1 |
On Mon, 09 Jun 2008 11:49:35 +0200 |
2 |
Luca Barbato <lu_zero@g.o> wrote: |
3 |
|
4 |
> Ciaran McCreesh wrote: |
5 |
> > On Mon, 09 Jun 2008 10:50:11 +0200 |
6 |
> > Luca Barbato <lu_zero@g.o> wrote: |
7 |
> >>> So how, specifically, is PMS "wrongly written", and why hasn't |
8 |
> >>> anyone who thinks so bothered to provide details? |
9 |
> >> - rewrite it as an rfc using a markup among xmlrfc, docbook, |
10 |
> >> guidexml. |
11 |
> > |
12 |
> > What technical reason is there to use a markup that's more work for |
13 |
> > those of us doing the writing? Writing XML is a huge pain in the ass |
14 |
> > compared to latex. |
15 |
> |
16 |
> More people can understand those markups |
17 |
|
18 |
I've yet to see anyone have any difficulties with Tex. |
19 |
|
20 |
> they are consistent with the gentoo documentation |
21 |
|
22 |
GuideXML can't even begin to cover our requirements. Simple example: |
23 |
try to rewrite the following in GuideXML: |
24 |
|
25 |
---START--- |
26 |
Global variables must only contain invariant values |
27 |
(see~\ref{metadata-invariance}). If a global variable's value is |
28 |
invariant, it may have the value that would be generated at any given |
29 |
point in the build sequence. |
30 |
|
31 |
This is demonstrated by code listing~\ref{lst:env-saving}. |
32 |
|
33 |
\lstinputlisting[float,caption=Environment state between |
34 |
functions,label=lst:env-saving]{env-saving.listing} |
35 |
---END--- |
36 |
|
37 |
> they look better on screen than on paper |
38 |
|
39 |
That's highly questionable. And PMS is sufficiently long that printing |
40 |
it out is the easiest way of reading it, no matter what format it's in. |
41 |
|
42 |
> tex is a great typesetting markup to write academic books. Right tool |
43 |
> for the right task. It address the problem "PMS is anything but |
44 |
> accessible" |
45 |
|
46 |
How does making PMS twice the file size and five times as complicated to |
47 |
edit make it more accessible? |
48 |
|
49 |
> >> - use EBNF when describing a syntax. |
50 |
> > |
51 |
> > Is there any indication that this is any clearer? EBNF gets messy |
52 |
> > when it comes to describing the whitespace rules, for example. |
53 |
> |
54 |
> It is less ambiguous than natural language. That solves the issue |
55 |
> "The syntax is underspecified" |
56 |
|
57 |
In what places is the syntax currently underspecified, and in said |
58 |
places, why is EBNF a better solution that tightening up the existing |
59 |
language? Please illustrate your answer with real examples. |
60 |
|
61 |
-- |
62 |
Ciaran McCreesh |