Been quietly following this...
On 21 Nov 2000, at 21:24, the Illustrious Paul Sokolovsky wrote:
> Hello Laura,
> Laura Michaels wrote on Thursday, November 16, 2000:
> LM> Return-Path: <lmemsm@...>
> LM> Delivered-To: paul@...
> LM> Received: (qmail 20230 invoked from network); Thu, 16 Nov
> 2000 14:54:24 +0200 LM> Received: from lmtp09.iname.net (HELO
> smv696-leg.iname.net) (18.104.22.168) by is.lg.ua with SMTP; Thu,
> 16 Nov 2000 14:54:24 +0200 LM> Return-Path:
> <lmemsm@...> LM> Received: from lists.sourceforge.net
> (mail1.sourceforge.net [22.214.171.124]) by smv696-leg.iname.net
> (8.9.3/8.9.1SMV2) with ESMTP id IAA13077 for
> <Paul.Sokolovsky@...> sent by LM>
> <lmemsm@...>; Thu, 16 Nov 2000 08:01:54 -0500 (EST)
> LM> Received: from glitch.crosswinds.net (glitch.crosswinds.net
> [126.96.36.199]) by lists.sourceforge.net (8.11.1/8.11.1) with
> ESMTP id eAGD1ru09902 for <pfalcon@...>; Thu,
> 16 Nov LM> 2000 05:01:54 -0800 LM> Received: from crosswinds.net
> (cwmail.crosswinds.net [188.8.131.52]) by glitch.crosswinds.net
> (8.9.3/8.9.3) with SMTP id IAA45993; Thu, 16 Nov 2000 08:00:01
> -0500 (EST) (envelope-from LM> lmemsm@...) LM> From:
> "Laura Michaels" <lmemsm@...> LM> Sender:
> lmemsm@... LM> Reply-to: lmemsm@... !!! LM>
> To: mingw-dvlpr-admin@... LM> Cc:
> earnie_boyd@..., pfalcon@...,
> pgarceau@..., lmemsm@... LM> Date: Thu, 16
> Nov 2000 08:02:18 +0500 LM> Subject: re: FAQ LM> Message-id:
> <3a13dada.d961.0@...> LM> MIME-Version: 1.0 LM>
> Content-Type: text/plain; charset="iso-8859-1" LM>
> Content-Transfer-Encoding: 7bit LM> X-UIDL:
> a1a0bf9da45977757d73dcfd384c9db0 LM> Earnie wrote:
> >>>You are subscribed as lmichaels@... and that
> >>>is where the
> >>>email would be sent to. If you do not see this message three
> >>>times then you
> >>>need to update your profile at sourceforge.net.
> >>>I give permissions to lmemsm@... to post to the
> >>>list so I'm seeing
> >>>from the list.
> LM> Got the message three times. I didn't see the message I sent
> get to the list LM> until it was forwarded. I'm e-mailing from
> LM> lmemsm@..., so if the permissions are set, it
> should LM> be going through.
> Please pay attention to line marked !!! above - there should
> mingw-dvlpr@... .
> LM> Paul Sokolovsky wrote:
> >>>But please note that
> >>>your FAQ currently contains lots of obsolete information,
> >>>including links and procedure in such important sections as
> >>>downloads and bug reporting.
> LM> If you'd e-mail me with what information has changed, I'll
> update it. Same LM> goes for anyone else. If you see something
> out-of-date or find a bad link, LM> please let me know.
> But why should I bother you and me with reporting, when I can
> change in the first place?
The idea here is to promote a team approach. This is probably
a cultural difference. Even so, for most of those of us who are
working primarily as members of a team, Lauras' approach is the
most favorable as it allows the team to work...any other
approach, even in the US corporate culture, is no longer
The western IT model has discovered, through decades of trial
and error, that when working with a project of this scope and
size, and given the specialties of the people involved, the team
approach is, in every case, the best approach.
There is an understanding that if someone doesn't know all
about a specific aspect of some project (any project), that
there is someone who does. It is also understood within the
Western IT development approach, that within the team, it is not
an individual, but a group that is most effective in scopes
(such as mingw for example) and the documentation of a given
> You think that such changes get out of
I do not believe it is about control. I do believe it is about
promoting a team (or sum of parts) over the promotion of an
The individual is far more important in a team approach than it
is otherwise...in fact, it is the very independence of the team
members that makes this team paradigm work, at least from the
Western IT perception.
