At 08:04 PM 4/4/02 -0300, Lalo Martins <lalo@laranja.org> wrote:

I propose to base the Zope3 ZMI on the Plone CMF skin (designed primarily by the talented Alexander Limi and Vidar Andersen with important coding by Alan Runyan), which can be seen at http://plone.org.

Perhaps we should take Lalo's suggestion further and collect a list of existing designs that show what real graphic talent has been able to accomplish. Some of the newer weblogs, wikis, forums, portals etc. out there are looking almost, well, beautiful. Here's an example from YABB (Yet Another Bulletin Board): http://www.yabb.info/community/. I'm sure there are many more that might provide inspiration. As for the perspiration, I realize this doesn't help get the job done. But I think Zope deserves to stand out above the crowd and it doesn't hurt to check out the competition. One design consideration is how much to rely on CSS. Looking under the hood (ie, viewing the source) for some of these "slick" designs reveals modest to sophisticated dependence on CSS. But I think the days of worrying how it's going to look on Netscape 4 should be just about over. What say?? Bill "The commandments of the LORD are right, bringing joy to the heart. The commands of the LORD are clear, giving insight to life . . . For this is the love of God, that we keep His commandments. And His commandments are not burdensome." (Psalm 19:8, 1John 5:3) <http://torahteacher.com/>torahteacher.com --- Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.343 / Virus Database: 190 - Release Date: 3/22/02

Stephan Richter writes:

I think we can safely rely on CSS. Common, Zope 3 is a new product, and if someone needs to make it backward Browser-compatible, he can always implement a less fancy CSS-free skin.

Argh! I can hear Guido tell us about his Netscape 4 now... There are still a *lot* of people using that one and it's ilk. Browsers at that level know just enough CSS to make it painful. ;-( I've spent more time making the Python docs work in NS4 than making them look good in more modern browsers, and the more advanced efforts I've made can't go in because the older browsers just don't come close! -Fred -- Fred L. Drake, Jr. <fdrake at acm.org> PythonLabs at Zope Corporation

I think this conversation is trending in the wrong direction. Zope 3 needs to make it possible to build YABB, interfaces which support all browsers while still looking slick, etc. However, it is important to note: Zope 3 is *not* a product. It is used to build products. The core ZMI is needed to the extent that it helps build or administer products. Thus, Zope 3 is not like YABB. Of course, Zope 3 can ship with one or more sexy sample applications, like YABB. But if we blur the line for Zope 3's ZMI, we'll be right back into the core problem of Zope 2's ZMI: audience confusion. One of the first questions to ask when building an interface is "What is the audience?" Giving a very focused, tough response can greatly boost effectiveness. With all this in mind, I think we can require developers to use standards-compliant browsers, and allow/facilitate them to build backwards-compatible interfaces. --Paul William Trenker wrote:

At 12:47 AM 4/5/02 -0500, you wrote:

...

I've spent more time making the Python docs work in NS4 than making them look good in more modern browsers

This is sad but true. I still have Netscape 4 here for testing as well. (I run into the same problem with other web technologies, like Macromedia Flash. Yes, probably 97% of browsers do have a Flash player installed, but what version? Many are still at Version 3.) Those "Download FREE upgrade" buttons may as well be invisible.

I don't think we should let Zope3 look ugly on old browsers, but isn't it acceptable for it to look more modest? With careful use of CSS, it is possible to let the older browsers "fail gracefully".

And it is amazing what can be done with very little CSS and lots of images. The page at, http://www.yabb.info/community/, has 2 styles, no Javascript, and looks very respectable. But look at all the GIF's.

If the web-centric, software-tools development community feels strongly about perpetuating old browsers (and old web standards) then isn't it about time for that community to provide tools to hide the details? When will Python, and for sure Zope, have built-in browser detection and formatting support driven by some sort of meta format that let's us application developers get on with developing applications?

mac OS X browsers: - IE 5.1 - mozilla 0.99 (=>1.0) i won't include OmniWeb because its CSS support is still flaky. jens On Friday, April 5, 2002, at 04:04 , Dan Pierson wrote:

On Fri, 2002-04-05 at 12:41, Paul Everitt wrote:

...

However, it is important to note: Zope 3 is *not* a product. It is used to build products. The core ZMI is needed to the extent that it helps build or administer products. Thus, Zope 3 is not like YABB. [snip] With all this in mind, I think we can require developers to use standards-compliant browsers, and allow/facilitate them to build backwards-compatible interfaces.

