1 |
My two cents: |
2 |
Do it like in linux kernel. |
3 |
|
4 |
The guys making the API change bear the burden of fixing anything it |
5 |
breaks, however, if something gets officially deprecated, don't go out of |
6 |
your way to support continued use. |
7 |
|
8 |
I for one would consider "ok, this method is not working, deprecate it so |
9 |
that it doesn't get any worse because we're going to be changing it later" |
10 |
is quite an appropriate way to bear the burden. |
11 |
|
12 |
Anyone getting grandfathered from doing it the old way escapes the burden, |
13 |
but by putting the entire community on notice via the GLEP process, well, |
14 |
grandpa got put out to pasture and after that point the community inherited |
15 |
the burden of not deepening the dependency on the freshly deprecated |
16 |
feature. |
17 |
|
18 |
Since this is a collective superproject it can also be said that we all |
19 |
bear the collective burden of cooperating towards a common goal. |
20 |
|
21 |
And yes, the personal attacks probably should die down. I'm not with |
22 |
devrel or comrel but even I can tell such sniping and whatnot can decrease |
23 |
overall productivity of the distro by bringing down morale. |
24 |
|
25 |
Solution? |
26 |
|
27 |
If anything broke because herds were removed? My suggestion is to leave |
28 |
behind herds as empty stubs somehow, officially deprecate herds (which the |
29 |
council already did from what I gather), and file a tracking bug to have |
30 |
herds removed entirely. |
31 |
|
32 |
Then just do what they do with new versions of gcc, and have existing tools |
33 |
that break when herds are removed have new bugs filed on them that block |
34 |
the tracking bug. The breakage cited in the first post would be a good |
35 |
example of a dependent bug that ought to block the "we still have herds" |
36 |
bug. |
37 |
|
38 |
What do you guys think? Just use a tracking bug to flush out anything |
39 |
still using herds like they do when new versions of GCC come out and break |
40 |
stuff? |
41 |
|
42 |
On Fri, Feb 12, 2016 at 2:02 PM, Michał Górny <mgorny@g.o> wrote: |
43 |
|
44 |
> On Fri, 12 Feb 2016 10:07:10 +0100 |
45 |
> Patrick Lauer <patrick@g.o> wrote: |
46 |
> |
47 |
> > On 02/12/2016 08:48 AM, Michał Górny wrote: |
48 |
> > > Dear Ignorant Patrick, |
49 |
> > Hello human! Your politeness module seems to have crashed. |
50 |
> |
51 |
> Please do not expect politeness when you insult someone. |
52 |
> |
53 |
> > > On Thu, 11 Feb 2016 21:15:34 +0100 |
54 |
> > > Patrick Lauer <patrick@g.o> wrote: |
55 |
> > > |
56 |
> > >> ... or why just changing stuff is not enough: |
57 |
> > >> |
58 |
> > >> A few days ago I was told that |
59 |
> > >> http://euscan.gentooexperimental.org/herds/ was displaying an empty |
60 |
> > >> list. Which is annoying because people sometimes want to see what |
61 |
> > >> upstream updates are available for their herd. |
62 |
> > >> |
63 |
> > >> Well, we renamed herd to project. Because reasons. |
64 |
> > > No, we didn't. Herd was collection a packages. Project is a collection |
65 |
> > > of developers. Project coexisted with herds for a long time. As it was |
66 |
> > > explained already in length. Multiple times. |
67 |
> > So you just pivoted the organization from A->B to B->A. |
68 |
> > |
69 |
> > I still don't see the advantage in that. Maybe I should have expressed |
70 |
> > my concerns more vocally, but in general I don't have time to worry |
71 |
> > about all the little things. |
72 |
> |
73 |
> You still have trouble understanding who did what. I'm tired of being |
74 |
> blamed for something that wasn't my idea. |
75 |
> |
76 |
> > >> I don't care how it is named, but this change broke euscan in a |
77 |
> > >> user-visible way. Now I could just try to rename things there too, but |
78 |
> > >> that won't work: |
79 |
> > >> |
80 |
> > >> euscan uses gentoolkit for parsing metadata.xml and herds.xml |
81 |
> > >> (Since herds.xml is basically unmaintained cruft at this point this |
82 |
> will |
83 |
> > >> break soon anyway ... but ...) |
84 |
> > >> Changing gentoolkit to use projects.xml instead of herds.xml won't be |
85 |
> a |
86 |
> > >> simple migration since the data organization changed. |
87 |
> > >> |
88 |
> > >> Now instead of looking up [metadata.xml] -> (herd name) -> [herds.xml] |
89 |
> > >> -> email it goes backwards: |
90 |
> > >> [metadata.xml] -> (maintainer type=project) -> email -> [projects.xml] |
91 |
> > >> -> Project name |
92 |
> > >> |
93 |
> > >> Since this involves XML and python's ElementTree library it's a |
94 |
> > >> nontrivial change that also removes a few now useless helpers |
95 |
> > >> (_get_herd_email has no reason to be, but we'd need a _get_herd_name |
96 |
> > >> helper instead. Err, get_proj ... ah well, whatever name works) |
97 |
> > >> |
98 |
> > >> And all that just so (1) gentoolkit output works and (2) euscan |
99 |
> updates |
100 |
> > >> properly. Both of which I don't really care about much, but now that |
101 |
> > >> I've invested ~4h into debugging and trying to fix it I'm a tiny bit |
