1 |
-----BEGIN PGP SIGNED MESSAGE----- |
2 |
Hash: SHA512 |
3 |
|
4 |
Michał Górny: |
5 |
> On Thu, 28 Apr 2016 19:41:06 -0400 Göktürk Yüksek |
6 |
> <gokturk@××××××××××.edu> wrote: |
7 |
> |
8 |
>> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 |
9 |
>> |
10 |
>> Brian Dolbec: |
11 |
>>> On Thu, 28 Apr 2016 15:39:05 -0400 Göktürk Yüksek |
12 |
>>> <gokturk@××××××××××.edu> wrote: |
13 |
>>> |
14 |
>>>> --- metadata.dtd | 5 +---- 1 file changed, 1 insertion(+), 4 |
15 |
>>>> deletions(-) |
16 |
>>>> |
17 |
>>>> diff --git a/metadata.dtd b/metadata.dtd index |
18 |
>>>> 7626a57..b608852 100644 --- a/metadata.dtd +++ b/metadata.dtd |
19 |
>>>> @@ -3,7 +3,7 @@ <!ATTLIST catmetadata pkgname CDATA ""> |
20 |
>>>> |
21 |
>>>> <!-- Metadata for a package --> -<!ELEMENT pkgmetadata ( |
22 |
>>>> (maintainer|natural-name|longdescription|slots|use|upstream)* |
23 |
>>>> )> +<!ELEMENT pkgmetadata ( |
24 |
>>>> (maintainer|longdescription|slots|use|upstream)* )> <!ATTLIST |
25 |
>>>> pkgmetadata pkgname CDATA ""> <!-- One tag for each |
26 |
>>>> maintainer of a package, multiple allowed--> @@ -13,9 +13,6 |
27 |
>>>> @@ explicit type) for Gentoo maintainers is prohibited. --> |
28 |
>>>> <!ATTLIST maintainer type (person|project|unknown) |
29 |
>>>> "unknown"> |
30 |
>>>> |
31 |
>>>> - <!-- Natural name for package, example: LibreOffice (for |
32 |
>>>> app-office/libreoffice) --> - <!ELEMENT natural-name |
33 |
>>>> (#PCDATA) |
34 |
>>>>> - <!-- A long description of the package in freetext--> |
35 |
>>>> <!ELEMENT longdescription (#PCDATA|pkg|cat)* > |
36 |
>>> |
37 |
>>> Isn't this almost obsolete? it's now xmlschema... And I hope |
38 |
>>> to have the new repoman with it out this weekend :) |
39 |
>> |
40 |
>> Does GLEP 68 explicitly declare metadata.dtd obsolete? I see that |
41 |
>> the example metadata.xml on the GLEP is missing DOCTYPE, are we |
42 |
>> getting rid of those too? |
43 |
> |
44 |
> No, and I don't know. |
45 |
> |
46 |
> metadata.dtd is still required by many tools, and as such it makes |
47 |
> sense to keep it. However, we may want to put some warning that |
48 |
> it's not very strict, and allows major structural violations due |
49 |
> to technical limitations. |
50 |
> |
51 |
After a discussion with ulm on IRC, we agreed that the following makes |
52 |
sense: "the format of the metadata is defined in GLEP 68. the syntax |
53 |
is defined in metadata.dtd. The xml-schema can be used for stricter |
54 |
validation checks." If you have no objections, I will update devmanual |
55 |
based on this description. |
56 |
|
57 |
> As for DOCTYPE, there was no formal decision on that. It's not |
58 |
> technically required, so the GLEP doesn't enforce it. However, I |
59 |
> don't expect it being gone anytime soon. One of the reasons behind |
60 |
> it is that repoman enforces it as part of metadata.xml validation. |
61 |
> If it is to be gone, stable repoman needs to accept it missing at |
62 |
> least for some time. |
63 |
> |
64 |
> There was a proposal to link new/additional schemas in |
65 |
> metadata.xml files. However, I personally don't think we should go |
66 |
> this way. DOCTYPE already proved troublesome (when we switched to |
67 |
> https), and maintaining any kind of schema reference in XML files |
68 |
> doesn't give us any real benefit (we don't enforce any hard |
69 |
> defaults besides languages there, and I don't think many XML |
70 |
> parsers would respect that anyway). |
71 |
> |
72 |
>> I understand that the DTD is more like a super-set, so anything |
73 |
>> that complies with GLEP 68 will comply with the DTD as well. |
74 |
>> However, there is a caveat here: for example the GLEP dismisses |
75 |
>> the list of possible values for <remote-id/> by saying "The list |
76 |
>> of available trackers and their specific identifiers are outside |
77 |
>> scope of this specification." but does not mention where these |
78 |
>> values shall be kept either. The moment we add a new remote-id, |
79 |
>> the xmlschema diverges from the DTD and stops being a subset. |
80 |
> |
81 |
> Well, there is good reason for not hardcoding the list in the |
82 |
> GLEP, and this obviously is to avoid updating the GLEP for every |
83 |
> single change. And yes, as you can see, I didn't point out a |
84 |
> specific location where remote-ids are to be defined to avoid |
85 |
> problems like the one noted below. |
86 |
> |
87 |
> As I see it, we can defer it to some project / wiki page, or |
88 |
> simply keep it in either of the schemas. Either way, we should keep |
89 |
> both schemas reasonably up-to-date. |
90 |
> |
91 |
>> Besides, the PMS says the format of metadata.xml is described in |
92 |
>> DTD. Even if we move to something else, doesn't metadata.dtd need |
93 |
>> to be kept around until the PMS is amended? |
94 |
> |
95 |
> The key point PMS is making is 'outside the scope'. The specific |
96 |
> location is just a minor problem, and it should be updated. I'll |
97 |
> take a look at it. |
98 |
> |
99 |
> Oh, and +1 for the patch. If nobody complains and nobody beats me |
100 |
> up to it, I'll commit it this weekend. |
101 |
> |
102 |
|
103 |
Please hold off the merge. I have found a few more things that require |
104 |
fixing. I'll send a new patch series for review. |
105 |
|
106 |
-----BEGIN PGP SIGNATURE----- |
107 |
|
108 |
iQEcBAEBCgAGBQJXJFJcAAoJEIT4AuXAiM4zcTMH+wcmYptb9MPo+7oSRx9ViS22 |
109 |
RVwUKiwdb55JmMHLXWu7ATQS+X1Jb5LqlNxIdpUz77zfP9+WFJJagJI9F1oi7a4s |
110 |
e9wi7a1J3H812Xo2JcRZGnp95OX2O/c7dyXCcG5VMwB8kqW8JMykb+QmI0opm9s7 |
111 |
ePX0e1lQ3mhaMy8thTRBZK9a1cA+B86PWBtQRtQtb3leotD06DDnO0sZchLYsDYZ |
112 |
A0qTDOZ+V5t9qJ0EtNJ8jb9T/7dz971noZUwdUDgkveU1Q+3o/IWBzZD765WaPYs |
113 |
dSlWhn8linRLk3UhHe95+vCL0NAuLLdEEPIiNNrLRG3orz/gfbz7kl0E8+JNW58= |
114 |
=DLlT |
115 |
-----END PGP SIGNATURE----- |