I agree. IMHO the browser compatibility requirements for the new ZMI should be summarized as:

Current IE Mozilla 1.0 Konqueror (KDE 3.0 version) ...what's the current state of Mac browsers...

No Flash

Anyone building or maitaining a Zope site shouldn't have problems running one of these. On the other hand, relying on IE-only features would be bad for those of us building only on Unix/Linux. Ditto for folks in the Mac world. What other OS's are important for site builders?

dan

On 05 Apr 2002 09:04:15 +0000, Dan Pierson <dan@control.com> wrote:

I agree. IMHO the browser compatibility requirements for the new ZMI should be summarized as:

Current IE Mozilla 1.0 Konqueror (KDE 3.0 version) ...what's the current state of Mac browsers...

That doesnt sound unreasonable as a list for browsers that will behave 100% correct. On occasions it is necessary (or convenient) to use a previous generation browser with the ZMI. Its would be good if they only suffered cosmetic problems, not functional limitations. Toby Dickenson tdickenson@geminidataloggers.com

I agree 100% with Toby. I don't care how it looks in NS4 or (insert old non-standard browser here), so long as the functionality is still there. I think the ZMI should also work 100% with w3m. If we do that, then we are basically already taking care or accessibility. I also vote to kill frames. However, IFrames might still be useful (for the tree view, etc). As for Javascript, it should work 100% with Javascript turned off. But, I'm happy to have some JS sprinkled in there as candy that makes things easier, or enhances the UI in some way. I am also in full support of utilizing CSS to the greatest extent practical. The best thing about CSS is that when done properly, non-CSS compliant browsers work just fine, they just can't display things as nicely. A proper CSS class structure takes planning, just like the Python classes do. We need to make sure that the CSS design is an integral part of the ZMI, not just purely evolved as an afterthought. In light of that, I would like to see a ZMI skin that is fully xhtml 1.0 compliant, and uses CSS2 to its full extend, and possibly some CSS3. To me that means that one could develop the html coding completely devoid of presentation (no tables used for formatting, etc). I would also like to avoid using images for this skin unless they convey some meaning (such as icons) or otherwise enhance the useability. IOW, no shims, rounded corners, etc. I doubt this could be the default skin because it would probably be pretty plain graphically and not display too well on older browsers. OTOH, you could simply replace/tweak the CSS to fully customize the layout of the ZMI. That would rock. If I had a clone somewhere I would volunteer him to develop it 8^). -Casey Toby Dickenson wrote:

On 05 Apr 2002 09:04:15 +0000, Dan Pierson <dan@control.com> wrote:

...

I agree. IMHO the browser compatibility requirements for the new ZMI should be summarized as:

Current IE Mozilla 1.0 Konqueror (KDE 3.0 version) ...what's the current state of Mac browsers...

That doesnt sound unreasonable as a list for browsers that will behave 100% correct.

On occasions it is necessary (or convenient) to use a previous generation browser with the ZMI. Its would be good if they only suffered cosmetic problems, not functional limitations.

Toby Dickenson tdickenson@geminidataloggers.com

_______________________________________________ Zope3-dev mailing list Zope3-dev@zope.org http://lists.zope.org/mailman/listinfo/zope3-dev

From: "Gary Poster" <garyposter@earthlink.net>

I agree that we ought to trash frames we ought to use strict xhtml 1.0 we ought to rely on CSS for all graphic elements (correlative) we ought to not use *any* shims or non-logical tables the site ought to work completely without JS or Flash the site ought to be usable and legible without any CSS

I disagree that we ought to use iframes (why lock out NS4 if we don't have to?) we ought to avoid Flash like the plague (see my XML file tree email) this shouldn't be the default skin (it *should*)

I further feel that it would be very easy to develop a CSS file folder that one could place in a skin that would deliver different CSS files based on browser-sniffing. The CSS could be cached by the browser if we always call it from the same

Hmm... If only someone had thought of that, say about a year ago. ( http://www.zope.org/Members/haqa/ZStyleSheet ) My question is this: Everyone is saying "The ZMI is bad, it's confusing, users don't like it". Could anyone show me evidence of this? Personally I love the ZMI in the current versions. I have also found there is a negligable learning curve for users who already know how to use Windows Explorer and similar products. I just don't see the need to throw out the ZMI - Are we in a baby/bathwater situation here?

address. While CSS support varies widely in all the various browsers, and I think we should *ship* with a CSS file that requires a standards-compliant browser such as those that Toby lists, I see *no reason* why we should design the html itself so that a NS4 user could not use the default skin simply by building him or her a new CSS file. It might not look nearly as slick as the compliant browsers, but you can still do some reasonable things with NS4 only...

