Astronomoers have found an object which could be called planet. Well it seems to be larger than Pluto and the size of Pluto is assumed to be the “minimum” required size for an object to have the chance to be called planet.
The official name is 2003UB313 yet, but unoffically called Xena. But really, everyone knows the 10th planet as Rupert, don't you?
Finally I got Bacula set up and running fine at work. Primarily I did this, because I did want to move our backup out of the main Windows 2K AD server. There I had some hardware problems with my backup tapes (2 HP DDS-3 tapes) with different SCSI controllers. It worked all nice twice or three times successively and than errors were returned. As this happened with different SCSI controllers and the tape drives worked fine in a different testing environment I decided to test these controllers and tape drives in a new computer.
Then I had to decide, whether to set up a second Windows server or set up a Linux server. One nice thing about a second Windows server would have been, that it could be a backup server of the AD. But a Windows license is more expensive than installing SuSE 9.3 from a ftp and as I'm running quite some services on the first Linux server right now, this would be the oportunity to move some things to the new server (e.g. when I'll switch to kolab2 I could do that on the second server). Further on I like open source…
So I looked for a nice backup solution, which is able to do remote backups of a Windows server. And I didn't want to do this using a Windows share as I'd like to backup home directories and want to save the ACLs too. So no chance for amanda, but I did ask for a quotation of arkeia.
Well, it's too expensive. I wouldn't have got the budget from nowhere and even if I would have, it would have felt wrong. So I remembered reading about Bacula in a recent edition of the German Linux Magazine.
First tests were quite ok. Much options, complex configurations but the manual is very good, so working slowly and deliberate I got a first set up. Using my two drives I faked an auto changer, so that both tapes would be used optimal. But trying to run a complete job, I got problems with the band width. The Windows server was just able to send about 100 KB/s which just isn't sufficient for backing up about 20 GB of data (even runnig over weeking would be a hard job). The queer thing was that in a first test I got acceptable 800 KB/s!
So what to do now? I installed the current beta vesion 1.37.30, which is considered stable and will be released with minor changes as 1.38 hopefully soon. And now I got about 800 KB/s to 1200 KB/s out of my Windows server and the Linux servers were even up to 2000 KB/s!
It's running for some days now (running a full backup every Friday and a differential one Monday to Thursday). One thing I'd still like to figure out is running the differential backups not compared to the last full backup but to the one before.
Yeah, I know, that's just a question to the mailing list… ;-)
You might thinks, that saving the state of the Windows server and the AD could be a problem. But fortunately the very good manual of bacula has a solution here. Actually bacula is able to call external programs before the backup job is run—and this even on remote hosts! So the trick is to call
ntbackup backup systemstate /F c:\systemstate.bkf
on job start.
For my backup system it's quite easy. All the important Windows based data is located on this one server (yes, we use roaming profiles and all shared data is located on network shares as well as all important personal data). All the databases of our measurement computers are copied once a day to the Windows 2K server (they have to be copied somewhere anyway, so I can do this to the server directly), and all the important configuration data is in the AD. So that' it.
Finally it seems that the old problems where some obscure problem of the combination of mainboard, scsi controller, other cards, driver, and/or operating system. I hate problems like this because they're so hard to find and solve.
If you don't mind that there isn't a nice graphical interface to bacula available (yet), I can really recommend this tool. Working in the bacula console is very fast and effective, I'd just wish some graphical browser for selecting what to backup or even more important, what to restore (if I'm just looking for one file or directory in a big tree). But hey, if you'd really need that, you could implement this on your own! ;-)
As it was a beautiful launch, but NASA has stopped all shuttle missions due to the debris falling off…
I heard a commentary on a German radio station today that one problem is, that NASA just isn't able to judge the seriousness of this event, because that's the first mission monitoring such a big amount of data of the start. Before they just didn't look how much debris fell off, etc. So right now they can't compare it to former event but can only judge on the current data with the last distaster at the back of their minds.
I hope they'll return without problems from ISS…
It was really magnificient to see this. We've watched at work and we did experience these problems with the stream as some comments say at Asa's blog entry as well. Both the realplayer stream and the wmf stream lost connections sometimes. Fortunately not at launch… ;-)
To be on-site and see this live is definitely something on my TODO list!
