It is an odd thought that sometimes wonders past my addled brain that a client for doing my blogging on would be really useful. Why? Well I can blog offline of course. Now you are going to say well then use a word processor and well you’d be correct but this would be a very short blog if I listened to you and did so. But that is not how I function. The logical easy path must require me to rebuild to modify to change something. Cutting and pasting from libreoffice straight into wordpress seems just too easy now doesn’t it? So what did I do instead I installed three blogging clients from the Fedora repo and well lets see how did I get on?

gscribble Project URL :

A desktop client for blogging for wordpress and blogger

Hmmm I liked the icon and well that is all. It wouldn’t work, couldn’t connect so I gave up. I know I should have tried harder but well I didn’t want to.

blogtk Project URL :

BloGTK is a weblog posting client written in Python for any platform that supports Python and the GTK widget set. BloGTK allows you to connect to your weblog with XML-RPC and create new posts.

Better it let me login and appeared to work but it was not that good to be honest. I am getting old and slow now and often make spelling mistakes and type-o’s and of course I like to see them as I type. But oddly this local run client was less functional than the actual wordpress environment itself.

lekhonee Project URL :

Wordpress desktop client frontend for Gnome

Imagine my surprise when I discovered I already had this client installed and apparently with login credentials. Now imagine how as I attempted to post a new blog realised the reason I had forgotten about it. I don’t like it. Again no in-line spell checking. Clunky interface.

With that I gave up. I wanted an offline client because whilst I am uploading loads of YouTube videos I wanted to work on things without the need for internet bandwidth and the answer is truly that simple use libreoffice or any native word processor.

This got me thinking some more. Why do I do these kind of things? For example EndNote. I hate Endnote always have and now even more than I use to. I have spent a long time finding alternate things to use. I have a MySQL database of journal information. I have a self coded program to organise and embed meta data into pdf files. This is quite nice in itself and based on things like ( because crossref is just brilliant. But with all that said and done when I need to make a publications list for my CV I ended up back at EndNote. Why? Well I’d stopped using all my own code and invested time in EndNote via Codeweavers Crossover. This was rather stupid as I will need to uninstall it shortly, it doesn’t work properly, it is horrible.

So where next?

Bibus - of course I had lots of things in Bibus - but NO bibus does not understand the DOI natively. Really. BUT we love the DOI I could write a blog on how much we love the DOI all my journal pdf files are saved as DOI codes. Yes no DOI. But wait a patch for DOI - yeah! Arrgh a patch no six patches and many, many hours later no way to get the DOI back out into the citation. For the love of God. In the end I had to find the source rpm from Fedora download install edit the spec file get the patches build a new rpm with the latest version of the Bibus code because the Fedora maintainer isn’t , probably because the project seems to be a little bit in a mess (paraphrasing their bugzilla post). Now I have a new rpm but still not a fully functioning product and do I want to invest more time? Ok well I can get my references out of EndNote and back into MySQL using Bibus and the DOI patches and the endnotexml import function. No, no - I can’t. Why because the title field is being split at the first “’” ok another hour recoding that and then I gave up and switched to RIS and yes another bug the DOI flag is in as DA not DO dooh - fixed done sorted still no way to cite DOI in the output - FFS.


Zotero - I use to love Zotero. Probably because it was all open and Firefox related but then I stopped using it. Went through hours of recoding styles so that I could get references out you guessed it for CV publication section and in the end stopped using it. Now I have the standalone version. I really want my data out of the sqlite datebase and into a mysql one so I can do other things such as link all those pdf files back with their parent references. But no they say NO we will never do that, no external database and NUTs to it.

I’m sure there is something out there. Referencer which I spent an age coding plugins for is no long in Fedora and well even after building a new version from source I must admit I am not impressed.

So do I write my own? - no not again my pdf pypdf thing is broken and not parsing pdfs so now my meta tag updater is broken which is annoying since once you’ve gone to the effort of getting all the data from it is nice to put it into the pdf so you don’t need to get it again and you’ll always have the data in the file. But I can quickly and easily write a MySQL parser to get my data out of the database in a sensible format irrespective of Bibus and I can use Zotero to help collect data as well so perhaps I can use many tools rather than one nice one and get to the same place? Who knows?

Now you are going to ask about searching for articles aren’t you? Well I have a search engine called recoll and I use that to search for information from my pdf files - hence my love of having the meta data in the file as well things like the title can quickly be resolved, etc, etc.

I now have comments open and will happily take them on Linux referencing software to use in an academic environment. If I see anyone mention “Papers” I will probably spill my coffee but other than that carry-on.