run jpm with profile: how to see console messages? - firefox-addon-sdk

When I run my extension with
jpm -p myuser
I don't see any console.log messages any more: that's almsot impossible to debug.
Same trouble described here : https://discourse.mozilla-community.org/t/jpm-run-with-profile-and-logging/3114/3
Any idea how to solve that ?

You might need to add a preference to the profile you're using to see all addon log messages, see this blog post for the gory details. IMO this is a jpm bug, when it copies the profile it's using it should set this preferences.

Related

gcloud components list not up-to-date?

Context
I will rely on a component shipped with Gcloud SDK CLI.
For migration purposes and other reasons, I want to know which version of the cloud-sdk starts shipping this component and avoid the "install the latest version".
My issue
In the official changelog: https://cloud.google.com/sdk/docs/release-notes, there are no references to the component I am looking for.
Tries
I have tried to run this command naively
for gcloud_version in 390.0.0-alpine 391.0.0-alpine 392.0.0-alpine 393.0.0-alpine; do
echo "---> ${gcloud_version}"
docker run --rm -ti google/cloud-sdk:${gcloud_version} gcloud components list
done
Unfortunately, every list do not show the component I am waiting for (even the latest version of gcloud, 393 at the time of writting).
Discovery
However, when I run gcloud components install MY_UNLISTED_COMPONENT it works ...
Not a very reliable way to find out which version has the component I want.
Do you know if:
this is an issue?
I can report this somewhere?
It is relevant to do it?
Thanks for your help!
From #DazWilkin
It would be helpful if you included the name of the public albeit unlisted component in your question. The Release Notes includes a "Send Feedback" option and you may want to provide this feedback there.
it's reasonable to expect it to be documented. I encourage you to send feedback via the release page and to consider filing an issue on Google's public Issue Tracker.

Typo3 broken: Frontend / Backend blank pages and no errors even after restoring backup

I was fixing things on a not well-maintained typo3 installation and it resulted in blank pages of frontend, backend and install tool.
I wanted to fix extensions because the powermail-extension was not working anymore even after updating to latest compatible version (tried emptying cache and database compare).
I deactivated an extension "compatibility6" to see if it was interfering and since then all pages have been blank including Install Manager. I reactivated the extension via PackageStates.php, emptied the cache but it did not change anything. I was not able to see any error messages even though I set the parameter in LocalConfiguration.php which made it hard to try to find out what is wrong. Then I restored a backup using vmware. Still, backend and frontend are blank.
I am using Apache and Typo3 7.6.23.
I do not know what might be causing the problem and would like to know what else to look for?
It means you have a PHP error there. The first thing would be to examine logs. If you see anything like "class not found", then your next action would be composer dumpautoload --no-dev.
Long time ago I had an article about blank pages in TYPO3 FE. Not up to date but still can be helpful: https://www.dmitry-dulepov.com/2009/03/blank-empty-page-in-typo3.html
Dmitry already wrote it: you have an error.
In production environment nothing is shown for not revealing any information about your system.
So the only information about the error can be found in the error log of your webserver.
What can you do to get more information about your error?
Enter the install tool and switch to develop mode. This will also enable deprecation logging into a file. Try to avoid filling your system with deprecation warnings by disabling deprecation logging.
Disable the default exceptionhandler by adding this to your typoscript setup:
config.contentObjectExceptionHandler = 0
Now you should see a call stack instead of blank screens.
Be aware that this might reveal confidential information from your system.
because of this you should use a copy of the live-system.

