Summary:
`strlen()` was used in Phabricator to check if a generic value is a non-empty string.
This behavior is deprecated since PHP 8.1. Phorge adopts `phutil_nonempty_string()` as a replacement.
Note: this may highlight other absurd input values that might be worth correcting
instead of just ignoring. If phutil_nonempty_string() throws an exception in your
instance, report it to Phorge to evaluate and fix that specific corner case.
Closes T15307
Test Plan:
Applied these two changes (on top `D25144`, `D25145`, `D25146`, `D25147`, `D25150`,
`D25151`, `D25152`, `D25153`, `D25154` locally applied) and `/conduit/` finally
rendered in web browser. Also the page /conduit/query/.
Reviewers: O1 Blessed Committers, valerio.bozzolan
Reviewed By: O1 Blessed Committers, valerio.bozzolan
Subscribers: speck, tobiaswiese, valerio.bozzolan, Matthew, Cigaryno
Maniphest Tasks: T15307
Differential Revision: https://we.phorge.it/D25156
Summary:
`strlen()` was used in Phabricator to check if a generic value is a non-empty string.
This behavior is deprecated since PHP 8.1. Phorge adopts `phutil_nonempty_string()` as a replacement.
Note: this may highlight other absurd input values that might be worth correcting
instead of just ignoring. If phutil_nonempty_string() throws an exception in your
instance, report it to Phorge to evaluate and fix that specific corner case.
Closes T15304
Test Plan:
Applied these four changes (on top of `D25144`, `D25145`, `D25146`, `D25147`, `D25150`,
`D25151`, `D25152`, `D25153`) and `/conpherence/` page finally rendered in web browser.
Reviewers: O1 Blessed Committers, valerio.bozzolan
Reviewed By: O1 Blessed Committers, valerio.bozzolan
Subscribers: speck, tobiaswiese, valerio.bozzolan, Matthew, Cigaryno
Maniphest Tasks: T15304
Differential Revision: https://we.phorge.it/D25154
Summary:
`strlen()` was used in Phabricator to check if a generic value is a non-empty string.
This behavior is deprecated since PHP 8.1. Phorge adopts `phutil_nonempty_string()` as a replacement.
Note: this may highlight other absurd input values that might be worth correcting
instead of just ignoring. If phutil_nonempty_string() throws an exception in your
instance, report it to Phorge to evaluate and fix that specific corner case.
Closes T15308
Test Plan:
Applied this change (on top of `D25144`, `D25145`, `D25146`, `D25147`, `D25150`,
`D25151`, `D25152`, `D25153`, `D25154`, `D25155`) and `/config/module/` rendered
in web browser.
T15308
Reviewers: O1 Blessed Committers, valerio.bozzolan
Reviewed By: O1 Blessed Committers, valerio.bozzolan
Subscribers: speck, tobiaswiese, valerio.bozzolan, Matthew, Cigaryno
Maniphest Tasks: T15308
Differential Revision: https://we.phorge.it/D25158
Summary:
`strlen()` was used in Phabricator to check if a generic value is a non-empty string.
This behavior is deprecated since PHP 8.1. Phorge adopts `phutil_nonempty_string()` as a replacement.
Note: this may highlight other absurd input values that might be worth correcting
instead of just ignoring. If phutil_nonempty_string() throws an exception in your
instance, report it to Phorge to evaluate and fix that specific corner case.
Closes T15305
Test Plan:
Applied these two changes (on top of `D25144`, `D25145`, `D25146`, `D25147`, `D25150`,
`D25151`, `D25152`, `D25153`, `D25154`)) and "Use Results -> Add Panel to Dashboard"
overlay dialog finally rendered in web browser.
Reviewers: O1 Blessed Committers, valerio.bozzolan
Reviewed By: O1 Blessed Committers, valerio.bozzolan
Subscribers: speck, tobiaswiese, valerio.bozzolan, Matthew, Cigaryno
Maniphest Tasks: T15305
Differential Revision: https://we.phorge.it/D25155
Summary:
`strlen()` was used in Phabricator to check if a generic value is a non-empty string.
This behavior is deprecated since PHP 8.1. Phorge adopts `phutil_nonempty_string()` as a replacement.
Note: this may highlight other absurd input values that might be worth correcting
instead of just ignoring. If phutil_nonempty_string() throws an exception in your
instance, report it to Phorge to evaluate and fix that specific corner case.
Closes T15300
Test Plan: Applied these three changes (on top of D25144, D25145, D25146, D25147) and `/diffusion/` finally rendered in web browser.
Reviewers: O1 Blessed Committers, avivey
Reviewed By: O1 Blessed Committers, avivey
Subscribers: speck, tobiaswiese, valerio.bozzolan, Matthew, Cigaryno
Maniphest Tasks: T15300
Differential Revision: https://we.phorge.it/D25150
Summary:
Fix numerous PHP 8.1 "strlen(null)" exceptions which block rendering the initial Account
Creation page in a fresh Phorge installation.