On Sat, Apr 06, 2002 at 10:52:27AM +0100, Adrian Hungate wrote:

My question is this: Everyone is saying "The ZMI is bad, it's confusing, users don't like it".

Which everyone? Not me.

Could anyone show me evidence of this? Personally I love the ZMI in the current versions. I have also found there is a negligable learning curve for users who already know how to use Windows Explorer and similar products. I just don't see the need to throw out the ZMI - Are we in a baby/bathwater situation here?

We're not throwing out the ZMI. We're writing something similar for Zope3. The name is ZopeTop. []s, |alo +---- -- It doesn't bother me that people say things like "you'll never get anywhere with this attitude". In a few decades, it will make a good paragraph in my biography. You know, for a laugh. -- http://www.laranja.org/ mailto:lalo@laranja.org pgp key: http://www.laranja.org/pessoal/pgp Brazil of Darkness (RPG) --- http://www.BroDar.org/ Python Foundry Guide http://www.sf.net/foundry/python-foundry/

If you're looking for a group that understands UI and HTML, check out www.37signals.com. They do excellent UI and design work that looks simple, clean, and slick. They're expensive, but I wouldn't be surprised if they donated their work just to say they designed Zope 3's UI, that would be a huge thing to put on their resume. I agree with the posts here that 3.0 should use all the technology available to it (CSS2+, XHTML, etc). It's frustrating to design for Netscape 4 and time people upgraded. I've personally stopped designing for NS4 altogether. Furthermore, will this design bleed over into the CMF? If I recall correctly, CMF is going to be built into Zope3. A new default theme for CMF would be nice. Sean Friday, April 5, 2002, 9:29:04 AM, you wrote: AL> From: "Casey Duncan" <casey@zope.com>

...

In light of that, I would like to see a ZMI skin that is fully xhtml 1.0 compliant, and uses CSS2 to its full extend, and possibly some CSS3. To me that means that one could develop the html coding completely devoid of presentation (no tables used for formatting, etc). I would also like to avoid using images for this skin unless they convey some meaning (such as icons) or otherwise enhance the useability. IOW, no shims, rounded corners, etc.

AL> This is *exactly* what Plone does now. Every widget and tab is controlled AL> and made exclusively with CSS. No graphics involved. (well, the icons of AL> course, but... ;) AL> -- Alexander Limi AL> _______________________________________________ AL> Zope-Dev maillist - Zope-Dev@zope.org AL> http://lists.zope.org/mailman/listinfo/zope-dev AL> ** No cross posts or HTML encoding! ** AL> (Related lists - AL> http://lists.zope.org/mailman/listinfo/zope-announce AL> http://lists.zope.org/mailman/listinfo/zope )

At 07:41 AM 4/5/02 -0500, Paul Everitt wrote:

I think this conversation is trending in the wrong direction. The core ZMI is needed to the extent that it helps build or administer products. Thus, Zope 3 is not like YABB.

Yes, your point is well taken. I hesitated bringing this up in the first place but the original request was for graphically talented designers and that got me thinking about the ZMI as an intrinsic demo of Zope's presentation capabilities. I remember when as an application designer looking for a web framework that, other than Zope.org's own portal pages, the ZMI was the first thing to catch my eye. As an app designer, when looking for tools out there, I can't help but get an impression of the potential of what the tool set can produce from how the tool set looks to me. When I'm new to Zope, trying to navigate my way through the tool set's screens I am, psychologically, getting an impression of the what I could produce for my own audience.

Of course, Zope 3 can ship with one or more sexy sample applications, like YABB.

Yes, good point. Let me add that when I started learning Zope the ZMI was the only sample that I knew how to become acquainted with. But that was probably because there were no other examples at hand when I installed it "out of the box". And I'm not ashamed to say that one way I've learned to do presentation with Zope is to study how the ZMI's presentation is implemented. Regards, Bill "The commandments of the LORD are right, bringing joy to the heart. The commands of the LORD are clear, giving insight to life . . . For this is the love of God, that we keep His commandments. And His commandments are not burdensome." (Psalm 19:8, 1John 5:3) <http://torahteacher.com/>torahteacher.com --- Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.343 / Virus Database: 190 - Release Date: 3/22/02

...

...

...

...

"JV" == Jens Vagelpohl <jens@zope.com> writes:

JV> as a data point, here is browser data for one of the sites i JV> own. the data is produced by webalizer. NS4 is becoming JV> negligible::

...

