BD Offline Study Pack: Pro Assets Zip That Works Without Internet
Build a Bangladesh-ready offline study pack that runs on low-cost Android phones and shared PCs. You will design a clean folder map, compress media without killing quality, add an index and schedules, and test on real devices. Two examples, a Dhaka mini case, metrics, and a step-by-step close included.
What this is, in plain English
An offline-first pack is a single zipped folder that works with zero internet. An asset is any file inside it, such as PDFs, MP3s, or flashcards. A codec is the rule set that compresses audio or video. OCR turns scanned pages into searchable text. A checksum is a short code that proves a file is intact after copying.
Compression without pain
PDFs
- Export text-based PDFs, not images.
- Embed fonts to avoid missing-character boxes.
- Target size: 75 to 150 dpi for images inside notes, keep text at full resolution.
- Run OCR on any scanned page so search works offline.
Audio
- MP3 mono at 96 to 128 kbps is clear for speech and saves space.
- Normalise peaks to −1 dB to avoid clipping.
- Keep tracks under 3 minutes to support short drills.
Zip discipline
- Keep each zip under 500 MB so Bluetooth or budget data plans can handle transfers.
- Use deflate compression, no exotic formats that old phones cannot unzip.
Integrity
- Generate a SHA-256 checksum list in /99_tools/checksum.txt. After copying, users can compare codes with a basic checksum app.
Make it searchable and navigable
Index
A simple offline index page links to the most-used items: “Start here”, “Day plan”, “Listening 2 min”, “Task 1 samples”, “Task 2 conclusions”, “Speaking shadowing”.
Anchors inside PDFs
- First page table of contents with page numbers.
- Bookmarks for sections such as “TF or NG drill” or “Data-to-language”.
Schedules
- Include a 14-day and a 28-day plan: 12 to 15 minutes listening and speaking, 20 minutes reading, 20 minutes writing.
- Each day links to exact files, not just topics.
Two worked examples
Example 1: Listening quick-notes pack
- Assets: 14 MP3s at 2–3 minutes each, 14 one-page cue sheets, one symbols sheet.
- Drill: “Shadow 30”, “List contour”, and “Clip and replace” with explicit targets: words per minute 110–150, fillers under 2 per minute.
- Why it works: tiny files, timed drills, measurable goals. Users do not need streaming.
Example 2: Writing Task 1 Academic set
- Assets: 20 annotated band-7 samples, a 4-page converter blueprint for change to, change by, gap, ranking, plus a percentage-point explainer.
- Drill: “4-sentence pack” with a three-number cap.
- Why it works: users learn pattern-to-sentence mapping without web lookups.
Mini case — Arif in Mirpur
Problem: patchy internet and a low-end Android made online courses unusable.
Intervention: Arif built a 420 MB zip using the folder map above, compressed audio to 112 kbps mono, OCR’d every scan, and added a 14-day index with daily links. He tested on a friend’s older phone and a shared PC.
Result: daily streaks held for 21 days, Task 1 time per response dropped from 25 to 15 minutes, and speaking fillers fell from 7 to 2 per minute.
Measurable tips
- Size budget: cap the whole pack at 500 MB; cap any single PDF at 15 MB.
- Access speed: open-to-first-page under 2 seconds on a 2 GB RAM phone.
- Audio loop: each MP3 loads in under 1 second inside a free player.
- Study cadence: 45 focused minutes per day, 6 days per week.
- Progress log: count completed drills, not hours. Aim for 18–24 drills per week.
Common mistakes
- Image-only PDFs that cannot be searched. Run OCR.
- Variable-bitrate stereo audio at 192 kbps that bloats size without clarity gains.
- Fancy zip formats that old Androids cannot open.
- Broken internal links because files were moved after you built the index.
- No “read me” or schedule, so users stall on day 1.
Edge cases and safe fixes
- iOS unzip limit: split the pack into 2 parts under 250 MB each.
- SD card permissions: place the zip in internal storage, then move folders; some phones block direct unzip to SD.
- Lab PCs with no admin rights: include a portable PDF reader in /99_tools/.
- Power cuts: print the 14-day schedule and two emergency drills per skill.
Tips and tricks
- Add a tiny search index PDF that lists every asset title with a page number or timecode.
- Put micro-cards at the top of each folder: “In 5 minutes” suggestions.
- Use color sparingly so printing is cheap and readable in grayscale.
- Keep a changelog in /00_READ_ME/ so users know what changed in v2.
- Test on the weakest device you can borrow. If it works there, it works everywhere.
To avoid
- Embedding web links as the only navigation. Assume zero internet.
- Overpacking rare accents or niche topics before core skills.
- Unclear file names like “final_new2.pdf”.
- Requiring specific paid apps. Link to free readers only.
Glossary
Offline-first: designed to work without internet access.
Asset: any file inside the pack, such as PDF or MP3.
Codec: method of compressing audio or video data.
OCR: optical character recognition that makes scans searchable.
Checksum: code that verifies a file’s integrity after copying.
Index page: a simple offline HTML or PDF hub with links to assets.
Next steps
Draft your folder map, pick 10 high-yield assets per skill, and export them using the size targets above. Build start_here.html, run OCR, generate checksums, and zip to under 500 MB. Test on two different phones and one PC. Ship v1, collect feedback, release v2 within one week.
- Actionable closing — How-to steps
- Plan the folder map and daily schedules in /00_READ_ME/.
- Collect assets and rename them with clear, ordered slugs.
- Compress PDFs and MP3s to the size targets; embed fonts and run OCR.
- Link everything from start_here.html with one-click access per day.
- Verify on two low-spec devices; fix any broken links and slow files.
- Protect with checksums and a printed quick-start sheet.
- Ship the zip under 500 MB and save a changelog for v2.
CTA: Block 90 minutes this week to build your v1 pack. Use the map above, hit the size targets, test on a weak phone, and share the zip with one study partner. Gather two improvement notes and publish v2 in seven days.