(Note: This is a long one)
Looking at the replies to this thread, and looking at the subject as a whole
from 30,000 feet, there are two thoughts I have on this:
1. On the topic of Network Printing within MPE, it's been woefully inadequate
since day 1, but for standard printing, and you don't wish to pay additional
dollars for a third party product, it does do the trick - just don't ask it to
do anything fancy. It's lack of LPD however, keeps it from competing in the
arena with other OS' network printing, including Windows NT (which has LPD).
2. Looking at the "what HP provides" versus "what the 3rd party market
provides" - again, from 30,000 feet - one has to ask: "Can, and does, HP wish
to be 'all things to all people'?" My answer has to an unequivical "No". It
cannot be all things to all people. HP tried that back in the 80s, and fell
flat on its face, as there was just too much to support. MM/3000 was an
example of a vertical market product, and it's now a "third party product" with
Exegesys handling it - again, HP cannot be all things to all people.
So, enter the third party market - and ODBCLink/SE is another classic example
of this. HP tried and failed to build an ODBC driver (PCAPI) that worked
properly. It tried in this case to be all things to all people, and could not
keep up. 32-bit Windows came along, and they still didn't have 16-bit access
right. They went to M. B. Foster (although they were not the only company to
have a working ODBC driver - they were just the ones that were very well known
at the time) for help. ODBCLink/SE is a stripped down, single-threaded driver
that does work, and lets one get their feet wet. One does not use this driver
for high-volume applications, but it does give the functionality to get
applications started. When an HP customer needs the seriousness of a
multi-threaded, high-performance driver, they go to the third parties that have
the drivers which handle this functionality. All of these third parties have
something different to offer, which arrive (hopefully) at a common goal -
robust, high-performance remote data access. Linkway works only in the
Image/SQL realm and does so admirably. MiniSoft goes direct to TurboIMAGE
through its own set of facilities, Foster gives both - and pricing is all over
the map -- but the thing here is, that all three strive to do the same thing,
remote data access - and the job of the third parties is to compete with each
other on the merits of their product, not compete with HP. There are also
non-ODBC connectivity products out there that might be better than ODBC is some
cases, such as LeeTech's CSF/AIM - again, another third party.
Network Printing? There are third party products out there that can do things
that HP's network printing cannot touch. HP's Network Printing could use some
improvements, and even with these improvements, it cannot touch ESPUL and other
products out there. Again, HP cannot be all things to all people - so the
third party market fills the void. When we (the third parties) see an
opportunity, it's our job to pounce on it, and do so quickly.
If HP keeps adding to MPE to fill these voids, several things are going to
happen. Support costs will rise, and HP will need to charge for this added
functionality due to a standard rule of running a business: You need to gain
revenue of some sort or another, to pay the people who are doing the work. One
thing that I've seen in the 3000 marketplace, on both sides of the desk, is
that we want it at low cost, or for free. Nothing in this world is free.
Samba/iX is distributed with FOS -- but HP can, if they wish, charge for
support - legally.
Network printing, ODBC access and other things that come bundled with FOS have
a cost (and a value) to HP -- but the third parties extend this functionality.
Right now, if you're looking for LPD, if you're looking for other extended
functionality, then you're going to be looking at a third party.
A case can be made for making Glance/iX part of FOS, because the tools to
measure performance on MPE are just not there. Glance/iX is an add-on product.
Should HP make this part of FOS, so that System Managers can adequately measure
performance on their systems? A case can be made for this. Of course, Lund is
out there with SOS, which not only does what Glance/iX does, but adds to it.
The same can be said for Disk Defragmentation. Windows 9x and Windows 2000
both come with a defragmenter. MPE has VOLUTIL's CONTIGVOL command, is what
many call "the poor person's defragmenter". Again, if one wants serious
defragmentation, Lund has Defrag/iX. Today's Windows NT 4.0 does NOT have a
defragmenter that works with NTFS, but one can purchase it from a third party.
Backup software (Orbit, HICOMP, etc.) is yet another example - each of the
aforementioned vendors do different things to accomplish the same task.
Comparing them would be like comparing apples and oranges, but they both will
back up the system. Again, they take TurboSTORE to the next level or two.
Application software example: Remember Visicalc on the 3000? Give me
QueryCalc any day of the week! VisiCalc, or even a ported early version of
1-2-3 could even compare with QueryCalc because QueryCalc can take data from my
database, and produce output that even Quiz cannot produce. Visicalc on the
3000 was, if my memory is not failing me, free or very low cost. QueryCalc has
a cost to it, but its cost far outweigh its benefits.
Security: Security/3000 and SAFE/3000 are the two third party products to fill
this void. HP's answer was again, late to the party, and inadequate to what
these vendors already provided.
Another tool: For years, HP's method for modifying a database was DBUNLOAD,
DBUTIL/PURGE, change schema, DBSCHEMA, DBUTIL/BUILD, DBLOAD. Oh, what a mess!
What a LONG process for big databases - talk about killing a weekend! Then
along comes Adager and DBGeneral which takes away the mess and does the job
thousands of times faster. Yes, HP came out with DBChange, but even this
product cannot keep up with the established two third party products. Mention
"database maintenance" and the first name you'll hear most often is Adager, and
the second is DBGeneral. DBChange is not even in on the radar screen, unless
you happen to be a user of it.
Heck, one can go on, and on, and on...
Overall, HP and other hardware/OS manufacturers should, and I think are
grateful, for the third party market. This includes Microsoft as well. There
is absolutely no company that can be all things to all people. Apple is a
company that tried it, and we saw what happened to them. Now they have a
thriving third party market as well. IBM was in this model too, a long time
ago - and found that the third party marketplace could do it better, so they
too, have a good third party market.
If HP were to try this again, we would probably see them tripping over
themselves, just like they did in the 80s. The third party market is here to
stay, and I think the HPs, the IBMs, the Microsofts, the Apples, and all the
others would do well to ensure that these companies are around to support them,
and not alienate them. The third parties are much more nimbler and can bring
product to market faster, because normally, they have a narrower focus, whereas
HP and others have an entire OS to support - a much larger focus.
Regards,
Joe
|