There's a nice article in the National Geographic Magazine about energy resources of today and what may be in future. It gives a nice overview about some of the options but the most important quote is this one of NYU's Martin Hoffert:
Without a big push from government, he says, we may be condemned to rely on increasingly dirty fossil fuels as cleaner ones like oil and gas run out, with dire consequences for the climate. ”If we don't have a proactive energy policy,“ he says, “we'll just wind up using coal, then shale, then tar sands, and it will be a continually diminishing return, and eventually our civilization will collapse. But it doesn't have to end that way. We have a choice.”
Even as fossil oil is getting more and more expensive, alternative fuels (like biodiesel or ethanol), or alternative devices with higher efficiency in using fossil fuel (like fuel cells) are even more expensive. And far worse, there's much development to be done to get competitive. The oil companies will invest in new technologies for sure—to exploit oil deposits which are unattractive nowadays. To develeop completely new technologies needs a much higher investment than to invest slowly in the improvement of a well-known one.
Recently you hear more and more talk about alternative fuel from big oil companies (let it be biofuels or hydrogen economy), though. It could be the push from some goverments (like the zero emission cars in California). Or they “just” want to get into a good starting position never mind how soon or late fossil fuels will be replaced anything else.
There's another point that's almost never talked about. What will happen if we use all the fossil fuels somewhere available (coal, oil, gas)? Can you imagine how much carbon there's stored inside and how much carbondioxide this will set free if burned? And even if you store the carbondioxide somewhere (one of the craziest ideas) how much oxygen you'll need to burn all of this fuel? Then you don't have to care about warming of the atmosphere any longer…
I'm not very satisfied with the search results by Google for my name. While almost all results have something to do with me, even if you don't enclose my name in quotation marks, I'd wish for some easy way to “remove” the results of posts to mailinglists, newsgroups, and forums.
Just try it using ?hl=en&q=Carsten+Schurig. My blog is currently the third entry (ok, that may have changed tomorrow). If you search using German (?hl=de&q=Carsten+Schurig) it's worse with rank 34 currently.
But Google isn't completely wrong, just not intelligent enough. Most of the entries ranking higher are from mailing lists or forum entries. Here I would wish, I could switch between “All Results”, “On Webpages”, and “Posts to Forums, Mailinglists, and Newsgroups”.
If I'm looking for a name, it would be a big usability enhancement, if I could switch easily between results in e-mails (and similar) and results found on “normal” web pages. Often results in mailinglists are just not what I want when searching for a person, but if you're active on public mailinglists using your real name, Google will return these as most MLs have public archives (and if not there's still gmane).
Actually it should be pretty simple to detect if a result is an e-mail, newsgroup post, or forum entry. E-mails and newsgroup posts should be easily recognizable because of the headers, but forum entries could be a bit more complex.
It should even be possible to tell if a search string is a name, or not with a high certainty. You could use a name database for sure, but if many results of e-mails are returned, it's even easier: if the search string is found in header entries followed by an e-mail address (whether obscured or not), it's a name almost for sure.
A nice example is searching at Vivisimo. This grouping there is pretty nice (although they don't recognize the search string as a name), but I'm not using Vivisimo very often, because it's just pretty slow…
Google, are you listening?
Alfons is contemplating whether he's a Geek or not. Just read his blog (and hey, he's not using blogger or something alike) and decide for yourself.
At least I'm pretty sure I count as geek. If you look at the definition at Wikipedia, that's pretty clear and I suppose I even fit in more than one of the mentioned definitions: I studied physics (you have to be a kind of Geek to do that, haven't you?) and I'm still interested in it (as well as maths) ;-), I like Star Trek (although I'm not a real Trekkie as I don't go to conventions and so on), I like reading SF and fantasy (my most favourite SF and fantasy writers are currently: Douglas Adams, Isaac Assimov, Terry Pratchett, J.R.R. Tolkien) but I'm reading more “serious” books too (Kafka, Nabokov, Sartre, Camus, Nietzsche). I'm fascinated by new technologies but fortunately I can refrain from buying every new gadget I find interesting and finally I know how to write a plugin for my blog system (having a blog with many computer related posts is a point as well). ;-)
And above all, I don't complain about being called a Geek…
First I did today was testing my idea I had yesterday evening. Windows can easily access WebDAV folders, and by this you can use a subversion folder almost as if it would be a normal Windows folder. Turning autoversioning on you can copy changed files to the folder and get a new revision.
Yes, unfortunately you can't edit files on such shares directly, because Windows isn't able to use the standard locking mechanisms defined for DAV.
So I had a local file to which I saved often and after saving I copied it to the WebDAV folder (it's really nice to have two monitors for such things ;).