>Once again, there's a CVSWeb which allows you easily to
> review all the changes,
CVS can not support the team communication that is necessary
for working on a shared project, such as the mingw faq. The
very best it can do is give an outline of what is going on. CVS
is not designed to support a team structure. It is designed to
support source revision structures, and has been show to do that
> including when the change was done, who
> did it, description of the change, and easy-to-look-at diff.
> to top-level CVSWeb for us is
> Specific page for mingwfaq.shtml is
>Example of diff I did while fixing user-reported
> (not even decided on my own) outfated stuff:
> Note that all this is essentially what you ask for: "let me
> know", but:
> 1. Person who makes effort for fixing immediately sees the
> result. For me, it's unquestionable benefit.
In fact, given the chrontab on the update, there is no
immediate turnaround. The only immediate turnaround happens
when you actually look at the faq on your own system. CVS is
updated as the last step in a very involved process which
includes a number of things starting with "What is incorrect or
needs to be changed...and why?"
CVS can not support such questions. It only records the final
update after someone, such as myself, has made the update to the
cvs repository at sourceforge. "diffs" are meaningless if there
is no means of tracking the reasons behind the updates which
have created such "diffs". Thus the need for someone to manage
some aspect of a project.
> doesn't depend on your perfomance
Nor should it...turnaround does depend on how connected a
specific someone is to the project at hand.
The example that immediately comes to mind is the testbed
source I use for mingw, CrystalSpace (CS). CS is a very large
3d rendering engine with an api that makes DX look like childs
In such case, I for one, do not have the time or the energy to
read every single diff when it is posted for projects such as CS
(CrystalSpace) simply because of the vast scope of that project.
The same is true when it comes to tracking whatever patches may
go through mingw...patches, again, are meaningless without some
sort of "raison de etre" and a well documented set of steps
which has been agreed to.
What is and "what will be" are two different things and have no
relationship to each other within the present. Vision is one
thing, technological expertise is another.
The scope of mingw is even broader than CS, and therefore can
not be controlled by a single person. Control is not at issue
> You still can review and edit changes (or even undo them if you
> are sure they are for bad), only after, not before the change.
> Unless you want be in total control, I don't see real difference
> (apart from benefits above).
The team needs to understand what is going on in order to
promote a fast turnaround. The only way a team can know what is
necessary is if they have some specific team leader, volunteer
or otherwise, who has understanding and vision of what the end
product is supposed to look like.
> Well, maybe you consider remembering visiting pages, and visiting
> them a burden? As Earnie proposed, we can setup email
> notification of CVS commits.
We already have too may notifications coming through the
developers mailing list. To add more to that seems extreme to
In fact, I would rather there was a specific patch notification
mailing list. The same applies in terms of any cvs updates that
might be added.
Again, I turn to the example of CrystalSpace. There is a
specific list for cvs reports and any patches that occur are the
responsibility of the type of developer in question (eg.
msvc/c++, win32gcc, borland, Mac, Solaris, Linux, etc.)
In the case of mingw, our current project, imho, patches and
cvs updates should be posted to a separate mailing list, and
anyone who wants to track those may have access to that mailing
> LM> Earnie wrote:
> >>>Early on, just after moving the source for the pages to CVS I
> >>>had asked if
> LM> we
> >>>should setup a mingw-patches list to which the differences of
> >>>such changes
> >>>could be discussed and approved before putting them into
> LM> Sounds like a good idea.
I agree with Earnie on this as it allows those of us who need
to be in touch with such changes to discuss them before they are
posted as "changes".
> To me, bad.
> >>>Given this, would you agree to be the "editor" in charge and
> >>>take control
> LM> of
> >>>the whole website, not just the FAQ?
> LM> Would really love to, but I'm working and going for my
> Masters LM> which is taking up most of my time right now. I had
> three web LM> sites of my own which I took down this year because
> I don't have LM> the time to properly maintain them.
It sounds like Laura wants to only deal with the faq at this
point in time. I can understand this from the standpoint that I
too have had to re-focus my energies from one thing to another
whenever I've started a new contract (employment) or project
[such as getting my degree(s)]. In the meantime, I do what I
can such as updating website, etc.
> >>> I could even setup a mingw-editor list if you think it's
> >>> needed.
> LM> This might help. If we could get a few volunteers for a list
> LM> like that and delegate responsibities, it might make
> maintaining the web site LM> easier.
