Adam Hauner asks, "Will be crash analysis data of 1.7b available again on ftp://ftp.mozilla.org/pub/mozilla.org/data/crash-data/?" Adam, we'll be doing pretty well if we can just collect some data and do preliminary analysis and bug reporting of the top crashers in time for some fixing within the 1.7 cycle. Generating clean reports for publishing on FTP would certainly be a lot lower on my priority list than getting the servers working and the basic reporting hooked up so that we can quickly get some bugs filed and fixed. Maybe in 1.8 or 1.9, but it's just not a priority as I see it.
René Pronk asks, "How many e-mails do people like David Baron and Boris Zbarsky get on an average day? What is the all time record most-emails-on-a-single-day of a single member of the Foundation?" René, that's probably a good question for Boris or David. You might be interested in the BZ's blog post where he did some analysis of time spent responding to bugs and review requests. If you want a real answer rather than my dodge, you could pose the question at bz's blog or dbaron's blog.
José Jeria wonders, "I see that mozilla drivers (i think, choffman for example) are setting the flag to block the next release sometimes. And then when no patch arrived, they unset the flag. Does the person that set the "+" flag then expect that sombody voluntary sees it and writes a patch? Its confused me a bit when I see the "nobody came up with a pacth" comment and then unsets the flag. Shouldnt drivers when setting the "+" flag assign the bug to somebody at mozilla.org?" José, I don't see the "+" flags getting unset. I see some of them get set to "-" if that's what you mean. The plus flag is a notation drivers use to mark a bug that they feel ought to be addressed for that particular release. Drivers work with our development community to try to get resources applied to those bugs. Because Mozilla depends heavily on volunteer contributors (only 5 or 6 of the nearly 80 people that landed code in 1.7 so far are employed by the Mozilla Foundation,) we don't always have the resources to get fixes for all the bugs we'd like fixed. We re-evaluate our plussed list regularly as we get closer to the release and some of them are inevitably set to minus meaning we wouldn't hold the release for those bugs.
Bogdan Stroe says, "There used to be a page called 'Upcoming landings'. I've seen the presentation about the future of Mozilla, but it would be good if the near future would be more predictable. So, are there any plans to restore that page?" Bogdan, that's a good question. The branch landing tool, which listed major upcoming landings, was used to make sure that we didn't land really big or scary changes haphazardly, without sufficient testing or coordination with other large landings. This was particularly useful when we were making lots of significant feature additions and rewriting large hunks of code. With fewer of these major changes, better communication between our developers, and improved tools within Bugzilla (like the flags system), we've been able get by without the branch landing tool. I'd actually like to see Bugzilla incorporate a bit more of this scheduling and coordination functionality because using a separate tool just never really worked very well without a lot of pushing from "management".
Jason Barnabe asks, "Where is the bottleneck in the project right now that prevents faster growth? Bug triaging, patch-making, reviewing...?" Jason, I'm not sure that "faster growth" is really desirable. We could certainly use more help in all of those areas but I'd rather see us making smart, targeted progress on simplifying and shrinking our codebase and featureset. Bottlenecks to faster (feature?) growth don't bother me much.
Andrew Wooldridge says, "Asa, after poking around a bit I found a nice tutorial on how to create a mozilla extension http://www.mozilla.org/docs/tutorials/tinderstatus/ but I dont see anything like that for creating a mozilla theme. Do you know of something like that? Are there any plans to create something like a developer.mozilla.org to have stuff like this?" There are certainly plans to improve code documentation at mozilla.org. Brendan spent some time talking about plans for a msdn library-like site during his presentation at the recent developer day, most directly related to your question, he notes the need for tutorials and called on the community to contribute to this documentation effort. Outside of mozilla.org, there have been several efforts, including the Theme development section of the mozillaZine Knowledge Base, the Creating a Mozilla Skin doc, the Basic Theme Project, the IBM Developerworks tutorial, and inside of mozilla.org, there's not much that I know of, though you'll probably remember our stalled MAD courses which would be nice to get going again. I think that extension and theme development tutorials and information would be really nice to get nailed down and then kept up to date at developer.mozilla.org or the mozillaZine KB.
mcsmurf asks "to comment 1: is Talkback working again (can't try it since i build myself)?" Yes, Talkback is working again. We still have a ways to go before we have all of the reporting set up and we still don't have it in the nightly builds, but we're making progress.
