I installed Koha some time ago but gave up. I now want to try again. I tried updating to v2.0.0 but the updater program threw up error messages. That doesn't worry me since I have no data stored in the existing version. But now if I try to run installer.pl perhaps unsurprisingly I get a message about Koha already being installed. So how do I uninstall Koha so that I can start again. ie what files/directories do I need to delete? And do I need to delete the existing files etc from Mysql? Roger -- Roger Horne, 11 New Square, Lincoln's Inn, London WC2A 3QB http://hrothgar.co.uk/
On Thursday 08 April 2004 16:23, Roger Horne wrote:
I installed Koha some time ago but gave up. I now want to try again. I tried updating to v2.0.0 but the updater program threw up error messages. That doesn't worry me since I have no data stored in the existing version. But now if I try to run installer.pl perhaps unsurprisingly I get a message about Koha already being installed.
So how do I uninstall Koha so that I can start again. ie what files/directories do I need to delete? And do I need to delete the existing files etc from Mysql?
Roger
Hi Roger, *) Drop the Koha database *) remove the web sites (rm -rf /usr/local/koha if you installed in the standard location) *) rm /etc/koha.conf *) rm /etc/koha-httpd.conf *) delete the koha related entries in the mysql users table Then your system should be comletely koha-free; I did this twice in the last few days :-)
On Thu 08 Apr, Frank Weis wrote:
Hi Roger,
*) Drop the Koha database *) remove the web sites (rm -rf /usr/local/koha if you installed in the standard location) *) rm /etc/koha.conf *) rm /etc/koha-httpd.conf *) delete the koha related entries in the mysql users table
Then your system should be comletely koha-free; I did this twice in the last few days :-)
Thanks (and to Paul Poulain)). I foudn that it was best first to open koha.conf (since it contained the paths to the other files) then to delete it and those other files, and then to drop the database. Your final point is the one that I discovered via trial and error: the configuration program refuses to work if there is already a kohaadmin entry. The installer script did not like Apache2 in two respects. First it seemed to treat the Apache user (which is Suse 9 is wwwrun) and group (which is www) as being the same. So an eror message came up to the effect chown wwwrun:wwwrun failed. The second is that it does not recognise the fact that people may want to use name-based virtual hosts. I suppose that it might be difficult to include that in a general purpose config program, but it might be worth considering adding instructions. (I already had a virtual hosts file which I did not need to change.) Roger -- Roger Horne, 11 New Square, Lincoln's Inn, London WC2A 3QB mailto:roger@number7.demon.co.uk
On 2004-04-12 10:49:25 +0100 Roger Horne <roger@hrothgar.co.uk> wrote:
The installer script did not like Apache2 in two respects.
It works fine with apache2 for me, but I guess that's to be expected. What fun would a bug be if the maintainer could reproduce it?
First it seemed to treat the Apache user (which is Suse 9 is wwwrun) and group (which is www) as being the same.
This is a bug for apache 1.3 too. I'll hunt and kill it.
So an eror message came up to the effect chown wwwrun:wwwrun failed. The second is that it does not recognise the fact that people may want to use name-based virtual hosts.
I don't understand what you mean by that. Can you explain what you think it should do, please? I use name-based virtual hosts myself. All it took on the simplest machine was symlinking /etc/apache2/sites-enabled/koha to the koha-httpd.conf (debian directory-based vhost config). The installer doesn't handle many complicated cases, but we could go happily insane before we get them all. -- MJR/slef My Opinion Only and possibly not of any group I know. Please http://remember.to/edit_messages on lists to be sure I read http://mjr.towers.org.uk/ gopher://g.towers.org.uk/ slef@jabber.at Creative copyleft computing services via http://www.ttllp.co.uk/
On Wed 14 Apr, MJ Ray wrote:
... The second is that it does not recognise the fact that people may want to use name-based virtual hosts.
I don't understand what you mean by that. Can you explain what you think it should do, please?
I use name-based virtual hosts myself. All it took on the simplest machine was symlinking /etc/apache2/sites-enabled/koha to the koha-httpd.conf (debian directory-based vhost config).
