Yeah! Thanks for the attr support, Dave! Now my site looks better.

is the new WebCore available somewhere? i looked at http://developer.apple.com/darwin/projects/webcore/index.html but WebCore is still at version -60.

Dotted borders! YES! I can't wait for the next build.

Thanks, David! Dotted borders... the smallest thing makes my small mind happy.

v62 has TABS!!! And that's not all...it also has AUTOCOMPLETE FORMS!!! Yes, yes, yes. They are only reachable through the Debug menu however as of now.

Great work Dave!

Go Hyatt! Go Hayatt! Go Hyatt!

I'd be careful posting stuff like that Mary - if you're just yanking our collective chains, you're likely to have a large mob at your doorstep, and not in good humor.

Considering that there have been no betas since v60, and Dave certainly hasn't mentioned anything, I'm not holding my breath. ;)

Anyone know about the status of OVERFLOW: Auto / Scroll on a Div?

I've just hit the first fatal flaw that has made me switch to another browser. Seems that some sites don't get their cookie data sent back to them. The particular one that is killing me is http://www.candlepowerforums.com/ it's not possible to log in. If I look at the cookies list, I see that they are indeed set, but the site doesn't see them. I'm back to OmniWeb for that site. There is a discussion going on there from other Safari users with the same problem. Some of them have even had to go back to IE.

I have made a bug report through the official bug button, just wanted to see if other people were having similar problems with other sites too.

Thanks,
James

Any thoughts on support for either Mozilla's or IE's contenteditable?

IMO that probably isn't going to happen for some time. In order to support contenteditable, you have to build an entire HTML editor. Mozilla has this. KHTML does not. You'd need a new team of people working on building an editor.

Oh oh oh! A bug I submitted being fixed! Cool. I hope that the new version properly puts the border width even if the border-type is set to none. I guess I'll see at the next release. Thanks!

Hello,

I see some people who seems to have access to non public Betas of Safari (like the v62).
Is there any way to become a beta tester of the non public betas of Safari?

Thanks in advance for the answer,
Regards

Um, Frodon, I'd place my bets that Mary is a troll. I'd love to be proved wrong in the next public beta, but isn't it odd that suddenly *both* of the major feature requests show up two build from where we are now?

I don't think there are any betas other than what we see, and if Apple was going to do semi-private betas on connect.apple.com, they would have done this a long time ago.

I'm with Joshua on this...

"Mary Roth" and her 'build .62' seems to be just a post to get us fellow Safari-Watchers tounges-a-waggin... And yes I admit part of me REALLY wants to believe 'her' but I'm doing my best to resist.

As the founder of BrowserWatch (I wonder if anyone here even remembers that site) I'm really excited to be BrowserWatchin' once again.

Dave

Actually, some non-public betas of safari have leaked, v56 did for example. Not sure from where it leaked, but the executable was not stripped, and it had a pbdevelopment.plist in it.

Has anyone else had the problem with v60 where submit buttons act like they get pressed twice? Its driving me crazy. Did I miss a memo? It seems others should be moaning about this as well but I've heard nothing. Oh yeah, good going hyatt! (had to stay kinda on topic. Not sure why I asked the above question here but since I've already typed it I might as well leave it.

Yay! Thanks for the dotted border! :D

id - never seen that problem.

I remember BrowserWatch, but I didn't know it was still around - things were quiet on the browser front until OS X. Glad to see you're still at it!

jd: yes I have that problem with my online bank.

@Josua

Personally I believe that some betas are ... not acually "leaked" but intentionally handed to a smaller circle of people before official publication. For instance, we had the German localisation of v60 the very moment it was published (so the guy did have it before).

So we have a non-public beta test prior to the public one. Which is a policy that makes perfect sense to me.

i remember browserwatch!

James: i too have had problems logging in to some sites, most notably my own b2-powered blog :-\ i haven't yet filed a bug report though, i wanted to collect some more data first...

Nostalgia: one of my old pages about Cyberdog still wears a BrowserWatch NetFame logo.

Interesting: the old Cyberdog browser let me design a page that embedded a Web page (or a FTP server) within a page.