The strlen() was used in Phabricator to check if a generic value was a non-empty string.
For this reason, Phorge adopts phutil_nonempty_string() that checks that.
Note: this may highlight other absurd input values that might be worth correcting
instead of just ignoring. If your phutil_nonempty_string() throws an exception, just
report it to Phorge to evaluate and fix together that specific corner case.
Closes T15279
Test Plan: After these code changes the account creation page got displayed (though without CSS and JS).
Reviewers: O1 Blessed Committers, valerio.bozzolan
Reviewed By: O1 Blessed Committers, valerio.bozzolan
Subscribers: avivey, speck, tobiaswiese, valerio.bozzolan, Matthew, Cigaryno
Maniphest Tasks: T15279
Differential Revision: https://we.phorge.it/D25137
Summary:
`strlen()` was used in Phabricator to check if a generic value is a non-empty string.
This behavior is deprecated since PHP 8.1. Phorge adopts `phutil_nonempty_string()` as a replacement.
Note: this may highlight other absurd input values that might be worth correcting
instead of just ignoring. If your phutil_nonempty_string() throws an exception, just
report it to Phorge to evaluate and fix together that specific corner case.
Closes T15286
Test Plan: Applied these five changes and `/project/view/1/` finally rendered in web browser.
Reviewers: O1 Blessed Committers, valerio.bozzolan
Reviewed By: O1 Blessed Committers, valerio.bozzolan
Subscribers: speck, tobiaswiese, valerio.bozzolan, Matthew, Cigaryno
Maniphest Tasks: T15286
Differential Revision: https://we.phorge.it/D25140
Summary:
`strlen()` was used in Phabricator to check if a generic value is a non-empty string.
This behavior is deprecated since PHP 8.1. Phorge adopts `phutil_nonempty_string()` as a replacement.
Note: this may highlight other absurd input values that might be worth correcting
instead of just ignoring. If your phutil_nonempty_string() throws an exception, just
report it to Phorge to evaluate and fix together that specific corner case.
Closes T15287
Test Plan: Applied these three changes and `/config/` finally rendered in web browser.
Reviewers: O1 Blessed Committers, valerio.bozzolan
Reviewed By: O1 Blessed Committers, valerio.bozzolan
Subscribers: speck, tobiaswiese, valerio.bozzolan, Matthew, Cigaryno
Maniphest Tasks: T15287
Differential Revision: https://we.phorge.it/D25141
Summary:
See Q59. Fixes rP935d7120ee32.
Call sites should be happy to use PhutilURI when possible.
Test Plan: visit any Repository page. Before - exception, now - data.
Reviewers: O1 Blessed Committers!, valerio.bozzolan
Reviewed By: valerio.bozzolan
Subscribers: speck, tobiaswiese, valerio.bozzolan, Matthew, Cigaryno
Differential Revision: https://we.phorge.it/D25139
Summary:
Fix numerous PHP 8.1 RuntimeExceptions caused by the deprecation of strlen(null).
The strlen() was used in Phabricator to check if a generic value was a non-empty string.
For this reason, Phorge adopts phutil_nonempty_string() that checks that.
Note: this may highlight other absurd input values that might be worth correcting
instead of just ignoring. If your phutil_nonempty_string() throws an exception, just
report it to Phorge to evaluate and fix together that specific corner case.
Closes T15264
Test Plan: Phorge homepage is displayed on PHP 8.1 after applying these changes
Reviewers: O1 Blessed Committers, valerio.bozzolan
Reviewed By: O1 Blessed Committers, valerio.bozzolan
Subscribers: speck, tobiaswiese, valerio.bozzolan, Matthew, Cigaryno
Maniphest Tasks: T15264
Differential Revision: https://we.phorge.it/D25132
Summary:
On a fresh installation with PHP 8.1, CSS does not load on the account creation page.
Closes T15283
Test Plan: After applying this change (and D25137), the account creation page is rendered correctly. See the screenshot in T15283
Reviewers: O1 Blessed Committers, valerio.bozzolan
Reviewed By: O1 Blessed Committers, valerio.bozzolan
Subscribers: speck, tobiaswiese, valerio.bozzolan, Matthew, Cigaryno
Maniphest Tasks: T15283
Differential Revision: https://we.phorge.it/D25138
Summary:
This change fixes the following recent JavaScript error:
Error: Empty ID passed to JX.$()!
The regression was introduced here:
90f9da643d
Closes T15272
Test Plan: - with this change I was able to open Conpherence Persistent Chat and Drop a file
Reviewers: O1 Blessed Committers, Dylsss, avivey
Reviewed By: O1 Blessed Committers, avivey
Subscribers: Dylsss, speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15272
Differential Revision: https://we.phorge.it/D25134
Summary:
When you visit the Members page of a Parent Subproject,
some actions are greyed out by design. And, if you click on
them, nothing really useful happens to continue with what
you want to do.
