On Thu, Mar 01, 2001 at 02:10:22PM +1000, Anthony Towns wrote: > Hi guys, > > Some questions, which you might or might not like to answer.. > > * What do you think are the three big things Debian should achieve > over the next twelve months?
In no particular order:
* Fish or cut bait with respect to non-free, and pass any constitutional
amendments necessary to permit us to make that choice;
* Release at least once, preferably twice;
* Be instrumental in compelling the US government to lift all restrictions
on exportation of cryptography
> * What do you think will be the three major problems Debian will face
> over the next year or two?
Sorry, I have to cheat and list four:
* accumulation of inactive maintainers
* accumulation of unmaintained packages
* inconsistent timeliness of published security advisories
* version skew between our many supported architectures hamstringing our
package pools
Arguably, these are all problems we have right now. I'd rather not see
them get worse, but I fear they will.
> * Debian has had four DPLs so far: Ian Murdock, Bruce Perens, Ian
> Jackson, and Wichert Akkerman. What do you think each of these DPLs did
> right, and are there any examples of things you would do differently
> in their place (with or without the benefit of hindsight)?
Ian Murdock:
I was not around for his tenure and so cannot say a great deal about his
style of project leadership. However, since I've worked for him for the
past several months I can say the following: Ian is very cautious about
ideology (in contrast to his successor), has a great deal of energy, and
knows exactly what we wants within a given scope that he is careful to
define. This kind of focus and drive, coupled with a real pragmatism (as
opposed to pragmatism-as-ideology that one finds in some pundits,
particularly those who criticize RMS) were, I believe, absolutely
critical to getting Debian off the ground in the first place. I was not
around to witness his mistakes as Project Leader (if any), so I cannot
speak to them.
Bruce Perens:
Bruce's biggest strength was almost certainly in drawing positive
attention to Debian. He issued press releases and was probably very
important in getting Debian onto people's radars, especially in the
press. On the downside, he probably let his temper get the best of him
in a public way a little too often. While I don't recall Debian
suffering from any bad press as a result of this, it probably wasn't too
good for morale. I am also not convinced that dropping out of the first
DPL election as soon as there was another candidate was a good idea. For
an election to be worthwhile, it should be contested. Contested
elections are a powerful reminder to elected officials that they are
there to serve, and not because they are entitled to the office. I'm not
sure Bruce felt an entitlement to the position of Project Leader, but I
suppose it is possible.
Ian Jackson:
Ian started out very well by recognizing and acting on the fact that
Debian had grown to the point where the Benevolent Dictator model of
leadership wasn't really sustainable. However, he very quickly retreated
into inactivity (or at least invisibility), which, in my opinion, swayed
too far the other direction, making the project seem -- and perhaps feel,
to some of our developers and users -- rudderless, or at the very least
without a spokesman. I think it's very important to have a figurehead,
if for no other reason than that it gives the press someone to interface
with when they're too lazy to find the right person to talk to, which
often happens. :)
Wichert Akkerman:
I think Wichert did an outstanding job for much of both his terms,
striking a good balance between the extremes of Bruce and Ian Jackson; he
was re-elected handily, and deservedly so. However, I think that into
his second term, he began to lose interest in the job a little bit. I am
also not persuaded that he was willing to apply pressure where it was
really needed in some problem areas (see my platform for more detail).
I'm not sure why, but if I had to guess I'd say that perhaps Wichert was
nervous about hurting someone's feelings, or making them so angry that
they quit. In the case of people who are so inactive that you wouldn't
notice if they DID quit (except for the positive effect that they can
clearly be replaced), I think you have to be able to get past this
squeamishness. If you couldn't guess, I am speaking here about delegates
rather than developers in general; if someone, e.g., mismanages a
particular package, the developers as a group are pretty effective at
applying pressure on someone to straighten up -- in the case of
inactivity, we have the well-established tradition of the NMU.
In the final analysis, however, I believe Wichert has been the best
Project Leader we have had to date. Perhaps Debian had a Golden Age back
in the days when 0.93R6 was first released, but that was before my time
(even though it was the first version of Debian I used -- there was a
considerable lag before 1.1).
> * The DPL is a fairly loosely defined position, so the successful
> candidate can probably use that position to provide leadership in a
> fair few areas. Bruce seemed particularly successful at publicising
> Debian, and Ian Jackson did a fair bit of stuff getting Debian
> set up procedurally. If elected, what sort of areas would you see
> yourself focussing on: technical, political, procedural, publicity,
> or something else? How would you see the areas you don't have the
> time or ability to handle being dealt with?
I don't perceive myself as having an outstanding degree of technical
facility; on the other hand, we have no shortage of people who do, and it
is pretty easy for purely technical skills to be channeled where they are
needed in this Project without the direct intervention of the Project
leader. One area I would like to be involved in is making sure that, in
the area of architecture support, we can get hardware in the hands of, or
near to, people experienced with it and desirous of working with it. While
ports are often started by people who already have a machine of the given
architecture, as that port matures, sometimes machines end up far from
those with the time and knowledge to take care of them. However, I don't
see this as being a problem yet, and it might not ever be.
