CNET Review of Mozilla 1.0 Release Candidate 2Monday May 20th, 2002Ezh pointed us towards a review of RC2 over at CNET. We've been sitting on this for a few days in the hope of getting some more reviews to tell you about but no more seem to have come in. Tell us if you find any. CNET's latest review, like the RC1 piece, is generally positive. Reviewer Rex Baldazo likes the fact that Java is easier to set up and approves of the new option to disable cookies in Mail & Newsgroups. However, he still hasn't read the bit in the Release Notes that says not to use the same profile with Netscape and Mozilla. The review stated that there will be a more extensive review of mozilla when 1.0 is released. Is there some way to insure certian important facts about Mozilla (such as not to use the same profile for both Mozilla and Netscape, information easily gained by reading the release notes) reaches the reviewers? Is there an offical packaage for reviewers of Mozilla sent out to potential reviewers? If not, such a package would be a good idea. Overall a favourable review, but slipups such as a known incompatiblity in sharing a profile between Netscape and Mozilla, shouldn't put a black mark on the reviews of Mozilla 1.0. Why does mozilla have to share the netscape profile. Couldn't some of these problems be avoided by making a COPY of an existing Netscape profile when Mozilla is installed? Likewise for Netscape if it is installed after Mozilla. Maybe have an option to share the profile, but by default shouldn't we be copying the profile. Afterall, Netscape is not Mozilla, and there are other browsers based on Mozilla, do they all HAVE to share a profile, could create bigger problems in the longer term. Beonex uses its own profile directory (/applications data/beonex). Netscape should do the same and NOT use the "/mozilla" profile directory. Mozilla should actually make it a policy to discourage this. Of course, if any mozilla-based browser finds another mozilla-based browser's profile, it should offer to import (not use) the other profile. I consider this to be a definete source of potential and serious problems and misunderstanding among users, and therefore a serious BUG. Why does mozilla have to share the netscape profile. Couldn't some of these problems be avoided by making a COPY of an existing Netscape profile when Mozilla is installed? Likewise for Netscape if it is installed after Mozilla. Maybe have an option to share the profile, but by default shouldn't we be copying the profile. Afterall, Netscape is not Mozilla, and there are other browsers based on Mozilla, do they all HAVE to share a profile, could create bigger problems in the longer term. "Is there some way to insure certian important facts about Mozilla (such as not to use the same profile for both Mozilla and Netscape, information easily gained by reading the release notes) reaches the reviewers?" Well, I just sent a message back using their feedback form, gently pointing out the release notes URL, and quoting the paragraph that says not to do that. "We did the thing they said not to do and it broke. Goodness!" *ahem* Even if they do consider it a major bug, it would be appropriate that the release notes did say "DON'T DO THIS CRAZY THING, YOU FOOL!" It's probably too late in the game now, but anyone ever thought of a small helper app before installation? That would be useful. Totally cleans out all prior mozilla/netscape garbage so the user has a nice fresh install. I don't know how many times I've upgraded Mozilla on top of itself (in both windows and linux) and ended up with some borked chrome window. Those of us that actually read the release notes know better, and do it because we're lazy, but if cnet's reporter doesn't bother and just expects it to "work out of the box", then what makes us think that the general masses will read the release notes. Imagine some poor sap somewhere that left ns6.0 on there and then tries mozilla1.0. What will happen? Maybe taking the typical warnings in the release notes and making them bigger/bolder would help out alot. Let the author know that there is an explicit notice in the release notes to not share between Moz and Netscape. He'll probably say something like 'this is a little disappointing' but he's saying that anyway, at least if we all email him (nicely...) we can get the facts across so he can get the facts to the readers. Having that note in the release notes is better than nothing, but the fact that Netscape and Mozilla use the same profile naming scheme but cannot share the profile is just an ugly bug if you ask me. Mozilla and Netscape should be smart enough to identify which version of a gecko based browser has created the profile and either reject it or convert it if it is incompatible. Honestly - does anybody really expect a significant amount of users to read 17 pages of fine print? The fact that Beonex http://www.beonex.com/ uses its own profile directory is a good thing. It solves all the problems that are associated with sharing profiles. Also as Beonex, Mozilla and Netscape 6/7 are distinct apps (although based on the same codebase) changing the preferences in one shouldn't affect the preferences of another. As Mozilla is not for end users, blah, blah, blah - we can make excuses for mozilla not playing nice with Netscape 6 browsers, but Netscape which is an end user product - they really should have used separate profiles just like Beonex do. It makes Netscape look bad if their browser breaks because someone out of curiousity installs Mozilla and then they find out that their copy of Netscape 6.x has broken somehow. The best way to solve this problem would be to move Netscape profiles to \Application Data\Netscape 7\ (or ~/.netscape7 on UNIX) and offer to migrate an existing Netscape 6.x/Mozilla profile if it if more recent than any Netscape 4.x profile on the machine. Alternatively, let Netscape keep \Application Data\Mozilla\ so that old Netscape 6.x profiles don't need reimporting and move the Mozilla.org releases to use \Application Data\Mozilla.org\ (~/.mozilla.org) as Mozilla has not reached 1.0 yet and has warnings plastered over the download site then people using Mozilla should be less surprised if they had to copy over their profile data manually (although an import feature should be written) There are many different issues involved here: 1) to what extend should profiles from different gecko-based applications be sharable at all? That means: to what extend should the program allow to use such a profile of a different app, provided the user explicitly wishes to do that. I think that profiles of apps that are based on essentially the same build might be reusable and it might be good if the user has the ability to share. However 2) Sharing should *never* be the default. That means, each app should let the user choose whether to create a new profile, upgrade/copy and existing one or share an existing one. Note that sharing also raises the problem of 3) Concurrency: several gecko-based apps or even several instances of the same one must be protected from sharing a