Weekend DiscussionFriday February 18th, 2000What *one* feature would you most like Mozilla to have by final release? If it's already on the slate to be in the final release, let us know why it's important to you. Just click the Responses link to join the conversation! Our forums are quite civil, and we'd love to hear your opinions. I'd like the scroll wheel to work. Of all the various things, this is the most important, because that is how I scroll to pages. I don't know if it is just not implemented or not, but on my Mac, the wonderfull wheel does not work. I know many other people are just as addicted to this method of page scrolling. Just my oppinion. Though I'm not using a Mac, I just wanted to comment that I was surprised and delighted to find that my scroll wheel under Linux was working perfectly. I'm using the ZAxisMapping thing, not imwheel. I'm so pleased with this, and the recent performance of Mozilla, I'm not even using Netscape anymore - I made additions to my .Xdefaults file to enable wheel scrolling in Netscape, but it never did work right. You Mozilla guys are cool. I'll wait as long as it takes for the end result. --Beej intellimouse Forward & Backwards button support. Thats what I want. On my new laser driven intellimouse explorer my thumb has forward and backward buttons and I wish mozilla supported them just like MS-Internet-Explorer. I'd also like to send in my vote on this one. I use microsoft's mousewheel thing and love how it works with IE. I've found that if I place the pointer over the scrollbar and click the center mousewheel button, it "holds" the bar and I can move up and down freely... IE's integration is far better allowing me to click ANYWHERE on the page and then move up, down , or diagonal. Hope it makes sense and It'd really make me feel more at home in Moz I would like to see XSLT support; a lot of the data I deal with cannot be adequately handled by CSS alone, but XSLT and CSS together are absolute dynamite. XSLT support in Mozilla would also make it more competitive with IE5. Steve One thing at a time, I think. Adding XSLT support at this point would probably be a bundle of work. I'd rather see that effort devoted to fixing that last 2% of the CSS/HTML spec that's still buggy. CSS2 and XSLT support should be top priorities for the /next/ release of Mozilla, but we shouldn't put time in on them until we've got the more basic stuff debugged. One issue here is: what does it mean that something will be in the `final release'. I don't disagree that bug-fixing is crucial, but that should be a given. The subject under discussion is: what is the feature any given individual most wants. It's not clear to me that it would be a great time-sink. Much work has already been done, starting from the basis of an XSLT processor that was already reasonably far advanced. Your assumption might be correct, but I think it would need demonstrating. XSLT, incidentally, is a W3C standard for transforming XML documents into other formats, e.g., XML or HTML. It enables much more sophisticated transformations of data than CSS, and would permit Mozilla to act more like a native XML browser even on data that was not marked up with presentation in mind. Of course, XLink support would also be nice in that context, but that's probably another topic. Does it automatically fill in the text when people type n/t? No, it doesn't, but it can go to the bathroom for you. I was serious about that question. I would like to know what people think is so good about XSLT. XSLT has the ability to completely restructure the input-tree, including adding tags, changing tags, removing tags, and changing the order of tags. Definitely a lot more powerful than CSS which just lets you control piddly little display attribute stuff like colors, fonts, margins, etc. Basically, with XSLT, you can take any sort of XML document, pick out certain parts of it, and rewrite those parts in another XML or HTML format. Barryp's description is good. One feature to add is the incredible power of XSLT's document() function. This enables you to pull in anything that can be referred to by a URI (or set of URI's), from simple XML documents to the results of CGI queries and other things. So, not only does XSLT give you total control over client-side subsetting and restructuring, but it also enables augmentation of the contents of the page, or dynamic client-side page generation, integrated with flexible document addressing (XPath) and transformation. Ok, I'll give you my situation using Microsoft's XSLT parser/engine... I know it's sad, but I write an intranet for 500 users on IE5: I have a TOC XML file for every section of my site, and one big one for the entire site. using XSL, I can have IE open the XML file and the translation file, and automagically output a table of contents "tree" that can expand three levels down. Since any update to the site triggers an update to the XML file, I have a way of getting everyone the correct table of contents without using frames (which suck), or serverside scripts (which would kill my wussy little server). The big deal that XSL has over CSS is that you can selectively display parts of the XML file, and you can include them in ways that CSS alone will not let you do (ie <p><xsl:attribute name="onclick">expandcollapse('<xsl:value-of select="docspec/@title"/>');</xsl:attribute></p> lets me include the title property of a docspec tag from the XML file as an arguement to the expandcollapse function in javascript. Also, there are conditionals, and quite a few other features that I use. This is one feature that I can definitely say kicks ass and fills that unavoidable niche of data reporting. i'm unsure as to whether or not this feature exists in mozilla already, but here goes: i think it would be excellent if all the different parts of moz that parse documents (ie. HTML, CSS, etc), were modularized to the point where you could pick and choose which ones you want to be loaded. (as i understand it, moz currently loads all parsers which leads to a big overhead - but i could be wrong) this feature would be great for helping third parties to make their own mozilla branded distributions. here's why: company X is would like to put mozilla into some special hardware for creating some sort of internet device. Their hardware is fairly limited, so they reduce the overhead of mozilla by removing certain parsers that won't be needed for whatever their particular requirements. what do you guys think? am i way off track here? I wouldn't say you're off track at all, in general terms. In specific terms, though, the issue is what should be in the release. The point you raise is important, but I would say it is secondary to ensuring the necessary functionality. Steve I actually had an idea I sent to Mozilla's "Blue Sky" section ages ago (middle of last year I think) where I suggested Mozilla went one further than that: Actually modularise the *entire* browser. A module would be a collection of shared libraries that implement a major feature (e.g. a mail module, a news module, an HTML editor module, even the protocols and parsers could be modules). You'd have dummy libraries to fill in return values that say it's not implemented and to add or delete a module, you simply replace the appropriate shared libraries with dummies or vice versa and restart Mozilla. A nice idea in concept (if I say so myself), but runs into nasty issues like dependencies and gazillion of many-platform many-version shared libraries having to be kept on Mozilla's sites (and mirrors) so they can be downloaded (via an intelligent, separate, installer program). Sadly, it never appeared on Blue Sky and I never even got an "are you mad ?!" e-mail back :-) However, the major advantage is that you build your own Mozilla combination out of modules. You have a core set of course that are compulsory modules and you can roll your own "Standalone", "Standard", "Professional" etc. by just choosing what modules you want. Of course the combinations would lead to a nightmare w.r.t. testing :-( Plus upgrading is just a simply case of loading up the intelligent installer, which spots which ones are out of date and downloads a .zip of the shared libraries that updates your modules. Sort of like "SmartUpdate", but better :-) I've got the original stuff I sent to Blue Sky, but the more I think about it, the more it's fantasy - it would require a major reworking of the source code to support it. couldn't even remotely consider using it without ssl. Can't be done until RSA's patent expires on September 29. I wouldn't be surprised to see it go in on Sep. 30, though. Just to clarify, netscape will be releasing a closed source version of mozilla once mozilla is complete. This closed source version will include all sorta of things inclusing ssl, verisign, AIM (EWWW). My favorite has to be one of my own requests: a hierarchical go menu. The bug is at http://bugzilla.mozilla.org/show_bug.cgi?id=21521 It would allow users to go back to a page, follow another link on that page, and still be able to easily jump to the page they were at before. As somebody commented, it would be like having a "sideways" button as well as back and forward. That would be cool - implementing backwards and forwards as a tree - so if you went back and then forwards to a new page, you'd just create a new branch on the tree. I'd like to see a built-in validator for (1) HTML 4.0 + CSS 1/2 and (2) XUL+CSS+DTD(+JS?) (chrome). The first would be very usable to webmasters who want to see if their pages are written correctly and to users who'd see if there are errors in the sites they view ;-) There should be some small button showing something like "bad HTML" and brings up a validator window showing the site's errors. The second (chrome) validator would be VERY useful for people working on skins, own components or localisation. There should be some validator window which checks for errors in chrome code and shows them (also in which line they occur - also in dtd or css files!). The "old" console window did show xul errors, if there was a dtd error, it only showed up the file - now those notifications are also gone if you redirect mozilla's output to a file :( I think the two validators could also be more or less the same feature... That would be very nice. There could be a "Validate" button next to the Shop@Netscape... I mean next to the "Check Spelling" button. It would be even better if the validator could offer recommendations for fixing the