| General > General Technical Chat |
| Migrating the forum to Discourse |
| << < (54/69) > >> |
| james_s:
--- Quote from: SilverSolder on April 26, 2021, 12:25:18 am --- Hopefully, if Dave ever gets tired of running the EEVBlog, it will be handed on as a going concern to "someone or something"... but it is hard to ignore that Dave is uniquely good at this, and any successor likely to mess it up (with tons of white space, for example!) :D [/quote There's only one Dave, but there are lots of interesting people out there and the forum has pretty much taken on a life of its own at this point. I think as long as someone is found who has the sense to not mess with a working formula it could survive just fine. --- End quote --- |
| EEVblog:
--- Quote from: SilverSolder on April 26, 2021, 12:25:18 am ---Hopefully, if Dave ever gets tired of running the EEVBlog, it will be handed on as a going concern to "someone or something"... --- End quote --- Hmm, so me up all last night projectile hurling into the bowl may not have been a myserty virus... I suspect I'd better watch my back, something is afoot.... :scared: |
| Ed.Kloonk:
--- Quote from: EEVblog on April 26, 2021, 06:54:29 am --- --- Quote from: SilverSolder on April 26, 2021, 12:25:18 am ---Hopefully, if Dave ever gets tired of running the EEVBlog, it will be handed on as a going concern to "someone or something"... --- End quote --- Hmm, so me up all last night projectile hurling into the bowl may not have been a myserty virus... I suspect I'd better watch my back, something is afoot.... :scared: --- End quote --- As long as your head doesn't start rotating on your shoulders. Keep an eye out for sudden bed-levitation. Failing that get the family to chant The power of Christ repels you. |
| MarkMLl:
--- Quote from: drussell on April 26, 2021, 01:28:41 am --- --- Quote from: aandrew on April 25, 2021, 11:21:22 pm ---...no way to archive the site as it is so it could be searched on something like the internet archive. --- End quote --- wget? --- End quote --- wget works well for static sites where every document has a unique URL, and there's a "root" document with links in a tree structure. It's much less use for "modern" websites where there is no explicit tree, and locations are either identified like https://www.eevblog.com/forum/index.php?action=unreadreplies;start=0 or like https://www.eevblog.com/forum/chat/migrating-the-forum-to-discourse/msg3557316/#msg3557316 It's no use at all where a page is identified solely by a site name, and all content is generated by inline scripting https://www.ebay.co.uk/myb/PurchaseHistory In principle, it's possible to scrape content by generating every possible message number and sending a wget request for that specific page. Both SMF and Discourse appear to be amenable to that to some extent, and I've done it with Hackaday in the past to get a complete content list (I was trying to organise my reading, it's a useful site but I'm 18 months behind on it). @aandrew mentioned the possibility of converting it to a mailing list, or IRC. I'd like to comment on that, and think it might be worth doing so in this message since anybody who isn't a comms nerd has already given up on it :-) Mailing lists and Usenet-style discussion groups have the useful property that every message has a unique ID that in principle is generated when the message is first created and never changed. That's why clever software can thread messages and subtopics intelligently, which makes using a large and busy discussion group much easier. Many email clients store messages in a flat file, which makes sorting and manipulation (i.e. "retrieve the message with this ID" difficult). Some email clients and most discussion group clients and servers store messages numbered sequentially. Looking "under the hood", client software can rapidly locate a message using that sequential number but not by the original message ID. A major problem here is that a user can post a link to a message stored on his client system or local server, but that link usually contains the sequential number rather than the ID so is useless to anybody else. In the UK there is a conferencing system called CIX, which at one time was heavily used by tech journalists and the likes. This had limited direct compatibility with mail/Usenet, BUT had the very important property that the unique message ID was guaranteed and any user could generate a message link containing it which was immediately useful to all other users and to the handful of offline readers (notably AMEOL, "A Most Excellent OffLine reader") which various people wrote to support it. I'd add here that MediaWiki, like SMF, stores pages in a MySQL database and that to some extent that can be backed up and manipulated. Discourse's choice is apparently PostgreSQL, which is probably superior in its scalability etc. and in any event isn't beholden to Oracle. Irrespective of backend, IRC messages carry no metadata and can't be meaningfully threaded. Good fun perhaps and OK for rapid-fire questions where a limited number of people are trying to talk at the same time, but much less generally useful. The bottom line is that for archival and forking purposes, there's no alternative to getting hold of the raw messagebase and then seeing what server software can be used to present it. I've known various people who've attempted to generalise this, and have done a limited amount myself. I'd like to emphasise that I am in no way suggesting a "palace coup" against Dave. SMF's presentation and Dave's site management are entirely adequate for what's going on here, although as a former CIX/AMEOL user I'd say that in principle the message threading could be improved. However if he did have any comments about scaleability etc. I think we'd all be very interested to hear them. MarkMLl |
| drussell:
--- Quote from: MarkMLl on April 26, 2021, 08:25:26 am ---wget works well for static sites where every document has a unique URL, and there's a "root" document with links in a tree structure. --- End quote --- The EEVblog forum has an index, so if one were to point a recursive crawler at it, (probably some incantation of something like wget -r -k -p https://www.eevblog.com/forum,) you should get a local mirror of the whole forum. This obviously doesn't let someone "continue" the forum but does provide a static archive of a site if it were otherwise disappearing completely. Obviously, common administrator etiquette dictates that you should never actually do that on a large site without first coordinating with the site admins, setting some bandwidth or rate limits, etc. |
| Navigation |
| Message Index |
| Next page |
| Previous page |