First, the dialogs used when copying to a WebDAV folder (and dialogs asking for overwriting and so on) are not Windows XP styled but look very old. That I can't understand…
Second, I had to add an unsecured access for this repository. For some reason Windows XP wouldn't accept my https connection, which IExplorer, Firefox and TortoiseSVN except without problems. I don't have an officially signed certificate obviously, but I imported this one when asked (and the dialog said, it's ok). It didn't work, explorer just kept complaining…
It didn't work as I hoped, so for real work with LabVIEW e.g. (direct editing) it isn't usable nevertheless. And I wouldn't like using subversion this way anyway. Because you commit too often (on every copy, or save if locking would work) and you don't get nice log entries (using some other client you don't get them by themselves as well, but you have a chance at least ;). So I will use this “temporary” repository for disaster recovery when working with some problematic but important applications like Diadem (it's faster to copy the file to the WebDAV share using D'n'D than using TortoiseSVN or doing a save as and you have old versions available), but not for more.
Today at work I was seething with rage: after working for some hours on some data, I lost the complete graph I had created (using some time consuming tricks to circumvent some incabalities of the program)! How that? Well, I know that with complex layouts the programs tends to crash sometimes and so I do save often. But it happened today, that the program gave just an error message on saving and wrote 0 bytes to the file, so that no working version has been left. I'm not able to send a file for testing to the support as I don't have it anymore!
But doing a “Save As…” anytime is really not convenient! (Maybe I should try using a WebDAV connection to a subversion repository with autoversioning turned on? Nice idea, just haven't tried if this is working yet…)
So you ask which application I use? Diadem by National Instruments. It has an unusual GUI, but it can access the database used by NI directly (all other programs can just using SQL commands…) and it has no problem with large data sets (analysing long term measurements can easily lead to datasets with more than a million points). The biggest problem I encounter are with creating and editing the graphs. Using many graphical elements leads to problems sometimes (not always easily reproducible) and the graphical cabalities are just developing so you have to use some tricks sometimes (which may increase the stability even further).
James Doohan (“Scotty” from Star Trek) passed away at the age of 85 today. I think because of the phrase “Beam me up Scotty” his characters name is the best known of the Star Trek world. Even many non-trekkies have heard this…
The post title is from a scene in Star Trek IV when Scotty tries to use a mid-1980s Macintosh (found in Moment Of Silence).
BTW: the Wikipedia entry had been updated already…
Just found on Spreeblick the link to the Trinity remembered web site. I'm not sure how biased the information on this site is, but the event should not be forgotten!
Ok, I have to admit it: I like reading Harry Potter and have finished volume 6. When I wrote about documenting my skin and packaging it I just forgot about HP6 arriving this saturday…
Yesterday there was a chance not to read too much, since I met with friends for a BBQ (I couldn't help to start reading, though ;-). I really tried to get a break today, but it just carried on and on (as I experienced it with the other volumes as well).
So now I can go on reading William Gibson's “Idoru” (German translation). As I'm reading it the second time, I couldn't put the new HP away! after reading this I'd like to get the Neuromancer trilogy. I've been reading just the first volume, but the German translation of 1987 is so bad, that I thought about returning this book to the shop more than once—sometimes I was remembered of newspeak in Orwell's “1984”… Luckily the translation of the “Virtual Light” trilogy is much better.
The URL for the alternative processes group at flickr is from now on http://www.flickr.com/groups/altprocess/. This is easy to remember but can't be changed.
Although the original article in Chronicle Careers is some days old now, I can't blog for myself without commenting this. I did find a reference to it in the Photoethnography Blog and after following some links I found these two very interesting blog entries about this article (read the comments!): Matthew G. Kirschenbaums blog and this entry on Planned Obsolescence.
I don't write about internal gossip in my company (even if there is some) and I 'm always very cautious blogging about work related topics. You'll find some things which fit pretty well with part of my hobby interests: part of my job is the administration of our small network. As we don't want to spend too much money, I can (if there's sufficent time) almost anytime implement some open source solutions. But I won't write many things about our fuel cell business, except things which we're publishing anyway (yes, I'm working on fuel cell systems primarily). I won't comment on strategies of my company as well. This is discussed internally and after it came to a decision, I think it's important to transport this with one accord.
I am publishing under my name. Everyone can use Google or blog specific search engines to find my blog. If you have to blog in secret, because you don't want your colleages to know, you should look for a new job anyway. I don't advertise this blog to my colleagues, though, but I won't deny it as well. One reason I don't advertise it very openly, is...
