posted by Dennis Heuer on Wed 22nd Sep 2004 06:41 UTC
IconEver used BugZilla? I'm sure you have. Frustrated? Dropped support for ever? Oh yes, I understand.

Many projects use bug trackers under which BugZilla is a famous choice. Once invented to help the Mozilla team out of its maintenance nightmare it made its way into user support. Some projects do not even provide a user mailinglist anymore. Everything is to be written into the BugZilla engine, even if it's not a bug but an idea, a comment or just a congratulation to the great product.

However, BugZilla wasn't made for that. It is not a support engine but a bug tracker. The user is forced to fill out complex forms with some unclear or just off topic questions only to receive an automated confirmation and wait. After some weeks, possibly, he receives another cryptic report only to find that somebody has changed the status of the bug message to "Not Replicable" or "Removed". No further word, no address of a person to discuss the case with. Is it completely wrong if one suspects that some projects push BugZilla to front intentionally to hide behind undisturbed.

Some time ago I'd posted a bug message to the BugZilla engine of the Mozilla project. A message dropped into my mailbox telling me that the bug message were a duplicate of another bug message. I compared the two messages and found that they were about very different subjects. The other bug message had already two other so called duplicates appended which weren't. Because there was no other way to contact the maintainer I opened a new bug message in which I argued about the different subjects of the four so called duplicates. My message was removed because "It includes four bug messages. Please send only one bug per message".

I reopened the message and asked the maintainer if he'd even read it. The message was removed again because of the same reason as before. I opened a new bug message and posted the same text in the hope that it would be read by another maintainer but the same maintainer removed my message and complained about my stubborness. I started a discussion through the BugZilla engine and asked him if he could forward the message to some other maintainer if he's not willing to look at the case. He refused to do so. I got somewhat angry and told him my opinion about his practice. He argued that he was only a volunteering assistant and not involved into the project. So, if he's only an assistant, he could've forwarded my pledge to a maintainer, I wrote. But he still refused to do so. I decided to stop supporting mozilla for ever. The error remained untouched for two further releases.

This is not worth it. BugZilla shows no pay off for the user, it only supports the developer. You may argue that the user is supported indirectly when the bug is resolved. However, the user needs some address to discuss details with. Mailinglists or user forums are better places for that because many users and developers can discuss the case together. While this is theoretically possible with BugZilla others first have to intentionally detect the bug message (before it's removed) which is unlikely in an ever growing bug tracker.

I've posted a bug message to the OpenSSH BugZilla engine. Except the confirmation I received nothing for about half a year. Two new releases produced the same error which hindered me to install OpenSSH from source. I sent a new bug message with same content. Again I received nothing for about half a year. I was frustrated because OpenSSH is not a simple game but a critical part of my platform that I wasn't able to update. I posted further requests, appended as comments to the existing bug message, but got no reply. Because there was no other contact address available I posted another bug message in that I complained about the project support. Suddenly about fifteen developers spammed my mailbox with uncensored comments. One only wrote that the GPL doesn't guarantee support. Ahh? I answered that support isn't even mentioned in the GPL and fully left to the project. And, there I asked for support and complained about it. He only repeated his opinion. As somebody was about sending me the whole X Window System as email appendix I skipped the discussion and switched to lsh.

For all this stress I had to create a user account. Because so many projects have forced me to create accounts I even had to create a separate passwords folder in my email client (yes, I know that browsers remember passwords but refuse to use this function!).

Over time, learning my lessons and still getting requests to messages I've posted last year, I moved over to refuse supporting a project that misuses BugZilla as a support engine or asks me to register. This may sound ignorant but I refuse to waste hours, if not days, for decrypting BugZilla reports, talking to assistbots and struggling with misunderstood developers without the chance for a productive contact to a friendly environment of human beings. Though this is not guaranteed by mailinglists or forums either, these platforms seem to evolve a different, more open and mature character over time. I definetly pledge for more projects following the example and banning BugZilla from the support area. It should only be used for real bug messages, and even then there must be another way to discuss these bugs if the nature of the problem needs more involvement or audience.

About the Author:
The author, Dennis Heuer, is a 34-years old german social scientist concentrating on human-computer-interaction (HCI) and e-learning.

e p (0)    107 Comment(s)

Technology White Papers

See More