Everything “else”…
I am far too pissed off to translate the whole article into English. Especially since no one reads my blog. Long story short, my EA Origin Account was hacked and stolen. Luckily for me, I only used it for free to play games so far and I lost nothing except for the time I invested into my virtual Springfield. I uninstalled Springfield and I will shun EA Origin, no matter how interesting their games might look.
This week I participated in a workshop on parallel programming. It was given by a guy from the HLRS. Since I myself worked at the University of Stuttgart before 2012, I got in some reminiscent mood; only a little.
The workshop itself was pretty good. It covered MPI and OpenMP together with some tools for development. In theory none of these are black magic, but it is good to try for yourself with well-prepared practical examples. That way you can get to know the stuff step by step.
Once again I missed a Sunday to do a post. Oh well. These of you who actually read ma post should know, that, even I am posting on each Sunday, I only make an interesting post every three weeks, or so.
Today, I have news from the VISlib: the VISlib is now finally released as version 1.0!
With that, the VISlib reached her final development state. We will only continue to fix bugs, based on that release, but we will not introduce new functionality. Instead we implement The.Vislib.Legacy-Project in the trunk of the repository of the VISlib. This project is a step-by-step migration towards TheLib.
The New Year started, but at the moment I am still enjoying my time off. The real world work starts again the upcoming week. Therefore, right now, there is not much to tell.
However, I am using a little bit of my time to make plans and to do some small maintenance at TheLib.
The exciting stuff will happen the next week …
I know, I know, the end of the year is not here, yet. However, it is the last Sunday 2013 and thus it is time for a Post at the end of the year.
Reviews of a year’s events are awful, especially if the year was not that great. Well, this year was not especially bad either, but I could have been much better. At least there was one pleasant surprise shortly before Christmas (I will write more about it early next year, as soon as it gets official).
Now it is time to look at the future. It is time to make plans for 2014 and to start working on their implementation. Form a business point of view many events and ideas are coming up and I am curious to see which of them will be successful. I don’t want to write too much about them yet, only, that I got ideas for research projects and funding project applications. When enough of these ideas work out, then science will remain interesting for me.
Considering my private programming projects, not much has happened; and that is an exaggeration. Well, we will see what will happen on that end.
But, now I wish everyone a Happy New Year!
There is not much happending on my website. Surprisingly, each year again, Christmas is coming around the corner.
Thus, I simply wish you a Merry Christmas!
(P.S.: Yes, I am a avowed LEGO fan)
This winter term I am visiting a lecture of Dr. Frank Furrer on Future Proof Software Systems. Very interesting topic and great lecture. Aside the normal lecture I also take the opportunity to discuss topics of my own research field with Dr. Furrer. Doing so reveals wonderful truths. For example the evolutionary steps of a software architect:
*sigh*
Last weak I wrote about legacy code. That topic prevails.
There are two fundamental views of legacy code:
The truth is not somewhere in the middle, like it is often. Instead the truth is that both types of arguments are right. Legacy code is a trash pile full of gold! The gold must be harvested and the trash must be removed. However, that, of course, needs a lot of effort to be achieved. One should not be greedy. With each project, one should remove a little bit of the trash and should reveal a little bit of the gold. Over time, everything will get better.
For TheLib we have that problem as well, of course! To implement the “small steps” I was writing about we started the.vislib.legacy project.
Whenever you work on a software, which is not trivial or small, than you have to work with legacy code. I talk about old codes. Either one wants to re-use them or one does not. Regardless, however, you always have to marry the new codes with the old ones, and that is never easy. I gets especially hard when have interfaces to the outside world, i.e. other software using at least parts of the software one is working on. Facing such tasks separates the men from the boys. :-P I am not talking about programming. Programming is easily learned and teached. I talk about software development, design, and architecture.
Current, I work with friends on the creation of the THElib, as successor and replacement of the infamous VISlib. The VISlib has several design errors which we cannot fix due to strong dependancies with other software projects. Of course, we are not rewriting THElib from scratch. She is based on large portions of the VISlib, streamlined and corrected. But here we face a huge foundation of legacy code we need to cope with. It is simply not simple.
Last week I updated my computer to Windows 8.1. The improved Skydrive integration is still completely useless, but luckily there are thrid-party tools for the job.
Whatever. Then I were copying some files onto my backup server and the copying dialog windows offended me with 2 BK/s. To make things clear: we are talking about 5 files, more than 500 MB each, altogether a bit over 18 GB. Thus, it is not the “small files are slow” issue here. After about half a day of checking the configurations of my server and my other computers and semi-guided googeling I finally found the answer:
Large Send Offload v2 (IPv4) – Disable
Large Send Offload v2 (IPv6) – Disable
The thing of this whole story which is really frustrating me, is that I knew about this. I already knew that these settings can be the reason for rancid network performance. *gnampf* It seems I am getting old.
