Re: [ql-users] Sbasic extensions toolkits
There are a couple of extensions around which allow renaming of proc/fn names. I think they do this by locating the token value in the BASIC tables and either swapping the corresponding names around or renaming them by patching. There's one in DIY Toolkit - can't remember its name, something like ALTER I think. IIRC the way these sorts of extensions work is something like: CHANGE_NAME_OF extension_name TO new_name finds the token corresponding to extension_name, then finds the token corresponding to new_name (and new_name should not already exist as a name in BASIC) and swaps either the tokens or names about, in effect creating a new dummy name and then swapping the names over. Not the best solution to this problem, but at least if this is done after the first extension is loaded to change a name then load the second extension it gives a bodge method of working around the problem of clashing names without resorting to permanently patching names in other people's extensions files. It should be possible to write a routine which loads an extensions file or toolkit and look for the names table within it and list the extensions out. That way, we can automate listing of existing toolkit names (I have copies of most non-commercial toolkits in my PD library). Rich - as you and your colleagues on the BASIC Reference Manual have documented the various toolkits for that, have you got listings of the keywords in the various toolkits listed by toolkit? That may help Wolfgang if he does decide to pursue this any further. -- Dilwyn Jones Usually every toolkit has a name. IMO the prefix should be derived from that. This can as well be the name of the author (anybody remember Beule-Tools from Peter(?) Beule?). As we're talking about SBASIC, I have spent much time thinking about this, and come up with a more elgant solution than my original suggestion. Command: PREFER command toolkit Example: PREFER cmp_string tkx In a case where two toolkits have the same username, a method should be created for a command to be identified with a toolkit by some simple abreviation. Thus, new toolkits would have some element or wand that says the toolkit name and version (so people can PREFER commands from different versions of the same toolkit) and use them in the same SBASIC instance. More elgant from the user perspective - no doubt a challenge to write. Just a suggestion. Dave
Re: [ql-users] Sbasic extensions toolkits
In a message dated 01/11/02 15:26:29 GMT Standard Time, [EMAIL PROTECTED] writes: Rich - as you and your colleagues on the BASIC Reference Manual have documented the various toolkits for that, have you got listings of the keywords in the various toolkits listed by toolkit? That may help Wolfgang if he does decide to pursue this any further. Ermm, no I haven't, sorry.. It should be possible to extract this information from the Q-Help files, I guess... -- Rich Mellor RWAP Software 35 Chantry Croft, Kinsley, Pontefract, West Yorkshire, WF9 5JH TEL: 01977 610509 http://hometown.aol.co.uk/rwapsoftware
RE: [ql-users] Sbasic extensions toolkits
Morning all, and a totally crappy morning it is as well ! Cold, miserable, foggy, traffic at a standstill due to some forward thinking private contractor to the local council (whatever they are called this week) starting work digging up the busiest set of junctions in the city - just in time for the morning 'rush'. 4 mile tailbacks - and I was in it Back on topic and off rant ! All this talk of having some way to rename extensions puzzles me, but perhaps I'm missing the point. Lets say I have a commercial compiled program which accesses an extension, say DJ_FILE_OPEN for example. If I use some utility to rename the extension, then my compiled program is going to barf - isn't it ? I like Dave's ides to use something like 'PREFER' in a manner similar to FLP_USE or RAM_USE etc, but I'm at a total loss as to how it could be made to work. When an extension with the same name as an existing one is loaded, what actually happens in QDOSMSQ ? Does the new one overwrite the old one or does it get pre-pended to the name list so that it is found first when searching for the name ? I suspect the latter myself, but I've never looked into it. Cheers, Norman. - Norman Dunbar Database/Unix administrator Lynx Financial Systems Ltd. mailto:Norman.Dunbar;LFS.co.uk Tel: 0113 289 6265 Fax: 0113 289 3146 URL: http://www.Lynx-FS.com - -Original Message- From: Dave P [mailto:dexter;spodmail.com] Sent: Wednesday, October 30, 2002 10:05 PM To: ql-users Subject: Re: [ql-users] Sbasic extensions toolkits Command: PREFER command toolkit Example: PREFER cmp_string tkx This email is intended only for the use of the addressees named above and may be confidential or legally privileged. If you are not an addressee you must not read it and must not use any information contained in it, nor copy it, nor inform any person other than Lynx Financial Systems or the addressees of its existence or contents. If you have received this email and are not a named addressee, please delete it and notify the Lynx Financial Systems IT Department on 0113 2892990.