The discussion on Planned Obsolescnece had some posts about hobbies not being reputable. And I think that's true for many people. Speaking about hobbies, which help to recreate from work and load the battery to be creative again, is often not well regarded with one exception: going to parties (and drinking a lot and maybe meeting people you normally wouldn't like meeting at all). But having hobbies alone (or in small groups) is often seen problematic. And worst is computers, computers at all. That's just freaks, seeing no light, having no social life at all…
Many people sitting in front of a computer in their spare time are just leading a different social life. With blogging it's more and more noticed on the outside of these communities, but I think there're some similarities to open source communities. Just look at most of the larger open source projects. That's many people working on some project (some even need some serious project management) interacting socially very often; just not face to face.
With blogs a similar community is more easily accessible to everyone connected to the net. You'll find blogs to almost any search string you submit to google nowadays (whilst the software project you'll only find if you're interested in some related topic most often) so people get to know these networks. Comparing it to a open source community isn't very good, because bloggers are usually not working on a project together (not counting colleagues and blogs about software projects) but are writing on things they're interested in and building networks as other bloggers find them and comment in some way—so I'll rather call it “blogging networks”.
But finding blog networks doesn't mean the bloggers are understood. E.g. in my company I installed a blog based on Nucleus a while ago. But for most of my colleagues blogging is an unknown concept to them. Writing an email to all (which will be lost at least for some much earlier often) is the concept right now, even if it would fit very well into the blog. This is a learning process, which will need time…
I think it's important what you write in your blog, but I never would say, someone is a problem because he/she has a blog! Then I could start saying, he/she is a problem because he/she is human. Definitely you can use a blog to get an impression of the person, and why shouldn't you if he/she advertises this openly in a letter/email or something else (actually I'd be glad to see someone having some fun in their spare time and communicating it).
if I'd advice him against of pro blogging, I'd tell him that right now there're many people who don't understand blogging and bloggers, but that's no reason to refrain from blogging. To be loyal to your employer is self-evident in my eyes, even to a former one. And I would advise him to use his real name. In Germany there're even some laws and it's not yet sure, whether they require a private blogger to publish his name, address and so on (I think they do but that doesn't mean I think that's sensible).
I think Mr. Ivan Tribble falls into the trap himself, as Daybreak comments on Planned Obsolescence:
As I mentioned elsewhere, what's odd is that this author commits pretty much every sin that the column claims bloggers commit. Actually, it reads rather like a bloggy rant. But really every day I read better blog enteries than this column.
I don't know if this entry is a “better” rant than Mr. Tribbles, but I did need to write that…
Uploaded new css files for the colour definitions as I made some stupid mistake in the first version: I did clean up too much and didn't pay attention to the ordering of the colours for links. To work correctly they have to be in some specific order unfortunately (:link, :visited, :hover, :active).
And I uploaded a black and white version too…
I finally had an idea for a name of my Nucleus skin which I happen to like: NuText.
For this I separated the colour code from the main CSS code for this skin. To demonstrate this use your browser functions to switch skins. The default theme is the old one now called “blueISH” and then there's another one called “orangeISH”. I plan to add a black and white one, resembling the original colours of the minima skin more or less.
Yesterday evening I've been to Episode 3. Quite nice. Most astonishing, it was a real OV. Until now, all English films I've seen in cinema here in Dresden were the German version with the English soundtrack. You can see this almost anytime because of the translated title, or even worse some translated subtitle (to German) for some part of the film (I remember some scene in one of the Harry Potter film, where a diary entry in English is translated to German in subtitles).
And what's more, even all the trailers were in English (but the only ones they had were from Sony film companies).
I did work again a bit on my skin and changed some of the inconsistences. E.g. now only the artice text is white, all the text around (labels, secondary information) is in grey. Further the comments are formatted slightly different now and stand out a bit more. The footer of the comments fits better the footer of the article now. If there would be any trackbacks, you could see that they're formatted accordingly…
And I promise, I'll release this skin soon! ;-)
I just updated NP_RandomImg to version 1.3, which adds an option for a section title (you can use this text for anything you like actually). So the “Random Cyanotypes” on my site is now no longer hard coded in the skin, but output using <%RandomImg(title)%>. Download NP_RandomImg-1.3.tgz if you'd like to use this. There has changed nothing else.