Webpage. Validation of webpages is not something that should be done by a browser. It is supposed to be part of HTML-editor packages like Homesite and Dreamweaver. Maybe a link that verifies a page @ W3C would be handy at times though..... And don't put another button in the chrome please!!!!!!!!! Jelmer I do not think Mozilla should have any "built-in" links to the W3C. If Mozilla has to link to a Website to validate Webpages, I think it would be better if Netscape created their own validator for this. Maybe they could use the Website Garage. Also, I do not think there is any reason an HTML validator would be inappropriate for an HTML editor. I assume one of those will be issued with Netscape. The button would only show in the editor, so it would not require space in the Web browser interface. I agree completely...one of the best ways to promote standards is to let people know when they're breaking them. Is there a BugZilla item assigned for this? If so, I'll vote for it. I do think that the indicator should be kept quite unobtrusive...perhaps something similar to the padlock security icon in the current Netscape. Just filed a bug for that: #28558 http://bugzilla.mozilla.org/show_bug.cgi?id=28558 Please add your comments if you feel this should be supported! Best to vote for it here: http://bugzilla.mozilla.org/show_bug.cgi?id=6211 YES, exactly that is the most important. And it would be nice if SVG too ! ...though it is now in drafts Sounds like a good idea, although it won't propably included in Moz for a while. But what should be really, really simple would be a button or menu entry with :"validate current URL in W3 validator" which would send the current URL to one of the existing validators. Although this solution wouldn't work automatically but only when specified. Am I on the right track? If it links to thw W3C, I will keep using Internet Explorer. Why the objection to the W3C? They /are/ the official body responsible for the standards that Mozilla is implementing, after all. Technically, every HTML document should have a link to their DTD's in the DOCTYPE specifier. Thinking about it, it seems like the 'best' approach would be to validate the document being displayed against the DTD specified in it's DOCTYPE declaration (which is actually what the W3C validator does, but only for W3C DTDs). As I said before, I believe the W3C is a closed, elitist organization, open only to the people willing to pay the high price to be a part of it. I believe it goes against much of what Mozilla claims to stand for. The W3C is to standards what Microsoft is to operating systems, and what AOL is to Internet service providers. I think the W3C validator would work, but other validators would work too, and they are not a part of the W3C. I still think Netscape should use its own validator for Mozilla. It would probably work better anyway. You call the W3C a closed, elitist organization, and including a mere link to them would cause you to continue to use Internet Explorer, made by, gasp, a closed, elitist organization? :) (Heck, at least the W3C has public mailing lists and working drafts posted every now and again.) If you had to choose between two browsers from people you did not approve of, would you choose the one you had to download, which very few people use, or the one you already have, which many people use? Also, unless there is a dramatic improvement, I am going to have to get more RAM to run the new Netscape and if I do that, I might as well get Windows "ME" and I will probably have IE 5.5 anyway. After all, it's me... The main thing I liked about Netscape is they were against those Microsoft principles I do not like, but they are seeming more and more like Microsoft with every step they take. It may soon be an issue of convenience rather than approval. I really don't see what you are trying to say here. What do you mean by "from people you did not approve of" ? What are the criteria's that browser makers have to meet before you will approve them? (just curious) Basic My main reason for supporting Mozilla is that I do not like the tyranny of Microsoft. I used to use Netscape because it was the best browser, and we all know what happened with that issue. I think ,if the Internet is directed by a "standards body", and that standards body is closed, then that creates an Internet that is built based on the rules of a few "elite" people rather than an open Internet. If Netscape supports this closed organization through unnecessary links, then they are removing my main reason for using Netscape, and being hypocrites, IN MY OPINION. Do I not have a right to be against the W3C? Is that against the W3C recommendations? How many times must I explain myself? Maybe I do not have the proper document type definition for disagreeing. I still plan to download the next version of Netscape and possibly build my own version if it is not too much of a hassle. I just do not think I can enjoy it if there is too much trash added, and this W3C thing would be added to the trash that is already confirmed, such as the Shop@Netscape button, the mail and news that shows Netscape ads, and all of the bundled software that just adds to the download time. In case you still do not understand, I will summarize it. >I really don't see what you are trying to say here. If Mozilla includes built-in links to the W3C, I am going to keep using Internet Explorer. >What do you mean by "from people you did not approve of" ? In my case, it is the W3C. If you need a definition of the words in that sentence then I just do not have the time right now. E-mail me instead. >What are the criteria's that browser makers have to meet before you will approve them? (just curious) I like when people ask me questions like this. My answer for that is complex. I will offer a few of the things I need for a huge browser download to be worthwhile. The browser has to render pages at least as fast as my current browsers, and should not require more memory than the current browsers require to render the same pages. It should render them the same way the older browsers do but I am trying to overlook that since it is the fault of the W3C, not Mozilla. It should have a good interface and be released in my lifetime. I have been patiently waiting and will continue to wait because I think Mozilla will deliver much of what is promised, though not all. Its installer should be customizable enough to let me choose where it is installed, and which parts are installed. Internet Explorer did not completely meet that criteria but I accepted that because I did not have much of a choice. There should be no commercialism such as banner adds, links to shopping sites where navigational buttons should be, or bookmarks I cannot get rid of. What about downloading the _Mozilla_ milestone that coincides with the Netscape release. It won't have any of that shop@netscape crap in it and if the only thing included in it that you object to is a link to the W3C then surely it's overwhelmingly open and free nature would outweigh a link to a body you don't approve of (at least enough to make it better than the MS alternative). Or you could grab the source at the same time as Netscape does for its release and remove the validator, or alter it to use someone else's. I don' thing I'm religious about this issue. So when I compare a an open source browser freely available to download and modify (which just happens to have a link to a semi-open organization) with a closed source browser from an elitist and closed company happily dominating 95 percent of the desktop world and feel really good about Mozilla, I don't feel like a zealot. I'm not 100% sure how the autocomplete will work in Mozilla, but the current Netscape ac bothers me. First it's not fast enough, often I must wait for it to complete the url (after which I realize that it would have been faster to just type it out). Second, and most important, it's dumb. I cannot figure out how it chooses which url to show when they contain the same characters. (ie typing "go" and you've visited google.com and go.com) I've tried visiting the site I want to display on a new install, and it works for a while, but chooses another site after awhile. So it's not the first site visited and it's not the most (recently) visited url either. So I propose it should work by urls with the most visits are given a higher priority. Preferably the user should be allowed to specify what url is more preferable (maybe by holding shift and enter when going to the site marks a site as a best choice).I'm not 100% sure how the autocomplete will work in Mozilla, but the current Netscape ac bothers me. First it's not fast enough, often I must wait for it to complete the url (after which I realize that it would have been faster to just type it out). Second, and most important, it's dumb. I cannot figure out how it chooses which url to show when they contain the same characters. (ie typing "go" and you've visited google.com and go.com) I've tried visiting the site I want to display on a new install, and it works for a while, but chooses another site after awhile. So it's not the first site visited and it's not the most (recently) visited url either. So I propose it should work by urls with the most visits are given a higher priority. Preferably the user should be allowed to specify what url is more preferable (maybe by holding shift and enter when going to the site marks a site as a best choice). Or even specifing the order of ranking in pref.js (or Mozilla's equivalent). -AU (PS: some (mostly newer) users find ac extremely annoying, so don't forget an "off" option (preferences would be best place, since its mostly newer users who have this complaint)) I'm not 100% sure how the autocomplete will work in Mozilla, but the current Netscape ac bothers me. First it's not fast enough, often I must wait for it to complete the url (after which I realize that it would have been faster to just type it out). Second, and most important, it's dumb. I cannot figure out how it chooses which url to show when they contain the same characters. (ie typing "go" and you've visited google.com and go.com) I've tried visiting the site I want to display on a new install, and it works for a while, but chooses another site after awhile. So it's not the first site visited and it's not the most (recently) visited url either. So I propose it should work by urls with the most visits are given a higher priority. Preferably the user should be allowed to specify