[parisc-linux] Metrics Gathering and Quality Measurements

[parisc-linux] Metrics Gathering and Quality Measurements

[Dunn RobertE Contr ESC Det 5 /NDSI]

Hi all,

	I have been following this list server for several weeks now.  I
have some questions.  I am in a Software Quality Assurance course in school.
We have a paper to prepare regarding SQA practices on a real project.  I
would like to ask a few questions regarding methods and methodologies in use
on PA-RISC linux port.

	My primary question is, are there any metrics being gathered on the
problems that I see reported on the list server?  Are the problems that are
reported being categorized and counted?  Is there any identification of when
the problem was introduced, when it was identified, and when it was
corrected?

	Are there any other concepts for porting linux to any other
platforms?

	Thanks in advance for any information you can provide.


Robert E. Dunn Jr.
Team ND
SND C2 SPO/NDSN
SenCom
Corporation
Database Manager
Phone: (719) 556-9448	DSN: 834-9448		Fax: 556-4417

Re: [parisc-linux] Metrics Gathering and Quality Measurements

[Michael Ang]

Currently the number of active developers is fairly small, so most bug
tracking is done by the developers themselves.  This is working
reasonably right now, since most bugs get fixed quickly and the number
of users of the code is small.

Once we get far enough that userspace applications work, the number of
developers and end users will likely grow quickly.  At that time it will
be important to scale our bug tracking as well.  Bugzilla
(http://bugzilla.mozilla.org/) would be a good bug tracking tool for us
to use.

Bugzilla is pretty slick.  You can report bugs, track progress, set
target fix dates, mark duplicates, etc.  All the data is stored in a
database so you can build interesting queries (such as listing how many
bugs each engineer has assigned to be finished before the next milestone
date).  Mozilla.org/Netscape wrote Bugzilla and they use it for all
their bug tracking -- there are a lot of good techniques we can steal,
er borrow, from them.

	- Mike.

Dunn RobertE Contr ESC Det 5 /NDSI wrote:
> 
>         My primary question is, are there any metrics being gathered on the
> problems that I see reported on the list server?  Are the problems that are
> reported being categorized and counted?  Is there any identification of when
> the problem was introduced, when it was identified, and when it was
> corrected?

Re: [parisc-linux] Metrics Gathering and Quality Measurements

[Alex deVries]

Dunn RobertE Contr ESC Det 5 /NDSI wrote:
>         My primary question is, are there any metrics being gathered on the
> problems that I see reported on the list server?  Are the problems that are
> reported being categorized and counted?  Is there any identification of when
> the problem was introduced, when it was identified, and when it was
> corrected?

Right now, we have no such data, but I'd be interested in seeing your
analysis.  Several of us have worked on Linux on different architectures
though.

- Alex

-- 
Alex deVries, Director of Professional Services, The Puffins at
Linuxcare
613.562.2759 tel
alex@linuxcare.com, http://www.linuxcare.com/
Linuxcare, Support for the revolution.

RE: [parisc-linux] Metrics Gathering and Quality Measurements

[Dunn RobertE Contr ESC Det 5 /NDSI]

	Right now, we have no such data, but I'd be interested in seeing
your
	analysis.  Several of us have worked on Linux on different
architectures
	though.

Alex,

	What file format would you like.  Linux people probably want Corel.
I have Word Perfect.  The school will want satan's Word 97 format.  I am
unsure about posting binary files to the list server.  I can e-mail it
directly to you.  Does anybody else want a copy?  It will be several weeks
before I will have it together.  I can tell you, however, based on what I
have learned thus far, there will be many negatives.  I will attempt to
identify any of the positive points that I can (for example: this list
server and bugzilla - thanks Mike -).  Any information the team can provide
will be most beneficial.

	Remember, my project deals with the quantification of quality.  This
includes the ability of the product team to quantify the quality of the
product.  My analysis will not address the quality of the product, it will
address the quality of the processes.  

	Some other questions:

	Has there been any risk analysis done?  I can see evidence of, at
least some informal, risk analysis.  The discussions about merging Linus'
product with parisc attempts to identify some of the risks that may be
encountered and some of the steps that may be taken to mitigate those risks.

	Have there been any prioritization of tasks?  Has the team
identified tasks that will take longer or have greater risk?  If they have,
have they also identified the earliest possible starting time for these
tasks?

	Is there a schedule?  It seems as though the project is just moving
in a forward direction with the goal of finishing when it can be finished?
I don't think I have seen any "future" milestones identified, at least, not
with a date associated to them.  There is plenty about accomplished
milestones.

	Are there budget issues or concerns?  I recognize that some of them
may not be discussed openly.  If any of that type of information is allowed,
then I will be happy to hear it.  I understand that Hewlett Packard in Fort
Collins is writing the checks for this project.

	What role is Linuxcare taking in the project now?  Have there been
any updates to the Puffingroup website in the last couple of weeks?

	Ok, I have rambled on enough for now.  I am looking forward to
answers to these questions.  I shall prepare my next list of questions.

	Again, thanks in advance.

Robert

Re: [parisc-linux] Metrics Gathering and Quality Measurements

[Alex deVries]

Dunn RobertE Contr ESC Det 5 /NDSI wrote:
>         What file format would you like.  Linux people probably want Corel.
> I have Word Perfect.  The school will want satan's Word 97 format.  I am
> unsure about posting binary files to the list server.  I can e-mail it
> directly to you.  Does anybody else want a copy?  It will be several weeks
> before I will have it together.  I can tell you, however, based on what I
> have learned thus far, there will be many negatives.  I will attempt to
> identify any of the positive points that I can (for example: this list
> server and bugzilla - thanks Mike -).  Any information the team can provide
> will be most beneficial.

I'd suggest text and postscript.  Send it to me, and if it's on topic
we'll host it on the PA-RISC Linux web pages.

>         Has there been any risk analysis done?  I can see evidence of, at
> least some informal, risk analysis.  The discussions about merging Linus'
> product with parisc attempts to identify some of the risks that may be
> encountered and some of the steps that may be taken to mitigate those risks.

There's certainly risk analysis that gets done on a project-to-project
basis, but it's a very informal practice.

>         Have there been any prioritization of tasks?  Has the team
> identified tasks that will take longer or have greater risk?  If they have,
> have they also identified the earliest possible starting time for these
> tasks?

Ish.  A lot of these get talked about off line.

>         Is there a schedule?  It seems as though the project is just moving
> in a forward direction with the goal of finishing when it can be finished?

That's pretty much it.  It's hard to guarantee any dates when the bulk
of development is done on non-paid time.

>         What role is Linuxcare taking in the project now?  Have there been
> any updates to the Puffingroup website in the last couple of weeks?

The Puffins at Linuxcare has a contract with HP to contribute to some
development. 

There will be updates very very soon.  We've been concentrating on
development in recent weeks.

- Alex
-- 
Alex deVries, Director of Professional Services, The Puffins at
Linuxcare
613.562.2759 tel
alex@linuxcare.com, http://www.linuxcare.com/
Linuxcare, Support for the revolution.