For example, see this page and click on "Join Project":
https://we.phorge.it/project/members/1/
After this change, the interface at least provides a link to
proceed to the next step of what you wanted to do - that is -
allowing to quickly list direct Subprojects and do something
more meaningful than that.
A/B for the "Join Project" view:
| Before | After |
|-----------|-----------|
| {F282337} | {F282331} |
A/B for the "Add Members" view:
| Before | After |
|-----------|-----------|
| {F282338} | {F282332} |
Just to be nice, this change adds some inline documentation
on some methods of AphrontDialogView used here.
Closes T15258
Test Plan:
- Create a Project (A) with a Subproject (B)
- Visit Members page of (A)
- Click on Join Project and click on the Blue Button
- Click on the Add Member and click on Blue Button
- in both cases it should get you to Subprojects list
Reviewers: O1 Blessed Committers, Matthew
Reviewed By: O1 Blessed Committers, Matthew
Subscribers: speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15258
Differential Revision: https://we.phorge.it/D25127
Summary:
Accessing a project's workboard URL of a non-existing workboard shows RunTimeException in PHP 8.1:
preg_match(): Passing null to parameter #2 ($subject) of type string is deprecated
Closes T15262
Test Plan: Page whether to create a workboard for a project was displayed after this change
Reviewers: O1 Blessed Committers, avivey, valerio.bozzolan
Reviewed By: O1 Blessed Committers, avivey, valerio.bozzolan
Subscribers: avivey, speck, tobiaswiese, valerio.bozzolan, Matthew, Cigaryno
Maniphest Tasks: T15262
Differential Revision: https://we.phorge.it/D25131
Summary:
A previous change added a couple of Countdown-related classes:
7ed35123a3
This change updates the related library map (arc liberate) in order to
fix any related regression (like PhutilMissingSymbolException).
Closes T15271
Test Plan:
- you run `arc liberate` and the file does not change
- you do not see any `PhutilMissingSymbolException` exception
Reviewers: O1 Blessed Committers, valerio.bozzolan
Reviewed By: O1 Blessed Committers, valerio.bozzolan
Subscribers: speck, tobiaswiese, valerio.bozzolan, Matthew, Cigaryno
Maniphest Tasks: T15271
Differential Revision: https://we.phorge.it/D25133
Summary:
Add a new RemarkupMetadata class and use with upload button as well as drag and drop pathways.
With this change, files are now attached automatically to the object. This this could solve
several issues, for example when a person uploads a file but only that author is able to see it.
Ref T15106
Test Plan:
- Drag and drop file, upload file with button. Check that both files are attached with "attachedFilePHIDs" values.
- drop file in a Task description (now works)
- drop file in a Task comment (now works)
- drop file in an edited Task comment (still not supported)
- This was already tested in Wikimedia Foundation: https://phabricator.wikimedia.org/D1203
Reviewers: O1 Blessed Committers, avivey, Matthew
Reviewed By: O1 Blessed Committers, avivey, Matthew
Subscribers: MCPCN, avivey, speck, tobiaswiese, valerio.bozzolan, Matthew, Cigaryno
Tags: #remarkup
Maniphest Tasks: T15106
Differential Revision: https://we.phorge.it/D25052
Summary:
This change adds the `countdown.create` capability to
allow to limit countdown creation.
This change comes from the Wikimedia Phabricator and
was originally proposed by v.bozzolan and implemented
by the kind user 20after4.
https://phabricator.wikimedia.org/T258599
Closes T15208
Ref T15081
Test Plan:
- go to the page /applications/view/PhabricatorCountdownApplication/
- play a bit with the "Can Create Countdowns" option
- note that it is respected
Reviewers: O1 Blessed Committers, valerio.bozzolan
Reviewed By: O1 Blessed Committers, valerio.bozzolan
Subscribers: speck, tobiaswiese, Matthew, Cigaryno
Tags: #countdown, #policy_archived
Maniphest Tasks: T15208, T15081
Differential Revision: https://we.phorge.it/D25101
Summary:
After this change I was able to generate again the Diviner
documentation from PHP 8.2 using the related command:
./bin/diviner generate
Closes T15255
Test Plan:
- run `./bin/diviner generate` - great success
- tested the advanced search in various ways - it still works
Reviewers: O1 Blessed Committers, Matthew
Reviewed By: O1 Blessed Committers, Matthew
Subscribers: Ekubischta, speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15226, T15255
Differential Revision: https://we.phorge.it/D25124
Summary:
We want users to use Ponder as primary support vector.
T15084
Test Plan: Generated diviner, saw that the reference works correctly.
Reviewers: O1 Blessed Committers, valerio.bozzolan
Reviewed By: O1 Blessed Committers, valerio.bozzolan
Subscribers: speck, tobiaswiese, valerio.bozzolan, Matthew, Cigaryno
Differential Revision: https://we.phorge.it/D25125
Summary:
Fix a regression introduced here:
96ae4ba13a
I reproduced this exception executing "svn commit" on a hosted repository.