102 |
> > >> IRRITATED. |
103 |
> > > You are completely incorrect, as you have been told already multiple |
104 |
> > > times. People would really appreciate if you spent at least a little |
105 |
> > > part of the time you spend complaining, inventing issues and insulting |
106 |
> > > others listening to what they're telling you. |
107 |
> > > |
108 |
> > > So let me repeat, again. euscan works. Want packages from Python |
109 |
> > > project? Then select the appropriate maintainer from the 'maintainers' |
110 |
> > > section: |
111 |
> > So you're saying I have no way to search by herd, err, project now. |
112 |
> |
113 |
> Yes, you have. You can use project's e-mail address to find |
114 |
> the project. And as I proved below, it works. |
115 |
> |
116 |
> > ... and metadata is now partially broken. |
117 |
> |
118 |
> Another of your unclear generic statements that mean nothing. |
119 |
> |
120 |
> > *ahem* This not of good idea sounding. |
121 |
> > > |
122 |
> > > http://euscan.gentooexperimental.org/maintainers/python@g.o/ |
123 |
> > > |
124 |
> > > Done. Was it that hard? Now the big surprise: you didn't have to create |
125 |
> > > some convoluted logic to get that! You don't need projects.xml to get |
126 |
> > > that! Of course, you'd know that if you would listen for a single |
127 |
> > > minute instead of throwing insults at others. |
128 |
> > If you had actually understood my criticism you would understand why I |
129 |
> > might be a tiny bit irritated. |
130 |
> > |
131 |
> > Some functionality is now actively *gone*, and that's not a feature. |
132 |
> |
133 |
> Yes, it's gone. However, it's relatively easy to bring it back. All you |
134 |
> have to do is enable filtering by type="". Which is definitely simpler |
135 |
> than having to process two disjoint data structures, one of them |
136 |
> requiring parsing additional XML file. But well, unnecessary complexity |
137 |
> was always considered a feature in Gentoo. |
138 |
> |
139 |
> > >> Please, next time someone has the brilliant idea of changing stuff |
140 |
> just |
141 |
> > >> to change it (I still don't see a reason why we had to change |
142 |
> > >> metadata.xml?), it should be required that support tools are fixed |
143 |
> > >> *before* the change, and working versions released. This avoids grumpy |
144 |
> > >> people and makes it harder for those that change things to |
145 |
> head-in-sand |
146 |
> > >> and claim everything works as expected when it obviously doesn't. |
147 |
> > > The fact is: things *work as expected*. If you have problem accepting |
148 |
> > > reality as it is, then it's your fault, not ours. Herds no longer |
149 |
> > > exist. Everything is based on *maintainers* now. Tools are not supposed |
150 |
> > > to magically turn project information back into herd-oriented design. |
151 |
> > Right, so gentoolkit returning bad info is a good thing. I find that |
152 |
> > hard to integrate into my understanding of the world ... |
153 |
> |
154 |
> Again. |
155 |
> |
156 |
> > Please don't redefine what 'expected' means to suit your limited |
157 |
> > usecases. It just causes friction and unhappy response from people that |
158 |
> > now have to spend lots of time figuring out how things diverge from |
159 |
> > their 'expected', which usually ends in *facepalm* omg how did that |
160 |
> happen. |
161 |
> > |
162 |
> > Plus the usual sequence of strongly-worded letters to the UN ;) |
163 |
> |
164 |
> I can't redefine it to suit unlimited usecases. I presume you're |
165 |
> capable of doing that. That's good news. I wonder why you haven't used |
166 |
> that ability to do something good. |
167 |
> |
168 |
> > > As I said before, please direct any further complaints directly to |
169 |
> > > the Council, and stop insulting the messenger. The Council has banned |
170 |
> > > herds explicitly before I even started working on GLEP 67. It was |
171 |
> > > the guideline I had to follow. |
172 |
> > > |
173 |
> > Hey thanks for publically demonstrating your unwillingness to cooperate |
174 |
> > with others. |
175 |
> |
176 |
> That's your opinion. Please do not expect people to be willing |
177 |
> and happy to work with toxic personalities like yours. So far you have |
178 |
> put a lot of effort on spreading confusion, making noise and doing no |
179 |
> good. I'd rather focus on doing something useful. |
180 |
> |
181 |
> > So now I know that in the future I will |
182 |
> > |
183 |
> > (1) categorically deny any change requests coming from you and |
184 |
> > (2) block/revert any changes that I don't like or understand. |
185 |
> > |
186 |
> > Nothing personal, just basic sanity. |
187 |
> |
188 |
> If you want to leave Gentoo, please do that without unnecessary drama. |
189 |
> That will reduce the load on ComRel and/or QA resulting from your CoC |
190 |
> violations and childish behavior, and reduce the discouragement you're |
191 |
> causing to other developers. |
192 |
> |
193 |
> -- |
194 |
> Best regards, |
195 |
> Michał Górny |
196 |
> <http://dev.gentoo.org/~mgorny/> |
197 |
> |