As a Debian developer and as a U.S. citizen, I overflow with political
opinions; however, I think it is important to maintain neutrality where
feasible when speaking as Project Leader, and I'd like to resurrect Ian
Jackson's practice of sending mail from "[EMAIL PROTECTED]" when wearing
the DPL hat, since I am likely continue partipating in the occasional
thread, or even flamewar, as a mere mortal. :)
Procedurally, as mentioned in my platform, I'd like to help expedite
formulation and revision of our existing procedures where necessary, only
taking a personal hand in drafting procedures when no one else will.
In terms of publicity, I'd like to see what I can do to get some fresh
blood into the Debian publicity team, since Nils Lohner must be getting
pretty lonely in that role. I'm happy to give interviews to the press,
appear at trade shows, etc.
> * Debian is becoming increasingly appropriate for commercial purposes:
> whether as a server in a place of business, or as a base for a derived
> distribution, or as a base on which to run proprietry software. What
> sort of things should we be doing to make Debian more suitable for
> these uses, for example, having someone people can call and tell
> their credit card number to, in case their server goes down and
> they need it back up right now, or making our non-free archive more
> readily available for people distributing no-cost (or trial-only)
> proprietry software (such as jdk 1.2, perhaps)?
I think we should use the partners area of the Debian website to provide
pointers to such off-site resources, in return for some consideration by
the support company / vendor in question. This "consideration" could be
monetary, it could be employment of Debian developers to work on Debian
packages (a la VA and Progeny), or anything else that can be agreed upon.
To avoid any conflict of interest, I'd be interested in appointing a
delegate, perhaps one of our many developers who is still in college and
thus (theoretically) not yet shackled to any particular commercial
interest.
> * Debian currently has a reasonable amount of cash thanks to both
> donations and awards. What should we be doing with it?
First and foremost, we should be screaming at SPI to produce reports on
this very subject. Deposits, withdrawls, etc.
Aside from that, there are some things I think Debian should NOT spend its
money on:
* trade shows; the day a trade show asks a non-profit to pay for a booth
is the day we should organize some sort of protest. On the bright
side, I think even the starchiest of suits understands that the .Org
Pavilion (or whatever it is called at a given show), is where the heart
and soul of Free Software resides.
* hosting; there are plenty of sites on the net that use Debian on their
servers, and if we have to start paying for our own bandwidth we'll be
bankrupt in a month. Fortunately, I don't think this is going to be a
problem.
Two things I can think of that would be good uses of money:
* Machines and their components (especially components), when efforts to
solicit donations fail;
* Expenses for trade shows; this includes materials for the booth (nice
signs, T-shirts, etc.), but NOT travel or boarding expenses for developers
(not even the DPL). If elected, I could be persuaded to expense travel
and accomodations for travel to trade shows (wink). No, seriously, the
day Debian has trouble finding sponsors for its Project Leader to
attend a show is probably the day we need to consider whether we belong
at that show. If I understand correctly, Wichert got flown to so many
trade shows earlier this year it literally wore him out.
> * Do you still use proprietry, non-free, or unpackaged software (or
> anything else not distributed by Debian) for anything? If so, what,
> and is anything being done to remedy this unacceptable situation?
Well, I get this out of the way and plead guilty, mostly due to laziness:
Non-free packages installed on apocalypse
fortify World-wide strong cryptography for Netscape
fortify-linux-x86 Fortify stub files for linux-x86 architectures
I think I can get rid of these, but I continue to commit the
greater sin of having Netscape (4.76) installed. I don't have it
open right now (instead, I'm using Konqueror and Mozilla).
I think the day will come soon when I can scrap Netscape, but I
confess I haven't done so yet. I will see about confirming that
the SSL support in Konqueror is good enough to let me do my online
banking and purchasing; if so, then I really do have no reason to
keep Netscape and fortify on my machine.
gimp-nonfree GIF and TIFF support for the GNU Image Manipulation Pr
libmagick4g-lzw Image manipulation library (non-free version).
These are in non-free only because of the Unisys patent; if you
examine their debian/copyright files you will find that are in fact
freely licensed. In my personal opinion, only failing the DFSG can
render something non-free. Brain-damaged patent laws are not the
fault of the free software developer.
gpg-idea IDEA (PGP 2.x-compatible) module for GNU Privacy Guard
This is another patent problem; gpg-idea is freely licensed (BSD +
obnoxious advertising clause).
gs-aladdin Postscript interpreter with X11 and svgalib preview su
Not sure why I have this installed instead of the free gs. Maybe
it was to test some goofy bug someone reported in X.
html2latex Convert HTML markup to LaTeX markup
Hrm, I don't actually use this. I should probably remove it. The
license is quite definitely non-free. Can someone suggest
alternatives? We could also attempt contacting the author, Nathan
Torkington. It was copyrighted 7 years ago, and the profile of
free licensing has risen considerably since then.
maelstrom An arcade-style game resembling Asteroids.
