951423 - [B2G][Calendar][Day View]Header displays truncated if date exceeds eighteen characters.

Status: RESOLVED FIXED

(Firefox OS Graveyard :: Gaia::Calendar, defect)

Description

[Ross Kuhlman ([e-mail])]

Attachment: logcat.txt

The calendar app has a 'Day' tab that allows the user to view all of their calendar dates for a single day. The date of the page being viewed is displayed in the header. The date is displayed in <Weekday> <Month> <Date> format.(ex: 'Monday December 16') If the date string contains eighteen or more characters, the display will be truncated. Only the first fifteen characters are displayed, followed by a '...' elipsis.

Repro Steps:
1) Updated Buri to Build ID: 20131216140918
2) Launch Calendar app.
3) Navigate to 'Day' tab.

Actual:
The header is truncated for any date that exceeds 18 characters in length.

Expected:
The header is never truncated.

Build Environmental Variables: 
Gaia      504cf9a988cd84760b4040706ccc41e508a97fd2
Gecko     55b75ce2263e7d2a8cbfdabc7e5206d9caa89f98
BuildID   20131216140918
Version   28.0a2
ro.build.version.incremental=eng.archermind.20131114.105818
ro.build.date=Thu Nov 14 10:58:33 CST 2013
RIL Type : Mozilla RIL
OS Version: 1.3
Device: Buri

Notes:
Repro frequency: 100%
See attached screenshot and logcat.

This issue occurs both with and without APZ enabled. Issue also occurs on current default 1.3 build.

Comment 1

[Jason Smith [:jsmith]]

Does this reproduce on 1.2?

Comment 2

[Ross Kuhlman ([e-mail])]

Attachment: TruncatedTitle.png

Comment 3

[Roland Kunkel [:RolandK]]

(In reply to Jason Smith [:jsmith] from comment #1)
> Does this reproduce on 1.2?

This reproduces in the latest Buri v1.2 build. It appears the issue is worse in this build, where any header more than 13 characters gets truncated with '...' ellipses.

Environmental Variables:
Device: Buri v1.2 COM RIL
BuildID: 20131217004001
Gaia: 4f53ba8b3628ac311253fc28dfdf66e7ba6832de
Gecko: 129ad3c335a5
Version: 26.0
Firmware Version: V1.2_US_20131115

Comment 4

[Jason Smith [:jsmith]]

can you provide a screenshot of this bug on 1.2?

Comment 5

[KTucker [:KTucker][Inactive 3/4/2016]]

Attachment: Buriv1.2.0Screenshot.png

Comment 6

[KTucker [:KTucker][Inactive 3/4/2016]]

Screenshot of the issue from the Buri v 1.2.0 Mozilla RIL has been attached.

Comment 7

[Jason Smith [:jsmith]]

Yuck. This is one nasty papercut - it makes the day view not very useful for determining the current day viewed. This has to be a regression because I don't recall truncation originally happening this bad originally.

Probably can't block on this for 1.2 at this point (even though this does look bad), but I'd really like to see this fixed in 1.3.

Dylan - Can someone on your team look into this?

Comment 8

[Evan Tseng [:evanxd]]

Hi Jason,

I think it is also a UX issue.
We should needinfo Harly for this.

Comment 9

[Harly Hsu[:harly]]

This is really a disaster as the user can not see the day in the day view. 
For UX side, we would like to propose the following solution:
1. Use abbreviation for month and day of the week, for example: Dec.19, Thu
2. Rearrange title, for example December 19, Thursday
3. Use number for month, for example Thursday 12/19
4. Move day of week out of header, and the header will only display December 19 (Need visual support)

Evan, could you study the feasibility of each one?

Comment 10

[Jason Smith [:jsmith]]

UX - Do you guys see this as a blocker for 1.2?

Comment 11

[Dylan Oliver [:doliver]]

CC'ing Gareth also to review the suggestions in comment #9.

I don't think we should block 1.2 on this but we can get a solution for 1.3.

Comment 12

[Jason Smith [:jsmith]]

Clearing needinfo on UX - comment 11 I think gives a path forward.

Comment 13

[don't use this account!]

Attachment: proposal-abbr_month_full_day.png

I did a quick test and for English it does fit properly with abbreviated month name + day + full day of the week. For pt-br it would also work "Dez 31, Quarta-feira" ("Wednesday" and "Quarta-feira" are both the biggest words)

I will do a pull request in a few hours. Just sending the attachment since I did not see any feedback about which option should be used.

Comment 14

[don't use this account!]

Attachment: pull request

Comment 15

[Roland Kunkel [:RolandK]]

(In reply to Jason Smith [:jsmith] from comment #12)

This issue appears to be occuring since the first Buri v1.2 build.

- First Broken -

6/21
Environmental Variables:
Device: Buri 1.2 MOZ
BuildID: 20130621031231
Gaia: e2f19420fa6a26c4287588701efaec09a750dba1
Gecko: 7ba8c86f1a56
Version: 24.0a1
Firmware Version: 20131115

Comment 16

[Evan Tseng [:evanxd]]

Hi Miller,

Anything I could help you for the needinfo from me?