Ah well, I am an amateur and my use of symlinks is pretty slim at the moment. What I wa referring to was the fact that the installer requires you to use two addresses, eg opac and library, listening on different ports, eg 80 and 8080. Suse 9 seems to set up things differently from debian. /etc/apache2 contains vhosts.d which contains a couple of templates. I have koha-httpd.conf in that directory and it includes my default host (beowulf) as well. The machine "listens" for beowulf, library and opac.
The installer doesn't handle many complicated cases, but we could go happily insane before we get them all.
I of course accept that. What I think might be useful would be for it to give pointers at that point to where people should look for further documentation, eg http://httpd.apache.org/docs-2.0/vhosts/ It does seem to me that Koha has a mixture of simple installation instructions (such as the main part of your installer) and ones which might perplex a linux guru (such as the virtually non-existent instructions for the z* daemon which I have yet to get running here (but then I am far from being any kind of guru). And while I am grumbling, I do think that Koha gives the impression in some respects of being amateurish and unfinished. (I know that one should not look a gift horse in the mouth but ...) Some months ago I tried the earlier version but gave up. So having successfully uninstalled the old vesion I tried again with v2. After the minor problems I have mentioned this went fine. So I tried to use it. As I said some months ago, I am a barrister. We have a Chambers library but all members also own their own books. So Room A might contain Snell on Equity belonging to Chambers as a whole and my room, Room B, might contain Lewin on Trusts, both of which are books on trusts. So it seemed to me that it would be a good idea to set up a number of different branches, eg one for Chambers and one for me. So I did that and set up a separate "book fund" for each of them. And set up several types, including law report (LR) and text book (TB). And I chose the budget based entry system. At the moment we have no electronic catalogue: everything will hve to be entered manually. So I followed the instructions and set up a supplier called "Already Owned". Now to enter a book I have to go to acquisitions and enter "al" as the supplier. Yippee, it exists. So I click on the "order" field and I get a form which enables me to search for the title. Since I have not as yet entered it, this naturally does not produce any result other than produce a "new search" page which enables me to "add new biblio". The form is http://library/cgi-bin/koha/acqui/newbiblio.pl?id=2&basket=3&sub= I completed this using: Title: Family Breakdown and Trusts Author: Longrigg Higgins [e 2 separate authors] Publisher: Butterworths Copyright Date: 2003 Item type: Text book [chosen from drop down menu] ISBN; 0406948291 Series [blank] Branch: Roger Horne [chosen from drop down menu] and in "Accounting details" I again selected Roger Horne as Bookfund and entered the cost as 0 (since I bought it before the current accounting period). I then clicked on OK. I then chose quick search, searching for Longrigg as author. This found the book, displaying the correct title and Authors, and copyright date, but claiming that there were 0 items and leaving the location field blank. There was no indication of the branch. (A customer searching a catalogue where the branches were several miles apart might find this annoying.) Clicking on the title produced a partial display of the Marc entries: 020 -INTERNATIONAL STANDARD BOOK NUMBER a International Standard Book Number 0406948291 852 -LOCATION/CALL NUMBER k Call number prefix (NR) TB 100 -MAIN ENTRY--PERSONAL NAME a Personal name Longrigg Higgins 245 -TITLE STATEMENT a Title Family Breakdown and Trusts 260 -PUBLICATION, DISTRIBUTION, ETC. (IMPRINT) c Date of publication, distribution, etc 2003 300 -PHYSICAL DESCRIPTION f Type of unit TB 650 -SUBJECT ADDED ENTRY--TOPICAL TERM a Topical term or geographic name as entry element [followed by a box containing] homebranch RH "TB" stands for text book so the 852 line would seem to be wrong. (If I go to edit I am given a drop down box for that line with text book, law report etc.) If I go back to the search window I am given the opportunity to search on "Subject". I am not entirely sure of the difference between "Subject" and "Keyword" but I assume that the latter means simply words in the Title, author, etc fields. Since the original form did not have an entry with the heading Subject the result of the search is, of course, empty. So I go back to the Marc entry page, http://library/cgi-bin/koha/acqui.simple/addbiblio.pl?oldbiblionumber=3 in this case, and modify the 650 line to add the word "divorce" and save it. Now when I look at the MArc entry I see 650 - SUBJECT ADDED ENTRY--TOPICAL TERM a Topical term or geographic name as entry element divorce So I search using the Subject line of the quick search form as "divorce": "Results 1 through 0 of 0 records [blank]" ... (And searching keywords for "divorce" also gives a nil result. I would try correcting this myself but I just do not know where even to start. (A quick search for eg "subject" through the perl scripts produces about 100 results.) </grumble> Entering books manually seems likely to take in excess of 10 minutes a title, which I consider to be unacceptable. Would it be possible for me to create a file of books and to run a script to enter them? (Many books are old and do not have ISBNs.) Roger -- Roger Horne, 11 New Square, Lincoln's Inn, London WC2A 3QB mailto:roger@hrothgar.co.uk http://hrothgar.co.uk/YAWS/
On 2004-04-15 11:41:55 +0100 Roger Horne <roger@hrothgar.co.uk> wrote:
to use two addresses, eg opac and library, listening on different ports, eg
These can be the same address, maybe the same port (not sure that I've tried that one) and even an existing name-based virtual host. Probably the wording isn't very good yet.
