www.digitalmars.com         C & C++   DMDScript  

D - Buglist ?

reply "Kris" <someidiot earthlink.dot.dot.dot.net> writes:
Many of us post what we uncover as bugs or inconsistencies. Some are purely
misunderstandings, whilst others are glaring errors.

Rather than Walter just trying to keep them all in his head, and/or perhaps
missing some posts, wouldn't it be more productive to sanction some site as
a "quasi-official" bug-list? This newsgroup doesn't cut it because a) bugs
are interspersed with other conversations and b) there's never any status
feedback (except by looking at the change-log upon subsequent compiler
releases).

There's at least one Wiki page that tries to address this, but without
general consent, especially from Walter <g>, it's somewhat of a lost cause
...

Perhaps there's some issue over negative publicity? I would certainly hope
not. How about it folks? Does Walter already manage a list like this? Am I
just wasting your time?

BTW: this is intended to help the release process; not hinder it.

- Kris
Apr 21 2004
next sibling parent "Unknown W. Brackets" <unknown simplemachines.org> writes:
"Kris" <someidiot earthlink.dot.dot.dot.net> wrote in message
news:c66gsp$1tgc$1 digitaldaemon.com...
 Many of us post what we uncover as bugs or inconsistencies. Some are

 misunderstandings, whilst others are glaring errors.

Now, I know this is my second post here, but I just thought I might offer a cent or two... I'm all the fan of forums, newsgroups, etc. After all, I am the lead developer of one. But, they have their limits... they just can't do everything. Projects like PHP (http://bugs.php.net/) and Mozilla (http://bugzilla.mozilla.org/) handle bugs in a very streamlined and clean way. In fact, Mozilla uses a forum as well (http://forums.mozillazine.org/) but after they are discussed there they are posted as bugs on BugZilla. The problem is that this takes organization, moderation, and the like. It is no one or two person job to handle this, and you have to have a few people there just to mark bugs as duplicates of other bugs. However, it can also make it a lot easier to organize the bugs, discuss the bugs, etc. RFE can also be done... and these systems provide much better management (what has happened to this bug, when it is expected to be resolved, etc.) than a forum would... simply because they are specialized. Now, I realize your post is simply about a listing of bugs, but that is one of the points of a bug listing system. Just my cent and a half. Thanks, -[Unknown]
Apr 21 2004
prev sibling next sibling parent reply J Anderson <REMOVEanderson badmama.com.au> writes:
Kris wrote:

Perhaps there's some issue over negative publicity?
  

-- -Anderson: http://badmama.com.au/~anderson/
Apr 21 2004
parent reply Stephan Wienczny <wienczny web.de> writes:
J Anderson wrote:
 Kris wrote:
 
 Perhaps there's some issue over negative publicity?
  


Maybe this time again?
Apr 21 2004
next sibling parent reply "Unknown W. Brackets" <unknown at.simplemachines.dot.org> writes:
"Stephan Wienczny" <wienczny web.de> wrote in message
news:c66n0s$290v$1 digitaldaemon.com...
 Thats what big W said whenever this discussion came up...
 Maybe this time again?

I find that a consistent and strong bug reporting system can actually be positive. It shows that development is moving... especially to those who have problems. All programming projects have bugs. In fact, all things do - even the specification could, or anything else. Even C++ could and moreover I'm sure still does have bugs. Knowing that your problem is being addressed makes people feel like the compiler is for *them*, and also significantly decreases reposting of the bug. While a publicly viewable list of known bugs can make it look "buggy" to industrial clients, wouldn't it look worse without one? Most people who would think a list of bugs makes the software less would also think that a language that doesn't have everyone and her brother using it is also not worth using I've found. But if the descision is to not do it for this reason, I will leave it at that. -[Unknown]
Apr 21 2004
next sibling parent Stephan Wienczny <wienczny web.de> writes:
Unknown W. Brackets wrote:

 "Stephan Wienczny" <wienczny web.de> wrote in message
 news:c66n0s$290v$1 digitaldaemon.com...
 
Thats what big W said whenever this discussion came up...
Maybe this time again?

