1 |
Yannick Koehler <yannick.koehler@××××××××.com> writes: |
2 |
|
3 |
> So far what I learned is that the whole system should be optionnal. |
4 |
|
5 |
if you're thinking "XML", yes. it should be. |
6 |
|
7 |
> It must retain support for existing configuration files for several |
8 |
> reasons such as working with current applications. It must not |
9 |
> gives a feeling of duplication but still provide a comprehensive way |
10 |
> for software to deal with the various configuration. |
11 |
|
12 |
ick. this sounds a lot like marketspeak. honestly, the day someone |
13 |
starts writing XSLTs for configuration files, I wish them pity. for |
14 |
one thing, adding a single option in a given config suddenly |
15 |
requires another DTD. another DTD which is in no way _really_ |
16 |
related to the previous one. maintenance becomes interesting. the |
17 |
XSLT work also has to change. |
18 |
|
19 |
second, we'd have the issue of verbosity. writing XML isn't |
20 |
something you do easily during emerge. imagine wanting to port from |
21 |
one version of the initscripts to another -- you might just want to |
22 |
add one single variable. we'll need to require a full XML parser |
23 |
just to do such a task, instead of adding a line to the end of a |
24 |
file. unless this gives us a very large benefit, I don't see it |
25 |
happening at all. |
26 |
|
27 |
third. most configurationfiles are flat. very flat. with a very |
28 |
loose order. changing this isn't a good idea in itself, it's not |
29 |
needed to change either. using XML to store a very flat hierarchy |
30 |
is usually overkill. |
31 |
|
32 |
and fourth, and this is very general. KISS. things work very well |
33 |
today. again, if we're going to change it, there should be some |
34 |
great promise of good things happening. again, I don't see it. :/ |
35 |
|
36 |
> The idea is evolving a little more. What if the system was not |
37 |
> going to create files, but was going to apply transformation rules |
38 |
> from a given folder when reading the file and then parsing it |
39 |
> internally into xml to then use inside some kind of application like |
40 |
> merge or configurator. |
41 |
|
42 |
whoa. this would make emerge optionally depend on a full XML |
43 |
suite. fair enough, 4Suite mostly works, but why? and if it is |
44 |
optional, will people use it? |
45 |
|
46 |
> So you'll have /etc which remains intact, a software that is |
47 |
> optionnaly installed with a bunch of transformation text file such |
48 |
> as xslt which would provide instruction to the software on how to |
49 |
> parse the original file and transform it into an xml one. |
50 |
|
51 |
honestly, that XSLT-file I want to see before I will give it much |
52 |
faith. I handcoded XSLT for a long while and I don't see this as a |
53 |
trivial task. how many developers know their XML and XSLT well |
54 |
enough to do this (properly)? |
55 |
|
56 |
> Then using xml tools merge software could be written and xml editor |
57 |
> used. The xml then become a software interface and never interfere |
58 |
> with existing etc folder. |
59 |
|
60 |
you'd have to interfere with /etc, otherwise inconsistencies would |
61 |
emerge (okay, bad pun :) and all hell would break loose. |
62 |
|
63 |
> What do you think on that one? |
64 |
|
65 |
I think it sounds like the debates that happen now and then on using |
66 |
XML in /proc on linux-kernel. it doesn't _give_ as much as it |
67 |
demands. by far. if anyone can prove the gain from such a change, |
68 |
I'd rethink my stance, but so far the idea scares me. it adds a lot |
69 |
of complexity and solves very few problems, if any. :/ |
70 |
|
71 |
and no, I don't work with XML anymore. I happily got away from it. |
72 |
two years of trying to make DTDs "fake" inheritance and deal with |
73 |
version changes now and then got the best of me. if I was going to |
74 |
rant about XML in general, and not about this specific issue, I'd |
75 |
say "XML is very often the wrong solution to the wrong problem". |
76 |
|
77 |
-- |
78 |
Terje |