jeroencoumans at gmx.net
Mon Jun 9 08:08:07 PDT 2003
Alexander E. Patrakov wrote:
> On Monday 09 June 2003 11:28, Jeroen Coumans wrote:
>>I'm not trying to suggest LFS is insufficient wrt. its intended
>>audience, I'm trying to suggest it's intended audience and it's actual
>>audience are not clear from the book.
> Maybe we should put some objective criteria (e.g. required skills) to the
> book. E.g. a person who did not recompile at least his/her distribution
> kernel will probably be disappointed with LFS.
Well, I wouldn't want to set such a hard limit, but I do think that
alot of seperate pages should be folded into each other so people will
not skip them so easily. (people often click "next" before they're done
reading). For example; the whoread, whonotread and prerequisites should
be 1 page; named "assumptions". Alex Groenewoud suggested such a
structural change at
and there's even a follow-up mail by Gerard stating it's a usefull
change for the next book's release (then 4.0). I'm gonna file a bug here
so it won't get lost again.
>>Yes, of course, I meant with that line that the books structure _does_
>>allow and include newbies, who are after all more than capable of
>>reading and following the book; they just don't understand what they're
>>doing. I'd like the book to move more in a direction where one has to
>>understand what they're doing in order to be able to proceed (like BLFS).
> I do follow this thread, but since I am not a newbie, I fail to understand
> what makes BLFS book unsuitable for newbies. This should be explained to me
> by means of words.
>>Yes I see that. I should refine my point by saying that the book should
>>be more difficult for newbie's than it is right now, thus more properly
>>reflecting the intended and assumed audience. Also, more mechanisms
>>should be built in for excluding it's non-intended audience. A more
>>proper "whonotread" page is in order, along with more pointers to other
>>good tutorials so people who lack the appropriate skill level can do so.
>>Integrating the "prereading" hint is in order.
> And maybe even a short exam on the web page (didn't pass => don't read the
Well, I'm trying to advocate adjusting the book's structure and logical
flow so newbies are more generally ruled out because they have a harder
time following it. I don't think these quizes are effective; esp.
because they have the potential to rule out people who could read the
book but will be annoyed by the arrogance of setting such an entry point.
>>(confusion being the "fbbg rule" vs. the "LFS is a guide, not a rule"
> I agree. Maybe: "The authors of the book know that there must be no problems
> when the book is followed exactly. Before asking questions about unexpected
> results of deviation, check that the book's way of doing things is correct.
> Mails to th esupport lists without indication of that will probably meet
> negative attitude."
Yes, this information should be in the book, preferably somewhere on
the assumptions page.
>>I think the actual percentage doesn't matter as much as the question
>>*why* the book continually fails to adress or defer or redirect these
>>people to more appropriate documentation, and why they do not find the
>>FAQ but do find the support groups.
> A dragon's solution: mention the mailing lists at the _end_ of the FAQ and
> nowhere else.
We don't want information to become less accessible, so I don't think
this is a proper solution. Previously I and other people have suggested:
* when someone subscribes to a mailinglist, mail the FAQ or a link
thereto in the "welcome" e-mail. IIRC, at the time this was deemed
unappropriate because it's assumed that people check out the FAQ before
they subscribe. We need a reality check here; it's the most unobtrusive
way to redirect newbies and I'm seriously for this proposal.
* put a link to the FAQ in the signature of each mail, like the
unsubscribe... message. This doesn't help against people who post for
the first time and ask a FAQ.
* post (a portion of the) FAQ regularly to the mailinglists. This has
the same disadvantages of the above, and is also annoying for regulars
and a waste of bandwidth. I still don't understand why this solution was
Based on my experience on these lists, I can say that people generally
find the mailing lists _before_ they find the FAQ. This is a structural
problem with *both the book and the website*. In addition to the above
suggestion I propose the following:
1. the book should be changed so that the "asking for help" pages are
all collapsed and presented on one page; with a clear hierarchy on what
information channel should be pursued first. (FAQ, search, mail).
Adjusting these is probably easiest when in transition for a new major
book release. Now is the ideal time.
2. the website should present this information with the same hierarchy
so people are clear and can rightfully be flamed when they still post
FAQ's to the mailing lists.
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe lfs-dev' in the subject header of the message
More information about the lfs-dev