I also agree with Laura here, as it allows a more organized
approach to updating the website, etc.
> Fairly speaking, I don't see how it will make maintaining
> easier - IMHO, it will make it more hard and complex.
I can't say "I agree" with this. Maintenance of pages is
simply a matter of having an itemized list of what needs to be
done, if anything.
It has nothing to do with making something either harder or
more complex in terms of maintenance. The exception to this
only occurs when it comes to allocating the necessary time and
energy to deal with whatever changes are determined to be
> it simply will be a bureaucracy, and any bureaucracy is known to
> make everything complex and waste resources.
Let's not confuse "bureaucracy" with "hierarchy". There are
major differences between the two.
> LM> For example, if we could get a person or a few people to
> volunteer LM> to update the documents to CVS, that would be _very
> Sorry, what do you mean? I am such person. I regularily
> mingw site (via CVS, of course). Note however that I do not
> volunteer - I'm techy person and don't have time for that, as any
> of us. Nonetheless - I *do* that. And I would really appreciate
> allowing me (and anyone else) doing that further.
> LM> It takes forever
> LM> on my machine to just access CVS.
If you need any help in updating your access speed, please let
me know, I may be able to help. I also have a 56k modem. DSL
would be nice, but that is not an option at this point in time.
> LM> I'd be happy to check over the web pages
> LM> for spelling/typograhical and readability problems
> There should be no problems with this.
> LM> and to check them for XHTML/HTML
> LM> validity.
> And that's of course good. I myself big adherent of standards
> used SGML before they did XML). But for me, the content and
> meaning in our case are primary, not the structure. For example,
> W3C XHTML validator currently complains on front page because it
> uses some dynamic URLs. I don't that something should be done
> aboy this, however.
> LM> Maybe we can also get a few people to check if links are
> It would be nice. But we do not have such people. However,
> some of
> us do it on occasion. And will do it in the future, if they will
> be able to *fix* them (not walk seven circles of bureaucracy).
Not sure what you are talking about here, Paul S. I have
updated links for the website on a number of occasions, and have
had absolutely no problems whatsoever.
> the other side, I will try to make everything to pay attention to
> such reports from community, and *fix* them *ASAP* (again,
> adequate access is required).
> LM> or to make sure the site is current with the latest news
> about mingw.
> This is done: infrastructure for centralized submission has
> set up, everyone is welcome to drop a note to admins (it's their
> poor fate to be in charge of news posting).
> LM> Perhaps
> LM> with a team in place and dividing up the work, it would make
> things more manageable.
> Don't want to push it to hard, but likely, not. Our primary
> aim is to
> provide adequate information, and second - to manage anything.
> So, we should set up easy-to-allow, easy-to-kick-off rules -
> likely, this will give better results *in our position* (no team
> in place).
Actually, there is a team in place...and has been since this
mailing list was created.
No person is an island...and anyone who has access to this
mailing list is part of the development team in some capacity or
another (whether they know it or not & whether or not they
acknowledge it)...if this is not understood, then it is not only
important, but imperative, that this misperception be corrrected
as soon as possible.
> LM> We could even go through one web site page a week on the
> list and discuss if LM> it's up-to-date or needs changing.
> I'm afraid, I won't be able to allocate much time to talks
> making things up-to-date. Nonetheless, I'd like to update
> out-of-date stuff and add new.
So, it sounds like, in some small way, you do want to be
involved with the website maintenance, Paul S.
Lauras' suggestion is a good one. Even so, those of us who can
contribute time only have a finite amount of time that we can
offer in a volunteer capacity.
> LM> What does everyone think of an idea like this
> LM> and are there any volunteers?
As I may have noted, Laura, I think your idea is an excellent
one. In terms of volunteering, the best I can do now is what I
have been doing...update the website when there is some request
to do so on an item-by-item basis.
If we choose to revamp the website in order to integrate both
the current mingw.org site and the proposed mingw.org site, that
will take some time as well as a good deal of planning and
availability of volunteers before it can even begin to happen.
> Please note that this is my (person who so far did most
> commits to
> CVS) opinion, and of course it doesn't apply to FAQ, which, per
> agreement, has been put into the original state.
> LM> Laura Michaels
> Paul Sokolovsky, IT Specialist
> MinGW-dvlpr mailing list
Nothing real can be threatened.
Nothing unreal exists.