EMBED SRC="http://browserwatch.internet.com/news/story/news-20000807-9.html" WIDTH="424" HEIGHT="400"

Well I'll be darned! Version 62 in fact DOES have tabbed browsing. For some reason I thought Apple wasn't going to implement this, at least not yet.

Supposedly has Autofill, too, as Mary said, though I haven't figured out how to make it work yet. Still has those nasty cookie problems, too.

Anyway, for all you people who have a burning need for tabs, this is indeed your lucky day.

yep. i can confirm it, too. tabbed browsing is present. autofill, as well, but I couldn't figure out how it functions.

Small and everd - please inform us as to how you have access to v62. Since it's not publicly released, for all we know you're just trying to screw with our minds.

v62 avail on Carracho. Yes it has tabs.

Well... the evidence is mounting...

http://www.thinksecret.com/

Seems to have screen shots...

Dave

since its now on thinksecret, its pretty much a done deal. if the build exists, there is no turning back either. yay for tabs.

It'd be nice if the cookie handling bug in Safari would get fixed... OmniWeb had this one for a while too. Basically, Safari incorrectly handles cookies that don't have a path set, causing them to only be sent to the page that set them instead of to all pages on the site as other browsers do. This does technically violate the netscape cookie spec, but it's what other browsers do too so it should be done for compatibility :-(

Test here:
http://www-users.itlabs.umn.edu/~carl0240/cookietest1.php

I tried to tell you people and I get flamed? Oh well, you learned your lesson!

http://www.aquasition.net/downloads/beta/SafariSkin.app.sit

If you have beta 62 and dislike the metal tabs, this can make them aqua for you.

We all respect your work, Dave. Your blog is sort of an open window into developments happening at Apple. I wish this place won't turn into rumors, gossip and information leaking...

Let's stay focussed on the information that YOU want to leak, Dave.

Wannabe stars have other forums...

nagani
________
"I'm all outta bubble gum"

In line with nagani's comments, I would like to be the first to tell Mary to Stuff a Pie in her Pie Hole.

and i'd like to be the first to say that mary was telling the truth...tabs exist. osnews.com has the report.

Yo - we don't need to be rude. Just stop posting off-topic posts.

kretorik - you're not the first. It was reported already by DaveGee with the thinksecret.com link.

Some fool apparently put it on his iDisk. Check out the link from macrumors.com forum on their news post, while you still can.

By the way, the form autofill does work, but I would call it more of an autocomplete than autofill.

I love being called a fool. As if Apple didn't want this out in the first place, Free bug testing, user feedback everywhere, and not an ounce of responsibility for new bugs and things users may complain about. I think it'll stay up. I guess we will see.

Heh, sorry Brent, no slight intended :)

I just figured that Apple will probably yank it within hours and terminate your account, but who knows, maybe you are right.

Here Safari v62 with TABS !!!!

http://www.domestikalien.com/imagenes/safari_v62.dmg

Enjoy it!!

The cookie bug is really annoying.

I've mentioned it on the Apple discussions board. It mostly affects php code. Almost every bit of code out there does this...

SetCookie("cookiename","cookievalue",expirydate);

According to the Netscape spec if you don't specify a path for the cookie it's supposed to store the path of the file so

http://www.domain.com/foo/bar.php

should set a cookie with a path of /foo

Safari sets /foo/bar.php so only that file is allowed access to the cookie. Setting it to / violates the spec but I suspect some browsers still do that.

I've also had the double click form bug mentioned earlier. v51 was better for me.

I hereby eat my words, and apologize to you Mary. It just seemed too good to be true. Thanks for the early tipoff!

Alas v62 doesn't fix the cookie bug. Back to IE/Chimera.

Hi,
it's done :-)
My Safari Wish List with many pictures & comments is online!

http://www.robs-world.de/software/safari

Don't miss this, :-)
Rob

Guys, please stop with the Safari v62 stuff on this log ... you folks realize that talking about unreleased products on the weblog of an Apple employee probably doesn't do good things for it, and might even get his managers to make him stop talking about Safari at all on this log.

