@Unixchick
I'm not colorblind but You have made MAM Green but I've not seen any betas jet, late April fool or...🤔

Sorry :

Just saw Adriverhoef's message in Beta Test Support Forum so they are more comming just around the corner hopefully 😁😁

PS .
Only Linux so far!!?

Two batches of 100 BETA-workunits each have been released.
The two batches were named BETA_beta30_9800033 and BETA_beta30_9800034.
Initially, the first batch was released between 2025-04-04T15:58:26 and 2025-04-04T15:58:42.
Initially, the 2nd batch was released between 2025-04-04T20:23:50 and 2025-04-04T20:25:05.
All workunits have been distributed to Linux clients.
All tasks, also the retries, errored out — of which a few examples can be found in the Beta Test forum — apart from the ones that are still in progress.

That's why there is a BETA-project. New projects need to be tested before any 'serious' batches are launched.
I'm glad that the team made it to the BETA Test stage, although the tasks are still failing at this moment ; I'm sure successful workunits are soon to follow.

Adri

We got an official update which I'll copy here

April 4, 2025

MAM1 Beta Begins: We have (finally) started MAM1 beta testing on April 4th, 2025. Linux build first.
Version 7.00 and 7.01 Crashes: First and second batches MAM1_9800033 and MAM1_9800034 of 100 work units each sent to 64 bit Linux clients crashed immediately with a segfault. Initially we thought this was a "regression" from pulling the wrong branch into the Docker container that we used to build MAM1 and graphics for Linux. However, version 7.01 intended to resolve the crash, which we had seen before in alpha testing; yet it did not. We will be pushing 7.02 to resolve these issues soon, and start releasing more batches then. Apologies for the oversight to all volunteers who ran with version 7.00 or 7.01 across these two batches.
Windows Build Status: We were able to produce a working build for Windows 8.1 Pro to integrate with the Visual Studio 2008 and Visual Studio 2005 project solutions for Graphics and BOINC dependencies. However, testing on Windows 10 revealed memory safety issues around checkpointing and graphics. We are working to fix these issues before we push the first Windows 64-bit build of the MAM1 application.
MacOS Build Status: Given a stable Linux build, we intend to finalize Intel-MacOS soon; we will work towards aarch64, amd64, arm64 once we have the Intel Macs covered.
Optimizations: Now that we have been able to build and ship the application based on the MCM1 code and have the ability to iterate on the build and release new versions quickly, we will begin our efforts to optimize and modernize the code.

I see that MAM Beta testing is active!
I run Windows boxes so I need to remain patient for my turn.
The question I have is: I have selected which profile the Betas are to be run under (Work) but I have not seen any means to select the maximum BETA WU's in queue (as there are for MCM, ARP and etc.). Due to the number of resends of MCM that usually end up as SERVER ABORTED on my system, I have basically set my queue to ZERO with at least one open processor.
How does the system determine how many WU's to send me?

It's good to see the beta30 (MAM1) items showing up, but a bit sad that the first batches failed out. (I got 60 across 4 systems, for what it's worth...)

According to their 4th April update (as posted earlier by Unixchick) this isn't a new version of the Mapping code yet. However. it appears that MAM1 is going to use a different VMethod [CV] (unless that's an alternative name for one of the two previous VMethods [presumably NFCV...])

Even with [effectively] the same code, a fair comparison with MCM1 performance will obviously be impossible (different data, and more recent compilations!).

However, once work on the optimizations starts it'll be interesting to see how much better it does. If newer code spends less time doing [possibly unnecessary] timing stuff, I'd expect that to be a major win regarding overall run time!

And, possibly not of interest to anyone else: seeing the failed beta logs has given me useful information that I can use to seed my analysis of MAM1 tasks when they hit production.

• The dataset has another of those long names -- I'll think up an alias to use in my database to save space :-)
  
• The parameter file seems to have a lot of comments in it now; useful to know what the comment format is, as those lines would break my parameter-watch script!
  

Cheers - Al.