Re: Serving images from databases

This WebDNA talk-list message is from

1998


It keeps the original formatting.
numero = 17051
interpreted = N
texte = >I need to jump into this discussion and make sure we're addressing the real >issue, which is a mechanism for WebCatalog to intercept browser requests >for binary files like pictures and such, and *not* the nice things we do >for PageMill owners. > >WebDNA was never designed to work with binary data -- its purpose is to >generate HTML text. Regardless of whether the NULL comes from a PageMill >file or a GIF file, WebDNA stops processing at that point. Without a total >rewrite+slowdown of the parser, that will remain unchanged. > >HOWEVER: Olin posted an excellent technique for allowing WebCatalog to >intercept picture requests and do useful things with them. Search >WebDNA-Talk at http://search.smithmicro.com for creative in the >subject line. The oldest message by Olin (at the bottom) contains a handy >tip regarding this subject. > >We have also noted the request for a special includeRAW tag that does no >NULL cleanup. Whether or not your browser can deal with such data will be >up to you -- HTML specs typically do not address embedded NULLS any >differently than WebCatalog does. Thanks Grant.I had asked earlier whether we were talking about binary data or text data, but no one said anything one way or the other, so I just assumed it was the wrong type of data to use the [include] tag for.There's a big difference between text data and binary data, but I don't know how to tell the difference simply by looking at the characters displayed in a file. Some text editors can open all kinds of files -- including binary files -- and once they are opened, there are characters displayed in the window of all these files. But just because there are characters displayed does NOT mean that data is text data ...I first experienced this situation while writing the code for the Rumpus admin interface in Typhoon. During that task, I had to deal with a file which was created by Rumpus. I opened it in Tex-Edit and it looked exactly like a tab-delimited text file ... so I thought Great, I can make WebCatalog change the data in this file because it's nothing more than a tab-delimited text file -- exactly the same as any other WebCatalog database file!But I was wrong.Even though the data I was viewing *appeared* to be nothing more than text, it was in fact binary data. I discovered the difference in these two data formats after making WebCatalog append or replace or delete records in that database and then seeing the data all screwed up afterwards. The results actually looked like WebCatalog was putting all kinds of strange and weird characters into that database file, sometimes in huge chunks, and I couldn't figure out what the heck was going on ...After contacting John O'Fallon (who wrote Rumpus), I learned that the data in that file was *not* just text, it was actually binary data. And since WebCatalog doesn't deal with binary data files, WebCatalog's attempted changes to that binary data ended up changing things alright, but not in any useful way, that's for sure.Thankfully, John changed the file format of that file -- from binary to text (or ASCII, I'm not even sure if I'm using the correct terminology here) -- and the Rumpus interface for WebCatalog and Typhoon was completed successfully.So ... to make a long story even longer, there's really no problem with the way WebCatalog is written, even with regard to PageMill. Instead, it's our own mistakes which are creating problems for us. After all, WebCatalog was never designed to [include] binary data, so we simply *cannot* expect WebCatalog to include that data properly.However, many thanks to Grant for his explanation, and also for putting this *special includeRaw* feature on the wish list, it might come in handy some day as WebCat users continue to push the envelope of WebCatalog's capabilities ... :)Sincerely, Ken Grome 808-737-6499 WebDNA Solutions mailto:ken@webdna.net http://www.webdna.net Associated Messages, from the most recent to the oldest:

    
  1. Re: Serving images from databases (Grant Hulbert 1998)
  2. Re: Serving images from databases ( 1998)
  3. Re: Serving images from databases (Christer Olsson 1998)
  4. Re: Serving images from databases (Grant Hulbert 1998)
  5. Re: Serving images from databases (Thomas Wedderburn-Bisshop 1998)
  6. Re: Serving images from databases (Dave MacLeay 1998)
  7. Re: Serving images from databases (Christer Olsson 1998)
  8. Re: Serving images from databases (Kenneth Grome 1998)
  9. Re: Serving images from databases (Grant Hulbert 1998)
  10. Re: Serving images from databases (Thomas Wedderburn-Bisshop 1998)
  11. Re: Serving images from databases (Bob Minor 1998)
  12. Re: Serving images from databases (Kenneth Grome 1998)
  13. Re: Serving images from databases (Christer Olsson 1998)
  14. Re: Serving images from databases (PCS Technical Support 1998)
  15. Re: Serving images from databases (Christer Olsson 1998)
  16. Re: Serving images from databases (Kenneth Grome 1998)
  17. Re: Serving images from databases (Sven U. Grenander 1998)
  18. Re: Serving images from databases (PCS Technical Support 1998)
  19. Re: Serving images from databases (Christer Olsson 1998)
  20. Re: Serving images from databases (Kenneth Grome 1998)
  21. Serving images from databases (Christer Olsson 1998)
>I need to jump into this discussion and make sure we're addressing the real >issue, which is a mechanism for WebCatalog to intercept browser requests >for binary files like pictures and such, and *not* the nice things we do >for PageMill owners. > >WebDNA was never designed to work with binary data -- its purpose is to >generate HTML text. Regardless of whether the NULL comes from a PageMill >file or a GIF file, WebDNA stops processing at that point. Without a total >rewrite+slowdown of the parser, that will remain unchanged. > >HOWEVER: Olin posted an excellent technique for allowing WebCatalog to >intercept picture requests and do useful things with them. Search >WebDNA-Talk at http://search.smithmicro.com for creative in the >subject line. The oldest message by Olin (at the bottom) contains a handy >tip regarding this subject. > >We have also noted the request for a special includeRAW tag that does no >NULL cleanup. Whether or not your browser can deal with such data will be >up to you -- HTML specs typically do not address embedded NULLS any >differently than WebCatalog does. Thanks Grant.I had asked earlier whether we were talking about binary data or text data, but no one said anything one way or the other, so I just assumed it was the wrong type of data to use the [include] tag for.There's a big difference between text data and binary data, but I don't know how to tell the difference simply by looking at the characters displayed in a file. Some text editors can open all kinds of files -- including binary files -- and once they are opened, there are characters displayed in the window of all these files. But just because there are characters displayed does NOT mean that data is text data ...I first experienced this situation while writing the code for the Rumpus admin interface in Typhoon. During that task, I had to deal with a file which was created by Rumpus. I opened it in Tex-Edit and it looked exactly like a tab-delimited text file ... so I thought Great, I can make WebCatalog change the data in this file because it's nothing more than a tab-delimited text file -- exactly the same as any other WebCatalog database file!But I was wrong.Even though the data I was viewing *appeared* to be nothing more than text, it was in fact binary data. I discovered the difference in these two data formats after making WebCatalog append or replace or delete records in that database and then seeing the data all screwed up afterwards. The results actually looked like WebCatalog was putting all kinds of strange and weird characters into that database file, sometimes in huge chunks, and I couldn't figure out what the heck was going on ...After contacting John O'Fallon (who wrote Rumpus), I learned that the data in that file was *not* just text, it was actually binary data. And since WebCatalog doesn't deal with binary data files, WebCatalog's attempted changes to that binary data ended up changing things alright, but not in any useful way, that's for sure.Thankfully, John changed the file format of that file -- from binary to text (or ASCII, I'm not even sure if I'm using the correct terminology here) -- and the Rumpus interface for WebCatalog and Typhoon was completed successfully.So ... to make a long story even longer, there's really no problem with the way WebCatalog is written, even with regard to PageMill. Instead, it's our own mistakes which are creating problems for us. After all, WebCatalog was never designed to [include] binary data, so we simply *cannot* expect WebCatalog to include that data properly.However, many thanks to Grant for his explanation, and also for putting this *special includeRaw* feature on the wish list, it might come in handy some day as WebCat users continue to push the envelope of WebCatalog's capabilities ... :)Sincerely, Ken Grome 808-737-6499 WebDNA Solutions mailto:ken@webdna.net http://www.webdna.net Kenneth Grome

DOWNLOAD WEBDNA NOW!

Top Articles:

Talk List

The WebDNA community talk-list is the best place to get some help: several hundred extremely proficient programmers with an excellent knowledge of WebDNA and an excellent spirit will deliver all the tips and tricks you can imagine...

Related Readings:

setting taxable to true (1997) Help name our technology! (1997) autosensing lanague selection (1997) WebDNA and CPanel Servers (2004) Trouble with [Math] (1998) Force a search at the default.tmpl page? (1997) [WebDNA] Triggers.db wiped completely blank (2011) [if] (2003) WCS Newbie question (1997) show all problem (1997) [OT] What System Software? (2003) Use of & in text areas.... (1998) RE: [WebDNA] Test (2008) New Command prefs ... (1997) wrong authentication (1998) Webcat2, WebCommerce, Mod 10 etc. (1997) [WebDNA] Retrieving a PDF from an email (2008) HideIf ip= OR ip (1998) Carts & Refering URLs (1997) Help! WebCat2 bug (1997)