I find that a consistent and strong bug reporting system can actually be positive. It shows that development is moving... especially to those who have problems. All programming projects have bugs. In fact, all things do - even the specification could, or anything else. Even C++ could and moreover I'm sure still does have bugs. Knowing that your problem is being addressed makes people feel like the compiler is for *them*, and also significantly decreases reposting of the bug. While a publicly viewable list of known bugs can make it look "buggy" to industrial clients, wouldn't it look worse without one? Most people who would think a list of bugs makes the software less would also think that a language that doesn't have everyone and her brother using it is also not worth using I've found. But if the descision is to not do it for this reason, I will leave it at that. -[Unknown]

As you can see this file from intel containing a list of bugs in intel C++. ftp://download.intel.com/support/performancetools/c/c_swdefects.pdf
Apr 21 2004
prev sibling parent reply James Widman <james jwidman.com> writes:
In article <c66ofa$2bf3$1 digitaldaemon.com>,
 "Unknown W. Brackets" <unknown at.simplemachines.dot.org> wrote:
 I find that a consistent and strong bug reporting system can actually be
 positive.  It shows that development is moving... especially to those who
 have problems.

Hear hear. Mozilla (and its components from which FireFox sprung) is certainly better than it would have been without bugzilla. Benefits of using bugzilla-like tools that are obvious to me: 1) bugs can be prioritized, so that the most important fixes are done more often than lesser ones. As a result more bug-free code goes out to the masses which leads more people to use it. 2) a lot of users don't have to spend their time typing a bug report with instructions on how to reproduce the bug, because some other user already did that (and it looks like D is going to have a *lot* of new users in the coming months). 3) new developers looking for a place to start can just look at a list of unassigned bugs without bothering the core developers who are probably dealing with more complex problems. 4) even if a bug isn't fixed, at least there's a database entry for it, and thus a common place to go to for information on when (or if) to expect a fix. This is probably also where people affected by the bug will learn useful work-arounds. 5) when a developer starts attacking a bug, he doesn't have to hunt for information relating to it. It's all in a single database entry, which may have been added to by other users. 6) the database tells you who took responsibility for each bug. That means less duplication of effort. It also tells you who knows the most about any particular bug, so if the assignee is to busy with other tasks at the moment, someone else can pick up the slack after a quick briefing. 7) users can be automatically notified by email when a bug gets fixed. ...The only *real* downside I've heard so far is that it sucks up a good portion of time for whomever decides to help maintain the database. But for everyone else, it's a time saver and productivity booster. Does that sound fair & accurate?
Apr 21 2004
parent reply "Unknown W. Brackets" <unknown at.simplemachines.dot.org> writes:
James Widman wrote:
 ...The only *real* downside I've heard so far is that it sucks up a good 
 portion of time for whomever decides to help maintain the database.  But 
 for everyone else, it's a time saver and productivity booster.

Right, but this can be done by people with no coding ability. In my past experience, people like this who want to contribute are much easier to come by than otherwise.
 Does that sound fair & accurate?

Yes. You seem to have a lot to say about bug reporting systems.... I'm planning to get some work in on one in about a month. (too much other work to do right now..) I plan to simplify moderation and make the look better, as well as use PHP and MySQL instead of Perl like Bugzilla does. I've already designed some of the database, as well... I plan this mainly because SMF (the forum I write) currently has no bug reporting system beside the forum, and I am not happy with any of the existing ones. But I still want to make a release out of it that can be used and extended by all. If you'd be willing to give me comments on what you might think bug reporting systems currently miss, or need more of, etc... I'd be very grateful. Maybe if everything went just right, my bug system could be at some point used for D ;). But that's just me being hopeful. -[Unknown]
Apr 21 2004
parent reply James Widman <james jwidman.com> writes:
In article <c674bc$2v55$1 digitaldaemon.com>,
 "Unknown W. Brackets" <unknown at.simplemachines.dot.org> wrote:
 Yes.  You seem to have a lot to say about bug reporting systems.... 

I just don't like the idea of important information falling through the cracks. ;-)
 ... and I am not happy with any of the existing ones.

What does Bugzilla lack? I wouldn't try to suggest that it has everything anyone could need, but in general I also don't like the idea of well-debugged finished products going unused.
 If you'd be willing to give me comments on what you might think bug 
 reporting systems currently miss, or need more of, etc... I'd be very 
 grateful.