what url is more preferable (maybe by holding shift and enter when going to the site marks a site as a best choice). Or even specifing the order of ranking in pref.js (or Mozilla's equivalent). -AU (PS: some (mostly newer) users find ac extremely annoying, so don't forget an "off" option (preferences would be best place, since its mostly newer users who have this complaint)) Another grip I have about the current auto complete in Netscape 4.x is that if Netscape happens to find the right URL you have often just typed another letter causing Netscape to start searching all over again. It would be very easy to fix, if there is a site that auto complete has found that matches all the characters typed in, then when/if another character is typed in check to see if it matches the current URL. I would like to see the IE style autocomplete implemented, as this can just be ignored, supports multiple option and is realtime. The worst thing about autocomplete in 4x is that you might type a perfect url in, then netscape adds something else to the end and you end up going to the wrong page. Lemming http://www.lemnet.com/ I agree, as I did before. I do not remember the last time I found the Netscape autocomplete useful. I think it has ruined correct addresses more often than autocompleting them. My only concern about the Internet Explorer autocomplete is it is horribly slow on my computer. I think maybe that is because it searches through all of the Windows shortcuts too, and even causes my CD-ROM to start going. If Mozilla had a more efficient version, that would be very good. To be more specific, the autocomplete should only place the root of the address in the address bar until the user adds to the address, at which point the next folder should be added to the address unless there are no more folders, at which point the filename should be added when the user starts typing it. I think Internet Explorer chooses which address to show first, based on aphabetical order. Order based on the number of visits would be more useful. An IE-style autocomplete that sorts based on the number of visits and does not require 30 seconds to finish an address would make the Mozilla autocomplete better than all of the other ones on my computer. IE style complitation cannot accept writing single words (to which it will add http://www. and .com). Perhaps it only accepts several words which is unacceptable. Try writing apple if you don't belive me - it will open a search page, usually saying it didn't find apple. Hi I want the Mozilla to have the right click menu "send Page" option. It is in Netscape 4x and i use it a lot. Its very useful, any page you are viewing and you want to send to someone just click the right button on the mouse select send page and it will open up messenger with that page in it. Love it. Very handy!!! I have not seen it in the lastest builds. PLEASE!! put it back. It sure would be nice to break away from bitmap-jail, and have vector-graphics in webpages. I was just checking out Adobe's SVG Netscape plugin at http://beta1.adobe.com/svgpreview_alpha/SVG/main.html ,and it can do some pretty impressive things. They did mention that the SVG had a big footprint in current browsers, but that future releases would be much smaller. must have drag and drop link between mozilla windows... Is there anything that can be "dragged" and "dropped" in Mozilla? It would be nice if the bookmarks had this feature. I have found it difficult to move bookmarks into the "personal toolbar folder". I had to edit the bookmark file with the text editor. I'd like the back button to remember where on the previous page I was. It's a pain to have to scroll down past two dozen slashdot comments just to get back where you were when you followed a link. I think this one is a really good idea. By the way, what makes mozilla so cool is how the mozilla team gets real feedback from their users like they are doing now. There are two basic issues with this. Though its good to note, that going back to anchors does work (I sumbitted that bug way back in m6 I believe). The problem with going back to a point in a page, is you would lose alot of the dynamic nature of many pages. If the page changed, it would be very difficult for the browser to intelligently guess as to where to point the browser. A system could be implimented so as to check, and if its the same to goto that point, but it would be strange loading, as if it was loading and discovered something was changed that got past its crc check, it would definatly freak out. well how about an approximation? Ie, maybe mozilla could remember what link you last pressed, then when you hit back, render and then put that link at the top of the screen, so that you can at least semi-find where you were? Dunno, maybe that makes no sense, W Anybody know anything about them implementing HTTPS? Will it be included in the SSL/encryption or is it something separate? I really want to be able to see how my Symantec beta site renders and works under Moz Similar to MSXML.XMLHttpRequest in IE5. This will allow for raw XML,XML-RPC or SOAP communication to server logic. If Mozilla is to be a fundament for next-gen web-based apps this is a must-have. Don't know if links work here.. <a>http://bugzilla.mozilla.org/show_bug.cgi?id=15119</a> I think XSLT and MSXML.XMLHttpRequest are extremely important, otherwise Moz will be lacking in XML support compared to IE. I'd also live to see psuedo(forgot the complete name) styles be implemented on more than just links... So I could use div : houver {background-Color:red} And then I'd love to see DOM2 mutation events implemented. Hehehe, In thinking of all the things I'd like to see in Moz I forgot you asked for only 1. So I'll just say that I'dike to see complete W3C standards support for all levels. Need ability to instantiate COM objects in client-side scripts. I'll vote /against/ this one. For one thing, it would only work on Win32 platforms. For another, ActiveHeX is a security nightmare...don't let it anywhere near Mozilla. ActiveX support might make Mozilla more useful for the other 95% though. It's been said before that ActiveHeX could be added as a plug-in, but it won't be in mozilla by default. <:3)~~ Um, maybe there is an activeX implementation for Unix, but i'm sure it's very raw. You won't be able to take a compiled binary for i386 and pop it over to a SPARC or anything. Also, if you are talking about MS COM ActiveX components, there's so much tied up in the win32 API, it would probably be a testing nightmare to check everything across platforms. Just stick with Java for cross-platform stuff. It'll save you a lot of work. Java won't save a lot of work because you have to deal with differencies among different implementations and versions of JVM. Ensuring that the user has a version of JVM that your stuff can work with is not easy. Using your own COM object is a lot more reliable. During a page download, I can push the right COM object, but I can't tell the user that he/she needs to update the JVM and come back to this page another time. I think you misunderstand the idea behind COM. You are not supposed to "take a compiled binary for i386 and pop it over to a SPARC or anything." COM provides a specification for an object interface. You are supposed to compile a binary for each platform you are going to deal with. COM allows you to rely on being able to call the same method on a COM object on different platforms. And i don't think you understand the idea behind Java. If you have a compiliant JVM it doesn't matter what OS you are running on, if you take a class from one JVM and it doesn't run in another JVM, it's the JVM's implementation that's the problem. And I know all about COM. Can you tell me where in the COM archetecture security providers is specified? How do you limit the access a particular piece of code that is the COM object? The origional poster said that ActiveX will lead to an assortment of security problems, and they are correct. I would definitely put a vote in to stop this 'feature'. -Chris Like a JavaBean, a COM component (and ActiveX components *ARE* COM components) is simply a package of logic. (XP)COM is just a modularization technique. Client-side / downloadable modules security has NOTHING (that's **NOTHING** again) to do with this. Signed archives and user permissions are one way to deal with this. If you believe code signing will prevent any sort of mishap from happening, you are the Ignoramus. What a joke: "Yes, by accepting this certificate, I trust that there is no bugs at all, none not a single line of code that will damage my local system". You think signing code ensures security? It ensures that the code hasn't been modified by the author, but big friggin deal, it still can do anything it wants on your local system, intentional or NOT. You wanna know why the majority of security holes with Java has been found in MS's implementation? J-Direct, baby! That's right, MS let java developers hook right into their activex technology, and look at the friggin mess that caused. If Mozilla wants to do that, they better be ready for people to find all sorts of problems. -Chris Both Java and COM are about building software from blocks("components"). Where that designers of the Java language decided Java-blocks would only connect to Java-blocks, the designers of COM make a choice for a much lower-level, machine-language like "connection" between "blocks": the virtual function call. These people wanted to solve a common set of problems, and they chose diffrent solutions. The discussion was about the functionality of both ideas. The approach to security is IMHO a different question(dilemma?): The ActiveX idea is: "We *do* allow binary code execution, what can we do to make it (as) safe (as possible)?" You seem to be unwilling to ponder a little over the second question(security). Millions of people use QT, Shockwave, Flash, QTVR, different other multimedia plugins and AX controls. My apologies for calling names, I take if back if you let me ;-) Throwing some solution away because if fails in *one* of your criteria, without looking at what it *does* offer, is ignorant imho. #155 Namecalling is inappropriate. Keep it civil.