Re: [ql-users] Sbasic extensions toolkits
Wolfgang writes, WL©_ASEARCH - ? ;) How about WL_IS_BEST_.. -groan- ;) Besides, I think a name or mnemonic should be helpful in reminding us what it does rather than who did it. Dont you? Sure I do - but then, MAT_xxx isn't really very clear either It is if my toolkit happens to be called MATrix TOOLkit. Anyway, it was just an example. (IMHO). If the list cleared all doubles anyway, we wouldn't even need prefixes... Since the whole purpose is only to avoid name clashes, perhaps using the author's name is the best way to do it - else, if you have two routines to compare strings, everyone will call them str_compare. Im sure just about everyone has written standard utilities called UPPER$, LOWER$ and the like. So as long as they do the same thing it shouldnt matter. However, this discussion seems only to have shown up the futility of continuing it. The idea of a list might be handy if it is easily available (eg on someone's web site) to check for names that have already been used. It is only necessary to list toolkits that are in the wild (eg general programming toolkits and special commands.) I think it would be a good idea to add parameter information to the list, as well as the name of the toolkit it belongs to, as both pieces of information will help to evaluate the likelyhood of a *relevant* clash. Per
Re: [ql-users] Sbasic extensions toolkits
Wolfgang writes: Toolkits will often be grouped for special purposes, such as array manipulation, string parsing, or to manipulate some low-level object (such as a database or timer, etc). Wouldnt it be more logical and aesthetically pleasing to use prefixes such as MAT_, STR_, DB_, CLK_, etc rather than WL_ or PJW_ ? Two things: 1 - don't forget the author's vanity :-) Maybe we should make the copyright symbol a legal S*Basic character, then we could have WL©_ASEARCH - ? ;) 2 - What if two people write similar keywords for the same thing? Youre right, but the chances of a clash are further reduced, and your list would still help things along. Besides, I think a name or mnemonic should be helpful in reminding us what it does rather than who did it. Dont you? Per
Re: [ql-users] Sbasic extensions toolkits
On 30 Oct 2002, at 14:28, P Witte wrote: WL©_ASEARCH - ? ;) How about WL_IS_BEST_.. Youre right, but the chances of a clash are further reduced, and your list would still help things along. Besides, I think a name or mnemonic should be helpful in reminding us what it does rather than who did it. Dont you? Sure I do - but then, MAT_xxx isn't really very clear either (IMHO). If the list cleared all doubles anyway, we wouldn't even need prefixes... Since the whole purpose is only to avoid name clashes, perhaps using the author's name is the best way to do it - else, if you have two routines to compare strings, everyone will call them str_compare. Wolfgang
Re: [ql-users] Sbasic extensions toolkits
[EMAIL PROTECTED] wrote: Sure I do - but then, MAT_xxx isn't really very clear either (IMHO). If the list cleared all doubles anyway, we wouldn't even need prefixes... Usually every toolkit has a name. IMO the prefix should be derived from that. This can as well be the name of the author (anybody remember Beule-Tools from Peter(?) Beule?). Marcel
Re: [ql-users] Sbasic extensions toolkits
On Wed, 30 Oct 2002, Marcel Kilgus wrote: Usually every toolkit has a name. IMO the prefix should be derived from that. This can as well be the name of the author (anybody remember Beule-Tools from Peter(?) Beule?). As we're talking about SBASIC, I have spent much time thinking about this, and come up with a more elgant solution than my original suggestion. Command: PREFER command toolkit Example: PREFER cmp_string tkx In a case where two toolkits have the same username, a method should be created for a command to be identified with a toolkit by some simple abreviation. Thus, new toolkits would have some element or wand that says the toolkit name and version (so people can PREFER commands from different versions of the same toolkit) and use them in the same SBASIC instance. More elgant from the user perspective - no doubt a challenge to write. Just a suggestion. Dave
RE: [ql-users] Sbasic extensions toolkits
Dilwyn, you must have so much stuff on your web site, that you have forgotton what you have :o) The source is available for download from your own web site. Cheers, Norman. PS. Most of the commands are 'specialised' names anyway and have not, so far, broken anything I use.. - Norman Dunbar Database/Unix administrator Lynx Financial Systems Ltd. mailto:Norman.Dunbar;LFS.co.uk Tel: 0113 289 6265 Fax: 0113 289 3146 URL: http://www.Lynx-FS.com - -Original Message- From: Dilwyn Jones [mailto:dilwyn.jones;dj.softnet.co.uk] Sent: Monday, October 28, 2002 7:36 PM To: [EMAIL PROTECTED] Subject: Re: [ql-users] Sbasic extensions toolkits Hi Wolfgang, to get this started I'll send (privately) the list and binary for DJToolkit which Norman Dunbar wrote to my spec. If needed I'm sure I can get the source to you too. This email is intended only for the use of the addressees named above and may be confidential or legally privileged. If you are not an addressee you must not read it and must not use any information contained in it, nor copy it, nor inform any person other than Lynx Financial Systems or the addressees of its existence or contents. If you have received this email and are not a named addressee, please delete it and notify the Lynx Financial Systems IT Department on 0113 2892990.
