cannot build opencobol-1.1 under cygwin
in [Cobol]
|
Prev: Spin-off
Next: In praise of compiler writers
From: Richard on 17 Apr 2010 18:21 On Apr 18, 12:53 am, "Pete Dashwood" <dashw...(a)removethis.enternet.co.nz> wrote: > > Given a choice I would prefer no UI at all and the job just gets done. > > > Regular jobs get scripted and cronned. > > Other than that, I think your points are fair. I'd like to know how you > psychically communicate with a computer so it doesn't need an interface,but > if you have managed to pull that off, "Good Luck!" Stuff can be started on bootup using rc.sysinit or rc.local and that is done using a text editor which may be GUI or text. Some tasks are started at intervals automatically using cron. Again a text editor is used to set this up. For example my backup runs every night to sync changes to an off-site machine. I never have to click options or make selections, it works out what has changed and sends me an email if there are problems (for example if connection lost). Other stuff (on my clients machines) runs continuously. A monitor program checks the log files and the input queues and, in the first place, takes action if the process is not functioning correctly, if this fails to correct then it sends out emails. It is about making computers do what they do best rather than having people do it. Of course the 'interface' is script files and logs and emails and that is an editor, when required, whether GUI or text based. Of course some stuff _requires_ a user interface, it is only my preference (as originally stated) that there be none. For example one system (a couple of decades ago it seems) was for warehousing from a client who could not change his invoicing system. The requirement was to take the printed invoice and key it into a nice UI so that the warehousing/despatch system could process it. I didn't do that (well actually I did do a UI as well), I wrote a program that emulated a printer that picked out all the relevant details, reprinted the invoice and created the warehouse order. We connected the machines using modems (pre internet) and they just 'printed' invoices.
From: Binyamin Dissen on 18 Apr 2010 03:47 On Fri, 16 Apr 2010 21:48:01 +0200 Fred Mobach <fred(a)mobach.nl> wrote: :>Richard wrote: :>> DRI's FDISK also had the ability to set different cluster sizes. FAT :>> is a particularly poor file system for random file access. So COBOL :>> indexed files were very slow. To access a specific block the system :>> had to start at the directory entry and follow the chain of FAT :>> entries to find the required clusters. By formatting the disk with :>> clusters that were 8Kb instead of 2Kb that MS-DOS did, random access :>> to largish (~1Mbyte) indexed files was 3 times faster. :>In the late '90-s I tested a COBOL program with an index-sequential file :>on MS-DOS and GNU/Linux on the same (multi-boot) computer. The Linux :>ext2 filesystem was very helpful in order to speed the program by a :>factor 3 compared to the execution on MS-DOS. :-) Was there a native ISAM in DOS? Were these identical COBOLs, or perhaps from different vendors each with their ISAM implementation. -- Binyamin Dissen <bdissen(a)dissensoftware.com> http://www.dissensoftware.com Director, Dissen Software, Bar & Grill - Israel Should you use the mailblocks package and expect a response from me, you should preauthorize the dissensoftware.com domain. I very rarely bother responding to challenge/response systems, especially those from irresponsible companies.
