HARTLAUB@DULRUU51.BITNET (02/08/90)
WOWwwwwh! There seems to be a lot of discussion out there. As many others I really like that. First of all: my first message was signed by U.D.O - that means: Udo Hartlaub at the Dep. of Virology at the University Ulm, West-Germany I didn't want to hide myself behind 3 letters; there was just the wrong file included into the message-file. Sorry. And if anything I wrote so far and I will write next sounds arrogant, it is/was not meant this way. As well I don't want to insult anybody; Sorry if there are some sentences which could be interpretated this way. The discussion has left the initial wish of Theresa Farrah. Even if there are individual solutions for her map/six-wish I see no problem for this special case. But I would prefer individual solutions. Since this is already a general discussion: I still wouldn't like the idea if the GCG-group is really thinking about adding more and more interacitve questions to the GCG-programms! If really something like this will happen - a prompt for every possibility of every programm - I would suggest a switch - like the "doc/nodoc"-switch - where you can choose between "extanded/normal/no-help" for interactive questions. But maybe an abbreviated form of the on-line COMMAND_SUMMARY-help included in the documentation (the text which appears after you have started a programm) could help those people who are not willing to read the manual or the on-line-genhelp. A new prompt "Include any optional parameters?" while you still can read the list of possibilities could satisfy those who don't want to type slashes or have forgotten the possible options. This prompt won't appear if there were already typed some qualifiers at the command level. >>I just got Lynna Stone-Infeld's mail: >(...)GCG should >include plain english explanations of all questions and command line switchs >for all of it's programs. These explanations should tell the user what the >answers mean to the problem at hand. Lisa Caballero recently explaned the >meaning of the, "Integrate how many adjacent diagonals?" question in GELOVERLAP >and WORDSEARCH. Her explanation was accurate and extremely usefull. Since, even >the newest user could understand the important role of the answer to the >stringency of search. Explanations such as Lisa's should be incorporated into >a GCG manual. Maybe there should be a new Tutorial Manual... But any of such solutions would still have a disadvantage: people have to read and people have to think first and then use the computer. That seems to the main problem - as Jasper Rees wrote. Concerning this discussion: how many of the real GCG-users are really out there ? I mean those who use it day by day. For real questions or purposes. It seems that most of the people discussing are "sysops" (as Michael Yablonsky wrote) who are more concerned about people working with GCG than using the programm by their own. It might be that there are myriads of GCG-users round the world, who have no access to networks (either because of missing knowledge or m:ssing opportunity). Those people - I guess - have no help by a sysop or a GCGuru who helps them every day with every problem. Two years ago I have started to learn this programm without even knowing what a directory is. This process worked - step by step - trial and error - with not understandable helps from people who where just two steps ahead - but in my eyes they were the real GCGurus. And I am still learning! I guess that this is a typical situation. Maybe some beginners know people who know more and who can help more than those I knew. I absolutely do agree that scientists are lazy and don't want to learn such simple things like basic VMS-commands. I, as well, didn't want that. But I had to. And at least Mike Yablonsky had to, as well. (My grandma - by the way- didn't want to bring her money to a bank, because she didn't want to learn how to handle checks - until some day she had to, because she couldn't pay some bills cash! Think about it!!) And I think that most people using GCG should be able to learn basic-VMS even if there minds are fixed by MACs (machines you don't find as much outside of Northamerica). But obviously they don't have to, if there is a nice guy (sysop, ...) who helps and helps and helps. ("Would you like some more help with this command or with that command...). Maybe such folks get too much help. I don't want to be misinterpretated: but a scientist who wants to publish sophisticated analysis of his/her sequence should know what the programm he/she used for this analysis had calculated. That's a lot! Starting a programm from the command-level with one or two qualifiers or creating a symbol is nothing in comparison ! Reading the opinion of Stephen Clark (just as an example) I can't help thinking: If the scientists in his institute are not willing to learn a handful (maybe two handfuls) of VMS-commands have they ever learnt what a output of programms as "peptidestructure" or "fold" means. Do they know which alignment/comparison-programm is the best one for a special question. Can I trust them if they write in a paper: there was no significant sequence-homology to sequences of the xyz-database? (At this point I better stop brain-storming) If they want to use simple prompt-prompt-prompt-prompt-prompt-prompt-prompt-prompt-prompt-output programms let them work on MACs or other PCs. There is lot of such software available. Some even for free as I learnt from Lisa Caballero. One of the big advantages of the GCG-package is that nearly all programms can be used at the command-level of the VMS-system. Of course this might be for some people a disadvantage. To make it easy the UWGCG-people had the great idea to introduce some help on different levels, including the definition of helpful symbols (for example the "up", "down", "home" -commands). That`s all right and it helps really a lot. There a lot of other good things, most people working with the GCG-package even don`t know. (Tell them!!! - I mean you operators and GCGurus out there and as well the GCG-Group. But you have to shout!) Everybody I talked to was able to make his/her own first steps with this help. >>Maybe these steps are easier if some explanations are easier to understand also for not native English-speakers, as Lynna Stone-Infeld suggested.