Re: [ql-users] Sbasic extensions toolkits
On Tue, 29 Oct 2002 at 18:11:58, Dilwyn Jones wrote: (ref: 004901c27f7a$21a515c0$a1075cc3blackpc) you must have so much stuff on your web site, that you have forgotton what you have :o) The source is available for download from your own web site. Just my job driving me towards a nervous breakdown...no time to do anything at all, even less enthusiasm for anything at all at the moment. Am absolutely DESPERATE to change jobs at the moment. Dilwyn - I have never know you _not_ desperate to change jobs (8-)# -- QBBS (QL fido BBS 2:252/67) +44(0)1442-828255 tonysurname.demon.co.uk http://www.firshman.demon.co.uk Voice: +44(0)1442-828254 Fax: +44(0)1442-828255 TF Services, 29 Longfield Road, TRING, Herts, HP23 4DG
Re: [ql-users] Sbasic extensions toolkits
Wolfgang writes: I've read this thread about names clashes in toolkits etc. with quite some interest. 3 - As a guideline, perhaps authors of future toolkits might want to envisage that they should, indeed, preface their extensions with, say, their initials, such as WL_ASEARCH instead of ASEARCH etc. This depends, of course, on your cooperation, but it might make future clashes less likely. What do you think? Toolkits will often be grouped for special purposes, such as array manipulation, string parsing, or to manipulate some low-level object (such as a database or timer, etc). Wouldnt it be more logical and aesthetically pleasing to use prefixes such as MAT_, STR_, DB_, CLK_, etc rather than WL_ or PJW_ ? Per
Re: [ql-users] Sbasic extensions toolkits
On 30 Oct 2002, at 1:10, P Witte wrote: Toolkits will often be grouped for special purposes, such as array manipulation, string parsing, or to manipulate some low-level object (such as a database or timer, etc). Wouldnt it be more logical and aesthetically pleasing to use prefixes such as MAT_, STR_, DB_, CLK_, etc rather than WL_ or PJW_ ? Two things: 1 - don't forget the author's vanity :-) 2 - What if two people write similar keywords for the same thing? Wolfgang
Re: [ql-users] Sbasic extensions toolkits
Hi all, I've read this thread about names clashes in toolkits etc. with quite some interest. It would seem to me that, at least for the time being, the path of least resistance would rather be to make sure that names just don't clash, rather than try to devise various -very ingenious- schemes by which this can be avoided at load time. There would be several aspects here: 1 - There must be a way of maintaining a list of currently existing keywords, list which could be sent out to anyone interested. I'm willing to compile and maintain such a list, provided I'm supplied with the toolkits that contain these new keywords. Please note that this would only be a word list, without any reference of what these keywords actually do. I'd gladly have somebody else do it - any volunteers? Isn't there already somethinbg available (commercially?) that contains the names of many toolkits (Sbasic reference manual?). 2 - There must be a way of making sure that names, old or new, don't clash anymore. I would propose the following in this respect: When I (or somebody else?) get the toolkits and compile these lists, we can already single out the clashes. Whenever possible, the authors of the programs could be contacted, in an attempt to have the names changed by them. If the source code exists, that should be possible. That way, at least a list of potential clashes can be published. Th writers of future software could get the list (total words + clashes) from me, or I could push it into this list periodically, if needed. Please note that there is absolutely no way that I could enforce any kind of order, I could only give the already existing names to software authors, if they still wanted to reuse old names, I couldn't help it. A problem exists when only the toolkit exists, without the source (and the author can't be contacted). Surely it is possible to change the names of conflicting keywords directly in the file they are contained in, with a file editor (as a blatant case of advertising, Wined comes to mind). This is entirely feasible, but for copyright reasons, we couldn't release the new binary file. However, a small basic program that made the changes in the toolkit could be released... 3 - As a guideline, perhaps authors of future toolkits might want to envisage that they should, indeed, preface their extensions with, say, their initials, such as WL_ASEARCH instead of ASEARCH etc. This depends, of course, on your cooperation, but it might make future clashes less likely. What do you think? (And, finally, I had already renamed SEARCH etc in newer versions of the fie and am sending his direct to François). Wolfgang