Accessing AEM 6.2 error logs over HTTP

In previous versions of AEM, certainly in CQ 5.6 and AEM 6.0, it was possible to tail the error logs over HTTP, without connecting to the server over SSH.
For example, I could get the last 1000 lines from the error log of my AEM author instance by calling:
http://localhost:4502/bin/crxde/logs?tail=1000
This seems to no longer be possible in AEM 6.2, this path does not resolve to anything.
Is there another way I could still tail the log over HTTP?
A colleague answered this question for me on a chat so I'm putting it here to make it easier to find in the future.
There's now a neat utility in the OSGi console that allows one to view the logs as well as configure the various loggers. You can find it at http://localhost:4502/system/console/slinglog
The Appender tab provides links to the various log files that can be used to load logs over HTTP.
Here's an example request it makes:
http://localhost:4502/system/console/slinglog/tailer.txt?tail=1000&name=%2Flogs%2Ferror.log
As you can see, both the log file name and the tail parameter can be specified. You can also use grep with both simple phrases and regular expressions.
This is a built-in feature of Apache Sling.
In addition FYI, you can also find the status-slinglogs where you can perform log file downloads in a zip and logger actions in a txt to your local at /system/console/status-slinglogs
http://localhost:4502/system/console/status-slinglogs
and the direct urls for the downloading these zip files are as below
http://localhost:4502/system/console/status-slinglogs.zip
http://localhost:4502/system/console/status-slinglogs/configuration-status-20170126-183246.zip (where as 20170126-183246 is and time stamp)
You should not be looking at log files via CRXDE lite.
log files in 6.2 are project specific - better to open them from a text editor.
see attached screenshot.
Hope this helps!
Regards,
Prince
You can curl the log with e.g.:
curl -u admin:admin 'http://localhost:4502/system/console/slinglog/tailer.txt?tail=4000&name=%2Flogs%2Ferror.log'
where 4000 is the number of lines you want to get.
I recently wrote a tool named "Log Tailer Plus" to solve exactly this problem. It's entirely free/open source - Take a look at a post describing usage here : https://blogs.perficientdigital.com/2019/05/14/introducing-aem-logtailerplus/
TLDR; You can grab an AEM package from here ( https://github.com/prftryan/LogTailerPlus ) install it to your machine, and access via http://localhost:4502/log-tailer-plus (if local) or http://server:port/log-tailer-plus
This tool will allow you to follow any number of logs at once by leveraging the out of the box logging endpoint(/system/console/tailer ) as well as dynamically checking active OSGI Logging Logger configurations. Currently, highlighting is supported, but only for relatively standard logging patterns (it's done via regex).
This is a new release, works on AEM 6.2+. Enjoy

Eclipse cannot debug. Please help I tried everything

Whenever I try to open the debug perspective, I get this message and the problem is that it disappears immediately before I can even check deny or allow. I can run the program, but I cannot debug.
I run OS X 10.8.5
Things I tried, but all failed
I have tried to uninstall and install Eclipse, but that did nothing.
I tried adding Eclipse to my firewall exceptions and checked "allow incoming connection" in Security and Privacy, but that did nothing.
I have tried following the instructions mentioned here.
But no help at all.
One "security" tempering I didn't saw in the link you reference in your question is:
removing Eclipse from the firewall list of apps
removing Eclipse from the plist files.
A plist is a special kind of text file that contains properties the application and other resources, typically the OS, use to retain and reuse information needed to run the application.
plist is a file type and has many uses, typically storing user preferences, but, essentially, it is an XML file. You can look to see if there are any caches, which are typically plist files, for the applications in question in /Library/Caches and /System/Library/Caches.
Checking on that particular error message, it seems to be caused by issues with the firewall configuration. I am not sure if it will help, but I would verify that the program is allowed to access your network. Check in System Preferences > Security & Privacy > Firewall > Firewall Options. You should be able to allow your app to access the network from there. Hope this helps!

Any good MongoDB admin UI for MacOS with auth support?

I've been using mViever for several months and it fitted me perfectly, however it doesn't support proper authentification (when mongod is launched with --auth parameter). So, now I'm trying to find any good replacement for MacOS with auth support.
Tried MongoHub, but it doesn't support most of common Mac hotkeys (even Enter doesn't work as submit), and by the way, it also has some bugs with authentification. To be shorter, which of these http://www.mongodb.org/display/DOCS/Admin+UIs to choose? I've tested some of them, but nothing really liked (except old good mViever).
I assume that's a typo and you are referring to mViewer. As you can see here, it does support authentication:
http://imaginea.github.com/mViewer/0.9/pages/getting_started.html
There were issues with auth in 0.9.1, but they have been fixed. See here:
https://github.com/Imaginea/mViewer/issues/75
For some reason, the 0.9.2 binaries are not available yet (even though the comment says October), but it's not too hard to build yourself:
git clone https://github.com/Imaginea/mViewer.git
cd mViewer
mvn clean package -Prelease -DskipTests
Once all that was finished, I took a look in the target folder:
cd target
ls mViewer-0.9.2.*
Here's my mViewer 0.9.2 session logged in with user "adam" to an auth enabled mongod:
I just released a minimalist mongodb data viewer for nodejs github.com/marcdelalonde/mongo-viewer
I have planned to add auth support very soon :)