I did add two features to rakaz' Trackback plugin for Nucleus. First you can now return the time of the post using <%time%> (like the <%date%> variable) and you can get the global id with <%tbid%> like it's possible for comments. With this you can create a permanent link to a trackback in your blog. You could see this working in my block, if there'd be some trackbacks, but unfortunately there aren't any (yet?) :-(.
This version is called 2.1.cs42, so you can easily see, that you don't have the original version installed. I hope, rakaz would like to add these new features… If you like to test my version, dowload the patch, unzip it and apply it to the original 2.0.1 version.
There're some things I'd like to (have) changed. Most important is IMO, that the date and time formats are used like defined in the template settings (but I just didn't find an easy way to retrieve this). Further on it would be better, to add the Trackback settings, to the skin templates (where you find the comment settings). This is much better fitting, as you could easily define different settings for different skins. Or at least make most of the options blog specific…
First I thought I wouldn't write anything about London today and I won't. I just want to say that I couldn't have expressed it better than Rainer in his blog entry.
I just cannot stand any terrorism, and extremism whatever reason they plead…
“Always use your best judgment and highest degree of care.”
So what's this about? Do it yourself:
As all my book are on shelfes in my working room currently (yes, that happens! ;), I went to my living room too, where always some books are scattered around. So here's the quote from the “living room” book:
“‘Are you serious, sir?’ he said in a small whisper which had the effect of silencing the pub.“
Now, does anyone know where theses quotes are from? At least the second one is easy (I don't want to know the exact edition/compilation). If I'd post the next sentence, it would be too clear…;-)
I wonder why that was known there not until today. Actually I didn't read anything in other technically (English speaking) blogs before today (maybe I just don't remember ;). But there's been news about it on German news portals end of last month (e.g. on 2005-06-30 at heise online).
What a coup of the french! They want to end the monthly migration of the members of the European Parliament to Strassbourg and use the saved 200 million € to fund the European Institute of Technology.
It seems, there's a chance now to establish this without any additional money paid by the member countries of the European Union. And the memebers of the European Parliament will be freed of a monthly burden…
Congratulations to this really good idea! Strassbourg is a nice place for such an institute too. Actually I don't care very much in which European country this institute will be established—hey we're living in the EU!
Security alert!? If you updated your Nucleus based blog (like me) to 3.21, you should be on the safe side again…
Actually all applications using XML-RPC for PHP have a problem and should update to a newer version or disable their XML-RPC interface.
There's a new GIMP development snapshot (2.3.2) out. Unfortunately it doesn't seem, that the next version of The Gimp will support editing of images with more than 8 bit per channel still—that's one of the features I really miss (and no: cinepaint isn't usable for editing large photographs). Including colour management is very good, though. I always have some problems giving pictures for printing somewhere, because a) they want CMYK (see more below) and b) if they print from RGB they're using Adobe Photoshop normally.
What's the problem with this you ask? Well, Gimp doesn't include information about the colour space used yet. So Photoshop opens the picture assuming the use of AdobeRGB—but it isn't! Gimp's always using sRGB right now (and should tell so). So the only colour management I really need, is to tell other programs the correct colour space.
Actually I think CMYK is the most overrated feature for amateur or even semiprofessional use. There always a RIP special for the used output device should yield the best results. As a matter of fact, these RIPs won't be included with The Gimp, and I think that's OK. These should be provided by the manufacturer of the output device or some specialized programmers, like the gutenprint ones.
But here it is again important, to tell the RIP the correct colour space you used working on the picture! (What you should have, is a simple way to show how a CMYK output device would alter the colours of your image.)
I think, converting to CMYK will get more and more unimportant for many (not all) users, because most of the output devices (right now some kind of inkjet printers for high quality photo prining mostly) will use more than CMYK anyway. And if you would just convert your image to CMYK, the printer (or a RIP inbetween) would have to convert it again to use all inks available…
There're some nice flickr tools out there. Herer are two:
I think, that flickrgraph is really good and interesting.
I finaly published my Nucleus plugin for more member information. Instead of MemberVita, I called it MemberInfo now.
As always I did add the wiki page and announced it on the Nucleus Support Forum.
It just adds to fields to the member settings page of a nucleus blog. First an url to an image and second a longer text entry which you can use to write something about you. On the member details page you can get an img HTML tag with the url inserted calling <%MemberInfo(image)%>. You'll get the text using <%MemberInfo(longtext)%>. But beware, this does only work on the member details page!
If set up correctly, it can look like this. ;-)
