Announcing Lucee 5.3.6.53 (Release Candidate)

Greetings, Lucites! Short and sweet indeed, just like (hopefully) our new-and-improved release cycle. We're actually still getting into a good rhythm with the new schedule--it's not quite as short and sweet as we've envisioned. Specifically, we're looking for the sprints to be 2 weeks max, with a ticket list in the range of 10 or so tickets (which will obviously vary based on the complexity of the tickets in a given sprint), with release candidate periods of about 2 weeks. But we're definitely moving the right direction.

Today weā€™re announcing the next Lucee release candidateā€“5.3.6.53. As I noted above, we pulled in quite a few additional tickets after the sprint started, in order to keep pushing for more and more stability, which in turn will help us fully get into that shorter/sweeter sprint rhythm. Hereā€™s the list of tickets covered:

|Ticket|Summary|
| ā€” | ā€” | ā€” |
|LDEV-2849|Error at ORM Page in Admin Panel||
|LDEV-2838|Upgrade OWASP ESAPI Extension to 2.2.0||
|LDEV-2835|add function queryLazy||
|LDEV-2833|NullPointerException - When viewing the info->bundle in Admin||
|LDEV-2829|java.util.ConcurrentModificationException with cfadmin action=ā€œsurveillanceā€ #2||
|LDEV-2827|Regression - LogBox ConsoleAppender no longer writes to System.out stream in 5.3.5||
|LDEV-2825|expanding path for this.customtagpath fails||
|LDEV-2817|request timeout can blocked datasource usage||
|LDEV-2815|setAutoCommit is called unnecessary||
|LDEV-2809|Regression - Class loading errors in 5.3.5 and 5.3.5 that werenā€™t in 5.3.4||
|LDEV-2804|Session/Client storage is limited to single access||
|LDEV-2794|Reflection looses cached method||
|LDEV-2786|RAM cache free fails||
|LDEV-2770|cfhttp fails with non ascci characters||
|LDEV-2769|use individual log in ExpressionUtil||
|LDEV-2768|Use single log instance in HostnameVerifier||
|LDEV-2760|PagePoolClear function returns java.lang.NullPointerException||
|LDEV-2757|documentitem evalatprint multiple output||
|LDEV-2756|onError fails on RequestTimeoutException||
|LDEV-2743|ESAPI functions break systemOutput under load||
|LDEV-2735|error listener for query fails||
|LDEV-2728|Lucee ignores exception thrown by OutputStream/Inputstream.close||
|LDEV-2714|getClassMetaData on ORMGetSessionFactory not found||
|LDEV-2711|remove unnecessary setAutocommit||
|LDEV-2688|NPE opening settings - > performance/caching in server admin||
|LDEV-2612|MySQL complains:requirements SSL connection must be established by default||
|LDEV-2611|PDF Extensions stops running my cf-code since version 1.0.0.76||
|LDEV-2604|BUG Exclusive connections for request. Persistent MSSQL Isolation Level.||
|LDEV-2559|arrayEach has too much memory overhead from duplicating pageContext||
|LDEV-2452|java.util.ConcurrentModificationException with cfadmin action=ā€œsurveillanceā€||
|LDEV-2422|Lucee sessions are set with NO timeout when using an empty onSessionStart()||
|LDEV-2334| supported CF SQL Data Types missing CF_SQL_NCHAR, CF_SQL_LONGNVARCHAR, CF_SQL_NCLOB and CF_SQL_SQLXML||
|LDEV-2295|Publishing a Webservice with Lucee. No longer accepting an array of components.||
|LDEV-1196|Provide a way to fully ā€œwarm upā€ a Lucee install for containerization|

Please grab a copy from the downloads site, give it a spin, and let us know any and all feedback.

Thanks for listening.

6 Likes

Thanks @IamSigmund, Iā€™m looking forward to this release. I noticed ticket [LDEV-1196] - Lucee is included in 5.3.6 but there have been no new commits listed in JIRA and no comments to indicate the missing bits were added, or how they were accomplished. Is there a way to get confirmation on how LDEV-1196 has been re-addressed?

Also, for you CommandBox users, ALL Lucee versions are now autopubished to ForgeBox. So you can spin up the new RC in box like so for testing:

server start cfengine=lucee@5.3.6-RC
2 Likes

@isapir is going to add a comment or two to the ticket.

1 Like

The number of regressions in the last few releases is mounting up, is it possible to list all the know regressions when asking people to go and test a RC?

There is a recurring problem with people trying out RCs, just encountering unresolved regressions which have been already reported during the sprint

Hereā€™s a rough jira list of known regressions
https://luceeserver.atlassian.net/issues/?jql=labels%20%3D%20regression%20AND%20status%20in%20(ā€œAwaiting%20Approvalā€%2C%20Backlog%2C%20"Being%20Investigated%2FTriaged"%2C%20"In%20Progress"%2C%20"Need%20more%20details"%2C%20New)%20ORDER%20BY%20created%20DESC

3 Likes

Hi Zac. First, yes, as I noted in Mondayā€™s post, weā€™re still not in the groove we need to be in with regard to our shorter sprints, and in particular with regard to getting caught up on regressions. That said, your list is not accurate. Itā€™s worth noting that our regression analyses are the most time-consuming part of the limited Lucee development schedule. They often take at least hours, if not days, and in most cases we end up determining with reporters that they are not, in fact, regressions.

Now, with all that said, we will certainly add this list of yours to our next standup meeting agenda. Stay tuned.

@IamSigmund Iā€™m still waiting for the comments and commit details to be added to ticket LDEV-1196. As of right now, JIRA shows no additional work at all has been done, it just appears the ticket was completed. The three commits on that ticket are from April of 2019 and the questions and concerns in the comments have been left unaddressed.

Mail-servers donā€™t verify in this version as well, as described in here:https://lucee.daemonite.io/t/email-not-sending-lucee-5-3-5-92/
Will this be fixed before a final release?

New bug filed, the test email should be sent with async=false, otherwise the test email is spooled and the error happens in the background

https://luceeserver.atlassian.net/browse/LDEV-2873

Have the async part been added? I am still setup to work on this. I can add that.

1 Like

Thanks for the information.

Thereā€™s another regression (in that list) since 5.3.5.80 which causes a 1m hang on shutdown with scheduled tasks. That extra delay sucks in production (and dev, especially with testing snapshots)

https://luceeserver.atlassian.net/browse/LDEV-2712