www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - Bug Reporting

reply Arcane Jill <Arcane_member pathlink.com> writes:
Could we clarify what the bugs forum is for, please?

I was under the impression that the bugs forum was for reporting bugs. That is,
if we find a bug, the correct and appropriate thing to do is to head over to the
bugs forum and report it, citing a minimal code fragment which demonstrates it.

It should not be necessary to justify to other posters that the bug is "worth
fixing". This isn't a competition. I'm just not interested in playing "my bug's
more important than your bug". That's a stupid game.

It should not be necessary to explain to other posters the context in which that
code fragment might occur in real code. That entirely defeats the point of
trimming it down to a minimal fragment for Walter to see.

It should not even be necessary even to /care/ whether it gets fixed or not. All
you're doing, after all, is conveying information about which Walter would want
to know, to Walter.

I am becoming more and more convinced that a discussion forum is not really the
most appropriate way to report  bugs. The traditional approach (write a bug
report, assign it a suggested priority, submit it, get allocated a bug number,
check back from time to time on its status) seems to work well enough for most
projects. Why is D doing things differently?

I ask because I've reported some bugs recently (some of which, I admit, were
very minor), and got positively chastised for doing so, almost as though I was
wasting Walter's time by bringing such trivial things to Walter's attention. (No
matter that some of them have given me an access violation in real code in the
past). I even got told that one report was "silly". (Why? Because I had reduced
it to three lines of code, as Walter had requested. Thus it was out of context.
Other posters said words to the effect of "well don't do that then". Huh?)

Now I will admit that I'm still not on my best form right now (my gran's funeral
is the day after tomorrow as I write) and maybe some things are getting to me
that would normally just wash over, but it's like you have to justify reporting
a bug now, and it's very disheartening.

I don't even /care/ that you can put statements after a return statement, for
instance. I'm not passionate about it. I can live with this. I'm quite happy for
other, more important, bugs to be fixed first. But does that really mean I
shouldn't report it? I thought I was being helpful. Seems instead like I'm being
a nuisance.

Maybe I just came back to the D fora too early. Maybe I should have taken a few
more days off. Ah well.

Jill
Jul 20 2004
next sibling parent J Anderson <REMOVEanderson badmama.com.au> writes:
Arcane Jill wrote:

Could we clarify what the bugs forum is for, please?

I was under the impression that the bugs forum was for reporting bugs. That is,
if we find a bug, the correct and appropriate thing to do is to head over to the
bugs forum and report it, citing a minimal code fragment which demonstrates it.

It should not be necessary to justify to other posters that the bug is "worth
fixing". This isn't a competition. I'm just not interested in playing "my bug's
more important than your bug". That's a stupid game.

It should not be necessary to explain to other posters the context in which that
code fragment might occur in real code. That entirely defeats the point of
trimming it down to a minimal fragment for Walter to see.
  
Not entirely. Sometimes its good to get a sense of ralivents about a particular bug.
It should not even be necessary even to /care/ whether it gets fixed or not. All
you're doing, after all, is conveying information about which Walter would want
to know, to Walter.

I am becoming more and more convinced that a discussion forum is not really the
most appropriate way to report  bugs. The traditional approach (write a bug
report, assign it a suggested priority, submit it, get allocated a bug number,
check back from time to time on its status) seems to work well enough for most
projects. Why is D doing things differently?
  
Partly because there is only one guy writting the software, who doesn't have time to reply to all emails sent.
I ask because I've reported some bugs recently (some of which, I admit, were
very minor), and got positively chastised for doing so, almost as though I was
wasting Walter's time by bringing such trivial things to Walter's attention. (No
matter that some of them have given me an access violation in real code in the
past). I even got told that one report was "silly". (Why? Because I had reduced
it to three lines of code, as Walter had requested. Thus it was out of context.
Other posters said words to the effect of "well don't do that then". Huh?)
  
Humm, I haven't read those emails. Often with newsgroups you need to take things with a grain of salt. Its so easy for people to either write a message the wrong way or to read it the wrong way, or to get into point-scoring (which sometimes is quite a constructive thing) as you have said. Bar, I'm repeating what has been said so often before.
Now I will admit that I'm still not on my best form right now (my gran's funeral
is the day after tomorrow as I write) and maybe some things are getting to me
that would normally just wash over, but it's like you have to justify reporting
a bug now, and it's very disheartening.