That crash happened because the PHP getenv() function can return false.
But, that is a very terrible value that blasts the non-string-empty check.
So, now the default getenv() value is skipped, without causing problems.
Closes T15253
Ref T15190
Test Plan: - I've run `svn commit` and I have not encountered any issue now
Reviewers: O1 Blessed Committers, avivey
Reviewed By: O1 Blessed Committers, avivey
Subscribers: speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15253, T15190
Differential Revision: https://we.phorge.it/D25122
Summary:
After this change, if you have a deprecated git version,
now you see a related information config warning:
{F276840}
Having said that the git version 1.8.3.1 - that for example
is provided by CentOS 7 - is surely problematic;
Some sources indicate all git versions *before* 2.5.0 as not
able to handle the '--' escape argument. The '--' arg
is used to separate normal git flags from user arguments,
so that they cannot be mistakenly exchanged. Kind of:
git <flags> -- <arguments>
The problem is, Phabricator/Phorge at the moment executes
this kind of git commands when you surf a git repository from
the web interface:
git cat-file -t -- <hash>:<file>
But, if your git version does not support "--", you can
get yourself into unhappy situations since "--" could be
just interpreted as a "wrong value", causing other
considerations and misleading exceptions, such as:
This path was a submodule at <repo>:<hash>
If you have seen the above error while surfing a git repo,
and if you have 2.5.0 or before, and if you are sure that
the mentioned path it's not a submodule, you probably
need a git update.
It is not yet clear whether Phabricator/Phorge should
support the possibility of this lack of support for "--".
In the meanwhile, just update your git on the server.
AFAIK no particular version is required on your clients.
Related information:
- https://unix.stackexchange.com/a/740621/85666
- b48158ac94
If you disagree, please do not simply ignore the
warning but share your experience in the Task.
Ref T15179
Test Plan:
- open Phorge, you see no warnings since you are
already up to date. Nice.
- Otherwise, you see a nice config message, and you
can ignore it as usual.
Reviewers: O1 Blessed Committers, Cigaryno, avivey
Reviewed By: O1 Blessed Committers, Cigaryno, avivey
Subscribers: avivey, speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15179
Differential Revision: https://we.phorge.it/D25089
Summary:
Phutil is not yet loaded during preamble, so we can't use it.
This change fixes a regression introduced here, fixing PHP 8.1 support:
{96ae4ba13acbf0e2f8932e950a92af0495f034d7}
Ref T15064
Test Plan: Included test script.
Reviewers: O1 Blessed Committers, valerio.bozzolan
Reviewed By: O1 Blessed Committers, valerio.bozzolan
Subscribers: speck, tobiaswiese, valerio.bozzolan, Matthew, Cigaryno
Maniphest Tasks: T15064
Differential Revision: https://we.phorge.it/D25114
Summary:
I noticed that - historically - setDownload() could also be used with
false, true, or other weird values to activate a download filename.
So, this change continues to give full compatibility to PHP 8.1 but
with extra validations.
This also adds a bit of inline documentation to put this more explicit.
For more context see the previous version:
{5952b0a31b6aac0718bc23aefe43560b9bfe8cc5}
{96ae4ba13acbf0e2f8932e950a92af0495f034d7}
Ref T15190
Test Plan:
- Drop a file in a comment, click, download
- See that it still works
Reviewers: O1 Blessed Committers, Cigaryno, avivey
Reviewed By: O1 Blessed Committers, Cigaryno, avivey
Subscribers: avivey, speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15190
Differential Revision: https://we.phorge.it/D25113
Summary:
This change introduces a builting query to easily list
Administrators of your beautiful Phorge instance:
{F277912}
Note that any builtin query can be disabled anytime or
reordered, so this is not a problem if you already have
a similar saved query. But probably you have not.
The goal is not to just "have a link" (this was already
possible) but to have a predictable link to be used in
any instance-agnostic document, including but not
limited to Diviner or Phorge's user interface messages.
Closes T15238
Test Plan: Visit /people/query/admin/ and enjoy.
Reviewers: O1 Blessed Committers, Cigaryno, Matthew
Reviewed By: O1 Blessed Committers, Cigaryno, Matthew
Subscribers: speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15238
Differential Revision: https://we.phorge.it/D25116
Summary:
The left part of a Removed Comment can contains User Badges that,
after this change, are more consistent with the new general softness.
| Before | Proposed |
|------------|-----------|
| {F277371} | {F277373} |
Probably five people in a million will notice this change, BUT,
these people will have a huge, deep, breath of relief, knowing that
Phorge takes care about their obsessive-compulsive impulses, to
have everything nice and consistent and neat. You are welcome!
Closes T15235
Test Plan:
- do something to deserve a Badge
- share a Comment somewhere (Maniphest?)
