home *** CD-ROM | disk | FTP | other *** search
- Please see the XRS*.DOC file for complete information on these settings!
-
- * * * * * * * * These suggested entries must be made in "CONFIG.XRS" for
- * * NOTICE! * * them to be effective. The guidelines below should help
- * * * * * * * * you optimize XRS and memory usage for your own machine!
-
- ALSO NOTE: This document only discusses a very few parameters available to
- be used in CONFIG.XRS, most of which are related to optimization of XRS on
- your specific computer. Please be sure to see CONFIG.NEW for recent new
- parameters not discusses here which may be useful to you, and CONFIG.DMO
- mentioned above, which is a 'model' for generic setup.
-
-
- The recommendations below are everything I know about tuning XRS, and I
- think I've done quite a bit of it over the past four years for people! I
- can't promise any of it will help if you don't read carefully, but I know
- that if you do, your copy of XRS will be finely tuned to your personal
- computer depending upon how much and what types of RAM you have available,
- what type of diskettes or disks you have, etc. These items all basically
- fit into three areas - things you should (or shouldn't!) do, things people
- commonly do wrong, and things most people want that are not defaults, but
- extend the flexibility of the program. XRS has grown from a very simple
- program to one of almost infinite possible variations, and I will soon have
- to write a free-standing "configure/setup" program, but for now this should
- suffice! Have fun, and let others know (via the "QMX_XRS" echo) how things
- work best on your computer if the 'normal' options don't work well for you!
- Please be sure to detail exactly *what* your hardware/software is so that
- everyone can help you analyze your sitution!
- Mike Ratledge
-
-
-
-
-
- Swap X
-
- Is one of the most important options - especially if you want to use an
- external editor rather than the editor which I have integrated into XRS.
- The 'X' parameter is optional, and specifies the drive letter where you
- want to swap out the XRS program if no or insufficient LIM/EMS memory
- exists. If it is left off, the default drive would be used. XRS leaves
- only a 3.5k "stub" of itself in low memory, freeing all the remaining
- memory for the external program (or DOS "Shell" on the <F10> hot-key).
- Note that the exact amount of memory swapped out will vary every time you
- swap, since XRS dynamically allocates up to all available low RAM and up
- to 2.1MB of LIM/EMS (and up to 96k of XMS memory if you use an overlayed
- version). If you don't have enough LIM/EMS available at that time, XRS
- will automatically swap to disk instead. If you have neither (say on a
- 2-360k floppy system), XRS will cancel the swap operation for future
- external program calls after informing you it is doing so. Unless you
- know this option won't be useful for you (you have a slow hard drive and
- / or don't have any LIM/EMS to swap), you should always enable this!
-
-
-
-
- PreLoad Summary
-
- This is probably *the* most important optimization option - especially if
- you tend to read large mailbags and/or jump around a lot (using the <J>ump
- menu-bar option after viewing a message)! This forces XRS to initialize
- it's "Heap Expander" routines, preloading up to 2MB of summary/index info
- in LIM/EMS and/or up to available memory less 80k in low (640k) RAM. If
- for any reason, this function fails (whether it is because you run out of
- available LIM/EMS or don't have sufficient low RAM), it automatically will
- free all space which was allocated and turn off the "preloaded" flag inside
- XRS, and XRS will use the SUMMARYx.XRS file instead of the data preloaded
- in RAM. If you have plenty of LIM/EMS memory, this allows you to preload
- the index for 30,000+ message for instantaneous lookup, if you have 640k
- and no LIM/EMS, and use an overlayed version, you should be able to preload
- information for 2500 messages. Again - unless you know you have small low
- RAM available (less than 640k) or seldom do any jumping around - and this
- includes reading by AREA, since it also uses the jump table - you should
- always enable this! If you have available memory and run off diskettes,
- you *should* try to limit your mailbag size and turn this on, since XRS
- will otherwise have to continually search your diskette for information.
-
-
-
-
- ;No EMS
-
- *Don't* use this unless you have a specific LIM/EMS problem with XRS.
- In every case to date, LIM/EMS conflicts have been resolved by updating
- to a newer EMM driver if available. If you do have a definite conflict
- and are unable to resolve it with a driver update, you *must* use this
- parameter, or somewhere along the way you will have problems! XRS does
- not require a LIM/EMS version 4.0 driver, though - any truly up-to-date
- LIM/EMS 3.2 driver will do!
-
-
-
-
- ;TimeAdjust 20
-
- If you have an 4.77MHz 8088 CPU or a CGA with no vertical retrace in the
- hardware (i.e. "Snow"), you probably don't need to adjust this. See the
- main documentation file for details. This only affects how often the
- background procedures 'signal' before any are allowed to execute.
-
-
-
-
- ;Packer LHARC.Exe
-
- 90% of the problems with this parameter are because people don't read the
- docs or notice you need to specify an extension (".EXE" or ".COM"). The
- .ZIP format is the *default* - if you want to use PKZip - leave this out!
- If you want to use LHArc (the older version) and PKZip would be found on
- your PATH (XRS always looks for it first) - use the above - if you want
- to use the 'new' LHA, try "PACKER LHA.EXE".
-
-
-
-
-
- Buffer 8192
-
- This parameter controls the variable (user-selectable) amount of buffer
- space allocated for the BATxMAIL.XRS file (the text of all the messages)
- and can be set anywhere from 512 up to 32768. If you have a slow hard
- disk, or run on diskettes, set it to a lower value, if you have a very
- fast ESDI or SCSI drive, set it up to the 32768 maximum - assuming you
- also have plenty (640k+) of memory, which you should is you have these
- fast (and usually large) drives! Depending upon how you set this option
- XRS may give spectacularly fast or "square wheel" clunky access to the
- messages if you jump around or read within message groups (which most
- people do). You may have to experiment a bit to find the optimal setting
- for your system, and you should always "tune" it if you move to another
- machine!
