[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: dpkg v2 ? I haven't heard about it, but ...

Ben Collins writes ("Re: dpkg v2 ?  I haven't heard about it, but ..."):
> Wrong, it's being written in C. Error in your ASSumtion.

I think I should have looked more closely at the list of people
involved.  Perhaps I'm confusing this new effort with a previous one.
As I say, I heard a rumour that it was to be done in C++ by people who
think C++ is The One True Programming Language.  I'm pleased to hear
that the dpkgv2 people think it's not (not that I really want to get
into the question as to whether it's a good answer to this particular
problem); I hope that you'll agree with me that there is no such thing
as The One True Programming Language.

> [response to my provocation deleted]

Well, you may think my attitude sucks, which I'm suppose from your
point of view it does.

I'm sorry; I have a habit of being overly rude, particularly when

I think I should explain why I'm annoyed, not by way of justification,
but in case you can find yourself agreeing with it.

The root cause of my annoyance is that certain people keep saying they
cannot understand my code.  I don't think the dpkg source is hard to
understand for competent programmers.  By `competent' I'm not
intending to refer to a small minority.  I think that if you've spent
some time writing your own C and reading and writing other people's,
and can with reasonable confidence write reliable and correct
programs, then you should have little difficulty with dpkg's code.  I
think I have some evidence to support this; about half of the patches
to dpkg proper from the NMU series are correct.

More to the point, I think that inability to understand dpkg's code is
a good indication of enough lack of ability that I wouldn't want those
people to be trying to rewrite it.  Furthermore, I think that apart
from perhaps adding documentation (eg, internal code documentation, or
comments), there aren't many areas in which the code could get much
clearer (though of course, there's room for improvement, and in some
cases room for quite a lot of improvement).

I hope that in fact the dpkg v2 team's comments in your spec. document
about some people finding the code difficult don't apply to the dpkg
v2 team, but are just reflections of other people's view.

Re dpkg v2:

I acknowledge that dpkg has been fallow too long; I'm now working on
it again actively.

I think that it's probably a good thing for both my dpkg and yours
that both projects exist.  Maybe we'll spur each other on.  It would
be no doubt be helpful if changes were made to make it easier to `mix
and match' components, and I shall try to do this.

So, when you say:
> Please, if a project like this has to be started to get you off your ass,
> then it was well worth it. But that fact also means I wont be quitting my
> efforts, it means I will strive harder. It's always been said that the
> threat of code forking and divergence are the best motivators in free
> software...let's see how well it motivates you.

Then possible fine, I agree :-).

I hope that us all getting annoyed with each other won't prevent us
working together constructively.  To this end I'll answer the other,
technical, questions people have posed, in a different email.


Reply to: