd forgotten how well the other browsers worked. </STRONG>

FYI, I haven't seen this at all. I've seen just one crash in the last couple of months with OW. It is curious why different computers running the same OS and same application behave differently.

[Diggory Laycock]

These days I find that most crashes (~80%) are not caught by the crash catcher, but are in fact some kind of crash where you just get the beach ball.

Is that this what is called a race?

The only times the crash catcher appears for me is when flash or Java is to blame.

[BZ]

For the person who keep talking about Bank of America online banking. Try changing your settings. I have been able to get into it for months. Of course, it seems much slower than in IE (SSL?) but it works.


BZ

[ImpishLM]

About changing OW's settings: the Compatability prefs are one the most important to play with.

For example, bidding on eBay doesn't work if I've set the compatibility prefs to emulate IE 5.1, but it does work if I set it to OmniWeb. In other words, set so that it doesn't try to make the eBay site believe you're using MSIE.

Other sites seem to require that I set the compatibility to IE 5.1, so I frequently accessing the that pref pane.

In short: the compatibility prefs can have a huge impact, and when a site doesn't work with one setting, try another. It's best to change settings BEFORE you go to a given site, so that OW has a chance to introduce itself has one browser or another.

Cheers!

[Gregory]

Sounds like it would be useful to have a site specific setting for what browser to present yourself as, or to 'emulate' rather than one setting for all sites... Would be nice if it could be done.

I'm not in favor of helping the "statistics" on what browser people use, when it isn't even accurate when users are sort of forced to play this pretend game.

[jasong]

I thought of a new toolbar option I would like to see. When it says Show Icon and text, Icon Only, and Text Only, I would like to have the text appear on the side of the button rather than underneath (IE 5 for Windows allows this). I have a 19 inch monitor that I run at 1280 x 1024 and higher, so I have more room on the toolbar that I have useful buttons, but I can always use more room top to bottom.

Just a thought.

-- Jason

[Rickster]

I thought of a new toolbar option I would like to see.

Yeah, that'd be nice. However, our toolbar is a standard Apple-supplied widget an we can't really make those kind of improvements without totally hacking it to pieces. I'd suggest sending your idea to http://www.apple.com/macox/feedback/ .

[cpac]

The last couple posts got me thinking - any chance of a toolbar pop-up menu from which you could set compatibility preferences? This might be easier to do than site-specific preferences and could make the process much faster for those who have to change frequently . . .

just a thought.

cpac

[JLL]

It seems that Omni Group have been working on fixing JavaScript so the sites that some of you visit work correctly, but only on one of the sites I visit (that have JavaScript menus) work - and that has always worked.

My main problem is e-boks where the menus don't drop down and thus making it impossible to use the site with OmniWeb.