give pointers at that point to where people should look for further documentation, eg http://httpd.apache.org/docs-2.0/vhosts/
OK, noted. I'll add that to the default config file produced.
perplex a linux guru (such as the virtually non-existent instructions for the z* daemon which I have yet to get running here
Some call me a guru and the z*daemon defeated me last time out. I will try again soon.
And while I am grumbling, I do think that Koha gives the impression in some respects of being amateurish and unfinished. (I know that one should not look a gift horse in the mouth but ...)
Well, it's a .0 release and I don't think we're any worse than most software I've had to deal with. Feel free to gaze and tell me just which teeth look rotten. The challenge for me is whether I can do well enough with the next releases to polish 2.0.x into something which seems better and learn things so that 2.2.0 is better. I'm just writing my "roadmap". Keywords: free software, accessable, interoperable, portable, usable, stable. 2.0 DRAFT ROADMAP ----------------- Month - aims May - find and kill as many packaging and installation bugs as possible; see if we can get CVS commits under control, else junk CVS; ensure first wave of docs (installation, configuration) are up-to-date; release. June - replace systems that are really desirable (backport MARC searching from 2.1? has z*daemon really been fixed? is anyone working on the biblio entry screens in 2.1?); remove broken parts (is full acquisitions really broken or just buggy?); make sure second wave of docs (end-user help) are up-to-date; release. July - see what strong user requests we've got and hopefully act on them, or at least make sure 2.1 addresses them; release if there's updates. August-September - ensure that fixes we've made have been carried forwards to 2.1; enter maintenance period: only release new version for security or data loss. October - 2.2 should be out, but be prepared for another 3 months just in case. I'd appreciate comments and offers of help with the above. Does the plan look reasonable? Anyone know where the koha wiki went? -- MJR/slef My Opinion Only and possibly not of any group I know. Please http://remember.to/edit_messages on lists to be sure I read http://mjr.towers.org.uk/ gopher://g.towers.org.uk/ slef@jabber.at Creative copyleft computing services via http://www.ttllp.co.uk/
MJ Ray a écrit :
And while I am grumbling, I do think that Koha gives the impression in some respects of being amateurish and unfinished. (I know that one should not look a gift horse in the mouth but ...)
Well, it's a .0 release and I don't think we're any worse than most software I've had to deal with. Feel free to gaze and tell me just which teeth look rotten. The challenge for me is whether I can do well enough with the next releases to polish 2.0.x into something which seems better and learn things so that 2.2.0 is better. I'm just writing my "roadmap". Keywords: free software, accessable, interoperable, portable, usable, stable.
/me just want to add that Koha is something very complex : it needs technical habilities AND librarian habilities. It has to be a lot tweaked before showing it's power. The tweaking is mandatory : without it, you would think that it does not fit exactly your needs.