-
-
-
-
-
- ;Virtualize 250
- ;VirtualJump 100
-
- The defaults for list virtualization (which is new in XRS 4.50) is to
- virtualize the list only if it is more than 500 messages, and to move
- the sliding window by 1 1/2 times the screen-size (i.e. virtual page
- size - or number of messages removed from one end and added to the other
- if you scroll past the current virtual list). Under Desqview, you may
- wish to make these smaller so that XRS runs in a smaller RAM footprint.
- The <J>ump list takes approximately 10k per 100 entries, so if you have
- plenty of RAM, you may wish to increase the defaults to avoid frequent
- paging - or eliminate paging in most cases altogether, building the whole
- list in memory each time instead. Virtual lists allows XRS to build a
- portion of a list for a 30,000+ message mailbag, for example, and to page
- entries in and out as required as you scroll through the list. This is
- not possible without virtualization, since that list would take 300k or
- more just for that one function (and it must be built in low RAM).
-
-
-
-
-
- ;No Clock
- ;No Seconds
-
- These two are usually not needed, but provide less "Snow" on CGA-based
- systems that don't do verticle retrace timing, or eliminate the constant
- on-screen clock ticking which throws "Screen-Speak" hardware/software for
- visually impared users off running all over the screen constantly.
-
-
-
-
- <INS> <F9> Reversed
-
- Almost everyone prefers to have the default C_Worthy functions of <F9>
- (toggle insert/overwrite) and <INS> (paste) reversed! Using this option
- doesn't optimize a thing, but makes those keys work more intuitively in
- the internal editor (and *only* effects them in the internal editor!).
-
-
-
-
- To Name Prompt
-
- Most people like the ability to quote/reply to a message, but redirect
- the response to someone other than the author of the original message.
- Again - this one doesn't really optimize anything, except it makes XRS
- more flexible. The reason this isn't the default is that previous XRS
- versions didn't have the option at all, and if I made it the default,
- people who typically never do this have to "bop through" another prompt
- for every message.
-
-
-
-
- Jump Headers
-
- This option forces more frequent access to the "BATxMAIL.XRS" text file
- during <J>ump mode (from the "read messages" menu-bar), but displays the
- header of each message as you scroll through the resulting point and shoot
- list. Note however that is you have a slow drive, or run on diskettes,
- this option is available for a single message by hitting <TAB>, so most
- likely you should turn the continual updating off. Note again that this
- optimization *only* affects using the "<J>ump" menu-bar selection. Since
- XRS does "relative jumps" inside the file rather than seeking from the
- start of the file, this option usually works well unless you have a very
- slow hard drive or diskettes only.
-
-
-
-
- ;list view
- ;page view
-
- These two allow you to force XRS not to use "Optimal Viewing" during
- message display. Normally, XRS looks ahead to see how many lines are
- in each message and picks one of these modes automatically (list view
- if it's too large to fit on one screen with the menu-bar, or page view
- if it fits). Page view displays a page-at-a-time if a message is more
- than a full screen - the mouse can only be used to continue to the next
- screen-full (and a marker keeps track of where you last were reading).
- List view displays everything in a true scrollable list, and the mouse
- can be used to skew the list up or down to read or review the message,
- however, you must use <ESC> or click the right button on the mouse to
- exit page view mode before the menu-bar appears. Different strokes for
- different folks. The optimal viewing is likely best - if you don't have
- a rodent, the up/down arrow keys and PGUP/PGDN, etc keys work in List
- view mode, anyway.
-
-
-
-
- ;Bundler RollUrOn.Exe
-
- Is *NOT* needed if you use the 'transparent' QWK format support via
- Rudi Kusters' "XCS" programs named "QWK2XRS.EXE" and "XRS2REP.EXE", but
- those two programs must be available somewhere on the PATH! This is a
- further attempt to "open up" the architecture and allow people to build
- exporting bundlers other than the internal "XRS_Pack" (FTS-0001) or Rudi's
- XRS2Rep.Exe (QWK) packers, so if you have need to build something else,
- please get in touch with me. (these two already cover "fidonet" and other
- similar network output and QWK flavor "*.REP" mail outbound mail packing)
-
-
-
-
-
- Thread only (this makes the "+" / "-" keys appear only for threading use)
- ;Threading (this makes them thread if available, otherwise next/back)
-
- The way the plus and minus keys worked in all previous versions were as
- the equivalent of the "<N>ext" and "<B>ack" functions. Version 4.50 has
- true multi-threading support, and using one of these two options enables
- either threading via plus/minus (if there is one - otherwise they act as
- before) or plus/minus *only* for threading. The behavior can be modified
- interactively from the <F4> hot-keyed configuration window. XRS displays
- blinking chevrons in place of the < or > thread symbols if either of these
- options are activated, so you know the plus or minus keys will thread and
- not read next or back as before.
-
-
-
-
-
- Always (Display Mailbag Handler on Exit)
-
- This is another new option in the 4.50 release, but one which you most
- likely would want to activate especially if you tend to read mail from
- multiple sources, or like to "put away" a partially read mailbag and open
- another, rather than use the default which XRS has always used (keeping
- the 'current' mailbag open and instantly available without unpacking). If
- you typically read mail from one source, and always read them sequentially
- then you probably do not need (or want) this.
-