I don't know what people here could need that Bugzilla doesn't have. So: does anyone want to suggest a list of requirements?
Apr 21 2004
parent "Unknown W. Brackets" <unknown at.simplemachines.dot.org> writes:
James Widman wrote:

 What does Bugzilla lack?  I wouldn't try to suggest that it has 
 everything anyone could need, but in general I also don't like the idea 
 of well-debugged finished products going unused.

Look at PHP's bug system. In some ways, Bugzilla is too complex, too ugly, and a bit bloated. It also has its set of bugs, like any project might.... which are reasons, again, that I don't want to use it. As well, I've not seen a bug system that does pagination - look at, for example, the mng bug on Bugzilla. That could take a long time to fully load for dial up users.... really, the comments should be shown in groups of about 25, which would in most cases cause absolutely no problems and solve a lot of things. And I said ugly. By this I don't just mean it isn't as pleasing as some - in the long run that doesn't matter - but it also can have somewhat confusing UI. I'm not trying to say Bugzilla is not a good system. I'm not saying D couldn't use it. I'm just saying, nothing, not even D, is perfect. -[Unknown]
Apr 22 2004
prev sibling parent reply J C Calvarese <jcc7 cox.net> writes:
Stephan Wienczny wrote:
 J Anderson wrote:
 
 Kris wrote:

 Perhaps there's some issue over negative publicity?
  


Maybe this time again?

This is the best reference I could find to Walter's dislike of bug lists: "I published a carefully documented bug list once in the past. It was a disaster." -- Walter, D/23404 I'm pretty sure there's a longer version of this story somewhere in the archives, but I haven't been able to locate it. -- Justin http://jcc_7.tripod.com/d/
Apr 21 2004
parent reply J C Calvarese <jcc7 cox.net> writes:
J C Calvarese wrote:
 Stephan Wienczny wrote:
 
 J Anderson wrote:

 Kris wrote:

 Perhaps there's some issue over negative publicity?
  


Maybe this time again?

This is the best reference I could find to Walter's dislike of bug lists: "I published a carefully documented bug list once in the past. It was a disaster." -- Walter, D/23404

IIRC, the problem was the bug list was used against Walter to show that his software is "buggy".
 
 I'm pretty sure there's a longer version of this story somewhere in the 
 archives, but I haven't been able to locate it.
 

-- Justin http://jcc_7.tripod.com/d/
Apr 21 2004
next sibling parent reply "Matthew" <matthew.hat stlsoft.dot.org> writes:
"J C Calvarese" <jcc7 cox.net> wrote in message
news:c678k4$403$2 digitaldaemon.com...
 J C Calvarese wrote:
 Stephan Wienczny wrote:

 J Anderson wrote:

 Kris wrote:

 Perhaps there's some issue over negative publicity?


Maybe this time again?

This is the best reference I could find to Walter's dislike of bug lists: "I published a carefully documented bug list once in the past. It was a disaster." -- Walter, D/23404

IIRC, the problem was the bug list was used against Walter to show that his software is "buggy".

That's my recollection. But the fact is that it is buggy, just as much as anyone elses. Frankly, the rate at which D is becoming popular, I think this concern is now irrelevant.
Apr 21 2004
parent J C Calvarese <jcc7 cox.net> writes:
Matthew wrote:
 "J C Calvarese" <jcc7 cox.net> wrote in message
 news:c678k4$403$2 digitaldaemon.com...
 
J C Calvarese wrote:

Stephan Wienczny wrote:


J Anderson wrote:


Kris wrote:


Perhaps there's some issue over negative publicity?

I think this is big W's reason against a bug list.

Thats what big W said whenever this discussion came up... Maybe this time again?

This is the best reference I could find to Walter's dislike of bug lists: "I published a carefully documented bug list once in the past. It was a disaster." -- Walter, D/23404

IIRC, the problem was the bug list was used against Walter to show that his software is "buggy".

That's my recollection. But the fact is that it is buggy, just as much as anyone elses. Frankly, the rate at which D is becoming popular, I think this concern is now irrelevant.

