MozillaZine

Full Article Attached Mozilla Bug Week

Monday October 22nd, 2001

Many of you are familiar with the Bug Day's that we started a few years ago. Well, with the number of people wanting to help out skyrocketing these last few months, gerv@mozilla.org and others have decided that a bug week was in order. They'll be running it from Saturday October 27th to Sunday November 4th, and they'll have plenty of smart people on hand to help folks learn the bug system, learn how to use the various other web tools, and of course, learn some tips on how to contribute code to the Mozilla effort. Click the Full Article link to get all the details.


#78 Re: Re: Learning from Mozilla

by strauss

Friday October 26th, 2001 11:44 AM

You are replying to this message

> "you need more professional accountability" is, on its own, a meaningless statement. <

It means quite a lot. As I said before, I'm sorry to learn that those words have no meaning for you.

> Your list of accusations are simply wrong. You will, of course, put this down to "defensiveness" and "denying the truth" and "playing fast and loose with the facts", but I think your baseless accusations are far more indicative of your attitude than our problems. <

You're denying defensiveness in the course of being defensive.

I could point out numerous recent examples of you playing fast and loose with the facts -- for instance, denying on /. that mozilla.org is at all dependent on Netscape, then coming back here within a few days and complaining that Mozilla.org is too dependent on Netscape; or claiming on NewsForge that using native widgets would have meant writing the whole browser over for each platform. The former is a contradiction, the latter a blatant falsehood. You, Asa, and other fervent supporters of Mozilla are distinctly slippery in your engagement of the world of facts. Your conclusion is always "whatever we have been doing has always been perfectly correct," facts and consistency be damned. I'm sure this is playing just about as well with your management as it is with me.

> What particularly irritates me is the way you assume you know exactly why Mitchell got laid off, and further (contrary to the evidence I see before me at meetings and on mailing lists every day) you assert she's no longer part of the project, and we are lying when we say she is. <

That's a perfect example of your habit of misrepresentation. I said that your insistence that Mitchell was still your chief, even after management fired her from that role, was a deliberate slap in the face of the people who are paying the bills, which could not help your relationship with them. Instead of responding to that, you responded to something completely different which I did not say. This is a normal experience in dealing with you, and Asa, and SmileyBen, and SubtleRebel, and the rest of the cult -- or the "flunkies," as Tanyel has taken to calling you. It doesn't play well to non-cult-members.

>> You need resources. You will have to pay them. <<

> Are you offering? <

A perfect example of a slippery non-response.

>> The volunteer model does not offer adequate reward for serious QA and support work. <<

> I disagree. I think we've done better for non-Netscape help in QA than in code. Netscape QA often admit they wouldn't have a prayer of triaging the incoming Bugzilla bugs without a lot of other help. <

Adequacy is easy to determine in this case. You admit you are buried under a mountain of incoming bugs which you do not have the resources to process, and which make your bug database's defect tracking lines relatively useless as a measure of product quality. Therefore, the current process and/or resources are inadequate. Your defensive response above chooses to ignore the actual point at hand in favor of yet another "everybody is doing a great job and things are going really well." They're not going well. If you can't use your bug database to generate useful quality statistics, things are going very badly. This is a problem to be solved, not to be waved away.

> Do you really think we can wander up to Netscape management and ask for ten people, and we'll get them? <

I don't recall saying anything about wandering. From your account, there are weekly meetings about project status. Have you made the case in the way that I suggested, backed with numbers and tied to schedule in a measurable way? You said they already suspect they are not getting adequate support from outside contributors, so pitching this as a more traditional in-house work mode already has the right spin. I suspect that the reason you don't make such a pitch is that your religious convictions about open source don't permit you to acknowledge that there are gaps in the open source approach.

> The amount of attention I pay to someone's comments on the project are in proportion to a) the number of helpful suggestions they offer, and b) the amount of constructive work (of any sort) they've put in so far. <

By (a) I should be high in your rankings. Unfortunately, your ingrained defensiveness makes it difficult for you to perceive critical comments as helpful. I acknowledge that my present state of exasperation with your slipperiness has made my tone less than conducive to such a positive perception, but I haven't gotten any farther with the polite approach in the past.

>> No open source project has every gotten to anything like a professional stage of development without large amounts of professional compensation expended for the commercial benefit of a particular company. <<

> I could name numerous counter-examples <

Feel free, if you can. There are very few open source projects that have ever reached such a level. The Linux kernel, Apache, GNOME/Eazel, OpenOffice and Mozilla have all followed the course I just described, in which companies paid people to work on them out of commercial interest in the results. I'm not so concerned with hacker-oriented stuff like emacs or the rest of the GNU tools -- GCC is the only one that even comes close to professional quality, and it's radically inferior to commercial compilers in everything but portability. Volunteer labor does not deliver professional results except as an ancillary contribution to a paid backbone.

> but, instead, I'll just ask you: why is it, per se, a problem that companies work on projects in their own interest, just as individuals do? <

Did I say it was a problem? I said it means that the volunteer model of open source doesn't reflect the reality of software development, which is that you have to pay people to get professional-quality software. Again, you don't seem to be able to respond to what I actually say.