Simultaneous editing of files
Tags: latex onedrivePosted in Getting published, Technical (ms word, tex), useful software, Web 2.0
Progress
In a post two years ago I discussed how difficult it was to find good tools for editing scientific texts by multiple authors simultaneously. As there is a great demand for this functionality it does not come to a surprise that the situation has improved considerably. I will describe two solutions here that in our group we now use on a daily base.
Scientists regularly write text with other authors. In the old days authors kept on sending emails to each other with as attachments the latest version of the manuscript. This is a sure recipe for chaos including as disasters where authors were editing an old version or where new versions got lost.
Cloud
With cloud servers like Dropbox there is no need any more to send each other new versions. They can all be downloaded from cloud server. This improvement however does not exclude authors working on different versions. On average scientists are not very good in understanding new digital products and with this cloud solution they regularly forget to synchronize their local copy with the cloud server. In addition cloud servers as Dropbox take up so much bandwidth of a desktop computer that it becomes problematic to have a live connection all of the time.
On-line
So we need an on-line service where the source files are located at just one location and where live editing is possible and not to slow. Google Drive seems to be a solution but their file format is a disaster. Under the interface it is all dirty and poor html unsuitable for scientists who share sheets, documents and presentation.
Acceptable solution 1: overleaf
Physicists and mathematicians need many formula’s in their papers. In such a case there is no alternative to Latex and its sisters Tex, AmsTex, etc. If you think that equations formatted with Equation Editor or MathType look great you are an amateur and you should go for Acceptable solution 2.
The site overleaf.com has a great Latex interface, where people can edit simultaneously the same file (left panel) and see the result as a pdf file (right panel). It works great and we are moving more and more all our papers in on that site.
Acceptable solution 2: onedrive
Many scientific journals allow manuscripts that are formatted as MS Word files. Microsoft has now created a cloud-based solution that is quite acceptable. On the Microsoft site onedrive.com files can be shared with and edited simultaneously by multiple authors with a MS Word browser interface. Users can also install a local program (client) that allows to edit a local copy of the file with an Office desktop application. In the latter case users should check whether their local copies of the files are synchronized.
Advice: use your onedrive file collection only for a limited number of files. In such a situation all the files are quickly synchronized and your computer does not suffer from a performance penalty with the onedrive client installed.
22 Dec 2015 17:52, Philip Harding
I claim to have no experience on simultaneous editing of manuscripts – for the simple reason that usually the manuscript is edited by one of the authors and then passed on to the next. However, I can certainly see some advantages of simultaneous editing, e.g., two authors working on separate sections of a manuscript.
Reading this post, I was rather surprised not to find the golden standard for concurrent editing – a version control system. Software developers share the following dichotomy all the time: Sharp deadlines must be observed, but both must continue work on the most recent version. For exactly this reason, version control systems were introduced.
One of the most popular ones, Apache Subversion svn permits simultaneous editing of all files in its repository. Before commencing an edit, the desired files have to be checked out. This ensures that the author always works on the latest version. After editing, the file formerly checked-out has to be checked in (committed). In the case that two or more authors commit a checked out file, svn checks for conflicts, which have to be resolved before finally committing. The files can be edited offline, but checking in and out require internet access. svn boasts a potent windows explorer integration via TortoiseSVN, so no need to use the command line.
A system which has gathered immense momentum in the past few years is git and popular hosting platforms GitHub or bitbucket. Programmed by Linus Torvalds for use with the linux kernel, it has become very popular with software developers and is especially powerful when dealing with a large number of files shared by a large team. One of the main differences to svn is that every member of the team has his/her own repository, which can be synchronized with the master repository only if so desired. Bitbucket permits small teams to set up up to 5 (?) private repositories for free.
So, why not try out git or svn on your next manuscript?