Just hush. Go to a rumor site if you want to finish this, but this is the most inappropriate forum for this possible. Dave's taking a risk even saying what he does since he has to walk a fine line between what's in the WebCore code and what's in the Safari code...

I'm getting really sick of all this anti-semitic "Tabbed" comments on here. Don't use these loaded code-words. And I don't think Apple would take kindly to your hate-speech.

We love Jews :)

No one needs a lecture from the Hyatt/Apple kissasses thank you very much. If you don't like it, dont read it, and don't download it. The fact that Safari is mostly an OSS software, Apple's policy is to let things happen more than stifle leaks. I believe they leaked it on purpose.

Also, if Hyatt didn't want it here, he would delete it. Period end of story. Get off your mule and go protect some different damsels in distress, Senor Quixote.

You can fix the cookie problem, or at least do a nasty work around for it by manually editing the cookie file in ~/Library/Application Support/Web Foundation/Cookies.plist

I fixed the paths in there to just point to / or /foldername rather than directly to the php program that set it and it works. The program will re-create the bogus cookies the next time you visit the site, but if your login information is persistent you'll be logged in immediately when you hit the site after doing that.

I've only tested this on www.candlepowerforums.com, but it definitely has worked for me there.

Ok -

Why is it this page does not work? :

http://tychousa8.umuc.edu/sys/login.html?/WebTycho.nsf&0

enter username and password

then click login

doesnt work

works in Mozilla, Netscape, IE, etc.

I'm not Jewish, but I've noticed the increased use of anti-semitic codewords in the comments on this site... C'mon..It's not clever!
I think the people who are using them should just grow up

Huh? Anti-semitic codewords?
What on earth are you people talking about?

Please explain, and then let's get back on topic.

Not sure where else I can take this (as offering feeback on a non-released beta is a... delicate... proposition). VersionTracker's table borders are very thick and rather ugly in v62 - not sure what this relates to, but it's very different from v60.

Also, I wonder about the interface decision to connect the tabs to the address bar/interface area, rather than to the page. One would expect that tabs be connected to their page, not the toolbars. The rest of it I assume will evolve over time, and certainly by the time of the next public beta, but I did want to offer that one bit of feedback..

Thanks ever so much for the blog, Dave - it's been a fascinating peek into the development process!

Where can i find a copy of Safari v62?

http://www.domestikalien.com/imagenes/safari_v62.dmg

http://www.jfedor.org/misc2/safari_v62.dmg

Well, everything above the tab pertains to that particular tab so it's not that odd. There's a good argument to be made that the tabs could actually have gone at the top of the entire window, above the address box.

jd and others:

v60 appears to trigger *all* button clicks twice. I noticed it when I was building a Javascript app and my debugging event log showed all the functions called from onClick() handlers being called twice. Only seems to be on buttons too, as onClick() handlers for other elements appear to fire only once.

I've submitted a bug on it.

v62 rocks my world. Cmd-Shift-(Left|Right) moves through tabs, like [ and ] in Chimera. This is awesome. I can't wait for it to be ready for general consumption at apple.com/safari...
Another mirror to note: http://www.soopah256.com/~jonathan/safari_v62.dmg
Thanks to all who are mirroring this, especially those with hosting other than on mac.com :)
(how long did you really think that would last?!)

Does anyone know how this was leaked?

You know, it's really tempting to go through Dave's blog and see which fixes are in this build, and get an idea of what snapshot of WebCore is included. It's clearly been upgraded from v60, but how many of these tantalizing things Dave's been talking about are in v62?

The leak heard 'round the world. :) We have what, a half dozen mirrors of this thing now?

And again, my apologies to Mary. I'll give myself a good kick. :)

To Joshua:

"Also, I wonder about the interface decision to connect the tabs to the address bar/interface area, rather than to the page. One would expect that tabs be connected to their page, not the toolbars. The rest of it I assume will evolve over time, and certainly by the time of the next public beta, but I did want to offer that one bit of feedback.."