>> The rest comes alone! Sometimes people are complaining that they have to type again and again the "pattern" in the find-programm. Really a lot of typing if they are looking for 25mer in some sequences. And a lot of proofreading. They were happy "(-8 and 8-)" when I told them how to use the command level and how to recall former commands. The same is with people who are complaining about the disaster of thousand files in their main-directory with .map .gap .xyz extensions. They are happy when I show them what a nice command the "delete" command is; and they are happy again - after thousand "delete" commands - when I tell them what a wildcard is, and they are happy again when I show them, that there is a /confirm and /exclude and /since qualifier. Let them make mistakes or let them use complicated commands or procedures like deleting file by file. One day they get bothered and - some scientists even think practically - think that a computer must be able to have a better solution. I myself I am still happy if somebody shows me how to make the VMS-work and GCG-work easier. But step by step! Remember: Every Guru needs his/her Guru. Thank you for still sitting in front of your stupid screen reading my stupid comments. Udo Udo Hartlaub TEL: (0731) 176-2687 Abt. f. Virologie FAX: (0731) 176-2038 Universitaet Ulm EARN/BITNET: HARTLAUB@DULRUU51 DFN/EAN: HARTLAUB@RZ.UNI-ULM.DBP.DE P.O.B. 4066 D 7900 Ulm West Germany
clark@MSHRI.UTORONTO.CA (02/08/90)
A few quick comments on Udo Hartlaub's message: /It might be that there are myriads of GCG-users round the world, who have no /access to networks (either because of missing knowledge or m:ssing /opportunity). There are 150 subscribers to this list, plus or minus a few. Anyone know how many sites use the GCG package? /Two years ago I have started to learn this programm without even knowing /what a directory is. This process worked - step by step - trial and /error - with not understandable helps from people who where just two steps /ahead - but in my eyes they were the real GCGurus. And I am still learning! /I guess that this is a typical situation. Maybe some beginners know people /who know more and who can help more than those I knew. Two years using the GCG package and haven't mastered it yet? Maybe it's kinda hard to learn? :^) /And I think that most people using GCG should be able to learn basic-VMS /even if there minds are fixed by MACs (machines you don't find as much outside /of Northamerica). I thought the best selling machine in Germany was the Atari ST. Have I been mislead? For those of you in the US who haven't seen one, it is very much like the Mac. /Reading the opinion of Stephen Clark (just as an example) I can't help /thinking: If the scientists in his institute are not willing to learn a /handful (maybe two handfuls) of VMS-commands have they ever learnt /what a output of programms as "peptidestructure" or "fold" means. Do they /know which alignment/comparison-programm is the best one for a special /question. Can I trust them if they write in a paper: there was no significant /sequence-homology to sequences of the xyz-database? I think it's fairly safe to say that you can't trust their interpretation of any computer analysis that's much more complicated than restriction enzyme mapping. (The reason it's safe for me to say this is because I haven't been successful in getting them to regard VMS mail as much more than a curiosity, so I know they won't be reading this newslist! Let me give you an example. A couple of weeks ago I showed one person [not only the head of lab, but the head of a division] how to use mail so he could retrieve the results of a fasta search that we had sent off to GenBank. The next day he asked me to get some information for him. About three days later he says, "What happened to those things you were going to find out for me?" Me: "I mailed them to you. On the VAX." Him: "Oh." Him, ten minutes later: "My password doesn't work anymore." Now of course not all people are like this, but the important thing to understand is that this is the kind of person that takes up so much of my time and is hopelessly frustrated by the GCG programs. But I am rambling.) /To make it easy the UWGCG-people /had the great idea to introduce some help on different levels, including /the definition of helpful symbols (for example the "up", "down", "home" /-commands). That`s all right and it helps really a lot. These kinds of symbols are the first ones I commented out of the start-up command procedure. I want people to understand what they're doing when they move around in the directory tree. If you don't understand directories and filenames, you're totally lost. Please let me know if you think I'm wrong in this regard. It just occurred to me that the reason Udo likes the GCG interface is because he only shows a few people how to use it, whereas I have to deal with many many more. Is this a realistic conclusion? Also, as Udo points out, probably the only people contributing to this discussion are those that are pretty well aquainted with VMS, the neophytes being oblivious to the whole thing, so we're missing the most important opinions of all! Stephen Clark clark@mshri.utoronto.ca (Internet) sinai@utoroci (Netnorth/Bitnet) "We should be quite remiss not to emphasize that despite the popularity of secondary structural prediction schemes, and the almost ritual performance of these calculations, the information available from this is of limited reliability. This is true even of the best methods now known, and much more so of the less successful methods commonly available in sequence analysis packages. Running a secondary structure prediction on a newly-determined sequence just because everyone else does so, is to be deplored, and the fact that the results of such predictions are generally ignored is insufficient justification for doing and publishing them." - Arthur Lesk, 1988