2.0 DRAFT ROADMAP -----------------
June - replace systems that are really desirable (backport MARC searching from 2.1? has z*daemon really been fixed? is anyone working on the biblio entry screens in 2.1?); remove broken parts (is full acquisitions really broken or just buggy?); make sure second wave of docs (end-user help) are up-to-date; release.
the z3950-daemon works in a real library since 3+ weeks in 2 of my libraries (Sorbonne & Chambery). No problems here. TJ Kotula sends me today some informations about the daemon loosing forks when he enters a lot of servers (>14 !!!). He sends me a suggestion to solve it (works for him). I'm working a lot on HEAD branch to : * clean some code * css the templates. * improve MARC searching & merge MARC & simple search. The search is almost done. It could be ported to 2.0 probably, but it's not a trivial job, as it uses CSS & some other things that does not exist in 2.0 DB (like "seealso" field in MARC parameters)
Anyone know where the koha wiki went?
Works fine for me ? (allthough a little slow) -- Paul POULAIN Consultant indépendant en logiciels libres responsable francophone de koha (SIGB libre http://www.koha-fr.org)
On 2004-04-15 16:28:20 +0100 paul POULAIN <paul.poulain@free.fr> wrote:
Anyone know where the koha wiki went? Works fine for me ? (allthough a little slow)
OK, it's back for me too now. Sorry all. http://www.koha.org/wiki/
On Thu 15 Apr, paul POULAIN wrote:
MJ Ray a écrit :
And while I am grumbling, I do think that Koha gives the impression in some respects of being amateurish and unfinished. (I know that one should not look a gift horse in the mouth but ...)
Well, it's a .0 release and I don't think we're any worse than most software I've had to deal with. Feel free to gaze and tell me just which teeth look rotten. The challenge for me is whether I can do well enough with the next releases to polish 2.0.x into something which seems better and learn things so that 2.2.0 is better. I'm just writing my "roadmap". Keywords: free software, accessable, interoperable, portable, usable, stable.
/me just want to add that Koha is something very complex : it needs technical habilities AND librarian habilities. It has to be a lot tweaked before showing it's power. The tweaking is mandatory : without it, you would think that it does not fit exactly your needs.
Well is it all that complicated? The main problem is the lack of documentation. I am a reasonable amateur perl programmer but i wouldn't have a clue where to start looking for any of the problems that I have found. Recently I was playing around with a totally different program (involving Court filing) and the writers had provided a program flow diagram which enabled me to modify, or tweak, the scripts in various ways. http://www.court-tech.org/incounter/documentation.php Roger -- Roger Horne, 11 New Square, Lincoln's Inn, London WC2A 3QB mailto:roger@hrothgar.co.uk http://www.hrothgar.co.uk/
Roger Horne a écrit :
/me just want to add that Koha is something very complex : it needs
technical habilities AND librarian habilities. It has to be a lot tweaked before showing it's power. The tweaking is mandatory : without it, you would think that it does not fit exactly your needs.
Well is it all that complicated? The main problem is the lack of documentation
We know that. As we know doc MUST NOT be written by someone that hacks the soft. So, if you want to volunteer to write it, be sure we will help you ! -- Paul POULAIN Consultant indépendant en logiciels libres responsable francophone de koha (SIGB libre http://www.koha-fr.org)
On 2004-04-15 17:59:07 +0100 paul POULAIN <paul.poulain@free.fr> wrote:
As we know doc MUST NOT be written by someone that hacks the soft. So, if you want to volunteer to write it, be sure we will help you !
End-user docs must not, but some docs MUST be written by the developers to give the sysadmins and technical people a start. There are a lot of undocumented scripts in there, even quite new ones. Developers MUST NOT forget this responsibility. -- MJR/slef My Opinion Only and possibly not of any group I know. Please http://remember.to/edit_messages on lists to be sure I read http://mjr.towers.org.uk/ gopher://g.towers.org.uk/ slef@jabber.at Creative copyleft computing services via http://www.ttllp.co.uk/
On Thu 15 Apr, MJ Ray wrote:
On 2004-04-15 17:59:07 +0100 paul POULAIN <paul.poulain@free.fr> wrote:
As we know doc MUST NOT be written by someone that hacks the soft. So, if you want to volunteer to write it, be sure we will help you !
End-user docs must not, but some docs MUST be written by the developers to give the sysadmins and technical people a start. There are a lot of undocumented scripts in there, even quite new ones. Developers MUST NOT forget this responsibility.
I entirely agree. I cannot see how anyone can contemplate writing any user guide unless he fully understands how the program works. I think that the guide needs to be written by a librarian who uses Koha every day. Roger -- Roger Horne, 11 New Square, Lincoln's Inn, London WC2A 3QB mailto:roger@hrothgar.co.uk
Roger Horne said:
I think that the guide needs to be written by a librarian who uses Koha every day.