From: Anonymous on 18 Apr 2010 12:19 In article <82tpr7FubrU1(a)mid.individual.net>, Pete Dashwood <dashwood(a)removethis.enternet.co.nz> wrote: >docdwarf(a)panix.com wrote: >> In article >> <84aeba39-2d2b-4fff-96b6-e35b7ad3fcf1(a)y17g2000yqd.googlegroups.com>, >> Richard <riplin(a)Azonic.co.nz> wrote: >> >> [snip] >> >>> In 1983 Paul Allen gave a talk at COMDEC (I think it was) about 'the >>> next version (2) of MS-DOS' where he promised various features such >>> as 'Help' (which did appear in MS-DOS 5). One thing that was >>> promised was command line editing and history (which was already in >>> Unix and DRI's CP/M-86 and other). >> >> My memory is, admittedly, porous, perhaps e'en moreso when it comes to >> two-and-a-half decades-old operating systems... but I seem to recall >> being able to get back at least the last command line Enter'd by >> pressing PF3. > >Yes, I thought that when I read Richard's post, but I saw little point in >responding, as my posts just seem like a red rag to Richard, and it is not a >particularly important part of the overall anti-MS diatribe. (note - I recall being taught that male bovines are colorblind and the hue of the cape is dictated by tradition... that being said, on to an attempt at humor and what may now be an obscure literary reference) As the Germans used to say, Mr Dashwood... aiee, Toro, Toro, andale, Toro! Vamanos a la Muerte Por La Tarde! DD
From: Fred Mobach on 18 Apr 2010 12:28 Binyamin Dissen wrote: > On Fri, 16 Apr 2010 21:48:01 +0200 Fred Mobach <fred(a)mobach.nl> wrote: > > :>Richard wrote: > > :>> DRI's FDISK also had the ability to set different cluster sizes. > :>> FAT is a particularly poor file system for random file access. So > :>> COBOL indexed files were very slow. To access a specific block the > :>> system had to start at the directory entry and follow the chain of > :>> FAT entries to find the required clusters. By formatting the disk > :>> with clusters that were 8Kb instead of 2Kb that MS-DOS did, random > :>> access to largish (~1Mbyte) indexed files was 3 times faster. > > :>In the late '90-s I tested a COBOL program with an index-sequential > :>file on MS-DOS and GNU/Linux on the same (multi-boot) computer. The > :>Linux ext2 filesystem was very helpful in order to speed the program > :>by a factor 3 compared to the execution on MS-DOS. :-) > > Was there a native ISAM in DOS? Were these identical COBOLs, or > perhaps from different vendors each with their ISAM implementation. The ISAM was part of the runtime system of the compiler. To create a fair comparation I compiled the program with MS-COBOL 3.0 in a MS-DOS environment, ran the same executable in MS-DOS and via DOSEMU in GNU/Linux. The difference were the partitions on the same disk. The MS-DOS partition at the outside of the disk and the GNU/Linux partition in the middle. So the MS-DOS had more than one advantage : - native instead of emulation on Linux - faster DISK throughput. -- Fred Mobach - fred(a)mobach.nl website : https://fred.mobach.nl .... In God we trust .... .. The rest we monitor ..
From: Pete Dashwood on 20 Apr 2010 07:58
docdwarf(a)panix.com wrote: > In article <82tpr7FubrU1(a)mid.individual.net>, > Pete Dashwood <dashwood(a)removethis.enternet.co.nz> wrote: >> docdwarf(a)panix.com wrote: >>> In article >>> <84aeba39-2d2b-4fff-96b6-e35b7ad3fcf1(a)y17g2000yqd.googlegroups.com>, >>> Richard <riplin(a)Azonic.co.nz> wrote: >>> >>> [snip] >>> >>>> In 1983 Paul Allen gave a talk at COMDEC (I think it was) about >>>> 'the next version (2) of MS-DOS' where he promised various >>>> features such as 'Help' (which did appear in MS-DOS 5). One thing >>>> that was promised was command line editing and history (which was >>>> already in Unix and DRI's CP/M-86 and other). >>> >>> My memory is, admittedly, porous, perhaps e'en moreso when it comes >>> to two-and-a-half decades-old operating systems... but I seem to >>> recall being able to get back at least the last command line >>> Enter'd by pressing PF3. >> >> Yes, I thought that when I read Richard's post, but I saw little >> point in responding, as my posts just seem like a red rag to >> Richard, and it is not a particularly important part of the overall >> anti-MS diatribe. > > (note - I recall being taught that male bovines are colorblind and > the hue of the cape is dictated by tradition... that being said, on > to an attempt at humor and what may now be an obscure literary > reference) > > As the Germans used to say, Mr Dashwood... aiee, Toro, Toro, andale, > Toro! Vamanos a la Muerte Por La Tarde! S�, Papa, pero a veces yo tambien soy un toro... Quieres ir pescar? :-) Pete. -- "I used to write COBOL...now I can do anything." |