Hello All,
I have not received any mails lately but I can see them on the glug
site. This is in response to Kenneth Gonsalves's mail quoted below.
"I am not laughing. What exactly is a bug? When, for example, in kword
I save a file in OOo format and OOo is unable to open it, is that a
bug? Or is it a failure of software - just bad software? As far as I
know, the approach to writing software is to first write a use case,
or all possible use cases that you can think of. Then write the
software. Then write tests to cover all the above use cases, and
release the software when it passes the tests. There will be unusual
use cases and edge cases which are not anticipated and for which
tests havent been written. Failure in these cases are bugs. But no
failure in the main use-case.
In the case of saving in OOo format and OOo not opening it, this is
not a bug. Because it is the main usecase of the particular feature.
A simple test - open an OOo file, save it in OOo format and open it
in OOo would have detected this, and the feature should not have been
released. Yes, bugs will be there - but when a main/common use case
fails - I dont think it can be called a bug and it is no laughing
matter either. I havent gone into kword code to find if there are
test suites, I hope there are ..."
Kenneth has echoed my thoughts exactly and put it nicely in words.
Having bugs in a software is possible but after more than 5 years if a
software behaves as if it was a beta and released just yesterday for the
first time, then this is a flaw in the software not some mistake or
attitude problem of the installing person. There is nothing hilarious
about a person complaining about such grave issues in the software. The
cell colour problem I had highlighted was not even connected to any
proprietary file
formats. It happens in Open formats.
M$ softwares may have more bugs, be less stable than nix based ones but
when they are in the stable working mode, at least they do the job they
are intended to do. And if FOSS fans feel it is below dignity to compare
libre software with M$ ones, then why is it that whenever a problem is
pointed out in the libre software, they start counting the bugs in M$
ones. Why can't they except that the particular software has serious
flaws that need to be corrected.
Free software is no longer the free blood donation camp. Today major
libre software and distro development is taking place through huge
sponsorship and donations and by corporate participation. The
programmers are compensated for their efforts, so while the institution
is giving away the software for free, the developer is not doing any
charity. Then why not expect quality work instead of a 'Fix it yourself'
community kit.
This is in response to Krish's reply to JTD's mail and I quote...
"actually the entire problem is that people don't realise the values of
running a script.
and by the way running a script is "nothing to do " with programming.
I think clicking on next next finish etc is not a real good way of
software setup."
Simplifying a procedure means that one can do more work instead of
wasting time running scripts. How would we feel if we got a mobile phone
where instead of pressing buttons, we had to create and run scripts
every time we wanted to dial a number or access the phone directory or
any other function. The same process is happening in the background but
the programmers have provided a beautiful interface to simply press a
button. If scripting is so important, why don't we start doing all our
home and office work in assembly language or machine code? This will
give us better control of the hardware. It is up to the programmer to
make software intuitive and easy to handle. It has no relation with
scripting skills of the vendor or user. Let each one do his job properly.
--
Regards,
Rony.
Knock Knock
Who's There?
Linux
Who Linux?
GNU/Linux