- delete that comment
- check that the badge is slightly softer
Reviewers: O1 Blessed Committers, avivey
Reviewed By: O1 Blessed Committers, avivey
Subscribers: speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15235
Differential Revision: https://we.phorge.it/D25112
Summary:
This small restyle makes any Removed Comment a little less
prominent than normal ones, with the goal of decreasing a
bit your in-page distractions and increase your individual
productivity in your business by at least 250 milliseconds
every 48 hours of hard work in front of your monitor.
| Before | After |
|---------------------|---------------------|
| {F274834,size=full} | {F274835,size=full} |
This implementation (which is called "Kasper on Diet")
contains these specific changes for Removed Comments:
- user icon visibility: reduced by ~50% (-> Kasper)
- black "trash" icon: reduced by ~50% (-> Diet)
- texts: visibility reduced by ~50%
- vertical padding: reduced from 16px down to 4px
Note that if your Phorge is under the Serious Business Mode,
it seems it is still technically possible to manually
activate the "Decaying Curse" proposal mentioned in the Task.
Closes T15192
Test Plan:
- Add a Comment "I love Phorge"
- Add a Comment "I love Phabricator"
- Mark the second Comment as Removed
- Call a person at your desk
- Plug that person to an eyeball tracker
If the general attention focuses first on a normal Comment and then
on the Removed Comment, this change works perfectly.
Reviewers: O1 Blessed Committers, Cigaryno, avivey
Reviewed By: O1 Blessed Committers, Cigaryno, avivey
Subscribers: speck, tobiaswiese, Matthew, Cigaryno
Tags: #comments
Maniphest Tasks: T15192
Differential Revision: https://we.phorge.it/D25096
Summary:
After this change, the page "/manage/" related to Diffusion
repositories has an "Actions" button for mobile users:
{F276113}
Before this change, that button was missing only from this
specific page, and so it was really difficult for them
(if not impossible) to edit repository's basic information:
{F276115}
Do not confuse the hamburger menu on the sidebar with this
"Actions". The hamburger contains queries like Active Repo,
All Repo, etc. Instead the Actions contains
Edit Basic Info, Allow Dangerous Changes, etc.
It seems the cause of this problem is originated from an
apparently incomplete rollback occurred in Sep 6 2017 from
Phabricator:
e1fd74ddb5
As you can see, the DiffusionRepositoryBasicsManagementPanel
was probably not completely restored from the previous
version, at least for the part related to the method
newActionList():
bb369c7b71
The method newActionList() was restored in every other
panel but no the one related to the Basics Management. So,
a new PhabricatorActionListView object was manually
created, but without generating an ID from the Celerity
utilities.
In short, this change assigns an ID to this Curtain, in
order to correctly generates a related "Actions" menu
again. Otherwise, on mobile the curtain (with all the
actions) is automatically hidden, without any alternative.
Closes T15067
Test Plan:
- Repository > Actions > Manage Repository
- Reduce the page until "Actions" appears (yeeh!)
Reviewers: O1 Blessed Committers, Cigaryno, avivey
Reviewed By: O1 Blessed Committers, Cigaryno, avivey
Subscribers: speck, tobiaswiese, Matthew, Cigaryno
Tags: #diffusion, #mobile
Maniphest Tasks: T15067
Differential Revision: https://we.phorge.it/D25109
Summary:
This patch changes a bit how your search results from Phriction
are described in the autocomplete component (the "Typehead"):
{F276843}
After this change, Phriction search results will also contain the word
"Wiki page" so that users can better understand what the result
actually is. Just like a Diffusion repository mentions that it's
a "Repository", and just like a Project mentions that it's a
"Project" and so on.
Before this change, the Typehead entries were only mentioning
the slug of that wiki page.
| Before | After |
| {F274820} | {F272278} |
It's unclear if the previous behavior was a mistake (since the
internal parameter of the Typehead is called "type", and so, it
is supposed to mention the application type, not the slug.
Anyway, as a compromise, the slug is still mentioned.
To be honest this is just an excuse to put the middle dot /
aka interpunct character in this project again. Yeah, here the
middle point was used as separator. The apparent reason is:
because the middle point was already in use elsewhere in Phorge.
The real reason is: I'm a lobbyist in a corporation that sells
interpuncts worldwide as our core business, and so, I designed
thisvil plan to put the following middle point - again - in
Phorge (evil laugh).
Closes T15213
Test Plan:
- Search "Change Log" in the up-right bar (or similar)
- You see "Wiki Page · change_log/" (or similar)
Reviewers: O1 Blessed Committers, Cigaryno, avivey
Reviewed By: O1 Blessed Committers, Cigaryno, avivey
Subscribers: avivey, speck, tobiaswiese, Matthew, Cigaryno
Tags: #phriction
Maniphest Tasks: T15213
Differential Revision: https://we.phorge.it/D25102
Summary:
This description was somehow broken since 2019:
{F276845}
Probably this was caused by a Remarkup breaking change introduced by this
specific XSS security fix:
{54bcbdaba94a3573e128c6498816dbfa41d3a9cb}
Maybe in the future the backtick operator will be supported again
in Remarkup. Or, maybe not. So, let's just workaround to make that
specific page "less exploded" again.
