1 - Strategy introspection and disabling.
2 - prepare-config (and others) create .wizard dir even
3 when not strictly necessary
4 - Bug out immediately if tags are not present in the master tip
6 - pending doesn't seem to get written out properly sometimes (or
7 it's being deleted); this makes it hard to --continue on the
8 event of an upgrade failure. Also, we seem to bounce back to
9 the production copy to check pending even when we run --continue
11 - Newline checks are /really really/ expensive on AFS; see if
12 we can minimize them or something. Right now, we're testing
13 a fix where we don't clone with --shared.
14 - Replace .split("\n") with .splitlines()
16 - Need to fix existing repo history? (not adding extra commits;
17 that'll be more difficult)
19 git rebase -i -p --root --onto COMMITID
21 This won't work if you need to change the very root of the
22 repository. You'll probably end up with conflicts and have
23 to manually resolve everything afterwards.
25 But usually you won't need --root --onto unless you really
26 fucked up the pristine branch. If you just need to change
29 git rebase -i -p COMMITID
33 - [SCRIPTS] MediaWiki 1.6.7, 1.9.3 and 1.10.0
36 "Connection to scripts.mit.edu closed" is confusing
37 the URL should be easier to copy and paste, which means we should
39 We should ... upgrade our autoinstaller
40 Apparently installing WordPress updates or themes never indicates
41 completion, and just says "Downloading..", and you have to guess
45 - phpBB or phpbb? (right now it's the former)
46 - need an upgrade story; srv needs more versions
47 - need a story about install/ contrib/
49 - Give users a "certificate" of their merge, which they can
50 use to reuse that merge commit if something unrelated fails.
52 - Human readable quota output
53 - Nice error message on --continue if you forgot to git add your
54 resolved file (look for conflict markers)
55 - The merge interface is a kind of major UI disaster; you won't
56 be able to use it unless you know how Git works. Also, the
57 merges can be quite difficult to resolve if upstream has made
58 large formatting changes like reindenting. We may also consider
59 providing a --rebase option, which seems to do better when
60 big problems like this show up.
62 - Better mass-* support for just one user (this includes letting
63 a user mass upgrade just his own autoinstalls)
65 - Show progress or something when upgrading
66 - Allow 'sticky notes' for future upgraders to notice
67 - .wizard/url semantics are subtly wrong: in particular, if we
68 explicitly configure a URL at install, we should be able to
69 detect this URL as baked in from the configuration
71 - Rerere support doesn't actually work
72 - "Version 3.0.0 doesn't exist; did you mean 3.0?"
73 - Be a little more intelligent when perform web checks; for example,
74 if we get a forbidden message, that probably means we go the right
75 address but it's blocked off; if we get a 404 message, that probably
76 means wrong address. Account Unknown is something particularly good
78 - Wordpress module can do something intelligent if we get redirected
79 to the installation page.
80 - wizardResolve* files seem to get left in tmp en-mass, and we don't
83 - [SCRIPTS] Wordpress needs to get rid of the siteurl hack, so that it
84 actually has a fully-qualified URL http://foo.scripts.mit.edu/blah.
85 This will also fix Wordpress's cron functionality. We should be
86 careful not to write over users who are on vhosts. We should figure
87 out who is still on twiddle paths. We should make sure the redirect
90 - Remerges aren't reflected in the parent files, so `git diff` output is
91 spurious. Not sure how to fix this w/o tree hackery.
92 - Sometimes users remove files. Well, if those files change, they automatically
93 get marked as conflicted. Maybe we should say for certain files "if they're
94 gone, they're gone forever"? What is the proper resolution?
96 - Parse output HTML for class="error" and give those errors back to the user (done),
97 then boot them back into configure so they can enter in something different
99 - [SCRIPTS] If you try to do an install on scripts w/o sql, it will sign
100 you up but fail to write the sql.cnf file. This sucks.
102 - [SCRIPTS] Web application for installing autoinstalls has a hard
103 problem with credentials (as well as installations that are not
104 conducted on an Athena machine.) We have some crazy ideas involving a
105 signed Java applet that uses jsch to SSH into athena.dialup and
109 - Tidy up common code in callAsUser and drop_priviledges in shell;
110 namely cooking up the sudo and environment variable lines
111 - Summary script should be more machine friendly, and should not
112 output summary charts when I increase specificity
113 - util.fetch() should use urllib under the hood, not httplib. Code
114 has to be changed. We should log if we get redirected.
117 - Add support for mypristine workflow
118 - Figure out why Sphinx sometimes fails to crossref :func: but wil
119 crossref :meth:, even though the dest is very clearly a function.
120 Example: :func:`wizard.app.php.re_var`
121 - The TODO extension for Sphinx doesn't properly force a full-rebuild
122 - Make single user mass-migrate work when not logged in as root. The
123 primary difficulty is making the parallel-find information easily
124 accessible to individual users: perhaps we can do a single-user
125 parallel-find on the fly.
126 - Don't use the scripts heuristics unless we're on scripts with the
127 AFS patch. Check with `fs sysname`
128 - Make 'wizard summary' generate nice pretty graphs of installs by date
129 (more histograms, will need to check actual .scripts-version files.)
130 - It should be able to handle installs like Django where there's a component
131 that gets installed in web_scripts and another directory that gets installed
134 - [SCRIPTS] ACLs is a starting point for sending mail to users, but it
135 has several failure modes:
136 - Old maintainers who don't care who are still on the ACL
137 - Private AFS groups that aren't mailing lists and that we can't get
138 to A question is whether or not sending mail actually helps us:
139 many users will probably have to come back to us for help; many
140 other users won't care.
141 - Whatever happens here should be used to improve user.email()
143 [ XXX: metadata.rst ]
145 COMMIT MESSAGE FIELDS:
147 Installed-by: username@hostname
148 Pre-commit-by: Real Name <username@mit.edu>
149 Upgraded-by: Real Name <username@mit.edu>
150 Migrated-by: Real Name <username@mit.edu>
151 Wizard-revision: abcdef1234567890
152 Wizard-args: /wizard/bin/wizard foo bar baz
156 Committer: Real Name <username@mit.edu>
157 Author: lockername locker <lockername@scripts.mit.edu>
161 [ XXX: doc/deps.rst ]
162 * Some parts of the infrastructure will not be touched, although I plan
163 on documenting them. Specifically, we will be keeping:
165 - get-homedirs.sh. which needs to be run as root on scripts. Store
166 in /mit/scripts/sec-tools/store/scriptslist
168 - parallel-find.pl, and the resulting
169 /mit/scripts/sec-tools/store/versions
171 [ XXX: doc/deploy.rst ]
172 GET APPROVAL BEFORE PROCEEDING ANY FURTHER;
173 THIS IS PUSHING THE CHANGES TO THE PUBLIC
175 NOTE: The following commands are to be run on not-backward.mit.edu.
176 You'll need to add daemon.scripts-security-upd to
177 scripts-security-upd to get bits to do this. Make sure you remove
178 these bits when you're done.
180 10. Run `wizard research appname`
181 which uses Git commands to check how many
182 working copies apply the change cleanly, and writes out a logfile
183 with the working copies that don't apply cleanly. It also tells
184 us about "corrupt" working copies, i.e. working copies that
185 have over a certain threshold of changes.
187 11. Run `wizard mass-upgrade appname`, which applies the update to all working
190 12. Run parallel-find.pl to update our inventory
192 [ XXX: doc/metadata.rst ]
193 * The repository for a given application will contain the following files:
195 - The actual application's files, as from the official tarball
197 - A .scripts directory, with the intent of holding Scripts specific files
198 if they become necessary.
200 - .scripts/dsn, overriding database source name