(n/t)by sacolcor Sunday February 20th, 2000 7:38 PM . Only on Win32, you said? Isn't support for it being developed on Linux and Unix (at least HP's version)? Actually, if Mozilla doesn't have an adequate support for ActiveX, it will actually make a job for some developers and testers a lot easier -- we'll just use IE as our UI host. <opinion> Support might be being developed for ActiveX on some other platforms, but I have my doubts that it will ever be widely deployed on them...it creates far too many security holes. Third-party implementations also create another layer of complexity for bugs to spring out of, since they usually differ slightly from the original. And aside from that, I don't think that it's a good use of time to try to support a technology that can be unilaterally changed by a single company. Mozilla's commitment is to support the public standards developed by the W3C...this does not include ActiveX. If MS were to turn control of ActiveX over to a standards body, and it was accepted, it might be worth another look...but I wouldn't hold my breath for that. IMHO, the developers that have committed to using ActiveX have already made the decision that they don't mind requiring their users to use IE, and are unlikely to be swayed by anything Mozilla does. In any case, I believe that as another poster pointed out, the decision has already been made not to support ActiveX in Mozilla. If a group of users really wants it, they can implement it via a plug-in, so that those of us who don't want it won't have to worry about it. </opinion> Why not a lightweight set of XPCOM interfaces to just interact with the DOM, call the XML parser, do HTTP I/O. What you said actually supports my arguments: 1) You said: "Third-party implementations also create another layer of complexity for bugs to spring out of, since they usually differ slightly from the original." Correct me if I am wrong, but JVM's for different platforms are implemented by different vendors. 2) You said: "If MS were to turn control of ActiveX over to a standards body, and it was accepted, it might be worth another look..." Sun hasn't done this with Java, but you support it. Why? With COM, you at least have control over the binary of the COM component. But you have no control over how JVM is implemented. I believe you are confusing my posts with someone else's, as I never mentioned Java. However, since you bring it up... Sun has made a number of mistakes with Java, the biggest of which is their insistance on maintaining control of it. That said, there are reasons that Java remains a superior option to ActiveX, at least for Mozilla's purposes. 1) Java was designed from the beginning with third party implementation in mind. There are very rigorous specifications and compatibility tests that are needed before a JVM is certified. (The fact that those tests aren't public is another of Sun's mistakes). ActiveX was designed to permit internet applications to make calls to the Win32 API. Third party implemtations and usability on other platforms are not issues that Microsoft places a very high priority on. Mozilla is intended as a multi-platform product; therefore Java is more appropriate for it to support. 2) ActiveX's security model is much weaker than Java's. ActiveX's signing mechanism protects against tampering, and can authenticate a program's author. It does not, however, provide any protection against bugs in the code, or intentionally malicious code. By requiring applets to explicitly request the privileges they desire, Java guarantees that it will be impossible for an applet to overstep its bounds. ActiveX died a long time ago - certainly as far as the internet is concerned. It may be clinging to a pale shadow of its intended existence in certain intranet environments, I suppose, but really it's a very niche technology. I definitely don't see any need for mozilla support. People using intranets will specify IE anyway if they want to develop with COM. IIRC Mozilla has a different system for web-application development (something to do with javascript and XUL, right?), best to stick with that. (Even though that's a flawed idea too, and the whole web-app thing is drastically exaggerated in importance because people seem to have forgotten that in many circumstances standalone apps are more appropriate, but anyway. :) --sam I think Yahoo uses ActiveX. I think Dialpad.com and AllAdvantage do also. Obviously MSN does. I could probably think of some other big Websites that use it if I cared. From http://www.dialpad.com/products/index.html "dialpad.com is the world's first free Java-based web-to-phone service". #193 Re: Re: Can't believe anyone's still using ActiveXby Tanyel Monday February 21st, 2000 12:16 PM Maybe I am wrong. What is this? http://sites.netscape.net/tanyel/dialpad.gif #248 Re: Can't believe anyone's still using ActiveXby sacolcor Wednesday February 23rd, 2000 8:11 AM Looks like a signed ActiveX component to me. Perhaps they're using both Java /and/ ActiveX? #249 Re: Re: Can't believe anyone's still using ActiveXby Tanyel Wednesday February 23rd, 2000 11:48 AM Yes, I believe that is what they do. I think Microsoft made an ActiveX support plugin so Netscape users could see their MSN news menu. Could that be used to provide ActiveX support in Mozilla? #258 Re: Can't believe anyone's still using ActiveXby sacolcor Wednesday February 23rd, 2000 3:19 PM I think so, in theory...the issues I'd be worried about are that: 1) MS would much rather convince people to move to IE than to devote any effort to improving the Netscape experience, so they're unlikely to do a really high-quality job on the plug-in. 2) Any third party plug-in would have to be reverse engineered, since the source code for ActiveX is closed. I would like to be able to use eight full bits of alpha channel translucency in PNG images. As programmers, we often have to composite foreground images with background when designers use edge antialiasing. This makes individual images hard to move or change, though. If real 8-bit translucency were activated, such images would be more easily reusable and movable. It would make for a fundamental improvement in the manner in which programmers translate the work of designers to this medium. I don't know if this is slated for the final release or not. It's been discussed at length in Bug #3013 and is apparently not considered a bug, but rather a feature request. It has been targeted for M20, but the way things are going I'd be surprised if there ever is an M20. If this is going to make it, it's got to be in by beta. Here's hoping. This bug is one prime example of how Mozilla has their priorities all out of order. 8-bit alpha channels would be a TREMENDOUS boon to developers and would dramatically improve the way websites are made. If you read through Bug #3013 you will see a big of ambivalence from certain Mozilla engineers as if they don't understand why this is important. Most of them, obviously, have never had to break a Photoshop file up into HTML. This is why I am getting more and more pissed off as I follow Mozilla's progress because so much attention is being paid to crap like XUL and Composer and meanwhile really important things like this bug or CSS2/DOM2 support are being cast to the wind. sad sad sad Your right...Mozilla needs to get the basics right first! I would think the order of operations for building the web (properly) would be something like HTML 4, CSS1, PNG, XML, CSS2, SVG, and XSL, etc. The order in which support for these things is added to Mozilla should reflect this. I mean...GIF is bad...the web needs to be built on truly open standards. Images are about as basic functionality to the web as you can get after HTML itself. Thus I would place a really high priority on full PNG support. Thanks for reminding me about this issue - imo it's a VERY important problem and I'll just go and add my mozilla bugvotes for it... (BTW, I also want SVG support, but for a first release I think it'll be acceptable if we just ship with Adobe's plugin.) --sam url is of the bug report : http://bugzilla.mozilla.org/show_bug.cgi?id=3013 url for the vote : http://bugzilla.mozilla.org/showvotes.cgi?voteon=3013 note: there were 59 votes when i added mine... Well, i haven't seen this feature in any browser, but i like it very much. The idea is to have kind of new type of bookmarks, which are stored in servers over the internet (may be yahoo, visto, netcenter, etc.) So if I move form one machine to another, i can request, use, and update my bookmarks in a easy way. Actually, when i go from work to home, and home to work, i need to send e-mails with new bookmarks i have found in order to have them in sync. May be, the same can be done with another mozilla setings, (skins, cookies, e-mail, etc.) So, what do you think? Best Regards... Another rare feature. Sometimes i would like to save a shoot of a web page, in order to print it later, or, tell to a friend to print it for me. This can be a valuable feature for people using devices without local storage (HD or Flash), as some of the internet appliances. May be, print to web-dav folders... Sorry if this is a non-sense feature, but i have been in this problem in the past, Best regards. See http://www.WebRoamer.com Roaming for Mozilla bug: http://bugzilla.mozilla.org/show_bug.cgi?id=17048 Being able to update the bookmarks through the Internet would be wonderful. There should be a choice between replacing the current bookmarks or merging the lists. My Netscape allows you to add a Bookmarks channel and then parse up to 100 bookmarks into this channel, for use from any Web-enabled computer that can let you log into your Netcenter account. Check out Backflip - http://www.backflip.com You can store all of your bookmarks from both Netscape and IE here, and maintain all of your bookmarks from this one place. This is possible in NS 4.x (don't know exactly from which version on) and is called Roaming user profiles. I use it constantly in a mixed environment of NT and Unix (both at home and at work) and it works great. N. I use mozilla with the sidebar set to display two panels. The first panel is my my.netscape web bookmarks. The second panel is my webmail allert and message count from my.netscape. This is a really siple concept and works rather well right now in mozilla. I've been working on exactly this project -- I've got it to the point where I can get it to display folders, but not the names for some wierd reason. The server's written in java. It connects to a database behind the scenes and sends out XML like bookmarks. I've written an extension datasource in RDF, and a protocol handler -- so right now, if you have a bookmark called "bookie://localhost" and you have the bookie datasource, it will contact the