Right. And we'll have to fight FUD whether there's a bug list or not. The slashdotting proved that. (Garbage collection is what makes Java so slow. C++ will always be the best; all other languages are for script kiddies. If Microsoft isn't interested, it won't succeed. Should've called it "P". The name "D" is already taken by 100 other languages. "D" stands for dumb. Etc.) -- Justin http://jcc_7.tripod.com/d/
Apr 21 2004
prev sibling next sibling parent John Reimer <brk_6502 NOSP_AM.yahoo.com> writes:
 
 IIRC, the problem was the bug list was used against Walter to show that 
 his software is "buggy".
 

Even worse, I think the bug list was published "verbatum" in an article reviewing his software... or something like that.
Apr 21 2004
prev sibling parent Stewart Gordon <smjg_1998 yahoo.com> writes:
J C Calvarese wrote:

<snip>
 IIRC, the problem was the bug list was used against Walter to show that 
 his software is "buggy".

If people are going to construe bug lists as personal attacks on the software developer, that's their problem. Bug lists are not supposed to be attacks on the software developer, nor on the software itself. They are supposed to be attacks on the individual bugs listed. In relation to the software, a bug list is constructive criticism. Further, it: - helps to keep the development organised - provides user confidence that a reported bug has been acknowledged - saves user time in diagnosing and re-reporting a bug that has already been reported. Stewart. -- My e-mail is valid but not my primary mailbox, aside from its being the unfortunate victim of intensive mail-bombing at the moment. Please keep replies on the 'group where everyone may benefit.
Apr 22 2004
prev sibling parent larry cowan <larry_member pathlink.com> writes:
In article <c66gsp$1tgc$1 digitaldaemon.com>, Kris says...
Many of us post what we uncover as bugs or inconsistencies. Some are purely
misunderstandings, whilst others are glaring errors.

Rather than Walter just trying to keep them all in his head, and/or perhaps
missing some posts, wouldn't it be more productive to sanction some site as
a "quasi-official" bug-list? This newsgroup doesn't cut it because a) bugs
are interspersed with other conversations and b) there's never any status
feedback (except by looking at the change-log upon subsequent compiler
releases).

There's at least one Wiki page that tries to address this, but without
general consent, especially from Walter <g>, it's somewhat of a lost cause
...

Perhaps there's some issue over negative publicity? I would certainly hope
not. How about it folks? Does Walter already manage a list like this? Am I
just wasting your time?

BTW: this is intended to help the release process; not hinder it.

- Kris

We need something of this nature. This is an attempt to define something appropriate to our current and immediate future needs, not necessarily long-term, and not subject to the problems of being officially recognized by Walter, just available to him as a resource. I feel it should have these characteristrics: 1. Moderated by a single person or small group having the ability to lock some of the content, and to delete any wrong or superfluous additions, and to purge or archive completed and rejected entries. I lean toward archiving purges to a separate (at least logically) info base. 2. It should be open to all users to add entries, further info, and comments subject to the actions of #1 above. 3. Documentation, specification, and implementation bug reports and feature requests should all be allowed with appropriate categorization. 4. Entries are not complete and should not be prioritized above "interesting" without a minimal set of demo code for the bug, further test code for checking correction if appropriate, and/or code to illustrate the desired capability. 5. If appropriate and possible, work-arounds, and ideas and code for correct implementation should be submittable. 6. Useable search capabilities must be provided for this to work, with various sectioning of where to search and with subsearches of initial response. 7. We don't need assignment of priority, but interest level indication by some "add a yea or nay vote or indicate a value/importance [1 to 10) vote" method would be useful. (Might need to lock to an IP or user to prevent multiple votes, but later if needed.) 8. It should be possible for this to handle A. compiler, B. phobos, C. peripheral libraries, and D. current project links. 9. It would be nice for projects to be able to tag entries with their requirements and desires. (classed as Necessary, Important, Useful, and Wishlisted ?) 10. Where appropriate, someone should be able to note that they are working to resolve the problem or need - with a web or enail link to them. Does this sound like a pre-configured, moderated Wiki with page templates specialized to our needs and perhaps specialized search capabilities? It doesn't sound like Bugzilla - more a consolidation and optimization of information and effort facilitator than a "Buglist". Let's keep the free and easy growth potential we currently have as long as possible. More, anyone?
Apr 22 2004