...

...

...

"PE" == Paul Everitt <paul@zope.com> writes:

PE> However, it is important to note: Zope 3 is *not* a product. PE> It is used to build products. The core ZMI is needed to the PE> extent that it helps build or administer products. Thus, Zope PE> 3 is not like YABB. So the question is, who make up the audience for the ZMI and what are the browsers that /they/ use? The general populace isn't going to be developing apps for Zope, it's the site admins, and I'd venture to say that the distribution numbers are quite different for that group of people.

...

...

...

...

"JV" == Jens Vagelpohl <jens@zope.com> writes:

JV> mac OS X browsers: | - IE 5.1 | - mozilla 0.99 (=>1.0) We (my wife and I :) are primarily a NS4 shop on MacOS9, and MacOSX comes with IE5, but we don't use it if we can help it. There it's Moz 0.9.9 which works great. -Barry

At 11:36 PM 4/4/02 -0600, you wrote:

I think we can safely rely on CSS.

Would you be comfortable with CSS Level 2? Current CSS standards even provide some of the dynamic formatting, like text rollovers, that used to require Javascript . I, for one, would limit the dependency on Javascript in the standard Zope3 skins. There are still problems with some browsers in this area. Even with the current browsers (Netscape 6.2, Internet Explorer 6, Opera 6.01) I'm still finding discrepancies. Some Javascripts work, some don't. And there's still the occasional Java Applet that works in one browser and crashes in another. Any thoughts on this? Bill "The commandments of the LORD are right, bringing joy to the heart. The commands of the LORD are clear, giving insight to life . . . For this is the love of God, that we keep His commandments. And His commandments are not burdensome." (Psalm 19:8, 1John 5:3) <http://torahteacher.com/>torahteacher.com --- Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.343 / Virus Database: 190 - Release Date: 3/22/02

William Trenker writes:

Would you be comfortable with CSS Level 2?

My own thought is that we could use CSS 2 and as much of CSS 3 as we can coax out of advanced browsers. We should use the advanced CSS features to avoid needing *Script, since that's evil and many users turn it off due to security and privacy concerns. -Fred -- Fred L. Drake, Jr. <fdrake at acm.org> PythonLabs at Zope Corporation

Hi, Alexander Limi here, I am responsible for Plone's design and XHTML/CSS. I just want to clarify some issues: * The goal of Plone is to be lightweight, to offload as much layout as possible to the CSS. Should be possible to use in low-bandwith situations (mobile 9600bps GSM connections etc). * Target browsers for Plone are * Mozilla/Netscape6 * Opera 5+ * IE 5+ * Konqueror * Netscape 4 * w3m * lynx These are the browsers I want to be able to guarantee that Plone works in. I even added a graceful degrading layer for Netscape4, since it is so common in the open source world. (it's broken at plone.org at the moment though, just returned from my Easter vacation, and somebody has checked in changes that stops it from working - I will locate said person and give him a good whacking ;) * Goal is 100% XHTML 1.0 compliance, and this will not break any browsers as some people seem to think. * Another goal is to use Javascript where useful, but always make it fully functional without it. * Plone should be very quick to load, close-to-no graphics, and is perfectly suited for the new ZMI (in fact I designed it with this in mind ;) * And - You Ain't Seen Nothing Yet(tm) - PloneNG will utilize a lot of nifty DOM tricks and probably XML/XSL to leverage the user experience on newer browsers. The point I really want to make here: cross-browser compatibility isn't that hard if you approach it the right way and accept some degrading for inferior browsers like Netscape 4. As for comments about Plone's slowness at the moment - this has nothing at all to do with the XHTML/CSS and general design - this is because of expensive method calls in the Plone code. What's *really* important about the new interface for Z3 is that the management interface is designed by someone who understands *both* HTML and UI design/usability to a high degree. I have been proposed as one of the persons to take on this task, but as I'm already investing a lot of my spare time in Plone, I would need something more than just something nice to put on my resume. Finding work in the interface/usability business is hard at the moment, and I can't afford to invest a lot of time in the Z3 interface without being compensated for it in some way. If somebody (Zope Corp or any other organization) wants to sponsor this effort, I'm sure we can come to an agreement on pricing. Discounts for the Zope crowd ;) I hope I don't come across as arrogant, but these are the realities - if I don't get paid, I can't afford to do it. Such is the state of business nowadays. Unfortunately. :( -- Alexander Limi