server and attempt to parse out the bookmarks. I ALMOST have it working. I'm so close... Will. Those are, by far, the most important things missing in the current Mozilla. Those who don't write a message saying that XSLT is the most wanted feature, don't know XSLT yet.. =) XSLT will change the web. It will make it lighter, more manageable, it will take search engines to a new level... I really do agree that XSLT support is important, and it would be great if it does make it into Mozilla's final release. (IMHO things like the rest of CSS2 are nice, but can wait for another release. Likewise MNG graphics.) On a completely unrelated note, does Moz illa support any form of URL for Internet Relay Chat? (It didn't last I checked.) http://www.w3.org/Addressing/schemes.html#irc That would be a good selling point for Moz and help make IRC a little less underground. Without the alpha channel, PNG files are not as good as JPEG files for 24-bit graphics and not nearly as good as GIF files for 8-bit graphics. Actually, I do not think full implementation of PNG support would make PNG files more practical than GIF files for 8-bit graphics, but we probably will not know until M329 when PNG support is finally finished. Actually, PNG is still better than GIF even for strictly 8-bit graphics. The gamma correction and 2-dimensional interlacing are great benefits. The fact, though, that GIF is still supported in more browsers than PNG is certainly a gating factor in PNG's adoption. You must be joking! PNG is smaller (better compression) and has more features which actually work than GIF. With working transparency it would eliminate tons of evil kludges used by web designers today. On the PNG-vs-JPEG issue I should perhaps remind everyone NOT TO USE JPEG FOR SCREENSHOTS. With few exceptions all screenshots will be smaller AND look better saved as PNG from a decent editor. JPEG 85% screenshots look like your font renderer broke. Well, i haven't seen this feature in any browser, but i like it very much. The idea is to have kind of new type of bookmarks, which are stored in servers over the internet (may be yahoo, visto, netcenter, etc.) So if I move form one machine to another, i can request, use, and update my bookmarks in a easy way. Actually, when i go from work to home, and home to work, i need to send e-mails with new bookmarks i have found in order to have them in sync. May be, the same can be done with another mozilla setings, (skins, cookies, e-mail, etc.) So, what do you think? Best Regards... With netscape 4.X you can easily configure mouse buttons. For instance, when I click the right button without moving the cursor, the browser goes back a page. If I move the mouse at all, I get the context menu. Also, I like the middle button being bound to "open link in new window." This would be a cool feature to include to add this: let us define the behavor for each button based on content of mouse location. ie if mouse is above text-box, 3rd botton past, but then if mouse is above a link, do a: open-link-in-new-window would be really cool to define this via XUL / chome... Clicking on a link while holding down the SHIFT button should open up a link in a new window. (This would match IE's behavior. How could it hurt?) The number one best feature of netscape 4.x is roaming access. I use it to keep my bookmarks in sync across my 2 computers at home and any computers that I choose to use around campus. I also use it to automatically keep my bookmarks uploaded and synced on my webpage so that I can access them from non-netscape browsers. Having the other config information in sync is also nice but the bookmarks syncing is where its at. I imagine that Mozilla could do an even more advanced implementation but anything would be nice. I can't really make the nightly builds into my default browser until such a feature exists. I think this is the very feature that someone earlier was asking for. Sharable bookmarks. I second the request for SSL. It is a must for me to be able to use Mozilla as anything more than a toy. SSL is 3rd party patented code by Verisign (i think) so they can't put it in mozilla. It will only be in the final Netscape release version of mozilla. SSL uses RSA (which is patented) and some symmetric protocols which are also patented. Depending on which SSL library (OpenSSL perhaps), you can compile a non-infringing shared library and link against that. If the server supports TLS (which is the "open" SSL standard), you could still access https and snews links. Then when the RSA patent expires, update libssl.so/libcrypto.so (or dll, etc.). Also, RSA is ONLY patented in the US. It is legal to use sans licensing fees everywhere else I know of (other laws may apply). The other way is to enable the secure proxy and punt the SSL to the proxy. This is how Lynx (the browser) does it. http://www.crl.com/%7Esubir/lynx.html, check the FAQ-o-matic under security. Posibility to save a page + all the pictures (like in IE5) I think the people at Netscape should be very concerned when people start saying "[ mozilla ] should be more like IE..." I have been using Mozilla since the beginning and have been singing its praises for YEARS now :-) (miaow) However, of all the things about Moz that I wish was fixed, it is caching. If I launch my browser and I have my prefs set to fetch a document only if it has changed then that is what it should do. If you select "view source" or "edit page" then Mozilla goes to the net for the page. Why? it should read it from the cache every time. Mozilla should also honour the settings in Preferences which state how how often the requested document in cache is compared to the network. It is unhelpful that current builds of Mozilla seem to relentlessly download pages evey time you launch the browser. This basic form of caching is further compounded by the distinct lack of viable offline caching. IE5 may be many things but you can't knock its implementation of offline browsing. If I want to study a ten page tutorial from Webreference or suchlike in offline mode, I have to use IE5 - period. even NN4 is too unreliable for this. As for Netcaster - ZZZzzzzzz... SOoooo, I want caching to work properly. Also, some people have Websites that generate random images. If I choose to save a file and Mozilla starts downloading all over again then it will probably save an image other than the one I was trying to save. I had the same problem with the old Netscape. Loading from a cache would fix this. Maybe the Mozilla Browser does not have the source in its cache; maybe the browser only caches what has already been parsed. Because of this, it has to go to net to get the source. Another possibility, the menu selections to View Source or Edit Source has not been updated to use the cache first. This is because the cache was not a priority in the early stages of developing the browser. It would be interesting to find out how the browser's cache really works. There.are.so.many.things.I.would.like.to.see in.mozilla.by.beta...but.cache.is.a.must!.This.baby.would.fly Wasn'y.Intel.supposed.to.be.doing.this?? Sorry.for.the...'s..seems.todays.build.has.a.small.glitch I'm pretty sure you can go to preferences, debug and put a checkmark in to enable the disk cache. Give it a shot and see. I believe that the way the bar to type urls work isn't satisfying. I mean, you can't use the shift+ctrl buttons to select all the text, it's really annoying and it makes it hard to type adresses as fast as with NN4. I've heard some talk of this, but I'd like to a better way to reject particular cookies. When prompted to accept a cookie I'd like the option to a) Accept it b) Reject it c) Accept all cookies from this server d) Reject all cookies from this server (or with this name from this server) I had heard some rumblings about a very fine grained security scheme for Mozilla in which you could accept/reject certain features based on the attributes of the page (e.g. don't accept window.open()'s, don't download images, etc.), but I haven't heard much about it since then. I can do something similar in IE5's "Security Zones", but it's not very fine grained (you only get 4 zones to play with - Internet, Intranet, Trusted, Restricted). Being able to disable the opening of new windows would make the Internet much less annoying. Netscape has always handled cookies better than internet Explorer, and improving that would only make things better. For instance: A window with each open connection, with it's status, bytes count, etc. Some way to retrieve the HTTP header information that was received with the object. The posibility to see the content of a cookie. HTML validation support, with reporting of errors ("tidy" integration? =) ). I disagree with the "possibility to see the content of a cookie". Most cookies are just ID strings used to key database information on the server side, and are therefore not human-readable. Being able to see cookie content would be useless. While the actual content of a cookie is pretty useless, the name of it is often useful. For example, ZDNet sets a cookie called useragent which I assume saves their backend software from having ot parse it out of the headers every time you pull a page. This is a pretty harmless cookie and is realy used in an appropriate manner, so I accept it. On the other hand, ad server cookies are used in a manner I disagree with, i.e. to track my surfing habits without asking me first. (The whole idea that you can set a cookie on an image URL is kind of silly anyhow, but that's a whole otehr story). I think that this is sneaky so i don't accept a cookie caled "adserverid", for example. My point is that the content of a cookie can be useful in determining whether or not you think its to your advantage to accept it or deny it. At any rate, from what I've read about the whole XUL interface it seems like it would be quite easy to show as much or as little info as you want on a dialog box. I agree...site based cookie handling would be a major plus. I'd also like to see site based banner ad blocking. I do agree they need to let you turn off window.opens. Some sites just start multiplying like bunnies and it ticks me off something fierce. Mozilla already has an ultra-enhanced version of IE's security zones, it just lacks a UI for it. You can create your own access zones, and specify sites which go into them, and set the levels of access sites in those groups have to ANY DOM function (e.g. window.open). This is already working. You have to edit your