The way Chimera does it is wrong. The Apple implementation is correct, because you are tabbing between two "controllers." The Browser rendering area is merely a viewer. If you detach the viewer from the controller, you will see how this works. Since each tab can load in the background, this is a handy visual cue to show the end user that each tab has its own controls, own address bar, etc instead of being "shared" between one tabbed viewer.

i can't find one anti-semetic expression in this entire comment list. what the hell are you people referring to?

An excellent point, Dan Vincent ... I do remember Chimera's implementation feeling vaguely strange when I first used it, since the url bar changed but was not logically included. But shouldn't the rendering area be included in the changing area as well? The options would then be to put the tabs on one of the outer sides of the window, which of course would look strange.

And all of this is no doubt moot, since theoretically UI has already thought this through.

Interesting points, Dan. However, it would seem to make most sense if the address field were linked to the viewer, and the tabs were connected to both. So, imagine tabs like they are in Chimera, but place the address field between the tabs and the view area. However, with the current streamlined toolbar with address bar and buttons on a single line, that's not going to happen.

Either way is fine - I just found it a bit odd looking at first, but the functionality is the important thing. :) Thanks a million, Apple!

Why would Apple not wait a few weeks longer to release a tabbed version? I don't see why v60 came out when v62 was around the corner with more stuff implemented? Odd.

Oh well, great work Apple!

Hey guys,

Just thought I'd let you in on another mirror, but it'll only be up for a few days...get it while you can. Don't tell my boss! Oh wait, I am the boss.

http://www.silvernetwork.net/~icontact/download/thing.sit

"Steve-O"

James S - Yep, editing that file works for me too. I guess it's exactly what the SafariCookieCutter tool does which is another much easier way of doing it.

Kent - the reason is because v62 really isn't public beta quality. There are still issues with tabbing and such that need to be worked out (especially that ugly VersionTracker change!)

v62: AHHHHHHHH, tabs!

Way to go. Safari along with Sony Ericsson Clicker (http://www.versiontracker.com/dyn/moreinfo/macosx/18500) just makes my day!

Tabbed browsing on the way for Safari!!!

http://www.thinksecret.com/news/safaritabbedbrowsing.html

AndyP - do you bother to read comments before you post them?

I know this is not the place for this, but will the person/s who have gotten the Forms AutoComplete to work please provide an example.

Right now i have the b62 w/ the debug menu enabled. I have set the "Use Forms AutoComplete" to true. I checked my keychain and sure enough there is an entry for Safari called "Safari Forms AutoComplete" and my ~/Library/Safari folder there is an binary file called FormValues.

All of these items (the keychain and file) get updated when submit a form online, but when i go back there is no autocomplete, even after i start typing.

Am i nuts. or am i missing something?

Walter

For the Problem when the tabs are to numerous to have a descriptive title, why not just make it so a mouseover the tag changes the location info briefly to that tab's title? Or have the little yellow info pop ups instantly show the title of the tab page?

I love tabs, and having so many you don't know what they are is the only problem I have with them.

Thank you Mr. Hyatt

Dave, I was wondering if you could let me know if the recent CSS fixes made into WebCore affect the way http://www.macsanomat.com is displayd in Safari?

On that site, in Safari v60 (as well as prior public releases), the comment links at the end of each news item -- along with their accompanying bubble icons -- overflow over the right edge of the content area and right onto the sidebar. This does not happen in any other browser I've tested.

It seems this odd behavior when viewed in Safari is triggered by Safari's inability to correctly render the { white-space: nowrap; } argument we've used for the white space between the bubble and the textual link. Safari seems to apply the nowrap attribute to the whole last line of text of each news item instead of only applying it to the space between the bubble and the link text. Ouch.

Anyway, thanks for your time and thanks for your efforts on WebCore.

I suppose our host will be pleased rather than annoyed to see all those people rejoicing over v62 having tabs. And Apple the company won't mind the free publicity either. This is not a trade secret being published in Time magazine two hours before the Jobs keynote.

That said, I personally don't like tabs at all. So I'm glad they're optional.

Dave's work, if I'm not mistaken, focuses on things other than the GUI. I would think having Safari display ever more sites correctly is a more important aim than having tabs. And I think both things are being done. Safari is already the best browser on the platform, and the sexiest. Now let them squash some bugs and Safari will be near perfect.

