

| Subject: | Re: Howto: make your own searchable vanagon archives |
| From: | Jake de Villiers |
| Date: | Sun, 11 Jan 2009 13:31:57 -0800 |
Wow Wes, thanks for doing this! You too John. :-) Now all I have to do is figure out how to use the file! On Sun, Jan 11, 2009 at 10:43 AM, pickle vanagon <greenvanagon@gmail.com>wrote: > A while ago, I promised that I would put together a file with some > easy-to-follow instructions about how to get the vanagon archives into your > gmail account "in a few days". First a few words on why this is necessary: > the web search interface for the archives doesn't use an indexed search, > which means that searching it can take 5 minutes or so, or, in practice, > hours or more, since it will often simply time out before returning a > result. Searching the archives with google is not a good substitute, since > google doesn't index most of the archives, so that a google search will > miss > many messages. (I will not cover this fact in these directions. If you > missed my previous examples of how the google search doesn't work, "search > the archives" to find them. Once you follow the directions below to get > the > archives in a browseable and searchable form, it will become obvious just > how bad google is for the archives, as you'll see that some searches that > get only a handful of hits with google should actually be getting > hundreds). There is one decent alternative to what I describe below when > it > comes to searching the archives: the "email" search interface, described > here: http://gerry.vanagon.com/info/searching.html. It is still painfully > slow and has its own share of issues, but it doesn't time out and is > reliable. > > Weeks later, I'm finally ready to deliver something less than what I > promised. It is possible to get all of the archives into gmail (I've done > it with my account) but I don't recommend it. It is slow and tedious (it > took me more than a week of constant uploading by an automated script) due > to gmail's slow imap import process. It can also make gmail mad at you (I > got locked out of my account for a day at some point) and is generally not > a > good choice for most people. For that reason, I'm not going to try to give > "step-by step" directions on how to get the archives into your gmail. > > But what I will do is almost as good (or maybe even better). By importing > the archives into the opera mail client (which is free), you will have > access to the entire archives of the vanagon list whether or not you are > connected to the internet. Plus, since opera mail has an "indexed search", > you can search the entire archives in seconds all on your own machine. The > best thing would be if the web archive interface was "fixed" to use an > indexed search, but until that happens, I consider this to be a pretty good > solution for most people. > > I've prepared the archives as an "mbox" file, which John Meeks has been > kind > enough to agree to host for listmembers to download on his vanagonauts > site. You can also import the mbox file into a different email client of > your choice, but I can't write instructions for every email client so > you're > on your own there. I chose opera as the "suggested" client because it > seems > to be the only readily available client that has an "indexed search" built > in (and is available for just about any operating system), but another > solution is to use another client (like thunderbird) in conjunction with > google desktop or a similar search program. Like I said before, you are on > your own if you want to pursue that route. > > One final thing: in spite of the fact that these directions can give us the > ability to have complete access to the archives, I think we should keep in > mind that most listmembers will not bother to download the file and go > through with this (which is quite reasonable, since the process requires > upwards of a gigabyte of free space to complete). As a result, I would > suggest that it is still probably not very constructive to answer peoples > questions with the answer "search the archives", since, for many intents > and > purposes, the web archives remain broken and google doesn't work either. > > > Here are my directions to get the archive into opera mail: > > 0. Download the gerry.zip file from > http://www.vanagonauts.com/files/gerry.zip > 1. The file gerry.zip is compressed and needs to be unzipped. (Some > versions of Windows may have this functionality built in?) The size of the > file gerry.zip is 162M. If you have a file that is much smaller than that, > it probably didn't download correctly. (For anybody that wants to know, > the > MD5 sum of the file is 408626910cc14145089907ba7f7c66c1) > 2. When unzipped, the file "gerry" (no extension) is 576 megabytes. It > contains every message in the vanagon archives (starting on April 2 1994) > through November 30, 2008. The unzipped file is an "mbox" file, which can > be imported or used by most mail programs. You are on your own when it > comes to importing this into your mail program of choice. I will show you > how to import the file into opera mail, which includes an indexed search so > that you can search the entire archive in seconds. > 3. Install opera (get it for free from www.opera.com) > 4. On the left bar in opera, click on the envelope button to go to opera > mail. Under the file menu, choose "Import and Export" > "Import Mail". > Choose "generic mbox file" and click next. Hit the "add mbox..." button > and > locate the unzipped "gerry" file. Under "Import into:", you can leave the > setting as "new account" or you can choose an existing account if you > already have an account setup that you would like to use. Click next, and > opera will start processing the file. This will take a long time. When it > finishes, you will have the archives in an easily searchable and browsable > location. > 5. If you setup opera as your email client for your vanagon subscription > (follow readily available online instructions), you will always have all of > the vanagon mail available for search, minus the "gap" between November 30, > 2008 and whenever you set it up. Maybe at some point in the future, I will > make an extra file to cover the "gap". > > Happy searching! > Wes > -- Jake 1984 Vanagon GL 1986 Westy Weekender "Dixie" Crescent Beach, BC www.crescentbeachguitar.com http://subyjake.googlepages.com/mydixiedarlin%27
Have something to add on this topic today? Revive this thread in the forum »