I don't even /care/ that you can put statements after a return statement, for
instance. I'm not passionate about it. I can live with this. I'm quite happy for
other, more important, bugs to be fixed first. But does that really mean I
shouldn't report it? I thought I was being helpful. Seems instead like I'm being
a nuisance.

Maybe I just came back to the D fora too early. Maybe I should have taken a few
more days off. Ah well.

Jill

  
I think an open bug newsgroup is a good thing. It means that things that aren't really bugs get answers very quickly (one less for post for Walter to deal with) and workarounds are given. Personally I like the way these newsgroups can head off in different directions, its really helps build the community feeling. If the bug-group heads of in a completely different direction I don't really want the NN (Newsgroup Nazis) telling people off for going of topic. One thing I would like to see in the bug group is a number system for bugs so that when you have an error you can quickly search the newsgroup for the solution. Of course that would require Walter to start printing out numbers with the error messages. -- -Anderson: http://badmama.com.au/~anderson/
Jul 20 2004
prev sibling next sibling parent reply J C Calvarese <jcc7 cox.net> writes:
In article <cdj5if$1gh9$1 digitaldaemon.com>, Arcane Jill says...
Could we clarify what the bugs forum is for, please?

I was under the impression that the bugs forum was for reporting bugs. That is,
if we find a bug, the correct and appropriate thing to do is to head over to the
bugs forum and report it, citing a minimal code fragment which demonstrates it.

It should not be necessary to justify to other posters that the bug is "worth
It's not necessary to justify anthing. No one's making you reply.
fixing". This isn't a competition. I'm just not interested in playing "my bug's
more important than your bug". That's a stupid game.

It should not be necessary to explain to other posters the context in which that
code fragment might occur in real code. That entirely defeats the point of
trimming it down to a minimal fragment for Walter to see.

It should not even be necessary even to /care/ whether it gets fixed or not. All
you're doing, after all, is conveying information about which Walter would want
to know, to Walter.
If you don't /care/ if it's fixed, don't bother /reading/ the replies.
I am becoming more and more convinced that a discussion forum is not really the
most appropriate way to report  bugs. The traditional approach (write a bug
report, assign it a suggested priority, submit it, get allocated a bug number,
check back from time to time on its status) seems to work well enough for most
projects. Why is D doing things differently?
I agree, but we'll have to convince Walter first. He's been stung in the past by people publishing out of context the number of outstanding bugs in a project or something like that. I think using bug tracking software like bugzilla or something similiar would be helpful for all involved, but it's not my decision to make. :(
I ask because I've reported some bugs recently (some of which, I admit, were
very minor), and got positively chastised for doing so, almost as though I was
wasting Walter's time by bringing such trivial things to Walter's attention. 
I don't know which particular posts you're referring to, so I hope it was an honest misunderstanding where someone was trying to help you out. (No
matter that some of them have given me an access violation in real code in the
past). I even got told that one report was "silly". (Why? Because I had reduced
it to three lines of code, as Walter had requested. Thus it was out of context.
Other posters said words to the effect of "well don't do that then". Huh?)

Now I will admit that I'm still not on my best form right now (my gran's funeral
is the day after tomorrow as I write) and maybe some things are getting to me
that would normally just wash over, but it's like you have to justify reporting
a bug now, and it's very disheartening.

I don't even /care/ that you can put statements after a return statement, for
instance. I'm not passionate about it. I can live with this. I'm quite happy for
other, more important, bugs to be fixed first. But does that really mean I
shouldn't report it? I thought I was being helpful. Seems instead like I'm    
> being a nuisance.
I think you should post your bug reports, but you need to realize that people may try to "help you out" by pointing out something that you might not realize (or that may be painfully obvious to you). For instance, if the compiler is adhering to the spec that might be helpful to know (assume of course that you care). At the point, I think it becomes a feature request. Feature requests are best handled in this forum, since they are usually less urgent that when compiler is acting erroneously.
Maybe I just came back to the D fora too early. Maybe I should have taken a few
more days off. Ah well.
Maybe you just forgot how much fun we are to be around. ;)
Jill
jcc7
Jul 20 2004
parent reply "Walter" <newshound digitalmars.com> writes:
"J C Calvarese" <jcc7 cox.net> wrote in message
news:cdjcja$1jls$1 digitaldaemon.com...
 I agree, but we'll have to convince Walter first. He's been stung in the
past by
 people publishing out of context the number of outstanding bugs in a
project or
 something like that.
It was a lazy author writing a review of compilers, who simply cut & pasted the bug list as his 'review'. It got published. Never mind that those bugs had been fixed before it went to print, the damage was done and went on for years afterwards.
Jul 20 2004
parent J Anderson <REMOVEanderson badmama.com.au> writes:
Walter wrote:

"J C Calvarese" <jcc7 cox.net> wrote in message
news:cdjcja$1jls$1 digitaldaemon.com...
  

I agree, but we'll have to convince Walter first. He's been stung in the
    
past by
people publishing out of context the number of outstanding bugs in a
    
project or
something like that.
    
It was a lazy author writing a review of compilers, who simply cut & pasted the bug list as his 'review'. It got published. Never mind that those bugs had been fixed before it went to print, the damage was done and went on for years afterwards.
So in other words (with a little twisting) you want to make the knowledge of particular bugs more difficult to find info about? -- -Anderson: http://badmama.com.au/~anderson/
Jul 21 2004
prev sibling next sibling parent Ant <Ant_member pathlink.com> writes:
In article <cdj5if$1gh9$1 digitaldaemon.com>, Arcane Jill says...
Could we clarify what the bugs forum is for, please?

 citing a minimal code fragment which demonstrates it.
I think a minimal complete program is better. Ant
Jul 20 2004
prev sibling next sibling parent "Ivan Senji" <ivan.senji public.srce.hr> writes:
I just want to say something to you AJ: The D specification has bugs, and
don't
give up on poining them out just because of the responce of other people.

To me also this doesn't makes sence(and similar things with throw...)

int func()
{
 label:
 goto label;
 return 0;
}

This is acceptable to the compiler although return 0; will never be
executed,
and it complains if return is removed, this doesn't make sence to me.

Some time ago i was being very annoying about operator overloading,
i was complaining because compiler didn't look for B.opSomething_r(A)
if A.opSomething(B) existed, and it was defined to work that way in the
spec. But now it is changed.

I did the same thing: took the peace of my code and simplified it so it
would
be as simple as possible to show the bug, and some responded: "why are you
doing that? Opearators aren't for that..."

Just don't give up. If you really believe that what you are saying makes
sense
(i do) then just mention this from time to time, and maybe Walter will
realize that what you are saying really makes sense and then the spec will
change!

Good luck :)



"Arcane Jill" <Arcane_member pathlink.com> wrote in message
news:cdj5if$1gh9$1 digitaldaemon.com...
 Could we clarify what the bugs forum is for, please?

 I was under the impression that the bugs forum was for reporting bugs.
That is,
 if we find a bug, the correct and appropriate thing to do is to head over
to the
 bugs forum and report it, citing a minimal code fragment which
demonstrates it.
 It should not be necessary to justify to other posters that the bug is
"worth
 fixing". This isn't a competition. I'm just not interested in playing "my
bug's
 more important than your bug". That's a stupid game.

 It should not be necessary to explain to other posters the context in
which that
 code fragment might occur in real code. That entirely defeats the point of
 trimming it down to a minimal fragment for Walter to see.

 It should not even be necessary even to /care/ whether it gets fixed or
not. All
 you're doing, after all, is conveying information about which Walter would
want
 to know, to Walter.

 I am becoming more and more convinced that a discussion forum is not
really the
 most appropriate way to report  bugs. The traditional approach (write a
bug
 report, assign it a suggested priority, submit it, get allocated a bug
number,
 check back from time to time on its status) seems to work well enough for
most
 projects. Why is D doing things differently?

 I ask because I've reported some bugs recently (some of which, I admit,
were
 very minor), and got positively chastised for doing so, almost as though I
was
 wasting Walter's time by bringing such trivial things to Walter's
attention. (No
 matter that some of them have given me an access violation in real code in
the
 past). I even got told that one report was "silly". (Why? Because I had
reduced
 it to three lines of code, as Walter had requested. Thus it was out of
context.
 Other posters said words to the effect of "well don't do that then". Huh?)

