ARender 4.0.7 is available!
A detailed list of changes (release note) is available at this address, where many more features/bugfixes will be described than in this summarized version.
If you currently run an ARender 4.0.x installation, we recommend you to proceed to upgrade to ARender 4.0.7, as many corrections have been brought to this cumulative release.
New for ARender 4.0.7 :
✔️ Reworked the way we relay video content from the backend to the end user
As you may have read in this article, in ARender 4.0.7, we reworked the way transfers were done in order to keep our effort to make ARender compatible to a wide variety of Desktop internet browsers.
The result is promising ! This new approach consumes less memory overall onto ARender servers, and proved to be more stable than the previous one.
This will lead us to improve furthermore the video experience in the following versions of ARender 4.
✔️ URL based accessors now follow redirects headers and status codes (Location)
We modified the default URL behavior to now be able to follow redirects by default. As this mechanism can allow to follow up links to reach documents that were not reachable before, we also updated the document title reflected to show the « followed » link. Depending on feedback received, this change in the title (now showing the source URL) could be changed back in the future version.
✔️ Restored (and improved) the way ARender 3 dealt with Tiff documents
Through ARender 4 releases, we already applied a set of bugfix to Tiff documents in order to reinforce our already wide support of this file format we gained over the years.
We found out recently that performance wise, we could gain at restoring the old mecanisms to handle Tiff files, while still applying the new set of fixes.
But the team did not stop here! We optimized this merge of « old logic » (to gain performance) and new one (to support more Tiff files) to produce the version included in 4.0.7 that is both faster than the old logic, generates overall less renditions of the Tiff document and widen the extended support of the format.
✔️ Change in the appearance of rendered rich text annotations in PDF documents (download with annotations)
In version 4.0.7, we improved how sticky notes and rich text annotations will render using our default « download with annotations » feature.
This, because of the much wider support of the rich text norm will lead to obvious visual changes, and we will wait for your feedback on this new rendering through our usual channels of support in case you’d like to see improvements to it.
✔️ Corrected and improved version 4.0.6
The goal of version 4.0.7 was also to stabilise the set of features brought in previous versions. On that topic here is a short list of the features that have seen corrections/improvements in version 4.0.7
If you want more details about the fixes, please refer to the full release note (with the Jira tickets IDs).
✔️ Benchmarked and ready
Before the release, we used the latest RC version which got selected for release promotion to update our sizing tool document.
We ran overnight benchmarks of our version 4.0.7 to generate input values for each of the typical mime type ARender supports.
Our team might ask you to fill in the sizing sheet, so we thought it might be a good time to explain what this sizing sheet does :
We recommend this third party benchmark tool to detect what is the typical performance you can expect from ARender on your on Premise installation. (reminder, if you prefer, ARender also has a SaaS offering)
The fact that we rely on third party tools also enforce the result to not be biased towards our product. The value to copy back into the sizing sheet is the « Multi core ; PDF rendering » as you may guess.
If you do not want/cannot use this tool, do not hesitate to talk to our team as we also possess internal accurate benchmark tools for our product that could give you a relative performance score.
This is where the benchmark ran for our version translates into required amount of CPU. From the values you will input as number of pages per document, types of documents, number of concurrent users, we will compute an amount of required compute units per second.
The amount of computed units then translates in an amount of your vCPUs required regarding the performance of the third party benchmark tool.
In the end, the sizing tool allows you to precisely know if your machines are cut for the ARender software, but as a general rule, a simple follow-up of our recommendations is enough to properly equip your ARender servers.