Pat Dunn asks, "My question is what is the future of the suite?" Pat, we continue to maintain the Mozilla application suite. From a UI and feature standpoint, it's on a, in Brendan's words, "sustaining engineering footing and should not be subject to radical change" so you'll see improvements to core code, with minor cleanup in UI and features, but probably (hopefully, as I see it) nothing earth-shaking in terms of non-gecko feature development.
Michael Ward says "You probably aren't the best person to ask this...but...I'm in my first year of college working towards a computer science major...I'm wondring how "good" you have to be to program something like mozilla?" Michael, since I'm not a programmer, you're right that it's a bit difficult for me to answer this. But when have I ever let that stop me ;-) Anyone that's been around the project for very long would probably agree that there are a few really strong hackers, a larger number of "OK" hackers, and even a few not so good hackers contributing to Mozilla. I've contributed a few patches over the years and I'd say that I'm definitely in the "not so good" category :-)
Paul Jankura says, "I'm a second year computer science student currently in a usability engineering course. Looking at the differences between Seamonkey and Firefox, Firefox looks like a case study in cleaning up a user interface (simplfying the menus, moving options from the pull downs to the Options box, adding customizability, etc). I was wondering who made these decisions and what were some of the specific goals and logic behind them?" Paul, I'd say that most of the credit goes to the three or four engineers that wrote the bulk of Firefox (first m/b, then Phoenix, then Firebird, now Firefox). The short answer is that Ben Goodger, Blake Ross, and Dave Hyatt had a lot of experience doing browser development on the old Mozilla application suite and they took everything they learned on that project and, without any of the various marketing, management, scheduling, etc. pressures, they set out to make a great web browser that wasn't going to try to be all things to all people. The goals have changed some over the years but I think that the original README for the project answers this question better than I can.
- CVS access is restricted to a very small team. We'll grow as needed, based on reputation and meritorious hacks to mozilla/browser.
- This will be a single process for the browser only. Mail clients, web editors, etc, will be out-of-process. Hooks for other apps will be provided eventually, although that is not an immediate goal.
- No profile manager UI on startup, although you can still select multiple profiles from the command line.
- The default theme will be Classic. Additional themes will be supported but will not be part of mozilla/browser.
- The toolbar(s) will be configurable. That includes moving the location bar where the user wants it (not just splitting it so it takes a whole toolbar width).
- The personal toolbar is the personal toolbar, not the whorebar.
- All wallet-like functionality will be rewritten from scratch.
- We will have a sidebar, but it may work differently from Mozilla's current one.
- There won't be 239 access points for Search and for Bookmarks!
- We may drop the throbber.
- The interface will not be "geeky" nor will it have a "hacker-focus". Nor will it be "minimal". The idea is to design the best web browser for most people. (This doesn't mean every feature has to be enabled by default.)
Tom Graham asks, "If you had unlimited resources, what element of mozilla would you invest in to improve and develop more?" Tom, I personally would favor investing in distribution. With unlimited money, I'd work out deals to see Mozilla applications shipping on every new computer. I'd mail out Mozilla and Firefox shaped USB keychains with the apps ready to run. I'd push for a massive internet and traditional media advertising campaign. I'd just go all out on getting Mozilla onto every single desktops because that's an area we must improve if we're going to stay relevant.
