Daynote mail 1999 week 10

Daynote mail: Week of 8 - 14 March, 1999

Daynote mail and replies -- Week 10

* Link to: last modified at 23:55 GMT+1, on 14.03.1999

Any quoted mail from reader feedback ends up here. This tends to reflect something of the ongoing discussions between myself and readers (and other web-daynote maintainers), provide tips, ask for help, and just be plain fun.

The sidebar "Daynotes"-link, beside each weekday, links to the corresponding day in the daynote file. The reverse linkage is also provided on the daynotes.

himself Mail your comments to: bo@leuf.com

Anyone who wishes correspondence to remain private should say so up front.

Quoted mail may be shortened and is usually based on my reply quotes. There may be some minor overlap between what's on the daynote page and what is given here in order to give correct context.

(BTW, week numbering is according to the Swedish calendar, which this year started January in week 53. "Current" weekday is of course based on GMT+1.)

Monday 08.03

Clas Kristiansson notes...

I've just sent a proposed "new" front page to Jerry Pournelle. No great changes, because I think it is good as it is. Simply a bit of polish. Since you may have more Swedish readers than the others in the "fourleaf", I send a Swedish site to you, apropos the others' convictions that Antiqua face (i.e. with serif) works fine as body text on screen.
http://www.idg.se/capdesign/serier.asp
I have tried to participate in the debate on WPTS, but the level is unfortunately not very high.
Fun that things are happening. (G)

The site can be interesting to look at even for those who don't know the language, although the actual article pages (with layout examples) I think Clas is referring to start at: http://www.idg.se/capdesign/artikelserier/caphtmok/index.htm

and

http://www.idg.se/capdesign/artikelserier/caphtmok/Fast.htm

Typography on the Web discussions can be interesting, if somewhat self-indulgant. For example, the above page suggests "freezing" the final rendered layout as bit images whenever the webauthor wants total control of the user display. This of course has some nasty consequences and side-effects, not least for e.g. content indexing, text copy&paste, and those users who for one reason or another cannot read the chosen bit-imaged text at the forced pixel size.

Still, it can be fun to look at.

Tuesday 09.03

I saw Clas' mail to Jerry Pournelle this morning; Jerry had published it as his lead on this week's Mail page (Monday). (Link adjusted from then "currentmail" to "week29".) It is quite true, as Clas implies, that serif (Times) faces render poorly on screen because of resolution constraints, although this depends how large they are set. Given the default font size that I know Jerry has on his screen display, I seriously doubt he would notice any significant difference in readability due to screen rendering between serif and sans.

At normal sizes however...

This line of text is (hopefully) rendered as (Times or default) serif at your default font size (typically 12pt).
This line of text is (hopefully) rendered as (Ariel or default) sans-serif at the same default font size (typically 12pt).

One thing that you might notice is that the serif (Times) may be noticeably smaller (and more compact) than the sans version, even at the same formal point size. For readability reasons, it should be the other way around. You might want to experiment and change your default fonts/sizes to improve screen readability in relation to your own screen. It is one reason I CSS-specify Century Schoolbook on many of my web pages, since it renders slightly larger and less compact. (And overriding my usual CSS-specified typeface, I have set today's text in Comic Sans. Just for another comparison.)

Clas Kristiansson comments about the URL sent earlier:

The pages are old and I fully agree with you. It is idiotic to place text as GIF images, even though I know that GP (major Swedish daily newspaper) has experimented with it. The reason I sent the link was more to show how linears (sans-serif) work better in body text on a screen, but I suppose the subject is beaten to death by now.
Guess who grew 10 meters... I had no idea he (Jerry Pournelle) would publish the letter. Thanks for the friendly words. I am very happy.
Starting to think... Maybe my book about Web Typography would work even on the US market, once it is done. Haven't seen anything like it. Americans tend to write at le-e-e-ngth (paid by word perhaps?), and I prefer the short and concise.

Bob Thompson had this to say apropos eFax (see Daynote page)...

Well, this is just another fax gateway service. The sender still has to send the fax using existing dial-up fax protocols. What I want is for IETF to define a standard for IP faxing. This would allow a sender to address a fax to an IP address or hostname (ideal for recipient companies who have full-time Internet connectivity) or to an email address (ideal for individuals, either at home or at their desks at work).
When a company bought a new fax machine, it'd have two connectors on the back. A standard POTS connector for a regular phone line, and an Ethernet connector to put it on the network. PC fax software would allow one to enter the recipient in three ways: standard phone number, IP address/hostname, or email address.
Existing fax protocols are unusable because, among other reasons, they are isochronous. But that's a holdover from the early days of faxing, when the scanning device and the output device were actually linked in tandem.
There's no need for that nowadays. A simple buffer is all it takes. The fact that IETF has not already defined a set of IP fax protocols makes me think the perhaps the situation is more complex than I appreciate.