Closes T15230
Test Plan:
- visit the page /applications/view/PhabricatorSearchApplication/
- check that now the Remarkup is rendered perfectly
Reviewers: O1 Blessed Committers, Cigaryno, avivey
Reviewed By: O1 Blessed Committers, Cigaryno, avivey
Subscribers: avivey, speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15230
Differential Revision: https://we.phorge.it/D25110
Summary:
This is a fix for PHP 8.1 deprecation of strlen(NULL), for these Phorge components:
- scripts
- aphront
- project
The strlen() was used in Phabricator to check if a generic value was a non-empty string.
For this reason, Phorge adopts phutil_nonempty_string() that checks that.
Note: this may highlight other absurd input values that might be worth correcting
instead of just ignoring. If your phutil_nonempty_string() throws an exception, just
report it to Phorge to evaluate and fix together that specific corner case.
Closes T15223
Ref T15190
Ref T15064
Test Plan: - check with your big eyes that there are no obvious typos
Reviewers: O1 Blessed Committers, avivey
Reviewed By: O1 Blessed Committers, avivey
Subscribers: speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15223, T15190, T15064
Differential Revision: https://we.phorge.it/D25105
Summary:
This change avoids some unnecessary uses of the strlen() function,
actually fixing some deprecation warnings in PHP 8.2.
In short, this is the suggested universal replace:
-if(strlen($v))
+if(phutil_nonempty_string($v))
And, if you know PHP, this is also another adoptable replace, but
only for cases where you are sure that the string "0" is not useful:
-if(strlen($v))
+if($v))
As usual the optimal solution depends on the contest.
Other similar patches will probably follow.
Closes T15222
Ref T15190
Test Plan:
- for the first time in my life, with this change, the unit tests are passed in PHP 8.2
- check with your big eyes that there are no obvious typos
Reviewers: O1 Blessed Committers, avivey
Reviewed By: O1 Blessed Committers, avivey
Subscribers: avivey, speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15199, T15190, T15222
Differential Revision: https://we.phorge.it/D25104
Summary:
This change fixes a regression introduced here:
rP5e2b3677157889104a7e540d7772a04f13164037
Thank you to the user @dadalha for auditing that commit
and sharing a stack trace to easily fix the issue.
In short, if you see this exception you are affected:
```
EXCEPTION: (InvalidArgumentException)
Value provided to "replaceQueryParam()" for key "tags" is NULL.
Use "removeQueryParam()" to remove a query parameter.
at [<arcanist>/src/parser/PhutilURI.php:341]
```
This change just fixes that specific problem.
Closes T15221
Test Plan:
1. Create a new Project
2. Create a new Milestone inside
3. Create the Milestone's Workboard
4. Visit the Milestone's Workboard and note that now it works
Reviewers: O1 Blessed Committers, avivey
Reviewed By: O1 Blessed Committers, avivey
Subscribers: speck, tobiaswiese, Matthew, Cigaryno, dadalha
Tags: #workboard
Maniphest Tasks: T15221
Differential Revision: https://we.phorge.it/D25103
Summary:
Hide the "hidden" fields on custom form previews.
Before this change, Phabricator's custom form preview doesn't actually hide the
hidden fields, instead it shows them at 50% opacity. It looks cluttered and
doesn't provide a very useful "preview" at all.
This just hides some fields via CSS.
Cherry picked from:
https://phabricator.wikimedia.org/rPHABbba62cf5243538af9e37cc1211a01d247294f9f7
Upstream Task:
https://phabricator.wikimedia.org/T209743
Test Plan: Tested in Wikimedia's fork. I believe it results in better UX.
Reviewers: #blessed_committers, O1 Blessed Committers, Cigaryno, avivey, Matthew, valerio.bozzolan
Reviewed By: #blessed_committers, O1 Blessed Committers, Cigaryno, avivey, valerio.bozzolan
Subscribers: avivey, Cigaryno, speck, tobiaswiese, valerio.bozzolan, Matthew
Maniphest Tasks: T15081
Differential Revision: https://we.phorge.it/D25037
Summary: Closes T15200
Test Plan:
Test 1:
I was able to run `arc unit --everything` without the error thrown from T15200
Test 2:
I also tried this and it still works:
```
php -a
require 'src/applications/metamta/exception/PhabricatorMetaMTAReceivedMailProcessingException.php';
$asd = new PhabricatorMetaMTAReceivedMailProcessingException("1", "TEST MESSAGE");
var_dump($asd->getMessage());
```
You get:
```
string(12) "TEST MESSAGE"
```
Reviewers: O1 Blessed Committers, Matthew
Reviewed By: O1 Blessed Committers, Matthew
Subscribers: speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15200
Differential Revision: https://we.phorge.it/D25099
Summary:
After this change, search results with resolved
questions are striked, just like any Closed Maniphest
Tasks is already striked, etc.
https://we.phorge.it/T15166
There is still the same thing to be done at Remarkup.