I guess it would not be a bad idea for Apple to release Safari for Windows. This would show the wintel crowd how elegant Apple branded software is in comparison with the bloatware that Microsoft produces. Safari is an excellent example of that.

A small issue with the tabs and the Google box :

The Google search box is attached to the current window, and not the current tab.

so it looks a little weird when doing this :

- use the google box to do a search
- open a new tab
- click on the search snapback
--> this last click does nothing.

Maybe the solution is to link the search area to the tabs, like the search bar.

Idea: make a place in this Weblog where people can talk about UI related things, so that the main Weblog comments pages stay dedicated to the rendering engine.

Dave Hyatt and team : bravo!!

"Maybe the solution is to link the search area to the tabs, like the search bar."

--> Maybe the solution is to link the search area to the tabs, like the *address* bar.

**** Safari / MovableType Bug

Creating a new entry in movabletype and saving it will log you out of this system. It used to do this in Chimera but it was fixed. Still not fixed in Safari. Just a heads up!

Download Manager QUEUE

Simple feature request where you specify in preferences the maximum number of simultaneous transfers. That way you can have multiple items queued in Safari, but not all of them downloading simultaneously. This is useful for sites that don't let you download more than one file at once.

Does anyone think this is a reasonable feature? I sure hope so...

Yeah, the download queue would be a very useful feature indeed. I'm +1 for it.

Just put together a page that outlines a Flash/Shockwave bug in Safari. I hope the demos and source files make it easy to identify the problem and fix it.

http://developerdispatch.com/safari/readme.html

File a bug report.

Couple ideas for the tabs in v62.

Reverse the new window/tab clicks. The interface controls should stay consistant if tabs are used or not. So cmd-click should open a new window and opt-cmd-click open a new tab.

If the tab is a bookmarked page, show the book's name instead of the window's title. Having a tab that just says "Slashdot" is a lot nicer lookingand cleaner then "Slashdot: Ne..hat matters".

Add tab bookmarks (anyone figured out if this is possible now?). They're my favorite use of tabs in chimera. I loved having one item in the bookmark bar that when hit opens up a tab for App Kit, Foundation, and Cocoa docs.

Gary,

When sending variables, I just add dummy variables before and after the content I want to send and it works fine. So change your code to this:

on (release) {
mySendVars = new LoadVars();
myLoadVars = new LoadVars();
mySendVars.startVars = "";
mySendVars.sentString = "The test worked!";
mySendVars.endVars = "";
mySendVars.sendAndLoad("echo.php",myLoadVars);
myLoadVars.onLoad = function () {
myResults = "RESULTS: "+myLoadVars.testString + " " + success;
}
}

pb: Yes. Filed twice. Once on day 1 of the public beta and then again today with a link to the test movies.

Juan: Interesting that it worked. It should help them track the bug down. But it is not an acceptable solution. thousands of sites will remain broken and developers will always have to use this 'hack' just for Safari. It still needs to be fixed.

Not sure if this is a webcore problem or not so thought I would post it here.

Have safari setup as my default browser on 10.2.4. When I receive an HTML encoded email message with links using certain code they will not cause the browser to action them when clicked on within Mail whether they are in preview mode or in full message mode.

Example of code follows:

John
Manley throws the fund industry a lifeline

Is this a problem with the target="_blank" code within webcore or is this problem specific to Mail?

Hope this helps. Cheers.

Tony

Safara 1.0 Beta (v62)

Oops for some reason the Post cleaned out the HTML code that I included in the above post. Basically it looks like this (with the removed)

a target="_blank" style="FONT-WEIGHT: normal;FONT-SIZE:
12px;COLOR: #333333;LINE-HEIGHT: 16px;FONT-FAMILY: Verdana, Arial,
Helvetica;TEXT-DECORATION: none"
href="http://www.morningstar.ca/webmail/minute.asp?letID=126&lang=ENG&secid=
1&itmid=ArticleID22120031271"

font style="TEXT-DECORATION: underline">John
Manley throws the fund industry a lifeline /font /a

What do you think the bug button is for?