 Now I will admit that I'm still not on my best form right now (my gran's
funeral
 is the day after tomorrow as I write) and maybe some things are getting to
me
 that would normally just wash over, but it's like you have to justify
reporting
 a bug now, and it's very disheartening.

 I don't even /care/ that you can put statements after a return statement,
for
 instance. I'm not passionate about it. I can live with this. I'm quite
happy for
 other, more important, bugs to be fixed first. But does that really mean I
 shouldn't report it? I thought I was being helpful. Seems instead like I'm
being
 a nuisance.

 Maybe I just came back to the D fora too early. Maybe I should have taken
a few
 more days off. Ah well.

 Jill
Jul 20 2004
prev sibling parent "Walter" <newshound digitalmars.com> writes:
"Arcane Jill" <Arcane_member pathlink.com> wrote in message
news:cdj5if$1gh9$1 digitaldaemon.com...
 Could we clarify what the bugs forum is for, please?

 I was under the impression that the bugs forum was for reporting bugs.
That is,
 if we find a bug, the correct and appropriate thing to do is to head over
to the
 bugs forum and report it, citing a minimal code fragment which
demonstrates it.
 It should not be necessary to justify to other posters that the bug is
"worth
 fixing". This isn't a competition. I'm just not interested in playing "my
bug's
 more important than your bug". That's a stupid game.
I've been online for 20+ years now. Just ignore anyone trying to pick a fight, it's the only way.
 It should not be necessary to explain to other posters the context in
which that
 code fragment might occur in real code. That entirely defeats the point of
 trimming it down to a minimal fragment for Walter to see.
True, but it can matter when prioritizing bug fixes. A bug that one constantly runs into and has no reasonable workaround will be higher priority than a contrived one that one would never run into with properly written code.
 It should not even be necessary even to /care/ whether it gets fixed or
not. All
 you're doing, after all, is conveying information about which Walter would
want
 to know, to Walter.
The discussions are frequently good. Posted bugs aren't always bugs, someone else might see an error in the code, or have some insight into how to solve it, or post a more finely tuned example, etc.
 I am becoming more and more convinced that a discussion forum is not
really the
 most appropriate way to report  bugs. The traditional approach (write a
bug
 report, assign it a suggested priority, submit it, get allocated a bug
number,
 check back from time to time on its status) seems to work well enough for
most
 projects. Why is D doing things differently?
The short answer is because it's easier doing it this way than administering a bureaucratic process. The latter is better if there is an actual staff tasked with Q.A., some managers, and a team of developers. But that's not the case here. I do note in the changelog each bug fixed, and I do an email to the submitter. (The latter often fails because posters understandably obfuscate their email addresses, oh well.)
 I ask because I've reported some bugs recently (some of which, I admit,
were
 very minor), and got positively chastised for doing so, almost as though I
was
 wasting Walter's time by bringing such trivial things to Walter's
attention. (No
 matter that some of them have given me an access violation in real code in
the
 past). I even got told that one report was "silly". (Why? Because I had
reduced
 it to three lines of code, as Walter had requested. Thus it was out of
context.
 Other posters said words to the effect of "well don't do that then". Huh?)
You're doing the right thing by reporting the bugs, even silly and trivial ones.
 Now I will admit that I'm still not on my best form right now (my gran's
funeral
 is the day after tomorrow as I write) and maybe some things are getting to
me
 that would normally just wash over, but it's like you have to justify
reporting
 a bug now, and it's very disheartening.
You don't need to justify yourself to anyone online who replies to your postings, not even me <g>. The n.g. is a free for all, with both the insightful and the insipid that comes with it. I think only about 3 or 4 posts have been yanked since it began 3 years ago, one for posting copyrighted code without a license and a couple from a spammer who discovered the news server. Given the 30,000+ messages posted, it has an incredibly good track record.
 I don't even /care/ that you can put statements after a return statement,
for
 instance. I'm not passionate about it. I can live with this. I'm quite
happy for
 other, more important, bugs to be fixed first. But does that really mean I
 shouldn't report it? I thought I was being helpful. Seems instead like I'm
being
 a nuisance.
You are being helpful, not a nuisance.
Jul 20 2004