prefs.js file manually though to get it set up. This is what I was talking about. I knew I had heard about this feature. Do you have an URL for how to set these in prefs.js? Are there plans to create a GUI for this feature? I would like java implemented. It seems Java for Windows and the Mac may be working now. --But nothing for linux. I had to vote for Java even over SSL support. That was a hard choice. As much standards as possible. I hope MathML, XSLT, CSS 1&2, XHTML, HTML 4.01 (Everyone mentions 4.0, but I haven't heard any mention of 4.01). I really like that you can put folders in the toolbars in netscape. I hope this continues to be a feature rather than have the sidebar replace it. I'm a math major and a developer, so these are my main concerns. The more features I can use in both Netscape and IE the better. -nuff said I can sort of understand the reasons behind dropping native widgets, but I still wish that Mozilla had a Mac-like feel on my Mac boxes, a GNOME-like feel on my Linux box, etc. I don't think it would be terribly hard to create such themes for each platform, and doing so would probably both lower the learning curve and raise user counts. Doing this is an interesting problem. When we think about showing a user interface, certainly the baseline condition should be showing widgets and such in the native "skin" of the host OS. A technology like XUL, while great, should allow for something akin to "transparency" to allow attributes of the host OS to show through. So, if no custom UI elements were specified, the host OS would draw the interface itself. I don't believe that recreating the appearance of the various OS's as XUL and GIFs is wise. If the OS changes, the skin would have to be changed too. I'd really like to see roaming implemented. The whole LDAP or HTTP based roaming has been a strong force in putting up with Netscape todate. The most important bug in Bugzilla is in my opinion bug #14983 http://bugzilla.mozilla.org/show_bug.cgi?id=14983 . This bug deals with to do with Mozilla's partial CSS 2 support. It causes a web page with 100% valid HTML, 100% valid CSS to *not* work in Mozilla. The page will work in a browser which only supports HTML (and not CSS) (like Lynx), a browser which supports HTML + CSS 1 and in a browser which supports HTML + CSS 2, but Mozilla will break this page. (Not supporting the CSS property will similar to supporting 'color' but not 'background-color' ...) *Please* vote for this bug! I just read this bug report. I am suprised it has gotton so little attention. I understand the developers resource problems. But, I have to agree that this one should either be fixed or the partial CSS2 support that is breaking these pages should be removed. I think everyone interested in standard compliance should vote for this to be resolved in some way. Although I am not a fan of IE, I think one feature which they have which they do would be full screen browsing. For those with small monitors, it would make browsing much easier. I use full screen browsing, or very near it, even with my 19 inch monitor. The main reason I switched to IE is that I can squeeze everything I need into a single toolbar at the top of my screen and, with 19 big open inches... it's awesome! I'll only use Mozilla if it offers the same customizable flexibility in an INTUITIVE way -- no matter what else it does! For instance: A window with each open connection, with it's status, bytes count, etc. Some way to retrieve the HTTP header information that was received with the object. The posibility to see the content of a cookie. HTML validation support, with reporting of errors ("tidy" integration? =) ). PS: Sorry, my first post was a reply to a unrelated message.. =/ I don't know how much the average user would appreciate this feature, but I'd have to imagine than any serious web developer would really go for it. I've been meaning to actually get into the Mozilla code and see what I could hack up for this for some time. I'd like to see the ability to edit cookies and headers, too. Testing caching, session mechanisms, etc. is just a royal pain without this. Does anyone working on Mozilla already have some of this code lurking about? It seems like stuff that would be terribly handy to have while developing a browser. The think I'd like is a home button next to the Back/Forward/Reload/Stop buttons. I like to have Mozilla with *minimal* chrome (so no sidebar as default please - it's an utter waste of space !) and I'd switch off the personal toolbar and the very bottom toolbar (the one below the status line). These should be defaults - space is ultra-critical for most users: I'd like text-only alternatives for the Back/Forward/Reload/Stop too (the Reload graphic, in particular, is *not* obvious to me - it could mean back or home !). Don't want the "Search" button at all - another space-waster. It seems to me that the default chrome leaves you with only about 60% of the window area for browsing, which is poor IMHO. Yes, we *can* tweak all the options, but not all the users are going to bother doing that - witness the number of people who didn't know how to get rid of the sidebar for example ! > no sidebar as default please - it's an utter waste of space !) you can hide the sidebar. it won't show up anymore : go to the View menu and uncheck Sidebar. (if i remember correctly, because i'm not at home) Yes, I know you can hide the sidebar - what I'm saying is that the *default* should be to have the sidebar hidden. IE 5 does not show the History window by default (its closest equivalent) - you have to pro-actively select it. The same should apply to Mozilla - the first comment from newbies running Mozilla will be "the browsing area isn't very big is it ? How do I get rid of that pointless sidebar ?" I rarely use the History window in IE5, but it's nice to be able to bring it up occasionally, select what I want and then get rid of it because it's such an enormous space-waster like Mozilla's sidebar is. I bet you that's what 99% of IE users and probably what 99% of Mozilla users will do. Hence, my opinion is that the sidebar should be off when you first install Mozilla. many applications have a modal dialog at first launch in order to ask you a few questions like "do you want to see tooltips at every launch ?" maybe mozilla could ask "do you want to hide the sidebar ?" and so on... the first time you launch it. it would be nice and i guess it's not very difficult to implement, thanks to XUL/Javascript. As long as we have a sidebar we ought to make use of it. I think that the search and search results panes of M13 ought to be combined. I ought to be able to type in a string to search for in the search pane and then see the results in that same pane. I would then want to be able to click on one of the results and have it appear in the same browser window. Please give us at least the option to have new windows open BEHIND the current one. It would save so much time and annoyance when websurfing. It's the ONE feature I'm most hoping for! I think you must mean "new windows *not* initiated by a user click" surely ? In other words, JavaScript window-opening code would be what you'd want opening behind your current window. If I clicked on the middle button (assuming that'll work in Mozilla...) or selected "Open New Browser" or whatever with my mouse, I'd always want the new window in front and I cannot see the logic in putting a user-requested new window behind other windows - in fact, it would be a poor feature ! But yes, those annoying JavaScript pop-up ad windows that aren't directly requested by a user click would be most welcome put behind other windows :-) I think he means, so he can open several new windows from one. For example, he gets to a page of links and decides he wants to see a few of them. He clicks each link with the middle button, then can go back and forth between them without having to fiddle with the back button and re-click links. I'm not sure how many people would do something like this, however. If you are at a page reading a story with a few links, I often opens the links while I'm there (so I will not have to go back searching for them when I'm done reading), but still wanting to read the whole story through. The first thing that always happens is that the new window opens in front, and the first thing I have to do is to refront the one I was reading at. So I easily see why it would be a nice function :) We need to add the option to have new windows open behind to the preferences, but leave the default the way it is. That way most users won't ever have to know it is there, but those of us (me included) that want such a feature, can turn it on. Also, what about using the keyboard (e.g. holding down shift while clicking "open in new window" opens behind, otherwise, opens in front)? If a link would open a new window, an indication would be nice. If the script itself opens a new window, I would like it iconified or have a way to disable just those (or potentially create a link even if it messes up someone's formatting). Several sites have annoying popus as the home page which is the main navigation page, so if I get rid of them they come back. What if you could filter out automatic window openings from certain sites (like you might reject cookies from certain sites..) Thus supreme offenders like Geocities can be cut off before they even start. Pressing Enter in a form should submit the form's data. Pressing Enter in a form should submit the form's data. #98 XP-COM to support multiple programming languagesby danmorg Saturday February 19th, 2000 5:15 PM If this forum is only for the user features in the Mozilla Communicator, I apologize for posting about a programmer feature. I would like to see XP-COM to support many programming languages including standard, open, and proprietary ones. There is so much code out there written in so many different languages. It would be great to connect Mozilla to legacy code. When I say 'legacy', I don't mean mainframes running old COBOL programs, but any production program that 'works' no matter what language, development tools, language, operating system, or archietecture it was written. COM is too tight with Microsoft Winddows and CORBA is too complicated. XP-COM used in the Mozila Communicator is the answer.