I sometimes wonder how long Fax will be with us at all.

Wednesday 10.03

Tim Werth answered my call on Outlook comments:

Bo,
I sent Jerry Pournelle some info on Outlook back when he had just started using it extensively (seems like last summer or spring, can't remember for sure). Jerry had some specific questions on how to do some things that most people don't usually use Outlook to do w/mailing lists. You might want to search his web site for references to Outlook, he had a lot of discussion there at one time when he was asking a lot of questions.
However, you know as well as I do it can be rather difficult to find things on Jerry's site. I don't think he has a "search" function built into it yet like Bob Thompson has done w/his site.
I have used Outlook since it first came out as Outlook 97 (which was v. 8.0) and have used many interim releases (8.02, 8.03, 8.04, all versions of Outlook 97) including Outlook 98 extensively (which is currently @ v.8.5.5603). I haven't played w/Outlook 2000, beta 2 much yet. I've been reluctant to load Office 2k, beta 2 on my main system but I think I will load it at home w/a good backup of everything soon. The general consensus is don't even bother w/Outlook 97. It has enough bugs that it's not worth messing w/at this point in time. Just install Outlook 98 or go straight to Outlook 2k. Outlook 98 is a free upgrade anyway so long as you have a licensed copy of Office 97 so there really isn't much reason to start w/Outlook 97 anyway. From the reviews I have read of Outlook 2k the main BIG change over 98 is that it supports VBA to some extent. All previous versions used some "VBA Scripting" or something like that. I've never messed w/it but a friend of mine has tried to use it a little and hasn't had anything good to say about the scripting so far.
The main thing I like about Outlook is that it integrates e-mail and PIM into one program. Before Outlook I never used Schedule + because I just didn't like keeping two memory hogs open on my machine at the same time (e-mail client and Schedule +). With Outlook I have access to my mail, calendar, contacts, to do list, and notes. While Outlook is still a bit of a memory hog that doesn't bother me as much now that I have upgraded both my computers at home and work and have 128 MB of RAM in both.
The main thing about Outlook that is confusing came about w/Outlook 98 when MS implemented the "Internet only" version and a "Corporate/Workgroup" version. Each version has some unique features that differentiate it from the other. For example you access most of your settings in the "Internet only" version thru Tools/Accounts whereas in the "Corporate/Workgroup" version similar info is accessed thru Tools/Services and there is no "Accounts" option under Tools. In order to access an Exchange server you will have to install the "Corporate/Workgroup" option. Each version has some things I like but both also have some annoying features as well.
There are a lot of little rules of thumb that I do every time that I install Outlook but this has probably gotten long enough. Hope some of this is useful to you, if you have any questions let me know.
Thank you for your detailed mail. I have a vague memory of those OL discussions, and will dig for them in due time. I will probably also get back to you on specifics.

I had Outlook97 on my system originally, but refrained from reinstalling it when I not long ago reinstalled my Office97 in a leaner and meaner configuration, because frankly it was not useful in the 97 version. For various reasons I had been unable to install an OL98 in its place. I am soon going to do a complete Office 2000 installation and go with that.

Apparently there is still some "VBA sscripting" left in OL2000, having to do with "forms".

Thursday 11.03

Clas Kristiansson about the seminar in Arvika:

Congratulations by the way. I requested an evaluation of your course on Tuesday. All were full of praise, but also said:
"We would have wanted this in the beginning of our course." and "He should have taught us HTML."
I am happy they even after the novelty thought what I said was sensible and insightful. My own conclusion was that they had learned a fair amount of HTML-syntax, but never really grasped the point of HTML. One doesn't really appreciate why it is so difficult to produce a visually pleasing HTML page until one realizes that it was not the visual that was the original idea.

Friday 12.03

More from Clas...