Comment 17

[Miller Medeiros [:millermedeiros]]

(In reply to Evan Tseng [:evanxd] from comment #16)
> Hi Miller,
> 
> Anything I could help you for the needinfo from me?

Hi Evan,

the needinfo was just because Harly asked for feedback/feasibility of each option and I ended up implementing what I thought that worked well for current cases. Just wanted you to confirm that current solution is good enough to consider the bug as "fixed".

Comment 18

[Evan Tseng [:evanxd]]

Hi Miller,

Got it.

The patch looks good.

There is just a little question about the integration test.
See it in https://github.com/mozilla-b2g/gaia/pull/15083/files#r8748601.

Thanks.

Comment 19

[Evan Tseng [:evanxd]]

BTW, one more thing we should notice.
After we send a PR to Gaia repo, we need to make sure Travis is good for the patch.

This patch seems have some issues in Travis.
See the below images.

But the issues seems to be not related with this patch.
So let us run it on Travis again.(I already did that)

Thanks.

Comment 20

[Evan Tseng [:evanxd]]

Attachment: Error on Travis - 1

Comment 21

[Evan Tseng [:evanxd]]

Attachment: Error on Travis - 2

Comment 22

[Evan Tseng [:evanxd]]

The "Error on Travis - 1" and "Error on Travis - 2" images mean that there are some issues happen on Travis for the patch.

In normal case, we could land the patch after r+ and Travis is good.

Comment 23

[Miller Medeiros [:millermedeiros]]

I saw that the Travis build failed but it looks like the error is on separate components and not related to the code I introduced. It seems that the build was re-run and now there is only [1 failing test related to the browser](https://travis-ci.org/mozilla-b2g/gaia/jobs/16553414) ("search navigate to invalid page (about:neterror) loads net_error.html from system" on "/home/travis/build/mozilla-b2g/gaia/apps/browser/test/marionette/navigate_test.js:87:16")

I simplified the integration test and added a comment explaining why I validate the content and container width before checking if both values match.

Comment 24

[Gareth Aye [:gaye] (back from PTO)]

Hi Miller -

Your patch looks good! Thanks :)

I'm a bit worried about the integration test. Yesterday I was trying to reenable some calendar tests that click over to the today tab, but there are still some flakes. Will you go ahead and run your integration tests 20 or so times on CI to make sure that we're solid? You can see an example here https://github.com/mozilla-b2g/gaia/pull/15083/files.

Also, in the future, if there's a failure that you didn't introduce on CI, filing a bug is great. Short of that, if you sign into Travis (oauth with GitHub) you can rerun the build. In general, I will not r+ something that fails on Travis (even if you didn't introduce the failure).

Comment 25

[Miller Medeiros [:millermedeiros]]

Attachment: pull request #2, rebased master branch

talked with gaye over IRC and he said that proper way to do it is to ask for PR directly to the master branch and somebody else will cherry-pick commit into proper branches.

CI still needs to finish but I think it should be good to go. Next time I'll open bugs for failing tests as well, wasn't aware of the process and heard about many failing tests on the #gaia channel (so I thought it was being ignored for now / know bugs).

Comment 26

[Miller Medeiros [:millermedeiros]]

travis build passed! https://travis-ci.org/mozilla-b2g/gaia/builds/16672108

Comment 27

[Gareth Aye [:gaye] (back from PTO)]

Merged onto master

https://github.com/mozilla-b2g/gaia/commit/61797fd7cf908f13f140b98f94ace038e2321c05

Comment 28

[Michael Henretty [:mikehenrty][:mhenretty]]

Had to disable day_view_test.js in travis for perma-red.
Bug 959901

It's not clear to me what caused this issue (might even be a gecko change), but please investigate and re-enable the test when you have the chance.

https://github.com/mozilla-b2g/gaia/commit/424a004947754238571efcabf153b52b7d826717

Comment 29

[Harly Hsu[:harly]]

Hi Miller,
The patch looks great, I am just wondering if all the languages Firefox OS supported will fit within the 18 characters limitation with abbreviation for month?

Comment 30

[Axel Hecht [:Pike]]

As this landed without a string ID change, the l10n communities will hardly notice the change.

Comment 31

[John Ford [:jhford] CET/CEST Berlin Time]

Uplifted 424a004947754238571efcabf153b52b7d826717 to:
v1.3: a1ee40b96be8ec2a2117c83987cd042534ab4141

Comment 32

[Ross Kuhlman ([e-mail])]

Attachment: 2014-02-21-15-02-24.png

This issue still occurs when the device is set to Serbian-Cyrillic. See attached screenshot.

Environmental Variables:
Device: Buri v1.3 Moz RIL
BuildID: 20140220004003
Gaia: a73dbe99db9e01f994f09c012e180a5596aa1847
Gecko: 32fd5d798477
Version: 28.0
Firmware Version: v1.2-device.cfg