I would like it if you click on mailto: link in the browser, a default mail handler would come up instead of built-in one. At the moment, it is kind of annoying because I have to use Lotus Notes as your email. Also, when I type in the Windows Run box: mailto:me@whatever.com, that only the Messenger come up, not the entire browser. But that's exactly how mailto is supposed to work! I usually see a lot of requests in Netscape newsgroups to call the default mailclient when clicking on mailto: instead of the built in. Actually that is why you defined it as a default, isn't it? And if I want to send a mail to somebody I don't want to open a browser window all the time. Otherwise I would have typed "mozilla". So I think this behaviour is pretty OK. A built-in tool to make different chromes would be cool. Sort of when games ship they include a utility to build game levels. May be not in the first release, but maybe in .1 release. This should be possible with the editor module. After all, Mozilla chrome (XUL) is just a type of XML. Here's one that no one has mentioned so far that's high on my list: File upload with FTP, both from the File menu and drag and drop like in N4. I know most people probably use command line FTP or some other FTP client, but I've gotten a lot of mileage out of that feature in N4. Also, students in HTML classes I have taught usually breathe an sigh of relief when I show them this feature in N4, after having demonstrated something as clunky and complicated as WS_FTP. Let's not take a step back; Netscape has had drag and drop FTP upload since version 2 (maybe before, for all I know)! Wow, I wasn't aware NS 4 had that FTP upload feature - pretty cool. But is there a way to rename/delete files? or do you have to use a regular FTP client for that? "...But is there a way to rename/delete files? or do you have to use a regular FTP client for that?" <br>No, N4 doesn't do those things, as far as I know. But I'd also like to see them added to Mozilla's FTP capabilities...delete, mkdir, and rename...Then I wouldn't even need another FTP client. But right now I'd settle for upload. Make sure it downloads files correctly. Many times I have had to re-download somethin using a non-graphical FTP client, because Netscape ruined the file on the way It seems that SSL is an absolutely necessary web standard that needs to be implemented. Yeah, this will probably be viewed as flame-bait by some, but here goes. I want to be able to open a Windows Explorer session, like double clicking on My Computer for example, and then type an URL in the Address location and have Mozilla display the page instead of IE. Here's hoping! You should be able to do that by mucking a bit in your windoze registry.. there's something about url or protocol handlers in there. Mozilla NEEDS a fullscreen mode comparable to IE's. I don't know how feasable getting rid of the taskbar is (if only IE can do that, because it's dealing with Windows code) but at the very least you should be able to get rid of everything but the taskbar. You almost can now by collapsing the toolbars and pressing Ctrl+Alt+S, but I would like to be able to do it with one button. n/t ... Intuitive enough for new users to be able to figure out on the first day!! I was referring to whether it's possible to hide the taskbar in win32, which the original poster queried (it is, it's easy), not as to whether Mozilla needs a fullscreen mode. [in which case i agree it should have one that you can access just as easily as IE.] Personally, it certainly isn't a priority for me, but I could see it being useful on occasions. --sam Type in "netscape -k" for kiosk mode or "netscape -sk" for super kiosk mode. I'd like to use the same profile (mail archives, history etc) on different platforms, just picking for instance a dir on my FAT32 drive for both systems (Windows and Linux). Netscape can do this, but it requires much effort and is not 100% reliable. So PLEASE, stick to the same file formats and filenames for every OS! This is another request that would be taken care of by roaming access. Even if the file formats were different roaming access could take care of making the proper conversions. For those of us who don't have LDAP or the ability to password directories on web servers, FTP support for roaming access would be great... For those of us who don't have LDAP or the ability to password directories on web servers, FTP support for roaming access would be great... For those of us who don't have LDAP or the ability to password directories on web servers, FTP support for roaming access would be great... Ok this isn't a "must have" but an idea in passing. What if the sidebar "snapped into place" whenever it reached a set interval. So when you were sliding it, it would be easy to find a "normal" sized window to browse with. ie. When the "browsing screen width" reached 1024, 800,640.. (of course there should be the option of turning the `snap to`off .. ' What an aweseome idea! And just to be complete, it'd be nice if the same feature would work for the vertical resizing, as well. Not sure I'd want this as my #1 feature, though. This would be very nice. Then it would be much easier to see if the Webpages look right in all of the different resolutions. Why do some web developers cling to the belief that most users browse in a maximized browser window? I only maximize my browser window when I am forced to by a short-sighted page designer (and then only very grudgingly). Lest you think this is because I have screen real estate galore, I should mention that I run at 800x600x24. The last time I checked, every Webpage I ever designed, for the public, looked fine at resolutions less than 640x480. They even look right on WebTV and those stupid little Windows CE calculators. The only platform that I was not ready for is the old Sega Dreamcast Web browser and I think Sega is not giving me much to work with anyway. Text files would not display right on that thing. First, many people do use a maximized browser window. I'm proving this with my completley scientific sample of looking aroung the computer lab right now. Maximized 12 Min 0,(we could argue chicken and egg forever but that doesn't really matter). Here's the reason I want the sidebar to "snap to". Primarily, it's for resolutions greater then 800X600. But the truth is that designers design for 640x480, 800x600 etc. That said, it makes sense to be able to easily find a 640x480 or 800x600 window. Whether it's full screen or not. Also what made me think of this suggestion was that i changed the size of the sidebar (Which 90% chance I won't use) and couldn't easily put it back into it's original position. This annoyed me :). Alright thats all. Hope I'm coherent. (Not you, but then you arn't forced to place the bar where it "Snaps in place" it just makes it easier to Similar to the way 'wget' works. You would be able to give a 'maximum link depth' and moz would download up to that depth for offline cache browsing. It would be even cooler to be able to download to a permanent cache area, so that it doesn't get wiped when you refresh the cache. a permanent cache would be fantastic: right click on an image, and have a link that says "store in permanent cache" (or call it some other fancy name). Then, with websites that you frequently visit, that always has the same graphics, you wouldn't have to redownload the same stuff each time. could speed things up nicely. If the site is frequently accessed, it should never drop out of the normal cache if you've got it to a good size. Every site can change it's graphics, so you should still check it hasn't changed, though you don't have to download the whole thing. One thing of benefit would be differing cache settings on a site-by-site or URL-by-URL basis, which is covered at bug #7380. Well, sometimes the cache gets full and you need to clear it. If you had downloaded a site or a page for offline browsing, you would then lose it when the cache was cleared. This is important for me as I live in the UK, and we still pay by the minute for internet access, so good offline browsing capabilities would be a real boon. So perhaps some kind of 'offline browsing cache' could do the trick...I don't know whether bug 7830 would cover this. I see all these M$ Users whinning about how cool their IE5 is. It sucks. Fortunately I have total immunity from using any M$ product in my contract :=). If possible I would like to be able to customise the shortcuts for most mouse clicks etc. Ctrl f x should open the location bar where I can type the location I want to visit etc. Ugh. Much as I love using emacs (I'm not being sarcastic, it's a great program), I really think Mozilla should stay away from the whole multiple-key shortcuts thing. It would confuse the heck out of any user not already familiar with emacs (just as I was confused when I first used emacs, "Why won't it save when I press Ctrl-S? #^#$^!!!") Mozilla should not require special training for use. I couldn't use emacs for anything better than basic text editing (no windows or frames, no special features while in c++ mode) until I got the O'Reilly book. He asked to be able to customize the key bindings, not for Mozilla and/or Netscape to set them to emacs-style as the default for a bunch of never-used-emacs windows users who would invariably be confused. Nothing run with customization. In Nav 4.5+ I can choose to view a two pane view rather than the 3 pain view. I could see how a 3pane view could be easier if you have a lot of newsgroups or mail folders, but some people don't have many folders. It's much easier for me to have the folders in combo box. Will the combo box be included in Mozilla Mail and News? I get the same thing no matter what that setting is at. I don't see any combo menu... Using the Win32 02-16 build. Mozilla needs a FullScreen Button as IE. And it should provide the ability to custimize the toolbar I thought I just saw Enable Disk Cache in the debug prefs...... I must be dreaming........ My must-have feature: I don't think Mozilla supports CSS2 dynamic fonts and OpenType (.eot, .eo1) format... although I haven't verified this. (If it does support it, I will jump around in glee, etc.) I think this is really important - the web has been waiting SO LONG for a proper embeddable font standard and it *still* hasn't emerged. (TrueDoc ain't it, and never will be. OpenType will be, as soon as it gets support.) Anyone know more about this? I'd be interested to see if there is a place I can vote for proper dynamic fonts [I couldn't find an appropriate bug in bugzilla and I did search for "font"...] I think this is probably the #1 most-wanted feature by web designers. Unless I'm wrong, IE actually supports this already, but if Netscape supports it too, life will suddenly become a bit nicer... specification link (this is part of @font-face): http://www.w3.org/TR/REC-CSS2/fonts.html#referencing Of course, if OpenType isn't an option, but we could still have TrueDoc specified by the CSS2 tag, then that would be not too bad since we could just specify both the .eot and the .pfr in the CSS2, however I am a cheapskate and don't want to buy the expensive tools you need for making .pfrs :) It's hardly in keeping with the "open source" philosophy... Is there anyone who keeps track with what's happening re Mozilla and fonts? I'd be interested in more information... --sam I would prefer to see TrueDoc implemented at least to the extent that it is in Nav 4 presently before worrying about OpenType. TrueDoc is the mature spec at the moment, with plenty of tools to support it. I agree that implementing CSS2 "@font-face" rule support would be nice, allowing us to remove those pesky LINKs from our HTML files. I would also like to see TrueDoc, unless something more generalized can be done in a reasonable time. I'm still not very experienced with TD and NS 4.7, but so far I'm disappointed at the lack of Linux support. I still haven't figured out yet if Mac support for TD is feasible. So, the upshot is that TD may be pretty limited (if it's just Windows), and it may even be better just to put the time into better Unicode support. We desperately need extensible cross-platform font support somehow. If TrueDoc is a "mature spec with plenty of tools", could you point us to them? (Seriously, maybe I missed them.) I'm particularly interested in a free or very cheap program for creating .pfr files. So far as I'm aware there are only two options, HexWeb (Mac only?) which costs $200 and some kind of "font wizard" from Bitstream which costs, er, $200. (Coincidence, or something more sinister?) As far as maturity goes, OpenType is supported to some extent in both major operating systems and is based on TrueType (which has been around a while) and Adobe Type 1 (which has been around a while longer)... and there is not only a completely free tool to make .eot fonts [from fonts that you own and have presumably already paid for, after all] but also a free (Windows-only) SDK for developers to use... Neither technology has been used to any extent on the Web. Basically, if TrueDoc is included at all in Mozilla, I would like to see it working via @font-face rules - that must be a relatively trivial change as it's syntactic and requires no extra functionality - but OpenType would really be a far better option at this point. (Even if only because it then creates a de facto standard, since the "other browser" supports it.) --sam I'm hoping that mozilla will have password protection for the user profile. This is important for people that share the same computer like the computer in my house. I do not know how Mozilla handles multiple profiles but the way netscape handles them is awful. It would be good if users could change profiles from a menu option rather than have a menu pop up every time somebody starts the browser. If the browser didn't ask for a profile name before it was opened, which profile would it start in? There is (at least in Windows) a way to have netscape start with a certain profile with a command line parameter. http://help.netscape.com/kb/consumer/19970621-1.html Is this possible with Mozilla? I would like to have an input-field for entering structured information, like html or xml in "wysiwyg" mode (styled with css). this would be very valuable. I dont know if this is possible by somehow embedding the editor in a form-field ? There has been some talk of supporting something like: <textarea type="text/html"> The infrastructure is all there, so it wouldn't be a difficult thing to add in. I think that Messenger should be integrated with the free Netscape WebMail service. I use WebMail and like it. Of course, I'm on a T3. I feel sorry for those people on 28.8 trying to use it. Mozilla should at LEAST be as easy to customize, and as flexible in doing so, as IE. So far, Mozilla hasn't measured up. With IE, I can fit everything I need on a single toolbar at the top of the screen if I want, whether in Full-Screen mode or not, and by merely dragging and dropping. In contrast, Mozilla is still rigidly restrictive. If this functionality exists, but I've not yet discovered it, then apparently it's not designed to be intuitive to the user. Get with it Mozilla -- or get left behind! XUL stands for eXtensible User-interface Language and is at the core of mozilla. It provides full customizability of not only toolbars, but also *any* part of the UI ! this is a lot more advanced than simple drag'n'drop of icons... yet, it's not ready for normal users (you have to edit XML and CSS files), but wait and you'll see: it rocks ! there's a good introduction to XUL written by a MZ reader for MZ readers :) http://www.mozillazine.org/talkback.html?article=1110
I, like many people, organize my bookmarks into folders. It's very tiresome to have to open the bookmark editor to move a bookmark to the desired folder every time I add a bookmark. It would be very nice to have an option for a window to pop up and ask me which folder it should be placed in. The most common comment I get when I show people Mozilla is that it doesn't have that bookmark window, like that other browser they use. Other than that, Mozilla is looking very cool, especially compared to 4.x in X :) Eric. There's a lot of in-depth dox there on mozilla.org, but it's sketchy in places & poorly organized. The better the dox are, the more developers there will be. Netscape for Windows has this I think but IE does it really well. Everyone knows you can press tab to move from widget to widget - in the browser window this means you can press Tab to jump from link to link. Press return and follow it ... It would be nice if the X/Gtk versions of Navigator were to have this as a feature. W3 - the emacs browser does it best of all in my view but it's missing a lot of other features ;-) you can already do that in moz, but you have to click into the web page window before you hit TAB. i guess this is a known bug (don't remember the number) I have tried Microsoft's Script Debugger for IE... It's not great but it works. I think a simular tool for mozilla would be great (and needed). I have tried Microsoft's Script Debugger for IE... It's not great but it works. I think a simular tool for mozilla would be great (and needed). I know this might seem pretty unimportant, but I just love the scrollbar in the Mac version of Communicator. There you have the buttons for scrolling up AND down right next to each other at the end of the scrollbar. For non-scrollbar mouse users, this very useful, instead of always moving all the way up or down the scrollbar to click that button. that's what the XUL technology provides. you can call it "skinnability". it will enable switching skins/chromes on the fly. read this nice article for a full description : http://www.mozillazine.org/talkback.html?article=1110
I've changed my mind. For God's sake, please just ship it as soon as possible. If I have to continue coding for that satanic abomination Navigator one day longer than necessary, I may slit my wrists. Navigator 4 just makes me want to quit the Web, it's so god-awfully bad. Do not worry. I have heard that Netscape 4.8 has many improved features. See for yourself: http://sites.netscape.net/tanyel/netscape2000.gif I would like to suggest to Chris Nelson to add one viewing mode (in addition to flat and threaded) to match what BetaNews.com discussions have. This way you could see the entire conversation, instead of endlessly clicking on threads. Feature Voting just like Bug Voting! Then we can give more of a democratic opinion what things we want the most! A feature to see the actual table boundaries (force border width to be nonzero and a visible color) and open a section in a new window there are too many pages which print as a thin column because of the junk on the sides. I would like to be able to extract rows, columns or cells in a new window. Someone mentioned this in one of the followups, but I'd like to emphasize this. I'd like to see a simple utility, like the N4 bookmark editor that lets you manage your cookies. You could mark cookies as good/keep, and then disable cookies so that only the ones you marked as good/keep are used. This would allow you to nuke the doubleclick lunacy, but still keep your slashdot (or other useful) cookies. That kind of thing. Hi I hope it is not too late to answer, because there is a feature I really really want to have in Mozilla: The possibility to reject all requests to some given site, or to some URL. For example, I don't want my browser to display the add banners from doubleclick.net. So all requests to *.doubleclick.net should not be send. I think such feature would be very great, for people who don't want to (or just can not) install junkbuster (http://www.waldherr.org/junkbuster/). It will make the internet more faster and cleaner!!! Michaël . I really liked the way that bookmarks were stored in a HTML file in NS4. I think that adding an Import/Export option to the bookmark editor so that you can import your NS4 bookmarks, and also export them as an NS4 HTML file is a must for many people upgrading from NS4 A feature to see the actual table boundaries (force border width to be nonzero and a visible color) and open a section in a new window there are too many pages which print as a thin column because of the junk on the sides. I would like to be able to extract rows, columns or cells in a new window. See http://bugzilla.mozilla.org/show_bug.cgi?id=24684 I am sad that the bug is closed; I don't know how style sheets would fix this. This one would make a dramatic browser improvement for me and our end-users :) There needs to be a QUICK (on a toolbar) way to nuke features on a per site/part of site basis. A quick list would be: JavaScript Java fonts & font resizing colors background pics opening new windows cookies blink graphics How many times have you been to a site that might have useful content if you could only read it....... but because the idiot who built the site thought putting three small tags in a row would be a good idea, or put purple text on a black field, or the whack-a-mole popups are so annoying you can't stand to go to the site, or the JavaScript crashes the browser, etc.... |