This was also my idea when I started the course. I was careful to point out that HTML only described the content, while the visual is set by CSS. I guess I was not clear enough. How did you manage it?
I don't recall exactly, I didn't properly store my own method meta tags. :)
Seriously, however, I started in a general way with the concept of "document" and what this means in different contexts, from paper to digital form.
Mainly, I made the point that in paper form, the "visual" is the only thing you get. Then I took up digital storage and the perspective that one does not then know in advance how the content will be used and interpreted. A useful trick was to take up a "blind-browser" that reads aloud document content (automatic indexing is a more usual example, but I think a bit too abstract to be useful here), and explain how content mark-up tags make it possible to convey document meta-information to someone who cannot see the visual conventions about how we see that it is new subheadings, item lists, etc.
From there I went on to show the basic principles of how an html document is constructed, and how (just a few) tags allowed a sort of minimalistic content mark-up, with examples of how the "blind-browser" might read the text differently depending on the tags used.
Maybe you should write a book about web design?
Who knows. I had contemplated making a publisher a proposal on XML around New Year, but other things came in between.
I like the idea behind DHMTL where the page is divided into three different parts (HTML, CSS and Javascript) but the problem is that so many only see the limitations, i.e that only a few web-browsers support it. Then there is another problem in the browsers' interpretation of HTML. The <P> tag is rendered "wrongly" (meaningless new line). <Hx> tags also give "incorrect" line breaks, etc... etc...
That part is "unspecified" -- browsers can do whatever they want.
Pity the code was not written by the typographically aware. :)
Then we should not even begin to talk about <FONT>, the only point of which is to format text rendering in HTML. How did this come about? Was not the idea to only describe content?
Browser-wars. Netscape and Microsoft. It was then largely about providing a mass of new features that almost exclusively concerned visual aspects.
Teaching HTML seems to become a process aimed at first clearing away missconceptions. Then one can teach the "correct" rules. Rules that don't agree with what HTML 3.2 actually looks like in practice. Food for thought.
Oh I don't know. With focus on the content tags it is not that bad. The problem is that it is not always clear in many tag description lists which are content ("logical") mark-up and which are visual ("feature") mark-up.
Starting to more and more long for somewhere on the Web where this is discussed in a constructive way, where one starts with the basic principles.
If it's lacking, we'll have to start one. "The Typo Gang".
From one thing to another. Found a very strange Javascript. Totally useless, but I got a fright when I saw it first. Worth a look. I'll say nothing more in advance, but check out www.forsmark.uu.se/~dt97sebr I think such things are fun.

Saturday 13.03

(busy with NT4 installation oddities)

Sunday 14.03

* Bob Thompson responds to my question about NT4 and PnP:

I've never had much problem from a resource conflict perspective getting NT installed. I suspect that at least some of your problems may be due to your BIOS/chipset settings for the PCI bus. There are a few things to keep in mind:

Although NT4 doesn't have full PNP support, it does have a pnp service that I believe interacts with CMOS settings for bus arbitration issues. I've never been able to find out any details about the Windows NT PNP service anywhere, including TechNet.

...

And he goes on with further details, and solicits advice from other readers.

Thanks for the advice. And maybe the readers will respond.
I use a (large) notebook, so the add-a-card method isn't really relevant. PC-cards (formerly PCMCIA) are supported well enough, although a brief experiment with a card-manager software add-on failed rather miserably last year. Not clear why, and actually, come to think of it, this was probably under Win95 in any case.
I'll be doing a base reinstall this evening, and with the last attempt fresh in my mind, try to do this one with better strategy. I also know more about memory requirements this time round and can plan accordingly.

Bob responds:

Okay, I just learned something. I was under the impression that vanilla NT was essentially unusable on a notebook--no power management, no support for PC Card devices, etc. etc. I thought that various notebook manufacturers, notably DEC, had done their own drivers to adapt NT to their particular models.
Yeah, well notebooks have developed a lot of non-vanilla stuff and the OEMs do need to include a number of NT drivers, and it's true that power management really not all there in NT. But given the machine has the oomph and the RAM, there's no inherent problem. What I have is apparently known in the US as Prostar 7800x, and by and large it's really nice to run with NT. (64M RAM).
I just dislike the way NT is playing shell games with its half-PnP and allocations. It takes too long to figure out what's going on based on the unhelpful event logs and sundry alerts.

Later...

Well, there's not much question that NT isn't notebook-friendly. I've never tried to run it on one. And I don't blame you for being annoyed with NT's lack of support for PNP and PCI arbitration.

Three installs and I still can't figure out what is going on. Tomorrow, I'll try again.

Back to top -- Week list