CIAO documentation features
which are missing or need improvement
Back to the Survey
9 - technical discussions 12 - ahelp pages usually need more content, such as brief descriptions of how the tools work, citations to ApJ (or other) papers on which tasks are based, etc. Online thread URLs have a nasty habit of changing. This can get very annoying! 13 - Not easy to find good detailed descriptions of standard products 15 - providing more examples under the ahelp pages of some of the more complex uses of certai commands. 18 - The web pages are too hard to follow -- to many hyperlinks. A simple linear thread would be better. 19 - Further expansion of the threads. For example, why do all the threads use PI bins instead of PHA? 21 - The "bugs" link at the bottom of every ahelp web page should go to the bugs list for THAT TOOL, rather than to the list of all tools. The cross-referencing between threads is often useful, but is sometimes just confusing, particularly when the cross-references send you several documents deep. In cases like that, it would be nice if the threads could be a bit more self-contained. 22 - Documentation which descibes what the scripts do, and what's called by the script would be useful. 23 - Some aspects of the current state of calibration are well-hidden to the user, specifically the areas where calibration is extremely uncertain are not as well-defined as they could be. 24 - The threads should offer much more explanation of the choices used. 28 - If I don't use CIAO for a while, I always forget the exact syntax for doing certain things (e.g., filtering on a spatial region). It always takes me a while to find different examples of it again. It would be helpful to have this information linked within every thread where it's relevant. 32 - The manual/online help could use a little more technical info; for instance, exactly how does sherpa go about subtracting the background, or simultaneously fitting BG and source; how does it deal with pixels on the boundary of a region defined in ds9; etc. 35 - More examples of analysis chains need to be given, particularly for grating spectroscopy. 39 - More warning regarding change of CIAO version while using same data 43 - brief usage messages, as common in linux w/ --help more top-down "ABC" guides, rather than "bottom-up" add-ons to web pages, which tend to fragment reading. 44 - *Printable* ABC Guides desperately needed. Threads are great once you know what you're doing, but a terrible way to learn. There needs to be something that you can print and read on the john. There's a certain cavalier attitude on the threads. All references to the 3 ms readout time of CC mode should be expunged. It's 2.85 ms, and thinking "Oh, that's close enough to 3, so let's just put 3 in 90% of the threads" really isn't good enough. 47 - The main thing missing in almost all the threads are discussions of the various assumptions and limitations that go into the analysis. Perhaps an expert user class of documentation needs to be considered. The threads seem to address the needs of the novice user fairly well already. Time to move beyond the lowest common denominator service level. 49 - More understandable examples on ahelp pages. 53 - threads! - Would like to know the general principles, and flowchart of SDP. Please don't ask me to treat x-ray analysis as a black box.. 59 - Information on calibration and potential problems with interpretation of the results. 60 - The documentation about script is not provided. 63 - The online ahelp pages take too many clicks to get to, and could easily do with more examples. More examples=better, always!! 64 - Please include more (and more diverse) examples in the ahelp. 69 - It would be nice to have a bit more information on the background dataset files in the CALDB. e.g. which datasets can be used with VF mode reprocessing on. As far as I can tell there is not much documentation on this and some of the header keywords in the files are misleading (e.g. header claims VF mode but the necessary columns are not there to use with a VF cleaned dataset). I also find the helpdesk search interface to be somewhat awkward. I find it very slow to use, and I often have to search to lots of irrelevant queries before finding anything usefull (which is anoying given that each one can take a long time to load). 75 - accuracy of threads (particularly hrc); lack of printable "cookbook" to browse 81 - more examples of correct analysis are needed, mainly as threads. Maybe authors of good papers could be invited to write a thread, explaining how (and why) they did various procedures in their analysis 82 - Some details of programs are poorly documented or difficult to find. 83 - It is hard to find the older versions of the documents e.g. threads for CIAO ver 2.2 84 - exceptions 90 - It is difficult when you have an idea of what you want to do, but don't know what task, thread, etc. will do it. Typing in keywords like "bin" doesn't produce output as useful as it could be. Even by somewhat familiar person as myself, some of the dictionary explanations are far too technical. While that technical detail needs to be there, there should also be a very basic laymans explanation. 99 - careful review of threads and ahelps by scientists. 102 - More up-to date informations. There seems to be a long lag between a "corridor rumor" on a problem/correction and the official announcement. We should not have to ask a question to get a (late) answer that says "Oh, this is a well known problem" when this "known problem" is nowhere described... 103 - the threads need to evolve to include more complex data analysis situations. 105 - I do not know. 106 - I'd like to see a more detailed thread on extracting and fitting radial profiles, including the use of an exposure map, and comparing profiles of a source with the psf to see if it is extended. 108 - more examples of common cases 111 - information on importance of corrections, e.g. CTI corrections... 113 - Consequences for data analysis of improvements in calibration 119 - add more threads, also on "basic" topics. Maybe it could be useful to collect them in a "how-to handbook" 124 - Scripts are often poorly documented (I mean the script code itself). There is little indication of what the steps are, and why things are being done. Also, scripts do not replace the need for step-by-step threads for what the script is doing. I don't want a black box, I want to know precisely what it is doing to my data. So, in general, I usually end up looking at the script, figuring out what it's up to, and writing my own with documentation of the steps, additional logging information, substituting tools with funtools or ftools, etc. 126 - The ahelp pages often lack information explaining all of the options for each keyword in the call sequence.
Back to the Survey