Yeah, this would work for some things. But usually such guides include screen shots and detailed instructions, which do not take local "tweaking" into account. A guide written by the users at Nelsonville, for instance, would have limited use for a user at Unido. So I would suggest that such a guide for librarians be written by someone using the default templates and scripts. Is there any library out there that uses the (2.0) defaults for day-to-day operations? -- Stephen Hedges Skemotah Solutions, USA www.skemotah.com -- shedges@skemotah.com
Stephen Hedges a écrit :
Roger Horne said:
I think that the guide needs to be written by a librarian who uses Koha every day.
Yeah, this would work for some things. But usually such guides include screen shots and detailed instructions, which do not take local "tweaking" into account. A guide written by the users at Nelsonville, for instance, would have limited use for a user at Unido. So I would suggest that such a guide for librarians be written by someone using the default templates and scripts. Is there any library out there that uses the (2.0) defaults for day-to-day operations?
If templates are the only problem, it's quite easy to solve : we just have to copy/paste screenshots with the default templates. But i'm not sure templates are the only problem. Note also that I 100% agrees other mails : i was speaking of end-user documentation. Technical doc must be written by one of the developper. -- Paul POULAIN Consultant indépendant en logiciels libres responsable francophone de koha (SIGB libre http://www.koha-fr.org)
On 2004-04-16 14:33:36 +0100 paul POULAIN <paul.poulain@free.fr> wrote:
Note also that I 100% agrees other mails : i was speaking of end-user documentation. Technical doc must be written by one of the developper.
Further, the basic technical docs should be written by the initial developer when they write the script and the end-user documentation (at least for librarians) probably won't happen fully until this has been done. -- MJR/slef My Opinion Only and possibly not of any group I know. Please http://remember.to/edit_messages on lists to be sure I read http://mjr.towers.org.uk/ gopher://g.towers.org.uk/ slef@jabber.at Creative copyleft computing services via http://www.ttllp.co.uk/
If templates are the only problem, it's quite easy to solve : we just have to copy/paste screenshots with the default templates. But i'm not sure templates are the only problem.
I've had one or two offers of installed systems on default templates to actually get these screenshots (at least one of which seems to have not happened). Major needs for this help outline what procedures need documentation, advise on level of detail, and help draft. HLT and Regula did most of this the last time. It is not exactly what you'd define as the glamorous part of open source work.... : ). Any help on this front is appreciated
Note also that I 100% agrees other mails : i was speaking of end-user documentation. Technical doc must be written by one of the developper.
Last time around the principal contributor (to the now in need of work version) was Pat. If people want to take a look at the "developer" section of the old manual, and throw notes on the wiki(and drop me a link), I'll start pulling that together... The FAQ is usually my attempt to digest the complicated and frequent questions on the mail list. Developer input on this is welcome, as it is very easy to miss the actual resolution to a problem when looking at the back posts. Nick The sometime docs person.
Well is it all that complicated? The main problem is the lack of documentation
A non-developer will be working on the documentation. More volunteers (hoping beyond hope....) welcome. We have some volunteers for proofing, and items graciously contributed (mostly about migration). Nick Been slack koha-wise, hoping to get some help in reforming himself.
Roger Horne a écrit :
I installed Koha some time ago but gave up. I now want to try again. I tried updating to v2.0.0 but the updater program threw up error messages. That doesn't worry me since I have no data stored in the existing version. But now if I try to run installer.pl perhaps unsurprisingly I get a message about Koha already being installed.
So how do I uninstall Koha so that I can start again. ie what files/directories do I need to delete? And do I need to delete the existing files etc from Mysql?
delete /etc/koha.conf & /etc/koha-httpd.conf delete /your/path/to/intranet and /your/path/to/opac remove your DB & user in mySQL if you want. -- Paul POULAIN Consultant indépendant en logiciels libres responsable francophone de koha (SIGB libre http://www.koha-fr.org)
participants (6)
-
Frank Weis -
MJ Ray -
Nicholas S. Rosasco -
paul POULAIN -
Roger Horne -
Stephen Hedges