Ref T15166
Test Plan: - go in the page /ponder/query/all/ and see that closed questions are now striked
Reviewers: O1 Blessed Committers, avivey
Reviewed By: O1 Blessed Committers, avivey
Subscribers: speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15166
Differential Revision: https://we.phorge.it/D25086
Summary:
If you want to create a new Herald Rule, sometime,
like for new "Object" Herald Rules, the Back button just
sends you to the very same page you already were.
The risk is that an employee could receive a specific
instruction, such as "Hey Alfreda, Go back, now"
and at this point Alfreda goes into an incredible loop
continuing to click the Back button until the duration
of the universe (since Phorge is so stable that it's able
to handle that "Back spike" forever - really).
https://we.phorge.it/T15184
Note that this also tries to avoid to change the base URI
to just go Back. For example, before this change, the Back
button was trying to send from /create/ to /new/, but
apparently they are just aliases.
Closes T15184
Test Plan:
- Visit the page /herald/create/?adapter=commit
- Click on "Object"
- Click on "Back"
- Verify that you only went back one screen, and not just forward in time
Reviewers: O1 Blessed Committers, avivey
Reviewed By: O1 Blessed Committers, avivey
Subscribers: speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15184
Differential Revision: https://we.phorge.it/D25087
Summary:
Note that the git version 2.5.0 is surely already assumed
as required as implicit fact, so it's just better to clarify
this in the documentation to prevent weird situations and
unclear related exceptions.
Note that if you have seen this Phabricator/Phorge error
in your installation:
This path was a submodule at R123:5378198ea123asdlol
This is probably that kind of weird errors that will be
fixed just updating your git version to 2.5.0, so that
the git escape sequence ('--') is correctly parsed by git.
Example of command that will fail with outdated git versions:
git cat-file -t -- <hash>:<file>
If you do not want to update git, but you want to fix,
you can convert that command to this one, manually patching
your installation:
git cat-file -t <hash>:<file>
Related material:
- https://we.phorge.it/T15179
- https://unix.stackexchange.com/a/740621/85666
- 6f9504c48e/Documentation/RelNotes/2.5.0.txt
Wontfix T15179
Test Plan:
- open Diviner page "Installation Guide"
- see that the git version is not explicit - yup!
Reviewers: O1 Blessed Committers, Cigaryno, avivey
Reviewed By: O1 Blessed Committers, Cigaryno, avivey
Subscribers: speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15179
Differential Revision: https://we.phorge.it/D25081
Summary:
This change contains a very minimal graphical change for Workboards
with a custom background.
https://we.phorge.it/T15186
After this change, if your Workboard has a custom background color,
the opacity of the floating menu of the Column is not 0.9 but 0.95,
so it's increased a bit. The problem with the previous value, is that
when you open the menu, you see too much text underneath, and this
could worsen the readability. Now also probably, but less I hope.
Closes T15186
Test Plan:
- Workboard > Change Background Color > Pick a nice color
- Backlog > Edit
- Enjoy the extra 0.05 of opacity of that menu
- NEVER ACCEPT IF YOUR EYES DETECT JUST A 0.04999998 OF EXTRA OPACITY
Reviewers: O1 Blessed Committers, avivey
Reviewed By: O1 Blessed Committers, avivey
Subscribers: speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15186
Differential Revision: https://we.phorge.it/D25088
Summary:
This expands a specific link in a specific menu of Workboard Columns.
You may never notice this difference unless you like to open links in another tab.
If you go to a Workboard, and you open its context menu, and you hover your mouse on the
Create Task action, you see this URL or a similar one:
http://example.com/maniphest/task/edit/form/default/
After this change, you see this URL or a similar one:
http://example.com/maniphest/task/edit/form/default/?tags=test
You see that the PhutilURI class was used to add the Project slug to the 'tags' query param so
that users can still open the URL in a new tab and have the form prefilled with the Project Tag.
Closes T15147
Test Plan:
- visit a Workboard
- open the context menu of a Column (the pencil icon)
- see that the Create Task link has the Project slug is in the URL of its prefilled form
Reviewers: O1 Blessed Committers, valerio.bozzolan
Reviewed By: O1 Blessed Committers, valerio.bozzolan
Subscribers: speck, tobiaswiese, valerio.bozzolan, Matthew, Cigaryno
Maniphest Tasks: T15147
Differential Revision: https://we.phorge.it/D25068
Summary:
This change fixes a typo that, in PHP 8.2, causes this exception:
Creation of dynamic property PhabricatorAuthPasswordException::$confirmError is deprecated
Closes T15201
Test Plan: - I checked that "error" was spelled with two "r"
Reviewers: O1 Blessed Committers, avivey
Reviewed By: O1 Blessed Committers, avivey
Subscribers: speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15201
Differential Revision: https://we.phorge.it/D25100
Summary:
Note that PHP 8.2 implemented this deprecation:
Using ${var} in strings is deprecated, use {$var} instead
This change fixes some known cases.