I remember participating in a licensing discussion about this game.
The copyright holders have presented us with a Catch-22; the game
engine is freely licensed, but it is against the license of the
non-free images and sounds to distribute them separately from the
game itself. So, we can have an unplayable game in Debian (because
we cannot legally package the graphics and sounds in isolation), or
the whole ball of wax can go in non-free. Not a good situation.
The game is pretty good; perhaps another effort should be made to
contact the authors.
mpg123 MPEG layer 1/2/3 audio player
This has a hand-rolled license. Is there a freely-licensed MP3
player available? I am aware of the patent issues but view them
with scorn.
pgp-us Public key encryption system (US version)
I haven't needed to use this in a long time, but I may keep it
around for a while longer until GnuPG has stabilized more, just so
I can rest easily that I can continue to decrypt old materials.
spim MIPS R2000/R3000 emulator.
I installed this when a person was offering to donate me an SGI
machine; that fell through and I can probably remove it. However,
it might be worth contacting the author again; given the footnotes,
in the copyright file, it may be possible that his desires are not
in fact in conflict with the DFSG.
wdg-html-reference WDGs HTML 3.2, HTML 4, and CSS references
It is my opinion that standards documents should be freely
licensed, and Debian should probably try to convince the W3C to do
so.
xautolock start a program if the X session is idle for some time
Aside from Netscape, this is the only package listen that I use
every day; I use it to launch xlock because I despise xscreensaver.
Its functionality could be trivially re-implemented as a very short
Xlib program, and in fact I have considered doing so before.
However, the notations in the copyright file indicate to me that
the author might be happy with the GNU LGPL; perhaps he did not
choose that license because he appears to prefer short licenses to
long ones. Whatever their virtues may be, I do not think the GNU
licenses can claim brevity among them. :)
xmame Multiple Arcade Machine Emulator
xmame-x X binaries for Multiple Arcade Machine Emulator
I don't even fool with MAME anymore. However, as I recall it used
to be freely licensed, except for the fact that (nominally) this
made it too tempting for people to bundle it with CD-ROM's full of
copyright game ROMs from big mean companies, and that the change to
a non-free license was attempt to hold off packs of lawyers from
Nintendo who would sue them along with the guys who actually broke
the law. I am not sure that was a good decision, and I know for a
fact it didn't stop people from violating the license both on MAME
an on the (unlicensed) ROMs themselves, having seen stacks of MAME
CD's available for sale at local computer fairs.
And now to confess the really big sin: I have Windows 98 installed on
/dev/hda1, as well as a bunch of games (which I paid for), and a non-free
piece of shareware called CoolEdit 2000, which I use to edit sound files
that I rip from CD's in conjunction with practicing guitar (it is easier to
figure out a guitar solo if you can slow it to half speed while keeping the
pitch the same).
I used to have more non-free software installed, but I purged it since I
didn't use it.
On the other hand, the other 3 computers I own have not a whit of non-free
software installed on them.
> * What do you think about the social environment of Debian? For instance,
> we tend to have a different flamewar every week; is this a
> demonstration of our firey passion for what we're doing, or that we're
> about to crash and burn? Should we be doing anything in particular to
> change the demographics of our membership; perhaps trying to increase
> the number of female developers, or documenters, or artists, or at
> least to make it easier for such people to contribute? If so, what?
> Are developers too removed from the userbase (with separate mailing
> lists and IRC channels), or is the mix still pretty good?
I am not convinced that the social environment is something that CAN really
be changed deliberately; it is the net result of the actions of the many
people within it. A single molecule in a gas cannot compel the motion of a
large number of its fellows.
I don't find the mailing lists to which I subscribe to be needlessly
contentious, or particularly unfriendly places. When people are idiots,
they get told so. Good ideas are usually affirmed as well. Incoherent
posts are often ignored.
I'm not sure there is a great deal Debian can do in terms of recruiting
members to do any particular task, and almost certainly not to
demographically impact our membership. I do not feel there are any racial,
national, religious or sexual prejudices manifest in the Project as a
whole, and none that could keep a rational individual from wanting to join
us. I think Debian is pretty committed to being a meritocracy:
"Are you willing and able to do good work for the Project? If so, welcome
aboard, otherwise, please just stand back and let us move on."
--
G. Branden Robinson | I suspect Linus wrote that in a
Debian GNU/Linux | complicated way only to be able to have
[EMAIL PROTECTED] | that comment in there.
http://www.debian.org/~branden/ | -- Lars Wirzenius
pgpOmFCaFijxL.pgp
Description: PGP signature
