<p>We have <ahref="http://wookware.org/talks/expocomputer/#/">an Overview Presentation</a> on how the cave data, handbook and website are constructed and managed. It contains material which will be merged into this website manual.
<p>We use a distributed revision control system (DVCS) for all the important data. On expo this means that many people can edit and merge their changes with the expo server in the Tatty Hut even if there is no internet access. Also anyone who is up to date with the Tatty Hut can take their laptop somewhere where there is internet access and update expo.survex.com - which will then get all the updates done by everyone on expo.
</p>
<p>In principle, survey notes can be typed into a laptop up on the plateau which is then synchronised with the Tatty Hut on returning to base.
</p>
<p>A DVCS is inefficient for scanned survey notes, which are large files that do not get modified, so they are kept as a plain directory of files 'expofiles'. The same goes for holiday photographs and GPS logs.</p>
<b>This password is important for security</b>. The whole site <strong>will</strong> get hacked by spammers or worse if you are not careful with it. Use a secure method for passing it on to others that need to know (i.e not unencrypted email), don't publish it anywhere, don't check it in to the website by accident. A lot of people use it and changing it is a pain for everyone so do take a bit of care.
<p>All the expo data is contained in 4 "repositories" at
expo.survex.com. This is currently hosted on a server at the university. We use a distributed version control system (DVCS) to manage these repositories because this allows simultaneous collaborative editing and keeps track of all changes so we can roll back and have branches if needed.</p>
<h3><aid="howitworks">How the website works</a></h3>
<p>Part of the website is static HTML, but quite a lot is generated by scripts. So anything you check in which affects cave data or descriptions won't appear on the site until the website update scripts are run. This happens automatically every 30 mins, but you can also kick off a manual update. See 'The expoweb-update script' below for details.</p>
<p>Also note that the website you see is its own Mercurial checkout (just like your local one) so that has to be 'pulled' from the server before your changes are reflected.</p>
60GB of stuff in 2017 which you probably don't actually need locally).
<p>If you don't need an entire copy of all 60GB, then it is probably best to use Filezilla to copy just a small part of the filesystem to your own machine and to upload the bits you add to or edit.
Instructions for installing and using Filezilla are found in the expo user instructions for uploading photographs: <ahref="uploading.html">uploading.html</a>.
<p>(do be <b>incredibly</b> careful not to delete piles of stuff then rsync back, or to get the directory level of the command wrong - as it'll all get deleted on the server too, and we may not have backups!). It's <b>absolutely vital</b>Use rsync --dry-run --delete-after -a first to check what would be deleted.
<p>If you are using rsync from a Windows machine you will <em>not</em> get all the files as some filenames are incompatible with Windows. What will happen is that rsync will invisibly change the names as it downloads them from the Linux expo server to your Windows machine, but then it forgets what it has done and tries to re-upload all the renamed files to the server even if you have touched none of them. Now there won't be any problems with simple filenames using all lowercase letters and no funny characters, but we have nothing in place to stop anyone creating such a filename somewhere in that 60GB or of detecting the problem at the time. So don't do it. If you have a Windows machine use Filezilla not rsync.
<p>(We may also have an issue with rsync not using the appropriate user:group attributes for files pushed back to the server. This may not cause any problems, but watch out for it.)</p>
<p>To edit the website fully, you need to use the disributed version control system (DVCM) software which is currently mercurial/TortoiseHg. Some (static text) pages can be edited directly on-line using the 'edit this page link' which you'll see if you are logged into troggle. In general the dynamically-generated pages, such as those describing caves which are generated from the cave survey data, can not be edited in this way, but forms are provided for some types of these like 'caves'.</p>
<p>Once you've downloaded and installed a client, the first step is to create what is called a checkout of the website. This creates a copy on your machine which you can edit to your heart's content. The command to initially check out ('clone') the entire expo website is:</p>
<p>or right clicking on the folder and going to commit in TortoiseHg. Mercurial can't always work out who you are. If you see a message like "abort: no username supplied" it was probably not set up to deduce that from your environment. It's easiest to give it the info in a config file at ~/.hgrc (create it if it doesn't exist, or add these lines if it already does) containing something like</p>
<p>Before pushing, you should do an <tt>hg pull</tt> to sync with upstream first. If someone else has edited the same files you may also need to do:</p>
<p>Simple changes to static files will take effect immediately, but changes to dynamically-generated files (cave descriptions, QM lists etc) will not take effect, until the server runs the expoweb-update script.</p>
<p>In Windows: install Mercurial and TortoiseHg of the relevant flavour from <ahref="https://tortoisehg.bitbucket.io/">https://tortoisehg.bitbucket.io/</a> (ignoring antivirus/Windows warnings). This will install a submenu in your Programs menu)</p>
<p>To start cloning a repository: first create the folders you need for the repositories you are going to use, e.g. D:\CUCC-Expo\loser and D:\CUCC-Expo\expoweb. Then start TortoiseHg Workbench from your Programs menu, click File -> Clone repository, a dialogue box will appear. In the Source box type</p>
<p>for expoweb (or similar for the other repositories). In the Destination box type whatever destination you want your local copies to live in on your laptop e.g. D:\CUCC-Expo\expoweb. Hit Clone, and it should hopefully prompt you for the usual beery password.
<p>The first time you do this it will probably not work as it does not recognise the server. Fix this by running putty (downloading it from <ahref="https://www.chiark.greenend.org.uk/~sgtatham/putty/">https://www.chiark.greenend.org.uk/~sgtatham/putty/</a>), and connecting to the server 'expo@expo.survex.com' (on port 22). Confirm that this is the right server. If you succeed in getting a shell prompt then ssh connection are working and TortoiseHg should be able to clone the repo, and send changes back.</p>
<p>The script at the heart of the website update mechanism is a makefile that runs the various generation scripts. It is run every 15 minutes as a cron job (at 0,15,30 and 45 past the hour), but if you want to force an update more quickly you can run it he</p>
<p>The scripts are generally under the 'noinfo' section of the site just because that has (had) some access control. This will get changed to something more sensible at some point</p>
<p>Clicking on 'New cave' (at the bottom of the cave index) lets you enter a new cave. <ahref="caveentry.html">Info on how to enter new caves has been split into its own page</a>.</p>
<p> Once you have pushed your changes to the repository you need to update the server's local copies, by ssh into the server and running hg update in the expoweb folder. </p>
<p>This is likely to change with structural change to the site, with style changes which we expect to implement and with the method by which the info is actually stored and served up.</p>
<li>Contents lists & relative links for multi-article publications like journals. Complicated by expo articles being in a separate hierarchy from journals.</li>
<li>Other people's work - the noinfo hierarchy.</li>
<li>Style guide for writing cave descriptions: correct use of boldface (<em>once</em> for each passage name, at the primary definition thereof; other uses of the name should be links to this, and certainly should not be bold.) </li>