Ref T15196
Test Plan: - I checked with my big eyes that everything is as it should be
Reviewers: O1 Blessed Committers, avivey
Reviewed By: O1 Blessed Committers, avivey
Subscribers: avivey, speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15196
Differential Revision: https://we.phorge.it/D25098
Summary:
This is a cosmetic modification for a specific Feed.
https://we.phorge.it/T15183
If you retitle a Diff (and you do nothing else), its Feed
is somehow too much verbose:
... retitled NEW_TITLE from OLD_TITLE to NEW_TITLE
In this specific case, the Feed now becomes shorter so that
the NEW_TITLE is not repeated twice:
... retitled NEW_TITLE from OLD_TITLE
Note that, if the title changes again, the Feed naturally comes
back as before so to mention the change and obviously without
any repetition. Example:
... retitled CURRENT_TITLE from OLD_TITLE to NEW_TITLE
Closes T15183
Test Plan:
- take a Diff named "OLD_TITLE" and retitle to "NEW_TITLE"
- visit the homepage and see "retitled NEW_TITLE from OLD_TITLE"
(this means the new behavior works)
- retitle the Diff from "NEW_TITLE" to "CURRENT_TITLE"
- visit the homepage again and see "retitled CURRENT_TITLE from OLD_TITLE to NEW_TITLE"
(this means the old behavior still works)
Reviewers: O1 Blessed Committers, avivey
Reviewed By: O1 Blessed Committers, avivey
Subscribers: avivey, speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15183
Differential Revision: https://we.phorge.it/D25094
Summary:
This change fixes the following page on your Phorge/Phabricator:
/readonly/config/
If you visit that page (with or without read-only mode), the
following crash will happen:
Call to undefined method PlatformSymbols::getPlatformServerSymbol()
Fixes T15185
Probably the regression was introduced in this point:
rP4d3f0dc7c727e6fdc0e8cba0e8e8f2b8ef1bd16c
This change also renamed "administrator" to "Administrator" in the messages
from that specific page, both because I have great respect for Administrators
and also because that is how this word is spelled in most other places in Phorge.
Test Plan:
- Visit the page /readonly/config/
- It does not crash anymore and you can see that beautiful message
that - it seems - no person in the world has ever seen since a
whole year. Take a good look at this page: it is a precious treasure.
Reviewers: O1 Blessed Committers, avivey
Reviewed By: O1 Blessed Committers, avivey
Subscribers: speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15185
Differential Revision: https://we.phorge.it/D25091
Summary:
The CTRL+click now opens the links in a new tab, like any other
normal link opened with CTRL+click.
Note that the middle-click was already working.
Closes T15157
Test Plan:
- visit a Workboard
- column > menu > mouse on "Create Task"
- CTRL+click: open in new tab (→ now works)
- normal click: open the pop-up (→ still works)
- middle-click: open in new tab (→ still works)
- column > single Task > mouse on "Edit"
- CTRL+click: open in new tab (→ now works)
- normal click: open the pop-up (→ still works)
- middle-click: open in new tab (→ still works)
Reviewers: O1 Blessed Committers, avivey
Reviewed By: O1 Blessed Committers, avivey
Subscribers: speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15157
Differential Revision: https://we.phorge.it/D25072
Summary:
Rest assured: external links remain evil, by default.
Don't adopt them randomly by induction.
Whether you believe it or not, this specific external
link merited some deep thoughts on Phorge:
- https://we.phorge.it/T15172
So, whenever you use a mouse, a finger, or whenever we have
a confirmation dialog or not to prevent onblur disasters,
this change is probably consistent with common expectations.
Having said, external links remain evil - by default.
Closes T15172
Test Plan:
- Drag & Drop a File on a Remarkup text
- click on the link inside the popup
- it opens in a new tab (without risk of form loss)
Reviewers: O1 Blessed Committers, avivey
Reviewed By: O1 Blessed Committers, avivey
Subscribers: speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15172
Differential Revision: https://we.phorge.it/D25077
Summary:
This modification improves a single link shown to the user.
The goal is to make it easier to change file permissions or
file name or whatever just after the upload of the file itself.
Before this change, doing that was difficult because clicking
on the popup would send you back to the binary file only.
Closes T15165
Test Plan:
- upload a file using drag & drop (for example in a comment)
- click on the popup's link
- you see the File page, not the binary file
Reviewers: O1 Blessed Committers, avivey
Reviewed By: O1 Blessed Committers, avivey
Subscribers: avivey, speck, tobiaswiese, Matthew, Cigaryno
Maniphest Tasks: T15165
Differential Revision: https://we.phorge.it/D25074
