# Nuking the @should taglet

**URL:** <https://talk.openmrs.org/t/nuking-the-should-taglet/26913>\
**Category:** Development\
**Tags:** testing\
**Created:** [February 19, 2020, 9:34pm UTC](https://talk.openmrs.org/t/nuking-the-should-taglet/26913 "2020-02-19T21:34:13Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![ibacher](https://talk.openmrs.org/user_avatar/talk.openmrs.org/ibacher/32/19751_2.png) [@ibacher](https://talk.openmrs.org/u/ibacher)\
**Post date:** [February 19, 2020, 9:34pm UTC](https://talk.openmrs.org/t/nuking-the-should-taglet/26913/1 "2020-02-19T21:34:13Z")

</div>

Basically, the purpose of this post is as a follow-on from [this post](https://talk.openmrs.org/t/rethink-unit-test-style-to-remove-duplication/10200) where the decision was made to stop recommending the `@should` taglet, but to maintain it’s presence in Javadoc. I would like to propose removing the taglet altogether from OpenMRS’s source code.

_Rationale_

1. It has not been the recommended way to create tests for years.
2. It is not clear that the IDE plugins required to make this annotation helpful have been maintained.
3. The Taglet API has undergone a major revision, following the sun-setting of the internal `com.sun.*` classes that it used to depend on, meaning we’d have to somehow maintain the current state of the taglet API and create a version for Java 9 or later.

_Way forward_

I’d propose we take the following steps:

1. Remove the `ShouldTaglet` class from openmrs-tools
2. Replace all occurrences of the `@should` taglet in OpenMRS Javadoc with `<strong>Should</strong>`, allowing us to keep the data provided by this taglet in a human-readable format.

Are there any objections to this or thoughts on a better way forward, etc.?

---

<div class="post-metadata">

**Author:** ![dkayiwa](https://talk.openmrs.org/user_avatar/talk.openmrs.org/dkayiwa/32/179_2.png) [@dkayiwa](https://talk.openmrs.org/u/dkayiwa)\
**Post date:** [February 19, 2020, 10:04pm UTC](https://talk.openmrs.org/t/nuking-the-should-taglet/26913/2 "2020-02-19T22:04:56Z")

</div>

Your proposal makes perfect sense!

---

<div class="post-metadata">

**Author:** ![gcliff](https://talk.openmrs.org/user_avatar/talk.openmrs.org/gcliff/32/8783_2.png) [@gcliff](https://talk.openmrs.org/u/gcliff)\
**Post date:** [February 20, 2020, 12:55am UTC](https://talk.openmrs.org/t/nuking-the-should-taglet/26913/3 "2020-02-20T00:55:23Z")

</div>

i absolutely agree with you @ibacher ,it will help avoid duplications & and making the OpenMRS’s source code more human readable which suits one of the aims of unit testing conventions ie good docs

---

<div class="post-metadata">

**Author:** ![ibacher](https://talk.openmrs.org/user_avatar/talk.openmrs.org/ibacher/32/19751_2.png) [@ibacher](https://talk.openmrs.org/u/ibacher)\
**Post date:** [February 20, 2020, 5:05pm UTC](https://talk.openmrs.org/t/nuking-the-should-taglet/26913/4 "2020-02-20T17:05:03Z")

</div>

I’ve created [TRUNK-5703](https://issues.openmrs.org/projects/TRUNK/issues/TRUNK-5703) to track this.