Jeff Walden asks, "When's the next BugDay? I have a feeling there either wasn't one this week (due to 1.7b) or I simply missed it." Jeff, every Tuesday is BugDay! We may be adding a second day or partial day (maybe a Friday or Saturday) for people that can't make Tuesdays. This last Tuesday, I was, indeed, overrun with release activities so I wasn't as active on IRC as usual, but I was there and so were a half dozen or so others who regularly help out on BugDays. Every Tuesday! Jeff went on to ask, "How does one accurately determine the number of bugs one has resolved (through a Bugzilla query, most likely)?" A query in the boolean charts for "status changed to resolved" AND "status changed by asa@mozilla.org" would give a list of the bugs I've resolved.
Samual Icky has a few questions:
"When will we see Pinstripe for Mozilla Thunderbird? Kevin mentioned in his blog we that should expect to see it late to mid-February." Not sure. You'll have to ask Kevin that question. I suspect that rsn is a reasonable guess :-)
"Will we ever see Sunbird?" You can see it today.
"Will any of the remaining Classic Mac OS bugs ever get closed. I know some one started that a few weeks ago but it appears the person stopped for one reason or another." Sure. We're workin' on it. There aren't that many left. If you'd like to test any of the remaining bugs and resolve them if they don't impact OS X or update the OS field if they do, that'd be great.
"Do you think the Mozilla Suite or Mozilla Thunderbird will ever have access to Mac OS X Address book? Alternatively, will the bird or suite ever support the Keychain?" Ever? Yes.
Marc Randolph asks, "do you know of a true streaming rss reader that fits in a Mozilla browser? I'm thinking of something that is exactly like what's shown at http://www.linkston.com/ ... except I want in my Firebi^H^Hfox sidebar." Marc, I don't know of any RSS extensions that currently allow you to aggregate feeds into a single list. RSS Reader Panel and Forumzilla are the two XUL feed readers I use. I think viewing all feeds in a single sortable list would be a nice feature addition for both of those apps. You might check out NewsMonster and Aggreg8 to see if either of those offer something closer to what you're looking for. I don't think they do but I haven't used them in a while either so I could be wrong.
Blonker got in just under the wire with this question: "I don't think this has been asked before. What do you think about the "UFO" Spirit saw this week? Reference info: http://news.bbc.co.uk/2/hi/science/nature/3520636.stm
Well, it certainly is an unidentified flying object. Unless we get a repeat, I don't think we'll ever have a definitive identification but I agree with all of the serious reporting, that it's very probably a meteorite or a man-made Martian satellite. My money's on it being a meteorite, but I wouldn't be surprised to find out that it was Viking 2.
Well, that concludes another installment of Ask Asa. Feel free to ask follow-up questions, to discuss or add to my responses, or to pose new questions. I'll look to the comments at this post for questions for the next installment.
Posted by asa at March 21, 2004 10:44 PM
Thank you very much for your answer Asa, I really appreciate that you take your time to answer our questions :-)
Posted by: José Jeria on March 22, 2004 12:28 AMReferring to Sam Icky's questions on Mac OS integration: Keychain and address book address would be something I would expect from Camino first. My question: When is the next Camino release planned?
A workaround for the address book would be: Both, Apple and Mozilla address book can access ldap servers. You can install Openldap with fink and install set up an ldap server, which then could be accessed by both address books.
Posted by: daniel. on March 22, 2004 12:39 AMAs you might know, I'm a programmer and not only a document writer. The reason why I have never tried hacking in Mozilla is because I find the source code overwhelming and even the UI code a bit too hard to grasp. I'm used to an integrated development environment like Microsoft Visual Studio, where you can debug, test and develop code with ease. What tools does experienced developers like Ben and Boris use when hacking Mozilla? Where would I start when trying to understand the code better (especially UI/features code, not Gecko)?
Posted by: David Tenser on March 22, 2004 12:51 AMPinstripe Thunderbird is > < this close :) I just have to get it reviewed to make sure my code changes don't break Qute.