profile and must use identical/compatible protocols to do so. 4) This all is connected to one major problem assiciated with profiles: currently it is impossible to specify the location of a profile at startup. The ability to do this would be part of a solution for both deliberately sharing and deliberately not sharing profiles ... Well the mailbox folders are all stored in a standard mbox format. The preferences defining the server info though, is not nessesarly standard. But being able to share an mbox would be possible. Other than that, the bookmarks file is pretty much standard, but not completly. A standard address book file would be nice, but I don't know if it is yet, I don't think so. With 1.0 coming I would be a good idea to have a link from the releases pages to a "Reviewers guide to Mozilla" ? He's also missed the third paragraph in the same section of the release notes as the profile bit http://www.mozilla.org/releases/mozilla1.0/#prefs He complains that the Cookie Manager window has no scroll bar, and he has to use the keyboard's up/down arrows to scroll through, when all he has to do is grab the edge of the window and make it larger. To those of us who've lived through far worse problems with Moz's chrome we know that the default size of some windows is wrong and have resized them months or years ago, but really this is one of those fairly inexcusable chrome polish things that the average Joe user (or reviewer) has no idea about. After all he expects scrollbars to appear when a window is too small, not disappear! "Need help? Do it yourself Mozilla's tech support hasn't changed either. Since Mozilla is freeware, it has no huge support department to help you debug problems. Mozilla.org hosts some newsgroups, but they're aimed at developers, not end users, so you're pretty much on your own. This Netscape help group offers the best Mozilla help we've found." This sentence seems to contradict itself. If an end-user is looking at 'debugging' problems, then these problem can be reported using Bugzilla, and one can use the newsgroups, MozillaZine and IRC to get in touch with developers/users to help you nail down the bug. When he says 'debugging problems' he doesn't mean actual browser bugs, he means things like 'how can I import my mail', 'why does this page look wrong', 'why doesn't this browser work on my 386sx 4 MB windows 3.1', etc. He's talking about technical support (of which there is none), not browser bug reporting (of which there is plenty). The last article used the same phrase. I don't have any quarrel with it, although I'm unsure exactly how much technical support the other companies actually provide for their browsers (Microsoft, Opera... or for that matter Netscape). I doubt any of those companies really offer free telephone support for their browsers. --sam "Need help? Do it yourself Mozilla's tech support hasn't changed either. Since Mozilla is freeware, it has no huge support department to help you debug problems. Mozilla.org hosts some newsgroups, but they're aimed at developers, not end users, so you're pretty much on your own. This Netscape help group offers the best Mozilla help we've found." This sentence seems to contradict itself. If an end-user is looking at 'debugging' problems, then these problem can be reported using Bugzilla, and one can use the newsgroups, MozillaZine and IRC to get in touch with developers/users to help you nail down the bug. > However, he still hasn't read the bit in the Release Notes that says not to use the same profile with Netscape and Mozilla. Yes, as we all know, Users Don't Read Documentation. Where's the news? Agreed. 'Anybody who used netscape 6 is potentially screwed unless they do something complicated', as a releasenote that nobody will read, isn't IMO really very good for a 1.0 release. At *worst*, surely it must be possible to detect that there's an existing profile and warn that you need to do something. I mean, even if it can't be fixed, at least it could tell you that you really do need to go read the releasenote. Oh well, I hope some workaround is done, though looking at the linked bug (btw the way releasenotes are linked to bugs *is* good), it doesn't seem too likely. At least - I presume - this kind of thing will be sorted out in one way or other for the netscape 7 release. --sam The funny thing is .... The release note get very long some of the time and when our eyes get tired, we start to overlook it or we start to quit reading the rest of it. We just fire up the gecko browser and use it. Then we forget about the problem and when we see bug or problem, we complain to the developer and so on. Isn't it amazing about how much of a short term memory do we have. Now, think about the developers and gecko people with short term memory too. FletchSOD The funny thing is .... The release note get very long some of the time and when our eyes get tired, we start to overlook it or we start to quit reading the rest of it. We just fire up the gecko browser and use it. Then we forget about the problem and when we see bug or problem, we complain to the developer and so on. Isn't it amazing about how much of a short term memory do we have. Now, think about the developers and gecko people with short term memory too. FletchSOD Why is Classic still the default, since it's obviously inferior to Modern (e.g. try using more than a couple tabs -- bug 109607)? If you look at the screenshot (link at the top right), they're using Classic. It's unfortunate if C|Net readers associate Mozilla with that mess. Tried using modern at work? I got sick of people asking what the heck kind of program I was running. I know this too well....but as I'm using Linux and running Windows (for testing sites in IE) in a vmware window they ask anyway so that doesn't make any difference ;) A couple of days ago I read a review of OpenOffice.org linked to from an OpenOffice.org mailing list where the reviewer made a complaint and said he's filed a bug about it. I was interested to see how the community had reacted so I tracked down the bug and was surprised to find that there wasn't much reaction and it dawned on me that nobody had thought: review->bug->fix->better next review. We are in an open community where reviewers can drop in at any time but we should take special note when reviewers take issue with something, especially when they file bugs that we treat them well so that mozilla is seen in a great light. Just a thought, David Illsley I wonder how important it is to mozilla.org to make a good impression. Reviewers will probably browse through it. There is still a lot of outdated links and info there, e.g. the roadmap still doesnt indicate that RC2 was released May 10th, there are still links to the long non-existant mynetscape channels in "Newsbot" etc. Mozilla already has somewhat of an image that it was cooked up by a bunch of hackers and I think the mozilla.org site reinforces that impression. For somebody with CVS access, putting the release date into the roadmap will take a maximum a 10 seconds, and this doesnt seem to be possible in more than a week? |