 |
 |
Apologies it was Alan who started being sensible. Re length on URL - there is an RFC somewhere which recommends that URLs should not exceed 256 characters. Matthew > -----Original Message----- > From: Theo van Veen [mailto:[log in to unmask]] > Sent: 14 June 2002 10:48 > To: [log in to unmask] > Subject: Re: Betr.: Re: result set model for srw > > > I did not use the word "sensible" so I assume you mix up my > comments with someone else. > With respect to escaping I think it will be sufficient to > prevent "%" and "&" ">" and "<" as part of URL parameters. > Internet Explorer does escaping automaticaly. It has more to > do with the fact that in combining javascript and XML it > makes things difficult. The user will hopefully never see the > resultsetid. > > I do not know the length limit of URL's. It depends on the > browsers and the server applications. I just want to > minimize the chance that a limit is reached. > > But in general: keeping things human readable will help > speeding up in realising stable applications and will lower > the threshold or barrier to use the URL's in other > applications for linking purposes. > > Theo > > > >>> [log in to unmask] 14-06-02 11:13 >>> > I was a little concerned by your comment "If a user is going > to use in a follow up query, then it should be sensible shouldn't it?" > > If by sensible, you just mean that it can be put in a URL without > > a) requiring escaping > b) exceeding the 256 recommended maximum length for a URL > > Then I agree with you > > We may need to include some limits on result set ids (e.g. > allowed characters and max recommended length) > > For a while I though "sensible" might mean human readable, > which I don't think appropriate here. > > Matthew > > > -----Original Message----- > > From: Theo van Veen [mailto:[log in to unmask]] > > Sent: 14 June 2002 09:43 > > To: [log in to unmask] > > Subject: Betr.: Re: result set model for srw > > > > > > There is no reason to assume that SRU confuses things. I do > > not think people are going to type in queries as URL's > > (although I sometimes do). The only thing that is important > > for SRU is that we keep the parameters URL friendly so that > > we do not need an extra level of escape sequences and that we > > keep them short. > > > > Theo > > > > > > >>> [log in to unmask] 14-06-02 10:14 >>> > > > > If we have (persistent) result set names, do we still > need session > > > > ids? > > > > > > > > --Ray > > > > > > One unanswered question to me (it might have been decided already > > > sorry): who invents result sets names? If the server just > generates > > > them, is there any obligation for the name to be > sensible? If a user > > > is going to use in a follow up query, then it should be sensible > > > shouldn't it? > > > > They are generated by the server. If this isn't clear in the > > current doc.s then it should be. > > > > Result set name is probably a misnomer - what this actually > > is, is a id for referencing the result set in order to > > maintain state. It isn't meant to be a nice easy name > > presented to the user! At the end of the day (although SRU > > used with thin clients and XSLT confuses the issue > > slightly) this is an on the wire protocol not a user > > interface definition! > > > > Matthew > > >