Flickr provides not only a large space for photos but also nice features around, including tags and other metadata allowing sorting of photos by different criteria - but one is very limited in the design of something like "frontpage" of such galleries. The solution is to use 3rd party JS solutions using Flickr API, like nanogallery. The purpose of this project is to learn to create such galleries, the best would be to add social features of flickr directly there, like comments or faves.

As live migration is widely used in many scenarios nowadays, spend one week time to study it and try to improve its efficiency, e.g. seamlessly migrate with less downgrading the guest. Another challenging part is memory migration, as it involves the dirty memory's detection, record and copy.

During testing virtualization , I usually met some issue and can not position the issue location (host or guest), so I would like to deeply learn QEMU code, card the work flow, understand communication mechanism between host and guest. Targets:

Learning design model can help to produce high quality codes, which will benefit our products. I'll do it mainly by reading "Design Patterns: Elements of Reusable Object-Oriented Software".

Learn Vim by reading book of Practical Vim

I'm planning to do some research about RV32E which has only 16 GPRs, but riscv-pk doesn't have RV32E support so that the RV32E aware GCC (https://github.com/kito-cheng/riscv-gcc/tree/riscv-next) cannot build riscv-pk. I'm going to study how the proxy kernel works so that I can do some modification to let it support RV32E.

since i am learning golang, i will make some contributions for the upstream project we used already @suse https://github.com/dmacvicar/terraform-provider-libvirt

Aim of this project is to create scenarios for manual deployment using limited feature set of mkcloud. Useful for situations when full mkcloud deployment is broken. It would help qam-cloud team a big time. We can reuse the result later in automation.

I planned to write a GUI for mokutil since I started the project. I guess now it's time to do that.

The theory behind the partitioning proposal of yast2-storage-ng is that all possible distributions of partitions in the disk are evaluated and the best one, according to this criteria, is chosen. But I have found several examples in which is hard believe that the result is actually the optimal distribution of partitions. So I want to invest some time checking if the error is on my side and the code is indeed proposing the best solution and, if that's not the case, improving the decision making of the code.

I will implement the ReaR upstream issue https://github.com/rear/rear/issues/1399

During regular L3 work I often don't find enough time to work on the command line disk partitioner parted which I maintain. Changes often directly affect yast-storage(-ng) and libstorage(-ng). @aschnell opens bugs faster than I can solve them. Upstream is often slow and often wants patches differently.

The YaST2 Network module manages network configuration including device configuration, DNS, Routing etc.. The current code is not very object oriented and in many cases there is not a good separation between the business logic and the presentation. The API is also not very clear an many people get confused for example between NetworkInterfaces (class that handles with the ifcfg-files and reside in yast-yast2) and LanItems.

While there are plenty of benchmarks that compare different IO multiplexing techniques such as epoll vs poll/select, there's really nothing out there that particularly measures epoll system call latencies under different scenarios. Design and implement a series of performance benchmarks for this call under the 'perf bench' framework.

PVH domains are a new guest type supported by Xen being as lightweight as possible (e.g. no emulation of legacy devices via qemu) while taking advantage of the hardware virtualization features of the x86 processor. As there is no BIOS for a PVH domain booting is a little bit different than for pure hardware virtualized guests. To be able to start such a guest a little shim is needed to gather some information about the environment (especially memory layout) before the standard kernel boot path can be entered. By using a boot loader like grub2 this shim can be avoided as the memory information is already known by grub2 and stored into the so called zeropage according to the multiboot protocol.