Posted by: Kevin on March 22, 2004 08:18 AMIn my question, by "faster growth", I didn't mean code size or feature set, I meant market share caused by being a better browser. Rephrasing the question:
Where is the bottleneck in the project that prevents Mozilla from being better more quickly?
Posted by: Jason Barnabe on March 22, 2004 09:22 AMAs someone who has been invited to be DevMo driver, let's comment on the themes documentation.
Brendan has made the iron fist rule that the first target audience of deveoper.mozilla.org would be extension developers. We haven't talked about themes editing at all yet, but I'd expect that to end up lower in priority. That doesn't exclude things, if someone finds a decent tutorial and someone to maintain it, d.m.o would prolly like to host it. I don't see us having time to hunt for something like that, though.
(This is a 1-out-of-5-or-so personal opinion)
You said: A query in the boolean charts for "status changed to resolved" AND "status changed by asa@mozilla.org" would give a list of the bugs I've resolved.
As I understand it, this isn't exactly the case. Indeed, this will show the set of bugs whose status you've changed, intersected with the set of bugs whose status has been changed to resolved. However, if you confirmed a bug (i.e., status UNCO → NEW), but someone else resolved it, that bug would show up on the list.
Is there any way to get Bugzilla to be more picky about these things? (E.g., to more accurately get those bugs whose status you changed to resolved.)
Posted by: louis bennett on March 22, 2004 03:41 PMAnother announcement? What would it be now?
http://www.space.com/scienceastronomy/opportunity_announcement_040322.html
Posted by: alfons on March 22, 2004 03:48 PMLouis, the query I described should give you what you want.
Posted by: Asa Dotzler on March 22, 2004 08:43 PMDavid Tenser:
>> Where would I start when trying to understand the code better (especially UI/features code, not Gecko)?
Sounds like you mean XUL/Javascript/CSS?
I personally would think the best place to start out would be with extensions. Once you can hack/make an extension, you've got some of the best first-hand experience on mozilla hacking. From their understanding the original code that you were overlaying with the said extensions, becomes a lot easier. Just my 2 cents.
-Jed
Louis, I neglected to say _why_ it works. Using the _AND_ with the boolean charts (in most cases) binds the two conditions to a single instance of that field. You can use the same technique to query something like "comment contains foopy AND comment changed by asa". This will turn up bugs where I said foopy, and not bugs where I made some not-foopy comment and someone else commented foopy.
This doesn't always work as you'd expect but it's pretty good for a number of these kinds of queries.
I double-checked with a direct SQL query and compared with the resolved query I proposed and both say 91 bugs resolved.
Posted by: Asa Dotzler on March 22, 2004 09:21 PMSorry... my bad. I'd done "add another boolean chart" instead. Thanks for double-checking. (-:
Posted by: louis bennett on March 23, 2004 07:21 AMYou say 5 or 6 of the people contributing patches for 1.7 were employed by the Mozilla Foundation. How many more are employed by IBM or Sun or another company supporting open source?
Posted by: Guy on March 23, 2004 04:34 PMYou considered running for Govenator. Why not president? You could get Mozilla installed in the Whitehouse. Perhaps an XUL interface for a "Bin Laden Finder". Or even better, tabbed browsing could increase efficiency in the capitol building from 0% to 0.5%. So what do you say Asa? Will you give me a reason to register to vote in the near future?
Posted by: Robert Accettura on March 23, 2004 09:45 PM"We may drop the throbber."
No! Say it isn't so! Tradition, man, tradition! ;-)
... and it is handy to see when a page is loading or finished...
Posted by: Jeff Wilkinson on March 24, 2004 08:49 AMWho is that gabrielle *g*? He (she?) always joins with the hostmask of the mozilla foundation office on irc.
Posted by: mcsmurf on March 25, 2004 11:21 AMDoes David Hyatt still contribute patches occasionally, or has he been fully enveloped by the Apple Safari? :)
Posted by: Albert on March 26, 2004 07:55 PM